Overview of recombineering in L. reuteri. (i) Electrocompetent cells in which RecT is expressed are transformed with a recombineering oligonucleotide. Transformation efficiencies of ∼ ≥10⁵cfu/µg DNA are required for high recombineering efficiencies. Wavy black lines with a red dot represent recombineering oligonucleotides with multiple non-complementary bases; the green circle represents a bacterial cell; pJP042 is the sakacin-based expression vector that contains recT; the black double helix represents chromosomal DNA; expressed RecT proteins are denoted in the cell. (ii) An oligonucleotide identical to the lagging strand contains multiple non-complementary bases that avoid the mismatch repair system resulting in increased recombineering efficiency. (iii) Viable cells are recovered on antibiotic-free plates and recombinants are detected by a mismatch amplification mutation assay-PCR (MAMA-PCR). Two oligonucleotides (blue) will yield a 1-kb fragment, whereas a third oligonucleotide (red) will only be extended by the polymerase when the mutations are incorporated in the chromosome yielding a second amplicon of 500 bp. As the recombineering oligonucleotide only targets one strand during DNA replication the colonies will be of mixed genotype. The red dot on the chromosome indicates that the mutations are incorporated. (iv) Single colony purification is performed to separate the wild-type genotype from the mutant genotype. (v) MAMA-PCR is repeated as described in section iii to identify a pure genotype mutant. A 1:1 ratio of wild-type and mutant genotypes is suggestive that during replication a single chromosome is being replicated per cell. (vi) The recombineering plasmid pJP042 can now be cured from the mutant strain by passaging bacteria without antibiotic selection to yield a plasmid-free derivative (vii).

Establishing and optimizing recombineering in L. reuteri. (a) The dsDNA sequence of the targeted rpoB region of L. reuteri is shown aligned with amino acids specified by each codon. On the left the leading and lagging strand are indicated and below the oligonucleotides are listed which result in different mismatches, all resulting in a rifampicin-resistant phenotype. The corresponding amino acid changes are listed on the right. (b) Titrations of the amount of oligonucleotides oJP133 (Ο) and oJP577 (Δ) were performed with 1, 5, 25 and 100 µg of oligonucleotide. Rifampicin resistant colonies derived from the control transformation are represented as a bar graph. Data are expressed as rifampicin-resistant cfu per 10⁹ cells. Data shown are the averages of three independent experiments and error bars represent standard deviation. (c) Assessment of recombineering efficiencies with 100 µg oJP133, oJP311, oJP813 and oJP577 in the wild-type strain (black bars), and in the mismatch repair deficient strains MutS1⁻ (light grey bars) and MutL⁻ (dark grey bars). Data shown are the averages of three independent experiments and error bars represent standard deviation.

Non-selected recombineering in L. reuteri. (a) One hundred colonies derived from transformation with 100 µg oJP577, which upon incorporation yields a rifampicin-resistant phenotype (RpoB H488R), were patched onto MRS agar without antibiotic selection (left—MRS), and onto MRS agar containing rifampicin (right—MRS + RIF). (b) Overview of a selection of genes mutated by non-selected recombineering in the L. reuteri ATCC PTA 6475 chromosome. The locations of the mutated genes on the L. reuteri ATCC PTA 6475 chromosome are indicated based on the closed genome of the closely related strain L. reuteri F275 (NCBI accession number NC_010609). The numbers displayed represent the last four digits of the locus tag (HMPREF0535_XXXX) for each of the genes targeted and the recombineering frequency shown in parentheses. The rpoB gene (DNA directed RNA polymerase; locus tag HMPREF0536_0828) and the ddl gene (d-Ala-d-Ala ligase, locus tag HMPREF0536_1572) are listed using their gene names based on experimental evidence. ND: not determined; ori: origin of replication.

Establishing recombineering in other LAB. (a) The dsDNA sequence of the targeted rpoB region of L. lactis is shown aligned with the encoded amino acids for each codon. On the left are the leading and lagging strand indicated and below oligonucleotides are listed yielding different mismatches, all resulting in a rifampicin-resistant phenotype. The corresponding amino acid changes are listed on the right. (b) Recombineering efficiencies obtained in L. lactis transformed with 100 µg oJP547 or oJP563 and the level of natural rifampicin resistant L. lactis colonies per 10⁹ cells (control). (c) Comparison of recombineering levels between L. reuteri (oJP577), L. lactis (oJP563), L. gasseri (oJP579) and L. plantarum (oJP492). All data shown are averages from three independent experiments and error bars represent standard deviation.

Converting vancomycin-resistant L. reuteri to vancomycin-sensitive by a single amino acid change. (a) The determined 3D structure of the d-ala-d-ala ligase (Ddl) protein of E.coli (PDB ID 1IOV; blue) and the modeled protein structure of Ddl of L. reuteri (green) were superimposed to locate the putative active site residues in Ddl of L. reuteri which, in E. coli, form a hydrogen-bonded network (Y216–E15–S150; boxed region). (b) Superimposing the determined structures of the Ddl proteins of the E. coli wild-type (blue) and a mutant (PDB ID 1IOW; grey) show that replacing a tyrosine at position 216 with a phenylalanine (Y216F) does not form a hydrogen-bond between F216–E15, and changes the enzymatic activity from a dipeptide ligase to a depsipeptide ligase. (c) We predicted the active site residues of Ddl-L. reuteri by superimposing the wild-type E. coli Ddl protein (blue) with the predicted protein structure of L. reuteri (green), and residues F258–E17–S186 in Ddl-L. reuteri correspond to the E. coli triad Y216–E15–S15, respectively. According to the E. coli model changing the phenylalanine at position 258 to a tyrosine (F258Y) in Ddl-L. reuteri would establish a hydrogen-bond between F258–E17 and may therefore yield dipeptide ligase activity which subsequently would result in a vancomycin-sensitive phenotype. (d) The dsDNA sequence of the ddl region of L. reuteri is shown aligned with the encoded amino acids for each triplet. Italic sequence represents the ApoI restriction site that is mutated by incorporation of oJP810. On the left are the leading and lagging strand indicated, and below is oJP810 showing the different mismatches with the resulting amino acid change (F258Y) on the right. (e) Susceptibility of L. reuteri wild-type (left) and the mutant RPRB3003 (right) to vancomycin using an Etest assay. The MIC is determined by identifying where bacterial growth intersects the Etest strip.