Biochemische und kristallographische Charakterisierung der eukaryotischen tRNA-Guanin-Transglykosylase (TGT)

In this work, based on the already known preparation conditions of the two separate subunits of the mouse TGT, an expression and a purification protocol for the combined preparation of the functional mouse TGT heterodimer could be established. For this purpose, an expression vector was first constructed based on a pETDuet vector. Since several E. coli strains proved unsuitable, an expression vector was finally developed using a V. natriegens strain. The removal of nucleic acid impurities while preserving the protein-protein contact of the heterodimer, which is based on non-covalent interactions, presented the greatest challenge. This problem was finally solved with a combination of using lithium chloride during affinity chromatography as well as a Phenyl-Sepharose™ matrix. In collaboration with the MarXtal laboratory of the Department of Chemistry at the University of Marburg, several initial crystal conditions were identified, some of which included the polyoxometalate TEW. All initial conditions were reproduced in-house and, in case of successful reproduction, were subjected to an optimisation study. The structure of the QTRT1/2 heterodimer could finally be solved de novo in the presence and absence of the additive TEW. In both cases, as in the crystal structure of the QTRT1 subunit, the α helix 0 can be observed. This indicates that the structural element is not a crystallographic artefact, but is also formed under physiological conditions in the absence of tRNA. Using the TEW condition, a structure in complex with queuine could also be obtained by co-crystallisation. Since the citrate-containing condition without TEW is readily unsuitable for soaking due to a citrate molecule bound in the active site, a method was developed to expose the active site in the first step and introduce a ligand into the active site by soaking in the second step. Thus, the binding pose of queuine in the binding pocket of QTRT1 was confirmed by comparing two different methods, co-crystallisation and soaking. In the crystal structure of mouse TGT in complex with Queuin, Van-der-Waals contacts could be observed between Val161 of QTRT1 and the 7-deazapurine scaffold as well as the dihydroxycyclopentene residue of Queuin. Val161 is replaced by cysteine in most bacterial TGTs. Moreover, the absence of a side chain at Gly232 creates enough space for the large-volume dihydroxycyclopentene residue in the binding pocket of QTRT1. In more than 90 % of bacterial TGTs, this glycine is exchanged by valine. Thus, both amino acids contribute to the substrate specificity of the eukaryotic TGT. The enzyme kinetic as well as mass spectrometric studies have shown that the mutation of Tyr 354 of QTRT2 to phenylalanine leads to a significant destabilisation of the QTRT2 homodimer. This leads to a shift of the equilibrium towards the functional heterodimer. Presumably, the mutation also has an effect on the functional heterodimer, but to a lesser extent. The mutation of Ser 41 as well as Tyr 354 of QTRT2 also leads to a lower turnover rate of the heterodimer. Based on the enzyme kinetic studies, it was shown that selectivity between bacterial and eukaryotic TGT can be achieved by addressing the ribose33 binding pocket with large-volume residues. In fact, inhibitors with lin-benzoguanine backbone substituted at position 2 showed hardly any inhibition of the eukaryotic TGT, whereas they inhibited the bacterial TGT partly in the single-digit nanomolar range. In addition to selected inhibitors, some known substrates of the eukaryotic TGT were also characterised enzyme-kinetically. Finally, a reaction mechanism could be postulated which explains the irreversible incorporation of queuine, preQ1 and 7 deazaguanine. Irreversible incorporation always occurs when the negative charge generated by the nucleophilic attack of the side chain of Asp279 on the C1 atom of the ribose sugar cannot be stabilised by the substrate in position 34.