The relBE2Spn toxin-antitoxin system of Streptococcus pneumoniae: role in antibiotic tolerance and functional conservation in clinical isolates

Abstract

Type II (proteic) chromosomal toxin-antitoxin systems (TAS) are widespread in Bacteria and Archaea but their precise function is known only for a limited number of them. Out of the many TAS described, the relBE family is one of the most abundant, being present in the three first sequenced strains of Streptococcus pneumoniae (D39, TIGR4 and R6). To address the function of the pneumococcal relBE2Spn TAS in the bacterial physiology, we have compared the response of the R6-relBE2Spn wild type strain with that of an isogenic derivative, Delta relB2Spn under different stress conditions such as carbon and amino acid starvation and antibiotic exposure. Differences on viability between the wild type and mutant strains were found only when treatment directly impaired protein synthesis. As a criterion for the permanence of this locus in a variety of clinical strains, we checked whether the relBE2Spn locus was conserved in around 100 pneumococcal strains, including clinical isolates and strains with known genomes. All strains, although having various types of polymorphisms at the vicinity of the TA region, contained a functional relBE2Spn locus and the type of its structure correlated with the multilocus sequence type. Functionality of this TAS was maintained even in cases where severe rearrangements around the relBE2Spn region were found. We conclude that even though the relBE2Spn TAS is not essential for pneumococcus, it may provide additional advantages to the bacteria for colonization and/or infection.

Figure 1. Changes in the growth profile of S. pneumoniae cells of the R6 or the R6ΔrelB2Spn mutant strains after inhibition of protein synthesis by SHT.

S. pneumoniae cultures, wt (○) and mutant (Δ) cells, were grown exponentially until an OD₆₅₀  = 0.2. Then, SHT (1.5 mg.ml⁻¹) was added, and incubation proceeded. Growth was monitored by determination of the OD₆₅₀ of the cultures treated (filled symbols) or untreated (open symbols) with SHT (A). At the times indicated, the numbers of cfu were determined by plating appropriate cell dilutions on SHT-free medium, and incubation for 36 h at 37°C (B). Percentages of viable cells from wt (x) or mutant (*) strains upon SHT-treatment (C) were calculated considering the number of cfu at time zero of the treatment as 100%.

Figure 2. Recovery of cell growth after removal of SHT.

SHT-amino acid starved pneumococcal cells (180 minutes of starvation) were washed twice and suspended into pre-warmed fresh medium, and incubation proceeded for the indicated times. Recovery of starvation was followed by turbidity of the cultures (OD₆₅₀) of wt (○) and mutant (Δ) strains (A). The number of cfu was determined by plating appropriate dilutions of the cells at the times indicated (B). Inset: Linear plot showing cell viability evolution after SHT removal in wt (○) or mutant (Δ) cultures.

Figure 3. Inhibition of protein synthesis by Erm treatment.

Cultures of R6 (○) or R6ΔrelB2Spn mutant (Δ) cells were treated with two concentrations of Erm, 1 µg.ml⁻¹ (black symbols) or 0.1 µg.ml⁻¹ (grey symbols). Growth was followed by absorbance at OD₆₅₀ (A) of untreated (open symbols) and treated (filled symbols) cultures. At the indicated times, appropriate dilutions of cells were plated and incubated as above (B) All experiments were performed at least three times. LIVE/DEAD-staining of wt and the R6ΔrelBE2Spn cells (C). The cultures were harvested 180 min after addition of Erm and stained with Baclight Bacterial viability Kit (Invitrogen). Live cells show green fluorescence, whereas dead cells fluoresce red. Percentages of viable cells in every assay, calculated from the results of Figure 3B are indicated in parentheses.

Figure 4. Presence of the reBE2Spn operon in the chromosome of isolates of S. pneumoniae.

Genetic organization of the relBE2Spn and yefM-yoeBSpn loci in the R6 strain; the position of promoters P_rel and P_mb, respectively, are shown (A). The primers used in the PCR amplifications are drawn as arrows, and the expected sizes of the corresponding PCR fragments are shown below. The PCR products detected using the oligonucleotide pairs relB2_p /relE2_c and yefM_N /yoeB_C (left panel) or the oligonucleotide pair relB2_p/relE2tga (right panel) were separated on agarose gels (B). DNAs from loci relBE2Spn (r) or from yefM-yoeBSpn (y) were isolated from different S. pneumoniae macrolide-resistant strains, as follows: CipR-67 (1); CipR-25 (2); CipR-22 (3); CipR-14 (4); CipR-23 (5); R6wt (C); Mr. Molecular weight standard, Hyper ladder I (BIOLINE).

Figure 5. Polymorphisms found at the relBE2Spn locus in clinical isolates of S. pneumoniae.

A genetic diagram of the chromosomal region flanking the relBE2Spn locus in the R6 strain shows that this region includes genes encoding Ldh (lactate dehydrogenase), GyrA (the A subunit of DNA gyrase), and SP1221 and SP1222 (putative type II restriction endonucleases). These orfs have been detected only in the pneumococcal strains belonging to Type I, whereas the putative cation channel protein was found in those strains included in both Type II and Type III. Other genes identified in this region were: relE2 (relBE2Spn locus encoding toxin RelE2), relB2 (relBE2Spn locus encoding antitoxin RelB2), and vicX (encoding VicX protein). The vicx gene comprises part of the TCS02 operon (vicRKX) essential for pneumococcal viability. In addition, a RUP element has been identified between the ldh and the sp1221 coding regions. RUP elements are an insertion sequence (IS)-derivative that could still be mobile . The collection of oligonucleotides used are also indicated by a number corresponding to the following oligonucleotides: 1, ldh_ter,; 2, SP1222; 3, relE2_C; 4, relE2_tga; 5, relB2_C; 6, rel2p5′; 7, relB2p; 8, relB2_N, 9, relE2_N. The genetic organization of the three types of relBE2Spn loci is depicted (A). Illustration of the three types of relBE2Spn locus genetic organization found out in a collection of S. pneumoniae clinical isolates (B) At the bottom, the region spanning the relBE2Spn promoter is depicted. The location of the IS1167 transposon and the mutation identified in strain 1531 (CipR-25) are highlighted.

Figure 6. Functional analysis of RelE2Spn mutants in E. coli.

Cell growth arrest subsequent to the expression of the relE2Spn wt gene or the mutants T34I or D39G E. coli TOP-10 cells harbouring plasmids pE2wt (•), pE71(D39G mutant) (▴), pE81(T34I mutant) (♦), pE600 (T34I mutant) (▪) or pEter as a control (o) were exponentially grown in TY medium containing 0.4% glucose and Km to an OD₆₀₀  = 0.15. Then, cultures received 0.4% arabinose and growth was measured by determination of the OD₆₀₀ (A) and by counting the cfu (B), this latter by plating appropriate dilutions on medium supplemented with 0.4% glucose and Km and incubated overnight at 37°C. All the experiments were performed at least in duplicate. The effect of the separate and of combined expression of the relE2Spn or relB2Spn in E. coli was also tested on solid medium (C). E. coli TOP-10 cells harbouring pB2wt and pEter (streak 1), pFSU2 (streak 2), pE2wt (streak 3), pE71 (streak 4); pE81 (streak 5); pE600 (streak 6) were streaked on TY plates containing Km and Ap, supplemented with 0.4% arabinose and with or without IPTG (2 mM). Arabinose induces the toxin expression, whereas IPTG (2 mM) promotes the antitoxin overproduction.

Figure 7. Molecular modelling of RelBE2Spn.

Structure of RelE from P. hirokoshii (PhRelE) (A) and structural model obtained for the pneumococcal toxin RelE2Spn (B). Residues important for protein synthesis inhibitory activity of PhRelE are displayed. In the case of RelE2Spn model, the Arg residues which aligned with those of PhRelE involved in protein synthesis inhibition are shown. Residues changed in the RelE2 pneumococcal mutants isolated here, T34 and D39, are indicated in red. The sequence alignment of PhRelE, RelE2Spn and RelE1Spn (a pneumococcal non-toxic homolog) is shown in the bottom of the figure. Residues in the PhRelE, involved in toxin activity are red underlined and the corresponding residues in RelE1Spn and RelE2Spn are highlighted in the same way. T34 and D39 positions in RelE2Spn are framed in green.