Background
Foreigner DNA fragments are usually needed to be introduced into bacteria. They can be constructed into replicative plasmids, plasmids are inherently unstable, limiting the application of this strategy in researches with high requirements for the stability of results. As a better alternative, exogenous DNA fragments should be irreversibly incorporated into chromosome or other stable DNA molecule inside the cell (Heap et al., 2012). For this sake, DNA fragments of interest could be combined with positive selection markers, typically using an antibiotic resistance gene, allowing desired recombinant cells to be selected and isolated after a single-crossover or double-crossover recombination. Positive selection markers can also be removed through specific procedure, such as FLP- or CRE-mediated recombination (Meyers, Lewandoski, & Martin, 1998), but frt or loxp site will leave behind inevitably (Lee et al., 2001). Generally speaking, for the sake of biosecurity or eliminating unanticipated effects, residual of resistance genes or frt and loxp scars on the genome is not expected during functional genome analysis and metabolic engineering research.
This issue can be overcome using a two-step selection method by combining positive-negative selection cassette. In particular, a positive-negative selection cassette could be introduced and recombinants are selected according to positive selection markers, then exogenous DNA fragments are transformed into cells to replace the above-mentioned cassette and the final desired recombinants are selected against the counter-selection gene.
Many positive selection markers are highly efficient so recombinants could be easily acquired in the first recombination event. In contrast, counter-selection markers are not so satisfactory. The main problem of counter-selection markers is that high background causes great difficulties for counter-selection, especially in the case of low recombination efficiency (Imam et al., 2000; Muyrers et al., 2000; Y. Zhang, F. Buchholz, J. P. Muyrers, & A. F. Stewart, 1998). Several research groups are being dedicated to developing and optimizing better counter-selection markers(DeVito, 2008; Ma et al., 2015; Wang et al., 2014; Wong et al., 2005) and some novel counter-selection marker genes have been reported. Among them tetA -sacB and inducible toxins system are the most outstanding (Khetrapal et al., 2015; Li, Thomason, Sawitzke, & Costantino, 2013). It is worth mentioning that inducible toxins system possesses the highest selection stringency (1.1×10-8 to 3.31×10-8) which even comes up to that of positive antibiotic markers in some strains (Khetrapal et al., 2015). Despite these achievements, excellent counter-selection markers are still scanty. In view of the situation, we are motivated to develop robust counter-selection marker and provide new strategy for optimizing counter-selection efficiency.
Lysis gene E of bacteriophage PhiX174 was discovered to have a role in lysing Escherichia coli in 1966 for the first time (Hutchison & Sinsheimer, 1966), and the subsequent researches shows that expression of this gene is sufficient to cause lysis of E. coli (Henrich, Lubitz, & Plapp, 1982; Young & Young, 1982). Lysis gene E codes for a 91-aa membrane protein with hydrophobic moieties at its N-terminal end that oligomerizes into a transmembrane tunnel structure(Blasi, Linke, & Lubitz, 1989; Witte & Lubitz, 1989). This specific tunnel structure is associated with the fusion of the inner and outer membranes of Gram-negative bacteria, leaving behind empty envelopes devoid of all cytoplastic content (Hajam, Dar, Won, & Lee, 2017; Witte, Wanner, Sulzner, & Lubitz, 1992). Non-living Gram-negative bacterial empty envelopes caused by E protein are knows as Bacterial Ghosts (BG), representing a potential platform both for potent candidate vaccines and for technical applications in white biotechnology (Hajam et al., 2017; Langemann et al., 2010; Won, Hajam, & Lee, 2017). BGs of many Gram-negative strains have been prepared successfully, including Salmonella typhimurium , Salmonella enteritidis ,Bordetella bronchiseptica , Vibrio cholerde ,Mannbeimia haemolytica , Pseudomonas aeruginosa , etc (Langemann et al., 2010). In view of the universal and potent killing effect mediated by lysis protein E in Gram-negative strains, we focus our effort on developing lysis E gene as a universal marker gene to try to expand the existing counter-election tool chest, which is especially useful for host strains lack of facile genetic tools.
This study reports on the introduction of lysis gene E of PhiX174 to create “seamless” genetic manipulations in Gram-negative strains. Under the control of the lambda pL promoter and AraC/PBAD inducible system, lysis gene E can effectively promote the death of E. coli and Serratia marcescens , and the selection stringency of lysis gene E is similar to the next best reported tetA-sacB double-negative selection system. By co-expressing lysis gene E and kilnegative gene (kil-sd-E counter-selection cassette) through a classic RBS (Ribosome Binding Site) sequence, the selection stringency frequency is even comparable with the best counter-selection system, inducible toxins system. By introducing araC gene harboring plasmid, the selection stringency frequency of kil-sd-E arrives at the level of 10−9 both in E. coli and inS. marcescens , This is the lowest selection stringency frequency for counter-selection reported so far, exceeding the best counter selection system, inducible toxins system, several fold.
Due to the universal lethal of expression of lysis protein E, counter selection cassette based on lysisE or kil-sd-E could be applicable to many, if not all, Gram-negative strains. And the strategy reported in this article of combining two short counter-selection genes could be used to increase the counter-selection efficiency of other negative markers, which can provide more powerful tools to the genome editing toolbox.