In this work the potential of the hydroxynitrile lyase from $\textit{Arabidopsis thaliana}$ (AtHNL) for the enantioselective synthesis of industrially important cyanohydrins was demonstrated by two different principles: reaction- and protein engineering. The enantiomeric excess of this enzymatic reaction is strongly compromised by a non-selective side-reaction resulting in racemic cyanohydrins and thus lowering the enantiomeric excess of the biotransformation. This non-selective product formation is influenced by pH, temperature and the water content of the reaction medium. For industrial applications aqueous or aqueous-organic reaction systems are used where the racemic product formation is suppressed by lowering the pH below pH 5 and running the process at $\le$ 10 °C. However, both approaches are not feasible with AtHNL, since the enzyme is rapidly inactivated below pH 5. In order to enable the use of AtHNL for the enantioselective synthesis of industrial important cyanohydrins two strategies were developed in this work. The first strategy concerned the suppression of the racemic product formation by reaction engineering. Thereby, a micro-aqueous reaction medium (buffer-saturated mono-phasic methyl $\textit{tert}$. butyl ether) was used to suppress the undesired side reaction, which resulted in a good enzymatic activity and high enantioselectivities for several industrially important cyanohydrins. Efficient cyanohydrin syntheses were reached in the established reaction system by application of whole cells and the use of immobilized enzyme (e.g., enzyme adsorption at celite particles, encapsulation in solgel, cross-linking of enzyme aggregates), espectively. Best results were obtained using whole recombinant $\textit{E. coli}$ cells as well as celite-adsorbed isolated AtHNL. For the synthesis of (R)- mandelonitrile both preparations showed comparable catalytic efficiencies, process-and storage stabilities, resulting in a maximal productivity of ~1.6 g (R)-mandelonitrile per g wet cell mass (after 6-8 recycling steps). The second strategy was the development of a stabilized AtHNL variant which is stable under acidic conditions and thus applicable in industrially established aqueous-organic two-phase reaction systems. Based on the highly similar HNL from $\textit{Manihot esculenta}$ (MeHNL), which is stable under these conditions, the protein [...]