Regulation of cytotoxin expression by converging eukaryotic-type and two-component signalling mechanisms in Streptococcus agalactiae

Abstract

Signal transducing mechanisms are essential for regulation of gene expression in both prokaryotic and eukaryotic organisms. Regulation of gene expression in eukaryotes is accomplished by serine/threonine and tyrosine kinases and cognate phosphatases. In contrast, gene expression in prokaryotes is controlled by two-component systems that comprise a sensor histidine kinase and a cognate DNA binding response regulator. Pathogenic bacteria utilize two-component systems to regulate expression of their virulence factors and for adaptive responses to the external environment. We have previously shown that the human pathogen Streptococcus agalactiae (Group B Streptococci, GBS) encodes a single eukaryotic-type serine/threonine kinase Stk1, which is important for virulence of the organism. In this study, we aimed to understand how Stk1 contributes to virulence of GBS. Our results indicate that Stk1 expression is important for resistance of GBS to human blood, neutrophils and oxidative stress. Consistent with these observations, Stk1 positively regulates transcription of a cytotoxin, beta-haemolysin/cytolysin (beta-H/C) that is critical for survival of GBS in the bloodstream and for resistance to oxidative stress. Interestingly, positive regulation of beta-H/C by Stk1 requires the two-component regulator CovR. Further, we show that Stk1 can negatively regulate transcription of CAMP factor in a CovR-dependent manner. As Stk1 phosphorylates CovR in vitro, these data suggest that serine/threonine phosphorylation impacts CovR-mediated regulation of GBS gene expression. In summary, our studies provide novel information that a eukaryotic-type serine/threonine kinase regulates two-component-mediated expression of GBS cytotoxins.

Fig. 1

Stk1 expression is important for survival of GBS in NIHWB. GBS were incubated in NIHWB and SI was calculated relative to the growth of each strain in human plasma over a 3 h period as described in Experimental procedures. Means were calculated from three independent experiments. Survival of WT A909 compared with LR113 or LR114 was significantly different, P-value <0.001 (ANOVA). Survival of LR113 compared with LR114 was not significantly different P >0.1. Addition of exogenous purines to human whole blood did not alleviate the sensitivity of LR113 and LR114 to NIHWB.

Fig. 2

GBS stk1 mutants are sensitive to phagocytic killing. Bacteria were opsonized in normal human serum and incubated with PMN at a ratio of 1:3 (bacteria: PMN) for 1 h. The per cent kill for each strain was calculated relative to growth of the same strain in the absence of PMN. Means were calculated from three independent experiments. P-values were determined using paired two-tailed t-tests. Per cent kill of LR113 compared with LR114 was not significantly different, P = 0.5.

Fig. 3

Stk1 expression is important for resistance of GBS to reactive oxygen species. GBS strains were exposed to H₂O₂ (0.015% for 1 h). SI was calculated as (cfu at the end of assay)/(cfu at time 0). All experiments were repeated at least three times in triplicate. Survival of WT A909 compared with LR113 or LR114 was significant different, P-value < 0.01 (ANOVA).

Fig. 4. Attenuated pigment production in GBS stk1 mutants

A. Pigment production in GBS serotype Ia strains. Absorbance profiles indicate reduced pigment production in the stk1 mutants LR113 and LR114 compared with isogenic WT A909. Note that the control strain IaΔcylE is completely deficient in pigment production (absence of the triple absorbance peaks) B. Pigment production in GBS serotype V strains. Absorbance profiles indicate reduced pigment production in the stk1 mutants LR122 and LR123 compared with the isogenic WT NCTC10/84.

Fig. 5. Stk1 positively regulates cylE transcription

A. Decreased cylE transcription in GBS stk1 mutants. RPA were performed on 25 mg total RNA from A909, LR113, LR114, IaΔcylE, NCTC10/84, LR122 and LR123 with a 250 bp antisense [α³²P] UTP-labelled cylE probe. Transcription of cylE was 2.5-fold decreased in the stk1 mutants compared with isogenic WT respectively, and was statistically significant (P < 0.1, Student’s t-test). B. Complementation with Stk1 restores cylE transcription to WT. DNA-free RNA was isolated from LR113 complemented for Stk1 expression (LR113/pStk1) and LR114 complemented for either Stk1 (LR114/pStk1) or Stp1 (LR114/pStp1). As controls, RNA was isolated from A909, LR113 and LR114 containing the vector pDC123. RPA were performed on 25 mg total RNA with a 250 bp antisense [α³²P] UTP-labelled cylE probe. Note that transcription of cylE was restored to WT (A909/pDC123) in LR113 or LR114 strains that express Stk1. C. Transcription of cylE is similar in WT and stk1 mutants that lack CovR. DNA-free RNA was isolated from ΔcovR strains derived from A909, LR113, LR114, NCTC10/84, LR122 and LR123. RPA were performed on 25 mg total RNA with a 250 bp antisense [α³²P] UTP-labelled cylE probe. Transcription of cylE was similar in all isogenic ΔcovR strains (fold difference < 2.0, P-value > 1, Student’s t-test.)

Fig. 6. Stk1 negatively regulates CAMP factor expression in GBS serotype V

A. CAMP factor activity is increased in GBS serotype V stk1 mutants. CAMP factor activity is represented by the triangular zone of lysis at the junction between GBS and β-lysin producing Staphylococcus aureus on blood-agar plates. Upper panel depicts CAMP factor expression in GBS serotype Ia strains. A909 represents WT and LR113 and LR114 represent isogenic Δstk1 and Δstp1Δstk1 mutants. Control IaΔcylE is a β-H/C-deficient GBS strain that is isogenic to A909. Lower panel represents CAMP factor expression in GBS serotype V strains. NCTC10/84 represents WT serotype V and LR122 and LR123 represent isogenic Δstk1 and Δstp1Δstk1 mutants. CAMP factor titer of each strain is denoted within square brackets. Note that CAMP factor activity is greater in serotype V stk1 mutants LR122 and LR123 compared with WT NCTC10/84. B. Transcription of cfb (CAMP factor) is increased in GBS serotype V stk1 mutants. DNA-free RNA was isolated from A909, LR113, LR114, NCTC10/84, LR122 and LR123. RPA were performed on 25 mg total DNA-free RNA from each strain mentioned above with a 250 bp antisense [α³²P] UTP-labelled cfb probe. Transcription of cfb was similar between A909, LR113 and LR114 (fold difference < 2.0). In contrast, transcription of cfb was 2.0-fold greater in LR122 and LR123 compared with WT NCTC10/84 (P-value ≤0.01, Student’s t-test).

Fig. 7

Increased CAMP factor expression in serotype V stk1 mutants requires CovR. Upper panel indicates CAMP factor expression in WT serotype Ia strain A909 and ΔcovR strains derived from A909 and isogenic stk1 mutants LR113 and LR114. Lower panel indicates CAMP factor activity in ΔcovR strains derived from WT serotype V NCTC10/84 and isogenic stk1 mutants. Note that the increase in CAMP factor expression observed in LR122 and LR123 are abolished in isogenic LR122ΔcovR and LR123ΔcovR strains.

Fig. 8

Transcription of covR is similar in WT and stk1 mutants. DNA-free RNA was isolated from WT GBS A909 and NCTC10/84, Δstk1 mutants LR113 and LR122, Δstp1Δstk1 mutants LR114 and LR123. RPA were performed on 25 mg total RNA from each strain with a 250 bp antisense [α-³²P]UTP-labelled covR probe.

Fig. 9. Stk1 phosphorylates CovR in vitro

A. Approximately 500 ng of purified N-terminal GST tagged-Stk1 (GST-Stk1, lane 1), 1.5 mg of C-terminal his-tagged CovR (CovR-His6, lane 2) and 1 mg of C-terminal his-tagged CovS (CovS-His₆, lane 3) were analysed on 10% SDS-PAGE and stained with Coomassie brilliant blue. M represents pre-stained SDS-PAGE Standards and M2 represents unstained SDS-PAGE protein standards (Bio-Rad). B. In vitro phosphorylation reactions of 500 ng of CovR or CovS were performed in the presence and absence of equal amounts of Stk1. Lanes 1 and 5 represent addition of 10 mCi [γ-³²P]-ATP to CovR and CovS respectively. Autophosphorylation of Stk1 is shown in lane 3. Open arrow represents a breakdown product of Stk1 on autophosphorylation. Lanes 2 and 4 represent phosphorylation of CovR and CovS in the presence of Stk1 respectively. C. In vitro phosphorylation reactions were repeated to include 500 ng of GST (26 kDa, lane 1), 500 ng of GST-Stk1 (95.8 kDa, lane 2), 500 ng of GST-Stk1 and CovR-His₆ (lane 3), 500 ng of CovR-His₆ (lane 4) and 500 ng of GST and CovR-His₆ (lane 5) respectively. Phosphorylation reactions were performed as described in Experimental procedures except for the addition of excess radioactive ATP (50 mCi of [γ-³²P] ATP). All samples were heated to 100°C, analysed on 10% SDS-PAGE, and stained with Coomassie brilliant blue (see panel on the left). Subsequently, the gels were dried and exposed to autoradiography (see panel on the right). The intensity of autophosphorylation of GST-Stk1 (lanes 2 and 3) is remarkably higher due to the increase in radioactive ATP in the phosphorylation assay. Note that CovR does not autophosphorylate (lane 4) and is also not phosphorylated by GST (lane 5). Phosphorylation of CovR was observed only in the presence of Stk1 (lane 3).