Figure 1.
LL-37 and Mg2+ have opposite effects on expression of CsrRS-regulated genes.
Expression of ten CsrRS-regulated genes in response to LL-37 or Mg2+ was quantified by qRT-PCR. Data represent mean fold-change ± SEM of gene expression in cultures grown in the presence of 100 nM LL-37 (grey bars) or 15 mM Mg2+ (black bars) relative to that in control cultures grown in unsupplemented medium (n = 3 – 5). Significant differences in response to LL-37 were found for all tested genes (P<0.008 for speA, sda1, ska, SPy0136, metB, SPy1414, grab; P<0.05 for slo, nga and speB). Significant differences in response to Mg2+ were found for all tested genes (P<0.02) except for speA and SPy1414.
Figure 2.
CsrS is associated with the cell membrane and contains a surface-exposed domain.
A) Western blot analysis of membrane and cytoplasmic fractions isolated from whole cell lysates of GAS wild type strain 854 (WT), isogenic csrS deficient mutant strain 854csrSΩ (SΩ), and 854csrSTM (TM) that expresses CsrS with 3 point mutations in the predicted extracellular domain. Specific antisera against CsrS, an unrelated membrane protein OpuABC, and CsrR were used to detect the respective proteins in both fractions. B) Biotin labeling via a disulfide linker of surface-exposed proteins in whole cells of wild type strain 854 (WT), 854csrSΩ (SΩ) and 854csrSTM (TM). After lysis of labeled cells, biotinylated proteins were captured on a NeutrAvidin column and then eluted by reducing the disulfide linker. Specific antisera detected CsrS in the eluted fraction, as expected for a surface-exposed protein, and CsrR in the flow-through, as expected for a cytoplasmic protein. As a control, wild type 854 cells were treated similarly, but without biotin labeling (NB). Results shown in both panels are representative of at least two independent experiments.
Figure 3.
LL-37 and Mg2+ signaling of CsrRS-regulated genes involves a cluster of negatively charged amino acid residues located in the CsrS extracellular domain.
A) Schematic representation of predicted CsrS protein domains. CsrS consists of two membrane-spanning domains (M), an extracellular domain (ECD), a cytosolic HAMP domain, a histidine kinase domain (HisK) and ATPase domain (HATPase). Acidic amino acids in the ECD replaced by uncharged residues in isogenic mutant strains are indicated in red. B) LL-37 stimulation of gene expression in wild type strain 854 (WT), isogenic csrS triple point mutant 854csrSTM (TM), and isogenic csrS single point mutants 854csrSD148N (D148N), 854csrSE151Q (E151Q), and 854csrSD152N (D152N). Expression of hasB, spyCEP, mac, and SPy0170 was measured by qRT-PCR. C) LL-37 and Mg2+ responses in M-type 49 strain NZ131 and its isogenic csrS triple point mutant NZ131TM (NZ131 TM). For panels B and C, data represent mean ratios ± SEM of gene expression in strains grown in the presence of 100 nM LL-37 (panel B; panel C, left) or 15 mM Mg2+ (panel C, right) compared to control cultures of the same strain grown in unsupplemented THY medium (n = 3 – 5). A broken line denotes a ratio of 1, which indicates no change in expression relative to that in unsupplemented medium. Asterisks denote significant differences between wild type and csrS triple point mutant (* P<0.05, ** P<0.006).
Table 1.
Effect of csrS and csrR mutations on expression of CsrRS-regulated genes in GAS strain 854 under standard growth conditions.
Figure 4.
LL-37 signaling depends on a functional CsrS to induce GAS resistance to opsonophagocytic killing.
Wild type strain 854 (WT), isogenic csrS mutants 854csrSTM (TM), 854csrSΩ (csrSΩ), 854H280A (H280A), and 854H280A,TM (H280A TM), and isogenic csrR deletion mutant 854ΔcsrR (ΔcsrR) were grown in the absence (open symbols) or presence (filled symbols) of 100 nM LL-37. Bacteria were then mixed with human peripheral blood leukocytes for 1 h in the presence of 10% human serum as complement source. Values represent the log of mean fold-change in cfu. Each symbol represents a single experiment performed in duplicate. When exposed to LL-37, wild type 854 showed a significant increase in resistance to phagocytic killing compared to untreated bacteria (P<0.001), whereas the isogenic csrS triple mutant (TM) did not (n.s. = not significant). Mutant strains 854csrSΩ, 854H280A, 854H280A,TM, and 854ΔcsrR were highly resistant to phagocytic killing in the absence of supplemental LL-37.
Figure 5.
A lysine residue at position 102 is essential for CsrR to transduce LL-37 signaling.
A) Regulation of CsrRS-dependent genes in 950771 and isogenic mutant strain 950771csrRR102K in response to LL-37 and Mg2+. Gene expression was measured in strain 950771 with non-consensus CsrR (csrR R102) and in its isogenic mutant strain 950771csrRR102K in which csrR residue 102 was changed from arginine to the consensus lysine (csrR K102); strains were grown in the presence of 100 nM LL-37 (left panel) or 15 mM Mg2+ (right panel). B) Comparison of baseline expression of hasB, spyCEP, mac, and SPy0170 in isogenic mutant strain 950771csrRR102K (consensus CsrR) with that in strain 950771 (non-consensus CsrR, csrR R102). Cultures were grown to early exponential phase in unsupplemented THY medium (unsuppl.). C) Mutation of csrR K102 in GAS wild type strain NZ131 prevents LL-37 and Mg2+ signaling. Gene expression was measured in NZ131 with consensus CsrR (csrR K102) and in its isogenic mutant strain NZ131csrRK102R (csrR R102) grown in the presence of 100 nM LL-37 (left panel) or 15 mM Mg2+ (right panel). For all panels, data represent mean ratios ± SEM of at least three independent experiments done in triplicate. A broken line denotes a ratio of 1, which indicates no change in expression relative to control. Asterisks denote significant differences in expression between parental strain and respective mutant strains shown in each panel (* P<0.05, ** P<0.008).
Table 2.
GAS strains used in this study.