Adaptation of Prokaryotic Toxins for Negative Selection and Cloning-Independent Markerless Mutagenesis in Streptococcus Species

Abstract

The Streptococcus mutans genetic system offers a variety of strategies to rapidly engineer targeted chromosomal mutations. Previously, we reported the first S. mutans negative selection system that functions in a wild-type background. This system utilizes induced sensitivity to the toxic amino acid analog p-chlorophenylalanine (4-CP) as a negative selection mechanism and was developed for counterselection-based cloning-independent markerless mutagenesis (CIMM). While we have employed this system extensively for our ongoing genetic studies, we have encountered a couple limitations with the system, mainly its narrow host range and the requirement for selection on a toxic substrate. Here, we report the development of a new negative selection system that addresses both limitations, while still retaining the utility of the previous 4-CP-based markerless mutagenesis system. We placed a variety of toxin-encoding genes under the control of the xylose-inducible gene expression cassette (Xyl-S) and found the Fst-sm and ParE toxins to be suitable candidates for inducible negative selection. We combined the inducible toxins with an antibiotic resistance gene to create several different counterselection cassettes. The most broadly useful of these contained a wild-type fst-sm open reading frame transcriptionally fused to a point mutant form of the Xyl-S expression system, which we subsequently named IFDC4. IFDC4 was shown to exhibit exceptionally low background resistance, with 3- to 4-log reductions in cell number observed when plating on xylose-supplemented medium. IFDC4 also functioned similarly in multiple strains of S. mutans as well as with Streptococcus gordonii and Streptococcus sanguinis. We performed CIMM with IFDC4 and successfully engineered a variety of different types of markerless mutations in all three species. The counterselection strategy described here provides a template approach that should be adaptable for the creation of similar counterselection systems in many other bacteria. IMPORTANCE Multiple medically significant Streptococcus species, such as S. mutans, have highly sophisticated genetic systems available, largely as a consequence of their amenability to genetic manipulation via natural competence. Despite this, few options are available for the creation of markerless mutations in streptococci, especially within wild-type strains. Markerless mutagenesis is a critical tool for genetic studies, as it allows the user to explore many fundamental questions that are not easily addressable using marked mutagenesis. Here, we describe a new approach for streptococcal markerless mutagenesis that offers a variety of advantages over the current approach, which employs induced sensitivity to the toxic substrate 4-CP. The approach employed here should be readily adaptable for the creation of similar markerless mutagenesis systems in other organisms.

Keywords: Streptococcus; counterselection; markerless mutation; negative selection; toxin-antitoxin; xylose.

FIG 1

Illustration of the counterselection approach. Putative toxin ORFs (illustrated in yellow) are transcriptionally fused to the xylose-inducible Xyl-S cassette. Xylose is a nontoxic sugar that is not metabolizable by streptococci, but is efficiently transported into the cells via still unknown mechanisms. Xylose will bind to the xylose repressor XylR and prevent it from inhibiting target gene expression via the xylose promoter operator xylA_O. To create markerless mutations via the two-step CIMM approach, the counterselection cassette is first created using OE-PCR to assemble the inducible toxin together with a positive selection marker, such as the aad9 gene conferring spectinomycin resistance. The assembled counterselection cassette is then introduced onto the chromosome by creating a typical allelic replacement mutation using positive selection. To remove the counterselection cassette and create a markerless mutation, a second unmarked OE-PCR fragment is subsequently transformed into the spectinomycin-resistant strain and then plated on xylose-supplemented medium to initiate negative selection. The resulting xylose-resistant colonies should all contain the expected markerless mutant genotype.

FIG 2

Xylose-inducible negative selection using different toxin genes. (A) A mixture of both endogenous and exogenous toxic ORFs were transcriptionally fused to the Xyl-S cassette and then transformed into S. mutans strain UA159. The resulting strains were tested for growth on agar plates ± xylose. (B) Several clones harboring the fst-sm counterselection cassette were sequenced and found to contain one of two different missense mutations within the xylR ORF. Both mutations targeted codon 7 of xylR, resulting in either A7S or A7T substitutions in XylR. (C) Several clones harboring the parE counterselection cassette were sequenced and found to contain one of two different point mutations in addition to the same XylR(A7S) mutation found in some of the fst-sm counterselection cassettes. The two unique parE mutations occurred either within the parE ribosome binding site (RBS) or codon 56 of parE, conferring a ParE(S56G) amino acid substitution.

FIG 3

Allelic replacement with the gusA ORF. (A) Illustration of the three genotypes encountered during construction of the ΔbrsM brsR-gusA reporter strains. The top panel shows the previously constructed brsRM-gusA strain that was used as a template during construction of the markerless mutant strains. Since this strain has a wild-type brsM ORF, expression of the brsRM operon remains in its basal state due to the inhibitory function of BrsM toward the operon activator protein BrsR. Consequently, this strain will not exhibit detectable β-glucuronidase activity due to low gusA expression. The middle panel shows the genotype of strains transformed with the different counterselection cassettes (+/−), replacing the gusA ORF from the parent brsRM-gusA reporter strain. The bottom panel shows the genotype of the expected markerless reporter strains created by replacing both brsM and the different counterselection cassettes with the gusA ORF. After deletion of brsM, inhibition of BrsR is relieved, resulting in potent autoactivation of the operon promoter and high levels of gusA expression. The markerless ΔbrsM brsR-gusA reporter strains will exhibit a dark blue colony phenotype due to the large amount of β-glucuronidase activity produced from the reporters. (B) The brsRM-gusA reporter strain (RM [left lane]) was spotted onto xylose-supplemented agar plates in successive 10-fold dilutions. Each of the adjacent lanes to the right represents the transformation results of strains harboring counterselection cassettes transformed with either H₂O (+/−) or DNA designed to replace brsM with the gusA ORF (R). The different counterselection cassettes are labeled as follows: A7S, Fst-sm(A7S) mutant; A7T, Fst-sm(A7T) mutant; RBS, parE(RBS) mutant; and S56G, ParE(S56G) mutant. The numbers on the right side of the image indicate the dilution factor of the cultures spotted onto the xylose plates.

FIG 4

Markerless gene deletions and insertions in S. gordonii and S. sanguinis. (A) Illustration of the three genotypes encountered during construction of the markerless spxB-renG reporter strains. The top panel shows the wild-type spxB locus of both S. gordonii and S. sanguinis. Both wild-type strains yield a blue precipitate on Prussian blue agar plates due to the H₂O₂ produced primarily from the spxB-encoded enzyme pyruvate oxidase. The middle panel shows an allelic replacement of spxB with the counterselection cassettes (+/−), resulting in a major reduction in H₂O₂ production. A subsequent transformation of these strains with spxB-renG DNA replaces IFDC4 and restores spxB to its original locus along with a transcriptional fusion to the renG ORF. The markerless spxB-renG reporter strains should produce similar levels of H₂O₂ to the original wild-type strains. (B) Representative strains of the wild type (WT), spxB mutant (ΔspxB), and spxB-renG reporter (spxB⁺ renG⁺) of both S. gordonii DL1 and S. sanguinis SK36 were spotted onto Prussian blue agar plates to observe the H₂O₂ production phenotypes of each. (C) Levels of luciferase activity were compared between the ΔspxB strains harboring the counterselection cassettes (IFDC4) and the markerless spxB-renG reporter strains. Results from the S. gordonii strains are shown in red, while S. sanguinis results are shown in blue. Luciferase data are presented relative to the cell-free background luciferase values, which were arbitrarily assigned a value of 1. Luciferase data were derived from five independent clones of each strain, which were averaged and are presented together with their corresponding standard deviations.

FIG 5

Introduction of markerless galK point mutations. (A) IFDC4 was used to engineer a markerless nonsense point mutation into the galK gene of S. mutans. After sequencing to confirm the presence of the expected stop codon, the mutant strain (*) was spotted adjacent to the parent wild-type strain (WT) in successive 10-fold dilutions on agar plates ± deoxygalactose (dGal). The numbers on the right side of the image indicate the dilution factor of the cultures spotted onto the agar plates. (B) Sequence results of the S. gordonii galK gene following IFDC4 point mutagenesis. The engineered stop codon is underlined, while the specific C→T mutation site is marked with an asterisk.