In this thesis a protein is characterised, which was formerly isolated from cell suspension cultures of Coleus blumei and identified as hydroxyphenylpyruvate reductase (Kim, 2004). Furthermore two crystal structures, one of the native protein and one complexed with NADP+, are presented in this work. Hydroxyphenylpyruvate reductase (HPPR) catalyses the reduction of hydroxyphenylpyruvates to the corresponding lactates. The enzyme is thought to be involved in the biosynthesis of rosmarinic acid in plants. After the putative HPPR cDNA was cloned and transformated into E. coli the protein was expressed by bacteria adding a His-Tag to the N-terminus of the protein. This His-Tag was used for purification of the H(P)PR. During characterisation of the putative H(P)PR some differences to the native protein were observed. The differences included a higher temperature optimum, different Km-values for the cosubstrates NADH and NADPH and most importantly differences in the substrate specificity. The Km-value for the proposed natural substrate 4-hydroxyphenylpyruvate (pHPP) with the native enzyme was determined as 10 and 80 µM respectively. For the same substrate the Km-value was 16.6 mM using the heterologously expressed H(P)PR. This is significantly higher compared to the values observed for the native enzyme. It should be mentioned that for pHPP substrate saturation could not be reached as high substrate concentrations could not be analysed by the available methods. For the heterologously expressed H(P)PR substrates that were better accepted than pHPP could be identified. The best accepted substrates were hydroxypyruvate with a Km-value of appr. 1 mM followed by glyoxylate with 2 mM. By determining two H(P)PR sequences from two Coleus blumei plants a mutation in the H(P)PR gene could be ruled out. As similar cDNAs/genes are known from different plants which do not produce rosmarinic acid as metabolite, doubts emerged if it is really the major role of the putative H(P)PR in vivo to be involved in rosmarinic acid biosynthesis. For further investigation two new protein sequences from two plants of the Solanaceae family were identified. Their amino acid sequence is 83.1% and 83.4% identical to the H(P)PR sequence. The protein was crystallised and two protein structures were determined. One is a structure of the native protein and in the other H(P)PR is complexed with NADP+. The overall structure shows the characteristic two-domain structure of the D-isomer-specific 2-hydroxyacid dehydrogenase protein family. The smaller of the two domains is responsible for the binding of the substrate, whereas the other one binds the cosubstrate. The catalytic cleft is located between both domains. For other structures a flexibility of the domains was reported as the domains move towards each other after binding of substrate and/or cosubstrate. For the two H(P)PR structures no significant change in the opening angles was observed. Comparison with similar structures revealed high similarities to the structures of a D-glycerate dehydrogenase from bacteria and a human glycerate dehydrogenase/ hydroxypyruvate reductase although the similarities on the basis of nucleotide and amino acid sequences are rather low. This similarity could give first indication to the main in vivo role of the enzyme. As no crystal structure with substrate was achieved by co-crystallisation, several substrates were docked into the catalytic cleft. For large substrates with a phenyl ring no clear binding mode was observed. However for smaller substrates a distinct binding mode was indicated, which is consistent with substrate binding modes found for other D-isomer-specific 2-hydroxyacid dehydrogenases. The results obtained from enzyme characterisation and comparison with similar protein structures indicate a different main role of the enzyme in plants. Nevertheless, it should be emphasised that an involvement of the enzyme in rosmarinic acid biosynthesis as a side function is still conceivable. A possible main function could be the involvement in photorespiratory processes. This could explain the appearance of similar enzymes in plants, which do not accumulate rosmarinic acid as secondary metabolite. From other plants a cytosolic hydroxypyruvate reductase (HPR-2) was described, using preferably NADPH as cofactor to reduce hydroxypyruvate to D-glycerate. HPR-2 is thought to be involved in photorespiration. Unfortunately no sequences or crystal structures of the HPR-2 are known to date. However, this theory is not supported by the fact that the mRNA, which was used to establish the H(P)PR-cDNA, was isolated from cell cultures of Coleus blumei that were grown in the dark. It is not expected that photorespiration occurs in the dark. Therefore genes coding for proteins involved in this process should not be expressed. Further investigation is necessary to prove the in vivo role of the cloned enzyme but first characteristics indicate the enzyme being a putative HPR-2.