Our team developed a novel approach for creating customized peptides that overcomes the above mentioned limitations by engineering synthetic Non-Ribosomal Peptide Synthetases (NRPSs). NRPSs are organized in modules recognizing one specific amino acid substrate and catalyzing the formation of a peptide bond between the amino acid substrate and the nascent peptide chain. Notably, non-ribosomal peptide synthesis does not require mRNA to direct the sequence of amino acid monomers incorporated into the growing peptide and is therefore not limited to proteinogenic amino acids (reviewed in [7]).

In order to show the applicability of custom NRPSs for synthetic peptide production, we created a set of synthetic NRPSs composed of varying modules derived from the Bacillus brevis Tyrocidine NRPS pathway. We successfully produced a number of di- and tri-peptides composed of proteinogenic and non-proteinogenic amino acids in E. coli. In order to simplify the detection, purification and quality control of those peptides, we invented an in-vivo labeling procedure for non-ribosomal peptides by fusion of the corresponding, synthetic NRPSs to an engineered blue pigment module (IndC).

Furthermore, we developed a procedure for improving the functionality of NRPS modules by shuffling single domains derived from different species and by engineering entirely synthetic domains. As proof of concept, we shuffled domains within the unimodular blue pigment NRPS IndC or introduced synthetic domains derived from consensus sequences across different species. To this end, we devised a novel cloning strategy termed HiCT (BBF RFC 99), which enables a cost-efficient and rapid assembly of synthetic NRPS module libraries.
We were able to engineer a library of IndC modules exhibiting varying blue pigment production efficacies in combination with different NRPS activating PPTases. Remarkebly, a subset of these synthetic IndC variants showed a broader PPTase specificity compared to their natural counterpart.
Furthermore, we identified the location of domain borders and optimized linker regions used for introduction of synthetic domains. By quantitative dynamic modelling, we obtained a more detailed understanding on the synthesis efficiency of the different IndC variants in different contexts and investigated the influence of indigoidine production on bacterial growth.

In order to enable the synthetic biology community to easily engineer custom NRPSs producing user-defined peptides, we developed a standardized framework for the production of non-ribosomal peptides. The framework consists of (1) the NRPS Designer, a software tool for the in-silico design of user-defined NRPSs, (2) a platform for standardized cloning and expression of NRPSs in different bacterial hosts and (3) a quality control procedure for the validation of NRP production. We documented detailed instructions for applying our framework in BBF RFC 100.

To exemplify the infinite application possibilities of NRPSs, we developed an sustainable procedure for gold recycling from electronic waste using the non-ribosomal peptide Delftibactin produced by Delftia acidovorans, which is known to precipitate elemental gold from gold-ion solutions [8]. Again we applied mathematical modelling in order to quantitatively investigate the feasibility of this approach for an industrial-scale recycling of gold. Currently, we are working on transferring the complete delftibacin production pathway into E. coli in order to increase yield and lower costs of delftibactin production.

We believe, that our project demonstrates the power of non-ribosomal peptide synthesis and that our standardized framework will enable the synthetic biology community to use this power to address many of the challenges of our century.