Synthesis, identification, and characterization ofpeptide natural products and non-natural derivatives

Natural products are a diverse class of metabolites produced by living organisms to serve higher functions like signaling, defense, and growth development. Exploitation of these compounds as pharmaceutical drugs or agrochemicals has been successful in the past. Nowadays, many drugs are either natural products or derivatives of natural products. However, with the ongoing antimicrobial resistance crisis, the need for new solutions is growing. Therefore, the discovery of novel natural products is experiencing a remarkable comeback. One source for new natural products is the entomopathogenic bacteria of the genus Photorhabdus and Xenorhabdus. Their genomes comprise many natural product-encoding biosynthetic gene clusters. Novel natural product discovery through bioinformatic analysis and metabolite detection upon activation of the biosynthesis-encoding genes is limited by low titers, unknown absolute configurations, and unpredictable biosynthesis differences like substrate specificity. Total synthesis can compensate for these issues and enable access to larger quantities of pure compounds. Peptide natural products like nonribosomal peptides can be synthesized using the well-established solid-phase peptide synthesis methodology. Therefore, the synthesis of natural products, especially peptides, can be much easier than isolation from its natural source. The compounds' conformation and absolute configuration can be validated easily and rapidly through analytical comparisons of the natural product and the synthetic standard(s). After successful structure validation, the pure synthetic compounds can be used for bioactivity testing. In this thesis, natural products were elucidated, and their bioactivity and functions were determined through their chemical synthesis. The first project of this thesis was to investigate the ecological function and bioactivity of several derivatives of the GameXPeptides and to identify their molecular target by chemical synthesis of suitable derivatives. The GameXPeptide class comprises cyclic peptides of four to six amino acids, found in most Xenorhabdus and Photorhabdus strains. The synthetic approach successfully combined solid-phase peptide synthesis and in-solution cyclization to yield head-to-tail cyclopeptides. Depending on the compounds' solubility, they were either purified via preparative high-performance liquid chromatography (HPLC) or filtration using C18 solid-phase extraction cartridges. The synthetic natural and evolution-derived derivatives (1-16 and 19-30) showed insecticidal activities in dependency of their structure and the target organism. In contrast, pull-down assays with the synthesized biotinylated derivatives (31-36) were not yet successful. Additionally, two new-to-nature, arginine-containing derivatives cyclo(D-Arg/L-Leu/D-Phe/D-Leu/L-Leu) (37) and cyclo(D-Arg/L-Leu/D-Val/L-Val) (38) were synthesized successfully. The arginine-pentapeptide 37 showed low antimicrobial activity against Mycobacterium smegmatis strain mc(2)155 with a minimal inhibitory concentration of 64 µg/mL. In the second project the structures of ishipeptide A (L-Leu/D-Leu/L-Val/D-Ala/L-Gln/D-Tyr/L-Leu/D-Val/cyclo(D-Ser/L-Leu/D-Leu/L-Glu), 39) and B (L-Leu/D-Leu/L-Val/D-Ala/L-Gln/D-Tyr/L-Leu/D-Val/D-Ser/L-Leu/D-Leu/L-Glu, 40) from Xenorhabdus ishibashii were elucidated. The linear ishipeptide B (40) was synthesized via solid-phase peptide synthesis. In contrast, synthesizing the cyclodepsipeptide, ishipeptide A (39), required a combination of solid-phase peptide synthesis and subsequent in-solution cyclization. Both compounds were synthesized successfully and used to verify the structures of the natural ishipeptides via co-injection and comparison of NMR (nuclear magnetic resonance) spectra. Especially the comparison of the C-Hα fingerprint region in HSQC (heteronuclear single quantum coherence) spectra was used to verify the structures and absolute configurations. Furthermore, Marfey's method was employed for ishipeptide B (40) and confirmed the configuration of the three amino acids D-alanine(4), L glutamic acid(12), and D serine(9). Besides the previously identified weak antimicrobial activity of ishipeptide A (39) against Micrococcus luteus, no further bioactivity was identified. Two new lipocitide derivatives, acetyl/D-Leu/L-Ala/D-Cit (56) and propionyl/D-Leu/L-Ala/D-Cit (57) from Xenorhabdus stockiae KJ12.1 were detected upon promotor exchange and activation of the lipocitide synthetase-encoding gene XEKJV2_07845 as part of the third project. Their structures were verified through co-injection with synthetic standards, which were synthesized by solid-phase peptide synthesis. Their acyl chain of C2-3 is much shorter than that of the previously discovered lipocitides from Xenorhabdus bovienii with C12-18 acyl chains. Both new compounds were lipocitide B derivatives with the C-terminal free carboxylic acid. The bioactivity of the new lipocitide derivatives 56 and 57 still needs to be investigated and compared to lipocitides from other Xenorhabdus species. Furthermore, the 4 chlorothreonine modifying-enzymes found encoded in Xenorhabdus stockiae KJ12.1 were over-expressed in the wild type, its Δhfq mutant and both lipocitide-producing strains, but no production of 4 chlorothreonine-modified natural products were observed. Hence it is not clear whether these enzymes can incorporate an additional amino acid through the termination module of the lipocitide synthetase. However, it can be assumed that they are not related, and the motif of the termination module is to epimerize the C terminal amino acid of the lipocitides since the bioactive function as an insect immunosuppressant has been shown. Another novel class of lipopeptides, the lipovasines, was discovered within the fourth project of this thesis. Activation of the nonribosomal peptide synthetase-encoding gene XEKJV2_07115 in Xenorhabdus stockiae KJ12.1 and heterologous expression of this gene in E. coli led to the discovery of six lipovasine derivatives (58-63). For the synthesis of the main product (R)-3-hydroxyoctanoyl/D-Val/L-Lys (59), the chiral (R)-3-hydroxyoctanoic acid was synthesized via enantioselective samarium(II)-iodide-mediated Reformatsky-type reaction of n-hexanal and 3 bromoacetylated chiral Evans-auxiliary. The main product was successfully synthesized using the (R)-3-hydroxyoctanoic acid and used to verify the natural lipovasine derivative 59 via co-injection. For the other derivatives, the structure and absolute configuration are assumed to be equal to the main product. Additionally, the gene XHOM_07515 from Xenorhabdus hominickii, which is located at the same position as the lipovasine-encoding gene, was investigated by promoter activation and liquid chromatography-mass spectrometry (LC-MS) analysis, but no novel compound was detected. However, heterologous expression in E. coli led to the discovery of a new peak, which is currently being investigated. This thesis's fifth and last project revolved around the biosynthetic production of C-terminal glutarimides by engineering nonribosomal peptide synthetases. The thioesterase domain of a dialkyl-indigoidine-producing NRPS (dIndS) showed the ability to cyclize C-terminal amino acids such as glutamine, asparagine, and diaminobutyric acid when fused to other nonribosomal peptide synthetases with the corresponding adenylation domain specificity. A subset of the biotechnologically produced C-terminal glutarimides (64-68) was successfully synthesized by solution-phase coupling of (S)-aminoglutarimide and the respective acyl or peptidyl residue. Even though the synthesized peptidyl-glutarimides were partly racemized, the main synthetic products were shown to be the biotechnological products. Investigation of the cereblon-binding potential of the synthesized C terminal glutarimides revealed only activity for compound myristoyl/L-Gln/L-Ala/L-cGln (68) with alanine next to the glutarimide and its diastereomer, which even showed a stronger binding potential. Since the absolute configuration of the diastereomer is not resolved, it is not certain whether the glutarimide or one of the other chiral positions of the diastereomer is epimerized. To further investigate the influence of the amino acid residue next to the glutarimide, the leucine derivative myristoyl/L-Gln/L-Leu/L-cGln (69) was synthesized. The leucine-containing compounds 69a/b have not yet been investigated for their cereblon-binding activity. Nevertheless, the biotechnological production of C-terminal glutarimides (64-68) has been confirmed and can, therefore, be optimized and employed to produce enantioselective C terminal (S)-glutarimides.