Biosynthesis and Engineering of Lasso Peptides

Abstract

Lasso peptides, typically 16-21 amino acid (aa) long, comprise a class of ribosomal peptide natural products that share a slipknot-like structure. The lasso structure features an N-terminal cyclized ring and a C-terminal tail that threads through the ring and is sterically prevented from slipping by bulky sidechains positioned on both sides of the ring. This highly constrained structure is extremely resistant to thermal, chemical, and proteolytic degradation.

Microcin J25 (MccJ25) is a well-studied member of the class of lasso peptides. It was isolated from a strain of <italic>Escherichia coli</italic> from the human gut microbiota. It is produced and released during periods of nutrient depletion to kill some species of Gram-negative bacteria including <italic>E. coli, Salmonella, and Shigella</italic> strains. It has been demonstrated that a four-gene cluster is responsible for the biosynthesis and export of MccJ25 - mcjA encodes a 58 aa precursor protein; mcjB and mcjC encode maturation enzymes that convert the McjA into the mature lasso peptide, mcjD encodes an exporter of MccJ25 and is also an immunity factor. A homologous gene cluster has been discovered for another member of the lasso peptide class, capistruin.

In my research, we applied a protein engineering approach to the study of MccJ25 for two aims. First, we sought to obtain fundamental insights into the maturation process of lasso peptide and the permissible sequence space of lasso peptides. Second, we sought to engineer and enhance the antimicrobial activity of MccJ25.

I constructed an engineered gene cluster for regulable and robust production of MccJ25 in <italic>E. coli</italic>, which serves as a platform for engineering the lasso peptide. Working together with the undergraduates Wai Ling Cheung and Jakub Rajniak, we performed truncation and mutation scanning studies of the leader peptide of MccJ25 and capistruin to determine the minimal requirement for the leader peptide and the role of the conserved threonine residue at the penultimate position of the leader peptide. Our current hypothesis is that the leader peptide, especially the penultimate threonine residue, serves as a recognition element in docking McjA into the maturation machinery. Homology analysis has shown that McjB exhibits sequence similarity to a family of proteases and McjC exhibits similarity to a family of enzymes catalyzing amidation reaction in an ATP-dependent manner. Using mutagenesis, we confirmed the key residues in McjB and McjC for MccJ25 biosynthesis. In addition, we demonstrated that fusions of McjB and McjC with flexible linkers are capable of producing MccJ25. Remarkably, even a 151 kDa tripartite fusion between the ABC transporter and the two maturation enzymes is capable of producing and exporting MccJ25. Furthermore, I demonstrated that the <italic>E. coli</italic> chaperonin GroEL and its cofactor GroES are essential for the folding of McjB and thus the biosynthesis of MccJ25.

I isolated functional variants of the lasso peptide MccJ25 by two parallel approaches - selecting variants from computational analysis and screening saturation mutagenesis libraries. The goals of these studies are to improve the inherent antimicrobial activities of MccJ25, gain insights into the peptide structure-activity relationship, and understand the sequence diversity possible within the lasso peptide framework. Among the highly ranked sequences from the computational design algorithm, we found that six of the eight variants selected, each containing two or three amino acid substitutions, were successfully produced by <italic>E. coli</italic> and three retained antimicrobial activity, though at a reduced level relative to wild-type MccJ25. By screening our saturation mutagenesis libraries, we discovered nearly 100 variants of MccJ25 that retain antimicrobial functions and found several variants with nearly 5-fold increase in potency toward MccJ25-susceptible strains. In these studies, we observe that a large amount of amino acid variation is tolerated by the lasso peptide fold. Given the extraordinary resistance of lasso peptides to proteases and thermal unfolding, they appear to be a promising scaffold for therapeutic peptide applications.

I have shown that unnatural amino acid can be incorporated into MccJ25 and participate in azide-alkyne cycloaddition chemistry. I have conjugated MccJ25 to a bacterial membrane-permeabilizing peptide by thiol-maleimide chemistry to carry MccJ25 across the membranes of a broader spectrum of bacteria. To explore alternative ways of using the peptide antibiotic MccJ25, I programmed the gene regulation of MccJ25 production and export to a cross-species quorum sensing circuit to allow the producing host to detect and kill other bacterial species. However, the detection system is not sensitive enough and requires more engineering efforts as future work. I also programmed MccJ25 biosynthesis genes into the M13 bacteriophage genome such that the non-lytic phage would infect bacteria and induce the biosynthesis of toxic MccJ25 within host cells. The killing efficiency of these engineered M13 phage requires further validation.