Genome editing of phylogenetically distinct bacteria using portable retron-mediated recombineering

Abstract

Advanced genome editing technologies have enabled rapid and flexible rewriting of the Escherichia coli genome, benefiting fundamental biology and biomanufacturing. Unfortunately, some of the most useful technologies to advance genome editing in E. coli have not yet been ported into other bacterial species. For instance, the addition of bacterial retrons to the genome editing toolbox has increased the efficiency of recombineering in E. coli by enabling sustained, abundant production of ssDNA recombineering donors by reverse transcription that install flexible, precise edits in the prokaryotic chromosome. To extend the utility of this technology beyond E. coli, we surveyed the portability and versatility of retron-mediated recombineering across three different bacterial phyla (Proteobacteria, Bacillota and Actinomycetota) and a total of 15 different species. We found that retron recombineering is functional in all species tested, reaching editing efficiencies above 20% in six of them, above 40% in three of them, and above 90% in two of them. We also tested the extension of the recombitron architecture optimizations and strain backgrounds in a subset of hosts to additionally increase editing rates. The broad recombitron survey carried out in this study forms the basis for widespread use of retron-derived technologies through the whole Bacteria domain.

Figure 1 –. Design and construction of recombitrons for editing across the domain of Bacteria.

a, Top: schematic of the recombitron operon with a donor encoded within the msd region of the ncRNA. Bottom: schematic of the retron-mediated recombineering process. Briefly, the retron RT produce multiple copies of the RT-DNA donor. SSAP and SSB proteins promote the binding of the RT-DNA donor to the lagging strand during bacterial replication installing the desired mutation. b, Unrooted phylogenetic tree of the bacterial species used in this work. The tree was constructed using a multiple sequencing alignment (MSA)of the 16S sequences of these species. The different colors represent the different bacterial phyla (Proteobacteria, Bacillota and Actinomycetota), class (Gammaproteobacteria), order (Pseudomonadales) or functional group (Coliforms). c, Left: Schematic of the operons used for optimizing recombitron architecture. Eco1 recombitron to make a deletion in the lacZ gene was used in all the configurations. Plasmid origin of replication and promoters are indicated in different colors. Right: quantification of precise editing rates of the lacZ locus and correlation with OD₆₀₀ of the cultures after overnight growth. Editing data were quantified by Illumina sequencing. Error bars are ± standard deviation for three biological replicates. d, Table summarizing the relevant features of the retrons selected to perform the retron-based recombineering experiments. Information provided in the column with an asterisk was sourced from Mestre et al., 2020. e, Schematic of the workflow used to analyze the portability of recombitrons. Briefly, the set of 10 recombitrons was cloned in a host-specific plasmid using a Gibson Assembly approach. SapI flanked donors were ligated in the retron ncRNA of the 10 retrons in parallel employing a Golden Gate reaction. Multiple colonies were screened using Sanger sequencing to obtain the different recombitrons with the proper donor in a specific plasmid backbone. Plasmids were naturally introduced, electroporated, or conjugated into the final host. Single colonies are grown until saturation, diluted (with the proper inducer if required) and grown again until saturation. Precise editing rates are measured by Illumina sequencing.

Figure 2 –. Retron-mediated recombineering in coliforms.

a, Schematic of the phylogenetic tree from Fig 1b, highlighting the group of bacteria evaluated in this figure, the Coliforms. b, PAGE analysis of RT-DNA production in E. coli bMS.346. c, PAGE analysis of RT-DNA production in C. freundii ATCC 8090. d, PAGE analysis of RT-DNA production in K. pneumoniae ATCC 10031. In b-d, each RT-DNA was prepped from cells once and their length relative to a ssDNA ladder with markers in nt for nucleotides is indicated (uncropped gels are shown in Supplementary Fig 3). e, Quantification of precise genome editing to make a 5 bp deletion in the lacZ gene in E. coli using the 10 recombitrons set. f, Quantification of precise genome editing to make a 5 bp deletion in the lacZ gene in C. freundii using the 10 recombitrons set. g, Quantification of precise genome editing to make a 5 bp deletion in the lacZ gene in K. pneumoniae using the 10 recombitrons set. h, Schematic illustrating the lengthening of a1-a2 regions of retron ncRNA to boost editing rates. i, Quantification of precise genome editing in C. freundii using longer a1-a2 region in -Ban1 and -Ebu1 recombitrons in comparison with wild type versions. j, Quantification of precise genome editing in K. freundii using longer a1-a2 region in -Eas1 and -Ebu1 recombitrons in comparison with wild type versions. k, Quantification of precise genome editing with -Eas1 recombitron in wild type and ΔrecJ/ΔsbcB double knock-out K. pneumoniae strain. In e-k, data were quantified by sequencing after 24 h of editing using Illumina NextSeq, circles show each of the three biological replicates, and errors bars are mean ± standard deviation. Additional statistical details are presented in Supplementary Table 3.

Figure 3 –. Retron-mediated recombineering in environmental Gammaproteobacteria.

a, Schematic of the phylogenetic tree from Fig 1b, highlighting environmental Gammaproteobacteria. b, Quantification of precise genome editing to make a 5 bp deletion targeting a non-essential intergenic region in E. amylovora Eat8 using a set of 5 recombitrons. c, Quantification of precise genome editing to make a synonymous 4 bp replacement donor (GAGT to ATCG) targeting the flgH gene in V. natriegens ATCC 14048 using a set of 9 recombitrons. d, Quantification of precise genome editing to make a synonymous 4 bp replacement donor (TTCG to CAGT) targeting the flgH gene in A. hydrophila ATCC 7966 using the 10 recombitrons set. e, Quantification of precise genome editing in A. hydrophila using longer a1-a2 region with -Eas1 recombitron in comparison with wild type version. f, Quantification of precise genome editing to make a 5 bp deletion in the lacZ gene in S. oneidensis JG2150 using a set of 8 recombitrons. g, Quantification of precise genome editing to make different edits types in the lacZ gene in S. oneidensis JG2150 using -Eco1 recombitron. A dead version of Eco1 RT was used as a negative control. h, Quantification of precise genome editing to make a 5 bp deletion in the lacZ gene in S. oneidensis JG2150 using lac or J23115 promoter. i, Quantification of precise genome editing to make a non-contiguous neutral replacement in mtrC gene and a contiguous neutral replacement in flgH gene in S. oneidensis JG2150 with -Eco1 recombitron. In b-i, data were quantified by sequencing after 24 h of editing using Illumina NextSeq, circles show each of the three biological replicates, and errors bars are mean ± standard deviation. Additional statistical details are presented in Supplementary Table 3.

Figure 4 –. Retron-mediated recombineering in Pseudomonadales.

a, Schematic of the phylogenetic tree from Fig 1b, highlighting the Pseudomonadales order. b, Quantification of precise genome editing to make a 4 bp deletion targeting the phzM gene in P. aeruginosa PAO1 using the 10 recombitrons set. c, Quantification of precise genome editing to make a 4 bp deletion targeting the ykgJ gene in P. putida KT2440 using the 10 recombitrons set. d, Quantification of precise genome editing to make a 5 bp deletion in a non-essential intergenic region in P. syringae B728a using the 10 recombitrons set. e, Schematic of the pORTMAGE-Pa1 derivative plasmids with wild type and inverted pBBR1 origin of replication. f, Quantification of precise genome editing using plasmids with wild type or inverted pBBR1 origin of replication in P. aeruginosa PAO1 using -Kva1 and -Eas1 recombitrons. g, Quantification of precise genome editing using plasmids with wild type or inverted pBBR1 origin of replication in P. syringae B728a using -Kva1 recombitron. h, Quantification of precise genome editing to make a 4 bp deletion targeting a glycosyltransferase gene in A. baylyi sFAB6437 using a set of 7 recombitrons. In b-h, data were quantified by sequencing after 24 h of editing using Illumina NextSeq. Circles show each of the three biological replicates, and errors bars are mean ± standard deviation. Additional statistical details are presented in Supplementary Table 3.

Figure 5 –. Retron-mediated recombineering in Bacillota and Actinomycetota.

a, Schematic of the phylogenetic tree from Fig 1b, highlighting the Bacillota and Actinomycetota classes. b, Top: Schematic of the dsDNA sequence of the targeted rpoB region of L. reuteri 6475 is shown aligned with amino acids residues specified by each codon. On the left the leading and lagging strand are indicated and below the retron RT-DNA donor with the single-point mutation that results in a rifampicin-resistant phenotype. The amino acid change (H488R) is listed on the right. Bottom: Schematic of the tow plasmid assay used for retron-mediated recombineering in L. reuteri. One plasmid encoded the retron RT and engineered ncRNA to edit rpoB gene in the pNZ9530 backbone and under the P_suc promoter. The other plasmid encoded the LrpRecT gene under the P_orfx promoter in the pSIP411 backbone c, Quantification of precise genome editing to make a single-point mutation in rpoB (H488R) gene in L. reuteri in the absence or the presence of sucrose and inducing peptide with -Csp1 recombitron. The experiment was performed for 6 passage rounds. d, Quantification of precise genome editing to make a single-point mutation in rpoB (H488R) gene in L. reuteri using a set of 3 recombitrons following single passage. In c-d, editing rates were calculated as % of rifampicin resistant colonies respect to total number of colonies in no- antibiotic plates e, Quantification of precise genome editing to make a 5 bp deletion in purA gene in S. suis P1/7 using -Ebu1 recombitron expressed under PCas or Pg promoters. f, Quantification of precise genome editing to make a 5 bp deletion in MSMEG_5894 gene in M. smegmatis MC² 155 using the 10 recombitrons set. g, Quantification of precise genome in M. smegmatis after 24, 48 and 72 h of editing and correlation with cell growth state (OD₆₀₀) using -Ebu1 recombitron. h, Quantification of precise genome editing to make a 4 bp deletion in beta-galactosidase gene in C. acnes KPA17202 using the 10 recombitrons set. In d-h, data were quantified by sequencing after 24, 48 or 72 hours of editing using Illumina NextSeq, circles show each of the three biological replicates (except in C. acnes in which only one or two biological replicates were performed), and errors bars are mean ± standard deviation.

Figure 6 –. Summary of retron-mediated recombineering across the Bacteria domain.

a, Heatmap depicting the absolute value of retron-mediated recombineering efficiency in the 15 bacterial species tested. b, Heatmap depicting the value of retron-mediated recombineering efficiency normalized to the best recombitron in each one of the bacterial species aasayed. In a-b, logarithmic scale was used to show the compiled editing data from fig 2–5 in a scale of greens (light green, low editing; dark green, high editing), untested recombitrons were shown in grey color and no editing recombitrons were shown in white. On the left a tree indicates the phylogenetic relationships of the bacterial species tested in the work. The retron clade (from Mestre et al., 2020) is shown above of the retron names.