Stains, plasmids, media, and growth condition
All Escherichia coli strain and plasmids used in this study are listed in Supplementary Table 1. E. coli cultures were grown at 37 °C in LB (both broth and solid) (Sigma, USA). Appropriate antibiotics were supplemented with the following working concentrations: spectinomycin (50 µg/mL), carbenicillin or ampicillin (100 µg/mL), chloramphenicol (25 µg/mL), kanamycin (50 µg/mL), and anhydrotetracycline (0–1 µg/mL). M63 minimal medium was used for positive selection of galK mutants. It is composed of 2 g/L (NH4)2SO4, 13.6 g/L KH2PO4, 0.5 mg/L FeSO4-7H2O, 1 mM MgSO4, 0.1 mM CaCl2, and 10 μg/mL thiamine, 0.2% glycerol and 0.1% 2-deoxy-d-galactose. Two percent agar was supplemented when making agar plates. X-gal (5-bromo-4-chloro-3-indolyl-beta-d-galactopyranoside) was used for screening lacZ mutants. Prior to use, each LB plate with appropriate antibiotics is plated with 40 µL of 20 mg/mL X-gal. All chemicals involved in this study were from Sigma, USA.
General protocol of DNA manipulation
All primers, important sequences, spacers, and 3′ extensions used in this study are listed in Supplementary Tables 2, 3, and 4, respectively. Standard protocols were used for DNA (plasmids and genomic DNA) purification, PCR, and cloning. PCR was performed using Q5® High-Fidelity 2× Master Mix (New England Biolabs, USA). The point mutation in dCas9 to create H840A Cas9n was made using Q5® Site-Directed Mutagenesis Kit (New England Biolabs, USA). DNA assembly was done by using NEBuilder® HiFi DNA Assembly Master Mix (New England Biolabs, USA) unless specified otherwise. DNA digestion was performed with FastDigest restriction enzymes (Thermo Fisher Scientific, USA) unless specified otherwise. NucleoSpin® Gel and PCR Clean-up kit (Macherey-Nagel, Germany) was used for DNA clean-up both from PCR products and agarose gel extracts. NucleoSpin® Plasmid EasyPure Kit (Macherey-Nagel, Germany) was used for plasmid preparation. Sanger sequencing was carried out using Mix2Seq kit (Eurofins Scientific, Luxembourg). DNA fragments were synthesized by Genscript while oligonucleotides were synthesized by IDT (Integrated DNA Technologies, USA).
All kits and enzymes were used according to the manufacturers’ recommendations. We diligently followed all waste disposal regulations of our institute, university, and local government when disposing of waste materials.
Multiplasmid system design and plasmid construction
All plasmids constructed in this study have been deposited to Addgene, individual Addgene plasmid number are listed below. Plasmids in the same testing system should be compatible with each other, and therefore they must have different origins of replication (ori). For this purpose, a combination of p15A ori, ColE1 ori, and ColDF13 ori was used.
Synthetic constitutive promoters J23119 (BBa_J23119) and J23106 (BBa_J23106), and the ribosome binding site (RBS) BBa_B0034 were obtained from the registry for standard biological parts in the iGEM Parts Registry (http://parts.igem.org/Main_Page).
The construction of GFP-based reporter plasmid: The plasmid was designed in silico to carry the GFP expression cassette, which is composed of a constitutive promoter J23106, a RBS BBa_B0034, a fast folding GFP variant GFP+21, and a terminator T0. The GFP+ coding sequence was codon optimized to E. coli. The whole cassette was synthesized by Genscript and assembled into the pCDF-1b plasmid (ColDF13 ori, Millipore, USA) replacing the MCS region by Gibson Assembly, and ended up with the plasmid pCDF-GFPplus (Addgene #172718).
The construction of CRISPR-Prime Editing plasmid: Firstly, we created pCas9n(H840A) from pdCas9-bacteria (p15A ori, Addgene plasmid #44249)36 by site-specific mutation of 10A of dCas9 to 10D using Q5® Site-Directed Mutagenesis Kit (New England Biolabs, USA). Secondly, we designed the 33a linker-M-MLV2 cassette in silico. Linker sequence: SGGSSGGSSGSETPGTSESATPESSGGSSGGSS. M-MLV2, a moloney murine leukemia virus (M-MLV) variant from a previous study19 with the following mutations D200N, L603W, T306K, W313F, and T330P compared to the WT M-MLV (GenBank: AAC82568.2). Thirdly, the cassette was codon optimized to E. coli, synthesized by Genscript, and then assembled into pCas9n(H840A) to replace the stop codon of Cas9n by Gibson Assembly, resulting in the plasmid pCRISPR-PE-bacteria (Addgene #172715). The fusion protein (cargo) Cas9n-linker-M-MLV2 is under control by a tetracycline inducible promoter.
The construction of PEgRNA transcript carrying plasmid: The empty PEgRNA plasmid was modified from the pgRNA-bacteria (ColE1 ori, Addgene plasmid #44251)36 by removing the 20 bp spacer, named as pPEgRNA (Addgene #172716). For construction of functional pPEgRNA there were three steps: firstly, a spacer and 3′ extension were designed in silico; secondly, amplification of the functional PEGgRNA cassette using the pPEgRNA as a template was performed concurrently with amplifying the PEgRNA backbone fragment using the primer set (PEgRNA backbone_F and PEgRNA backbone_R); lastly, the functional PEgRNA cassette was assembled into the PEgRNA backbone. Sanger sequencing was used for validation. Spacers and 3′ extensions were designed both manually and using PrimeDesign37.
For introducing the second nick, we constructed pnsgRNA (pSC101 ori, kanR) by replacing the sfGFP expression cassette in pVRb20_992 (Addgene plasmid #49714)38 with the sgRNA transcript cassette from pPEgRNA. We first amplified the plasmid backbone of pVRb20_992 and the sgRNA cassette with primer sets of pVRb_backbone_F and pVRb_backbone_R, and sgRNA_cassette_F and sgRNA_cassette_R from pVRb20_992 and pPEgRNA, respectively. Then these two fragments were Gibson assembled and later validated by Sanger sequencing, resulting in pnsgRNA plasmid (Addgene #172717). Spacers for introducing the second nick in the nsgRNA paired with the related PEgRNA were designed using PrimeDesign37. This plasmid is also used for delivery of the second PEgRNA.
High throughput electroporation of multiple plasmids
In vivo assay of strains carrying multiple plasmids were performed from freshly transformed E. coli DH10β strains. A HT Nucleofector™ System (Lonza, Switzerland) together with 96-well Nucleocuvette plates (Lonza, Switzerland) were used for high throughput electroporation. Before electroporation, the 96-well Nucleocuvette plate was transferred from −20 °C to ice for 10 min. Twenty microliter of electrocompetent DH10β or MG1655 E. coli cells with 10% glycerol were added into each desired well, 0.5 µL of each plasmid (about 30 ng) was subsequently added. A total amount of plasmid DNA of <100 ng per transformation normally performed well. The program used in this study is X_bacteria_14, with the code GN-100. After electroporation, 180 µL of fresh LB broth were added into each well. The cultures were then transferred into a 96-deep well plate containing 200 µL of fresh LB broth (making the transformation culture in total 400 µL) for recovery for 1 h at 37 °C and 300 rpm.
Illumina deep sequencing-based genome-wide on/off-target evaluation and analysis of “escapers”
For on/off-target evaluation, one or two Sanger sequencing validated clones of each designed editing events were selected; while for escapers examination, ten clones with induced CRISPR-Prime Editing systems targeting pCDF-GFPplus, still showing GFP-fluorescence, were randomly picked. Together with necessary control strains, they were inoculated in a 50 mL tube (Greiner Bio-One, Germany) containing 10 mL LB broth without any antibiotics. After incubating at 37 °C, 300 rpm in an INNOVA 44R incubator shaker (Eppendorf, Germany) for 24 h, 5 mL of the culture was used for genomic DNA plus intracellular plasmid DNA isolation with a Blood & Cell Culture DNA mini Kit (Cat No./ID: 13323, Qiagen, Germany). While a NucleoSpin® Plasmid EasyPure Kit (Macherey-Nagel, cat. no. 740727.250) was used for the WT pCDF-GFP plasmid isolation. The genomic library construction and illumina paired-end sequencing were carried out by Novogene Co., Ltd. (Beijing, China), using the NEB Next® Ultra™ DNA Library Prep Kit (New England Biolabs, USA) with a target insert size of 350 nt and six PCR cycles.
The illumina reads obtained from the sequenced samples were trimmed using Trim Galore (v. 0.6.4_dev, Cutadapt v. 2.10) with the switches –length 100 and –quality 20. All mutation calls were performed using breseq (v. 0.33.2, bowtie2 v. 2.3.4.1)23,39 with default parameters. For plasmid-based editing, the E. coli DH10B genome sequence NC_010473 is used as the reference, while for chromosome-based editing, the E. coli MG1655 genome sequence NC_ U00096 is used as the reference, both along with the relevant plasmids. Mutation calls that existed in all samples as well as the parental strain were not counted as off-target effects.
Editing efficiency evaluation using a fluorescence-based colony counting assay
Fifty microliter electroporation culture (400 μL in the cases of a second nick is introduced) of each strain was plated onto appropriate antibiotics containing LB agar plates supplemented with and without inducer, respectively. All plates were covered by aluminium foil and incubated at 37 °C for 24 h. After cultivation, total colonies were counted by a Doc-It imaging station (Fisher Scientific, USA) with a trisection protocol. Non-fluorescent colonies in each zone of all three zones were further counted with and without a Blue-Light Transilluminator (Safe Imager 2.0, Thermo Fisher Scientific, USA). The editing efficiency was calculated as: the number of non-green colonies in each zone/total number of visible colonies in the same zone. The graphs are generated by Prism (version 8).
Editing events confirmation by Sanger sequencing
Eight to 24 primarily identified positive clones of each strain were picked, and inoculated into 5 mL LB broth with proper antibiotics. After overnight (~16 h) cultivation, cultures were subjected to plasmid isolation using the NucleoSpin® Plasmid EasyPure Kit (Macherey-Nagel, Germany) or colony PCR using Q5® High-Fidelity 2× Master Mix (New England Biolabs, USA) if a chromosomal region was targeted. The isolated plasmids and the cleaned PCR products were Sanger sequenced using the Mix2Seq kit (Eurofins Scientific, Luxembourg) with proper primers. The obtained sequence traces were analyzed and visualized using SnapGene (GSL Biotech, USA).
Colony forming unit (CFU) assay
All CFU experiments were performed using E. coli DH10beta electrocompetent cells. The plasmids of interest were measured using a NanoDrop UV spectrophotometer (NanoDrop2000; Thermo Fisher Scientific, USA) and diluted to 100 ng/µL. For each transformation, 100 ng of each plasmid were used and added to 50 µL of electrocompetent cells. The cells were then incubated on ice for 30 min and subsequently electroporated using 1 mm cuvettes and a MicroPulser electroporator (Bio-Rad, USA) using the Ec1 program. Four hundred microliter of LB medium were subsequently added and the cells were incubated in a thermoblock at 37 °C while shaking for 1 h. The transformations were then plated in appropriate dilutions on LB plates supplemented with the corresponding antibiotics with and without ATc. After incubation overnight at 37 °C, the colonies were counted and the real CFUs were calculated by multiplying the counted CFUs with the corresponding dilution factor.
Growth profiling in liquid culture
All growth experiments were performed in a ELx808 plate reader (Buch and Holm A/S), set to 37 °C, constant shaking, and measurement of OD630 every 20 min. Measurements were taken for 24 h. All cultivations were performed in 96-well microtiter plates with F-bottom and a lid. For inoculation of the precultures, four colonies were picked from each non-induced LB plate (12 in total) and used to inoculate wells of a 96 deep well plate filled with 1 mL of LB supplemented with the corresponding antibiotics (spectinomycin, 50 µg/mL ampicillin, 100 µg/mL; chloramphenicol, 25 µg/mL; and ATc, 200 ng/mL). The preculture was incubated overnight at 37 °C at 250 rpm. The next morning, the OD630 of the preculture was measured using the ELx808 plate reader. The cultures were diluted 1:2 to obtain values below 0.8. For each well, the necessary volume for inoculation of 200 µL of microtiter cultures with a starting OD630 of 0.05 was calculated. The corresponding volumes were then added to one microtiter plate with 200 μL and the corresponding antibiotics only, and one with both antibiotics and ATc. Both cultivations were run in ELx808 plate readers in parallel using the same run protocol.
μ
max and the maximum doubling time calculation
For determination of μmax and the maximum doubling time based on the Monod equation, the natural logarithm of the OD630 values was plotted against the time in hours. In this plot, the exponential phase can be easily determined based on the linear progression. For each cultivation, the exponential phase was determined using this graphical approach, and linear regressions were calculated for those time points. The slope of the linear regression corresponds to μmax in h−1, which can subsequently be used to calculate the maximum doubling time by dividing ln(2) by the determined μmax value.
Reporting summary
Further information on research design is available in the Nature Research Reporting Summary linked to this article.
