Doctoral Thesis
Refine
Year of publication
- 2014 (3) (remove)
Document Type
- Doctoral Thesis (3) (remove)
Has Fulltext
- yes (3)
Is part of the Bibliography
- no (3)
Keywords
- Proteindesign (3) (remove)
Institute
Within this thesis the protein engineering, immobilization and application of enzymes in organic synthesis were studied in order to enhance the productivity of diverse biotransformations. Article I is a review about Baeyer-Villiger monooxygenases (BVMO) and provides a detailed overview of the most recent advantages in the application of that enzyme class in biocatalysis. Protein engineering of a former uncharacterized polyol-dehydrogenase (PDH) identified in the mesothermophilic bacterium Deinococcus geothermalis 11300 is described in Article II. Article III covers the combination of one PDH mutant with a BVMO in a closed-loop cascade reaction, thus enabling direct oxidation of cyclohexanol to ε-caprolactone with an internal cofactor recycling of NADP(H). Article IV and Article V report a process optimization for transamination reactions due to a newly developed immobilization protocol for five (S)- and (R)-selective aminotransferases (ATA) on chitosan support. Furthermore, the immobilized ATAs were applied in asymmetric amine synthesis. In Article VI, an ATA immobilized on chitosan, an encapsulated BVMO whole cell catalyst and a commercially available immobilized lipase were applied in a traditional fixed-bed (FBR) or stirred-tank reactor (STR), and were compared to a novel reactor design (SpinChem, SCR) for heterogeneous biocatalysis.
In this thesis several methods of protein engineering were applied to explore and increase enantioselectivity and thermostability of certain carboxylesterases and to better understand the relationship between sequence, structure and function. For example, we were able to confirm the observed conservation of motifs like GX/GGGX and GXSXG, which was reported earlier. Yet, even more details were revealed and some were designated in numbers. However, the numbers may vary when even more sequences will be available, but the trend should remain the same. The power of the ABHDB lies in the information available throughout the very diverse and quite large superfamily. Structural equal positions can be easily compared and analysed regarding mutations, correlated mutations, prevalence etc., and visualization is simplified through direct output with YASARA software. The ABHDB was the first structural alignment of such a large number of known enzymes of the alpha/beta-hydrolase fold superfamily. With methods of rational protein engineering we were able to show that there is little flexibility of the GGG(A)X motif for the eukaryotic enzyme PLE 1 and the natural motif appears to be a good solution for high activity and enantioselectivity of PLE 1 in the conversion of tertiary alcohol esters. In a focused directed evolution approach, we were able to identify variants of BsteE with moderate, but significantly increased enantioselectivity in the kinetic resolution of tetrahydrofuran-3-yl acetate, and hence, were able to proof that the concept of ‘small but smart’ libraries is an efficient way to find improved mutants, while the screening effort was reduced. Moreover, we were able to show that the domain exchange enhanced the thermostability of BsubE, while expression level and activity were maintained or increased, respectively. Despite the great achievements and possibilities at present, we are not yet in the position to directly modify the gene to alter the structure in a complete predictable fashion to improve functional properties as imagined by Ulmer (1983). Nevertheless, substantial changes can be targeted and as demonstrated in this work, several broadly applicable methods are at hand. Furthermore, bioinformatics tools play an essential role in planning of experiments, analysis and interpretation.
Carbamoylasen und Hydantoinasen werden im „Hydantoinase-Prozess“ großindustriell zur Synthese enantiomerenreiner Aminosäuren eingesetzt. Durch Verwendung einer Hydantoin-Racemase kann eine Ausbeute von theoretisch 100 % erreicht werden. In dieser Arbeit wurde untersucht, wie die beteiligten Enzyme mittels Protein-Design, an neue Substrate angepasst werden können. Dabei wurde die Carbamoylase aus A. aurescens einem gene-shuffling mit den eng verwandten beta-Ureidopropionasen aus S. kluyveri und A. tumefaciens unterzogen. Weiterhin wurde ein durch Dockingexperimente gestütztes rationales Design durchgeführt und auf dieser Basis gezielte Mutationen im aktiven Zentrum der Carbamoylase eingebracht, sowie fokussierte Bibliotheken des aktiven Zentrums erzeugt. Es konnte die Akzeptanz von β-Aminosäurederivaten als Substrat in dieser Carbamoylase erreicht werden. Das aktive Zentrum einer Hydantoinase aus Ochrobactrum spec. wurde ebenfalls mittels rationalem Design vergrößert um eine Akzeptanz von 5,5-disubstituierten Hydantoinen zu ermöglichen. Die Aktivitätsmessung gegenüber den neuen Substraten wurde in einem mehrstufigen Assaysystem durchgeführt. Dieses System basierte auf einem auf Ammoniummangel beruhenden Wachstumsassay, einem NADH Assay mit Glutamat-Dehydrogenase als Kopplungsenzym, sowie auf direktem HPLC Nachweis.