Development of a tunable wide-range gene induction system useful for the study of streptococcal toxin-antitoxin systems

Abstract

Despite the plethora of genetic tools that have been developed for use in Streptococcus mutans, the S. mutans genetic system still lacks an effective gene induction system exhibiting low basal expression and strong inducibility. Consequently, we created two hybrid gene induction cassettes referred to as Xyl-S1 and Xyl-S2. Both Xyl-S cassettes are xylose inducible and controlled by the Bacillus megaterium xylose repressor. The Xyl-S cassettes each demonstrated >600-fold-increased reporter activity in the presence of 1.2% (wt/vol) xylose. However, the Xyl-S1 cassette yielded a much higher maximum level of gene expression, whereas the Xyl-S2 cassette exhibited much lower uninduced basal expression. The cassettes also performed similarly in Streptococcus sanguinis and Streptococcus gordonii, which suggests that they are likely to be useful in a variety of streptococci. We demonstrate how both Xyl-S cassettes are particularly useful for the study of toxin-antitoxin (TA) modules using both the previously characterized S. mutans mazEF TA module and a previously uncharacterized HicAB TA module in S. mutans. HicAB TA modules are widely distributed among bacteria and archaea, but little is known about their function. We show that HicA serves as the toxin component of the module, while HicB serves as the antitoxin. Our results suggest that, in contrast to that of typical TA modules, HicA toxicity in S. mutans is modest at best. The implications of these results for HicAB function are discussed.

Fig 1

Construction of the Xyl-S1 and Xyl-S2 cassettes. (A to C) Illustration of the genetic organization for the B. megaterium xylRA locus (A), the Xyl-S1 cassette in pZX9 (B), and the Xyl-S2 cassette in pZX10 (C). (D and E) The sequences of the Xyl-S1 cassette (D) and Xyl-S2 cassette (E) are shown. Promoter elements are shown in red, while xylA operator (xylA_O) inverted repeats are shown in blue.

Fig 2

Luciferase activity from Xyl-S1 and Xyl-S2 reporter plasmids in three species of Streptococcus. (A) S. mutans was transformed with pZX8 (wild-type B. megaterium xylose cassette luciferase fusion; dashed line), pZX9 (Xyl-S1 cassette luciferase fusion; black line), and pZX10 (Xyl-S2 cassette luciferase fusion; gray line) and assayed for luciferase activity over a range of xylose concentrations (Con.). Luciferase measurements were made 2 h after the addition of xylose to the growth medium. Luciferase activity was normalized by dividing luminometer values by optical density values (RLU/OD₆₀₀). (B and C) The experiment was repeated in S. sanguinis (B) and S. gordonii (C) using pZX9 (black line) and pZX10 (gray line). All data points represent the average values from three independent experiments ± standard deviations, which were generally <10%.

Fig 3

Growth phenotypes of Xyl-S–mazF expression constructs. (A) As described in Materials and Methods, mazF was fused to both Xyl-S1 and Xyl-S2 using POE-PCR and the amplicons were directly transformed into S. mutans. The resulting plasmids were verified and then retransformed into the wild-type (WT) and ΔmazEF backgrounds. Dilutions of the transformation reaction mixtures were spotted onto selective medium with or without xylose. Constructs exhibiting little or no toxicity supported colony formation after transformation, whereas toxic constructs yielded no transformants. (B and C) Next, three separate wild-type transformants of the Xyl-S1–mazF fusion were cultured, diluted, and then spotted onto agar plates containing either no xylose (B) or xylose (C). (D and E) Likewise, three ΔmazEF clones harboring the Xyl-S2–mazF fusion were diluted and spotted onto agar plates containing either no xylose (D) or xylose (E). The S. mutans wild-type strain containing no plasmid was used as a control. This experiment was performed three times with similar results.

Fig 4

Characterization of the hicAB locus. (A) The organization of the hicAB locus in S. mutans. Open reading frames surrounding the hicAB operon are listed by their NCBI locus tag designations. (B) Important regulatory elements within the hicAB operon. Putative ribosome binding sites are shown in gray, whereas other regulatory elements are shown in bold. (C) The intergenic region upstream of the hicAB operon was inserted upstream of a promoterless luciferase open reading frame in two separate reporter constructs. One construct contained an intact hicAB promoter, whereas the other contained a deletion of the −10 sequence. The resulting luciferase activities of the two constructs were compared. Luciferase activities was normalized by dividing luminometer values by optical density values (RLU/OD₆₀₀) and are presented relative to the luciferase activity obtained from the −10 deletion construct, which was arbitrarily assigned a value of 1. Data represent average values from three independent experiments ± standard deviations. *, P < 0.01 (2-tailed paired t test).

Fig 5

Growth phenotypes of Xyl-S–hicA expression constructs. Two types of hicA expression plasmids were assembled using Xyl-S cassettes. (A) Constructs labeled as Xyl-S::hicA_ORF consisted of a hicA ORF translation fusion to the Xyl-S cassette ribosome binding site. (B) Constructs labeled as Xyl-S::hicA_GENE consisted of a transcription fusion of the entire hicA gene and the Xyl-S cassette. Therefore, hicA gene expression was driven by both the Xyl-S cassette promoter and the hicA promoter, while translation of the hicA ORF was controlled by the hicA Shine-Dalgarno sequence. (C and D) Wild-type strains harboring Xyl-S1–hicA_ORF translation fusions were diluted and spotted onto agar plates without added xylose (C) and in the presence of xylose (D). (E and F) ΔhicAB strains harboring Xyl-S2–hicA_ORF translation fusions were diluted and spotted onto agar plates without added xylose (E) and in the presence of xylose (F). (G and H) ΔhicAB strains harboring Xyl-S1–hicA_GENE transcription fusions were diluted and spotted onto agar plates without added xylose (G) and in the presence of xylose (H). The S. mutans wild-type strain containing no plasmid was used as a control. This experiment was performed three times with similar results.