CRMAGE: CRISPR Optimized MAGE Recombineering

Abstract

A bottleneck in metabolic engineering and systems biology approaches is the lack of efficient genome engineering technologies. Here, we combine CRISPR/Cas9 and λ Red recombineering based MAGE technology (CRMAGE) to create a highly efficient and fast method for genome engineering of Escherichia coli. Using CRMAGE, the recombineering efficiency was between 96.5% and 99.7% for gene recoding of three genomic targets, compared to between 0.68% and 5.4% using traditional recombineering. For modulation of protein synthesis (small insertion/RBS substitution) the efficiency was increased from 6% to 70%. CRMAGE can be multiplexed and enables introduction of at least two mutations in a single round of recombineering with similar efficiencies. PAM-independent loci were targeted using degenerate codons, thereby making it possible to modify any site in the genome. CRMAGE is based on two plasmids that are assembled by a USER-cloning approach enabling quick and cost efficient gRNA replacement. CRMAGE furthermore utilizes CRISPR/Cas9 for efficient plasmid curing, thereby enabling multiple engineering rounds per day. To facilitate the design process, a web-based tool was developed to predict both the λ Red oligos and the gRNAs. The CRMAGE platform enables highly efficient and fast genome editing and may open up promising prospective for automation of genome-scale engineering.

Figure 1. Schematic cartoon of CRMAGE system.

CRMAGE consist of two plasmids, pMA7CR_2.0 expresses λ/RED β-protein and CRISPR/Cas9 protein that are inducible by L-Arabinose and aTetracyline respectively. The β-proteins are co-expressed with dam, which gives a mutS mutator phenotype, and cas9 is expressed in a operon with recX, which blocks the repair of double strand breaks. The second plasmid (pMAZ-SK) contains an aTetracyline inducible sgRNA used for selection against the wild type sequence, as well as a self-eliminating circuit that targets its own backbone to enable plasmid recycling and sequential recombineering. Upon L-rhamnose induction (and aTetracycline for cas9 induction), a tracrRNA that combines with two crRNAs, arranged in a natural CRISPR in order to target the origin (Ori) and the antibiotic marker (Kanamycin).

Figure 2. Wild type killing efficiency.

The graph represents the killing of a wild type strain compared to a galK mutant over time. For each time point the cells were plated and the ratio between WT/galk* was calculated.

Figure 3. Contribution of RecX in CRMAGE efficiency.

The graph shows the effect of recX expression on the CRISPR/Cas9 killing activity after λ Red recombineering. In pMA7CR_2.0 (represented by the dark grey bar on the right) cas9 was expressed in a synthetic operon together with recX. In pMA7CR2, cas9 was expressed alone (represented by light grey bar on the left). The presence of recX positively contributes to the negative selection (dark grey bar on the right) with p-value < 0.1 according to a T-test analysis.

Figure 4. Efficiency of CRMAGE.

Panel (A,B) show the efficiency of CRMAGE for gene recoding using pMA7CR_2.0 compared to the pMA7 control using only λ Red. (A) The efficiency of CRMAGE for introducing codon substitution stop codons in xylA and lacZ. Additionally, a stop codon was introduced in galk* linked to a secondary silent codon substitution used by CRMAGE as counter selection. (B) The efficiency of substituting a weak RBS in front of GFP with a strong one (TCCTCC > AGGAAG) while disrupting the PAM sequence used for negative selection. Panel (C–D) shows the efficiency of multiplex CRMAGE. (C) The efficiency of a single round of CRMAGE for simultaneously introducing the two mutations mentioned above (the RBS exchange on the left and stop codon substitution on the right). (D) CRMAGE multiplexing efficiency (pMA7CR_2.0) of the two mutations compared to the control (pMA7) using only λ Red. The data result from the analysis of 15 to 40 positive clones per each mutation that were re-streaked and tested if they were carrying also the second modification.

Figure 5. Self-eliminating proprieties of pMAZ-SK plasmid used for negative selection.

The graph shows the rate of plasmid curing over time. The population was induced with L-rhamnose (to induce self-killing crRNA) and aTetracycline (to induce Cas9 expression). Cells were plated on LB plates with and without kanamycin. Plasmid loss was calculated as the fraction of kanamycin resistant colonies compared to the total number of colonies.

Figure 6. CRMAGE Web-Tool.

The program guides the user through the necessary steps to create a CRMAGE mutated oligos, considering also the option of using a degenerate codon if the desired mutation does not contain a PAM itself. It consists of three major steps: (1) Enter (copy/paste) the wild type DNA sequence; (2) Select the base to mutate and which base to change to Step 3: Choose the Open Reading Frame and the size of the CRMAGE mutation oligo. A list of potential gRNA sites are shown, and the user can pick silent mutations to destroy the indicated gRNA PAM site (pointed by the red bracket). Finally the program outputs the CRMAGE oligo sequence (shown by the red arrow).