Amide Bond Formation in Nonribosomal Peptide Synthesis:The Formylation and Condensation Domains

Nonribosomal peptides are of outstanding pharmacological interest, since many representatives of this highly diverse class of natural products exhibit therapeutically important activities, such as antibacterial, antitumor and immunosuppressive properties. Understanding their biosynthesis performed b...

Full description

Saved in:
Bibliographic Details
Main Author: Schönafinger, Georg
Contributors: Marahiel, Mohamed (Prof. Dr.) (Thesis advisor)
Format: Doctoral Thesis
Language:English
Published: Philipps-Universität Marburg 2007
Subjects:
Online Access:PDF Full Text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Nonribosomal peptides are of outstanding pharmacological interest, since many representatives of this highly diverse class of natural products exhibit therapeutically important activities, such as antibacterial, antitumor and immunosuppressive properties. Understanding their biosynthesis performed by multimodular mega-enzymes, the nonribosomal peptide synthetases (NRPSs), is one of the key determinants in order to be able to reprogram these machineries for the production of novel therapeutics. The central structural motif of all peptides is the peptide (or amide) bond. In this work, two different amide bond forming catalytic entities from NRPSs were studied: The condensation (C) and formylation (F) domains. Firstly, the N-terminal F domain of LgrA1, belonging to the biosynthetic machinery required for the production of linear gramicidin was biochemically characterized. Using F-A-PCPLgrA1 in in vitro experiments, its acceptor substrate specificity towards the template-bound branched aliphatic amino acids valine, isoleucine and leucine was identified. From sequence alignments with other formyltransferases, N10-formyl-tetrahydrofolate (N10-fTHF) was expected to serve as formyl donor in these reactions. This molecule was chemoenzymatically produced and successfully used in the assays. Interestingly, its isomer N5-fTHF was also accepted even though the apparent product formation speed was 18-fold slower. The necessity of a formylated starter unit for the inititation of the nonribosomal biosynthesis was then tested with the dimodular F-A-PCP-C-A-PCPLgrA1-2 enzyme. It was shown that no dipeptide was produced, unless a formyl donor substrate was provided – in which case the formylated dipeptide could be detected. Obviously, the N-terminal formylation of linear gramicidin is critical for its bioactivity, where it functions as an ion channel in a head-to-head dimeric complex that is able to penetrate bacterial cell membranes. Secondly, the bidomain enzyme PCP-CTycC5-6 was used as a model system for C domain studies. Its previously discovered ability to cyclize the PCPTycC5-bound hexapeptide DPhe-Pro-Phe-DPhe-Asn-Gln was further investigated. The head-to-tail connectivity of the cyclic product was proven by MSn-spectrometry, and the substrate specificity for this reaction was probed in the context of six other oligopeptide substrates, three of which were accepted. Mutational studies were furthermore carried out to scrutinize previously suggested models for the C domain catalyzed reaction. According to one model, a conserved histidine residue of the C domain is involved as a base in the catalytic process. Even though the according alanine and valine mutations produced here led to well-folded soluble proteins, their cyclization activity was found abolished. Crystallization screens with the wild-type apo-PCP-CTycC5-6 enzyme afforded one promising condition which was further optimized in collaboration with the crystallographic group of Prof. Dr. Essen (Marburg). Thus, for the first time, the structure of a bidomain enzyme from nonribosomal peptide synthetases was solved – shedding light on so far unknown aspects of inter-domain communication. The relative arrangement of both domains was interpreted as a state in which the apo-PCP domain seeks interaction with either a different nonribosomal domain or a phosphopantetheine transferase. As expected, the highly variant so-called linker region that connects both entities was found unstructured, yet weakly interacting with both domains’ surfaces. Interestingly, a buffer-derived sulfate ion was seen in direct proximity to the proposed active site of the C domain. It is hypothesized in this work that this tetrahedral anion resembles the transition state of the amide bond forming reaction. Consequently, the transition state would be stabilized by the electrostatic environment of the C domain rather than by chemical base catalysis.
Physical Description:116 Pages
DOI:10.17192/z2008.0068