Substratspezifität und Funktionalität von Epimerisierungsdomänen in der nichtribosomalen Peptidsynthese

A broad variety of structurally diverse and pharmacologically relevant natural compounds is synthesized by multimodular nonribosomal peptide synthetase (NRPS) assembly lines. Each module within these megaenzyme systems comprises a number of domains that have distinct catalytic functions and interact specifically to ensure the incorporation and modification of one building block. In many cases the products contain D-amino acids, the majority of which is generated in an equilibrium-forming reaction by module integrated epimerization (E) domains from the corresponding peptidyl carrier protein- (PCP-) bound L-intermediates. Within the scope of this work a chemoenzymatic approach was developed in order to investigate the native substrate specificity of four E domains derived from different synthetases. For this purpose, corresponding PCP-E bidomains were isolated at gene level. Two of these constructs (TycB3- and FenD2-PCP-E) originated from elongation modules containing so-called peptidyl-E domains, while the two other (TycA- and GrsA-PCP-E) were derived from initiation modules accordingly including aminoacyl-E domains. The recombinant apo-proteins were overproduced and modified with a selection of peptidyl-CoAs utilizing the promiscuous 4´-phosphopantetheine- (Ppant-) transferase Sfp. Enzyme-bound peptidyl-S-Ppant reaction products of the subsequently occurring stereoinversion catalyzed by the E domains were cleaved chemically and analyzed for their L to D-ratio. Individually calculated kobs values and final L/D-equilibria observed for each reaction showed that all four E domains tolerate various peptidyl-S-Ppants. Apparently, the substrate tolerance of an E domain does not correlate with the specificity of the module (activated amino acid) it originates from. The epimerization efficiency can equally be affected by alterations of the substrate at the C-terminus (directly thioester-bound amino acid) and the N-terminal part. All tested E domains converted N-methylated substrates with great activity. The most interesting observation was the ability of aminoacyl-E domains to epimerize peptidyl-substrates. Additionally, it could be demonstrated that the condensation (C) domain of TycB1 is able to elongate peptidyl-substrates transferred from TycA. In several systems the C-termini of synthetases harbor E domains, which in these cases seem to facilitate the ordered interaction with the following enzyme and the directed transfer of intermediates. Therefore, the second part of the work presented here aimed at elucidating this additional functional role. Bimodular derivatives of the tyrocidine synthetase B were constructed at gene level and the overproduced recombinant proteins were characterized. Initially, sequent tryptic proteolysis and high-resolution mass spectrometry were optimized for the direct interrogation of intermediates attached to the TycB3-PCP domain of the wild-type TycB2-3 and the E domain exchange protein TycB2-3-AT.CAT/EtycA. The E domain exchange significantly affected dipeptide- (Phe-Phe-) formation. In addition, the two mentioned proteins and a version of TycB2-3 fused to the communication-mediating (COM) domain of TycA were applied in product formation assays with the acceptor TycB1 to corroborate E domain impact on intermodular NRPS action. Significant functional differences in terms of in trans interaction and misinitiation were observed between the two types of C-terminal E domains. Peptidyl-E domains seem to be optimized for regulating the progression of peptide bond formation, epimerization, and intermediate transfer to the downstream module, whereas aminoacyl-E domains impair upstream condensation and cause misinitiation. These perceptions will be of fundamental importance for the successful generation of novel bioactive compounds by biocombinatorial rearrangements of NRPSs involving E domains.