In this study we present an electrotransformation protocol optimised to enhance the electro-competency of clinically relevant Gram-negative bacteria using the cationic cyclic peptide PMBN. We showed that PMBN enhanced the electrotransformation frequency of several Gram-negative bacteria in a dose dependent manner, achieving improvement of up to 100-fold. We also demonstrate how this protocol is readily applicable to enhance commonly used molecular biology methods involving DNA and DNA-protein complex delivery into bacterial cells.
The study of large genetic elements in bacteria (e.g. operons, genomic islands and virulence or antibiotic resistance plasmids) requires the successful delivery of large plasmid constructs in a clinically relevant host background and this still remains a challenge in the field. Cloning and introduction of such large constructs is relatively straightforward using laboratory K-12 E. coli or by a yeast assembly system24. However, the delivery of large constructs into most clinical strains can be a challenge, as demonstrated here. Conventionally triparental bacterial conjugation25 could be used, but this requires re-engineering of the host vector and is restricted to recipient strains that lack conjugation exclusion factors26,27. Here we have demonstrated successful delivery of a large plasmid construct (25 kb) encoding a gene cluster responsible for the synthesis of a heterologous O antigen in a pathogenic E. coli background. We have also demonstrated a 5-fold increase in the yield of chromosomal mutants in a widely studied pathogenic S. Typhimurium strain. The improvement in the mutagenesis rate we observed is likely attributed to two factors: a) enhanced DNA delivery into the cells, as demonstrated by our study findings and b) PMBN could also function as a membrane permeabilising agent to counteract TolC mediated efflux of L-arabinose, thereby allowing more DNA template delivery as well as higher-level of induced expression of the lambda-Red proteins, resulting in an increased allelic exchange rate.
Our improved mutagenesis protocol using PMBN could also enable construction of chromosomal gene mutant libraries. This could be done by: 1) acquiring a target gene knockout with a positive-negative selection system28 by the standard lambda-Red mutagenesis method; 2) generating target gene templates with random mutations via error-prone PCR; 3) performing a second allelic exchange reaction using our PMBN enhanced mutagenesis protocol to deliver target gene DNA amplicons with random mutations; and 4) selecting putative flawless allelic mutants on an appropriate negative selection media. An improved yield of mutants by employing our PMBN method will enable efficient construction of a mutant gene library at the chromosomal level so that the study of the mutant alleles is not subject to plasmid copy number or expression differences. Lastly, we have shown that PMBN enhanced electrotransformation can be used to optimise the delivery of DNA-protein complexes (e.g. for TIS library construction) and we expect that this could also potentially be extended to developing a delivery system for ribonucleoprotein (RNP)-mediated CRISPR genome editing29 in bacterial cells for efficient targeted mutagenesis.
Interestingly, our findings with PMBN and pathogenic E. coli and S. Typhimurium strains, did not extend to a clinical P. mirabilis strain. P. mirabilis is naturally resistant to PMBN-induced antibiotic sensitising, in contrast to various E. coli and S. Typhimurium strains that were shown to be sensitised towards a set of antibiotics following treatment with PMBN12. PMBN was shown to readily bind to E. coli and Salmonella membranes at a level of 3-5 µg/mg dry weight bacteria, whereas P. mirabilis membranes do not bind PMBN13. This would imply that the enhanced electrotransformation efficiency we observed requires PMBN to bind to the membrane and we thus predict that this protocol would have limitations with other naturally polymyxin-resistant bacteria such as Proteus vulgaris, Morganella morganii, Providencia stuartii and Serratia marcescens, but could be used to enhance electrotransformation frequency of polymyxin susceptible bacteria, such as Klebsiella pneumoniae, Klebsiella oxytoca, Enterobacter cloacae, Enterobacter agglomerans Acinetobacter calcoaceticus, Pseudomonas aeruginosa and Pseudomonas maltophilia12.
We also observed that growing pathogenic E. coli in the presence of PMBN made the electrocompetent cells more vulnerable to high voltage. We speculate that this is likely due to PMBN enhancing membrane damage induced by electroporation (thought to transitionally cause formation of aqueous pores on membranes and even membrane rupture30) by binding to lipid A and disrupting the asymmetric bilayer of the bacterial OM. In future, the survival rate of PMBN-grown cells could be improved by exploring conditions that promote efficient repair of cell membrane damage in the cell. It is important to note however that the enhanced electrotransformation frequency by PMBN is likely not solely due to its membrane disrupting effects, as short exposure of bacterial cells to PMBN at early/mid-exponential phase instead reduced electrotransformation efficiency. This, combined with the fact that PMBN treatment had no effect on plasmid chemical transformation frequency shown here and previously10, while it increased the electrotransformation frequency of treated mid-exponential phase cells, suggests that bacterial growth in the presence of PMBN may be required to induce membrane modifications. Indeed, PMBN was previously reported to be sensed by the bacterial two component PhoPQ system31 and induce PmrAB-mediated gene expression32 to modify lipid A in the OM. Such modifications might be related to the enhanced electrotransformation frequencies we observed and a mechanistic investigation into PMBN-induced OM modifications constitutes a future investigation. Irrespective of the exact nature of the PMBN induced membrane changes, they seem to be transient and reversible as cells no longer exposed to PMBN revert to lower electrotransformation frequencies similar to PMBN untreated cells.
To our knowledge, this is the first study to show enhanced DNA delivery into clinically relevant Gram-negative bacteria using a cationic membrane binding peptide. The method presented here can be readily adapted for improving other genetic manipulation techniques involving the delivery of exogenous DNA into natural isolates. Therefore, we propose PMBN as a powerful Gram-negative membrane electropermeabilisation adjuvant for achieving enhanced DNA delivery into bacterial cells by electroporation.