The SaeRS two-component system regulates virulence gene expression in group B Streptococcus during invasive infection

ABSTRACT Group B Streptococcus (GBS) is a pathobiont responsible for invasive infections in neonates and the elderly. The transition from a commensal to an invasive pathogen relies on the timely regulation of virulence factors. In this study, we characterized the role of the SaeRS two-component system in GBS pathogenesis. Loss-of-function mutations in the SaeR response regulator decrease virulence in mouse models of invasive infection by hindering the ability of bacteria to persist at the inoculation site and to spread to distant organs. Transcriptome and in vivo analysis reveal a specialized regulatory system specifically activated during infection to control the expression of only two virulence factors: the PbsP adhesin and the BvaP secreted protein. The in vivo surge in SaeRS-regulated genes is complemented by fine-tuning mediated by the repressor of virulence CovRS system to establish a coordinated response. Constitutive activation of the SaeRS regulatory pathway increases PbsP-dependent adhesion and invasion of epithelial and endothelial barriers, though at the cost of reduced virulence. In conclusion, SaeRS is a dynamic, highly specialized regulatory system enabling GBS to express a restricted set of virulence factors that promote invasion of host barriers and allow these bacteria to persist inside the host during lethal infection. IMPORTANCE Group B Streptococcus (or GBS) is a normal inhabitant of the human gastrointestinal and genital tracts that can also cause deadly infections in newborns and elderly people. The transition from a harmless commensal to a dangerous pathogen relies on the timely expression of bacterial molecules necessary for causing disease. In this study, we characterize the two-component system SaeRS as a key regulator of such virulence factors. Our analysis reveals a specialized regulatory system that is activated only during infection to dynamically adjust the production of two virulence factors involved in interactions with host cells. Overall, our findings highlight the critical role of SaeRS in GBS infections and suggest that targeting this system may be useful for developing new antibacterial drugs.


SaeRS is required for invasive infection
To analyze the role of the SaeRS regulatory system in GBS, we first used loss-of-function mutations constructed in the NEM316 (CC-23) wild-type (WT) strain.The ΔsaeR is an in-frame deletion of the saeR gene, while the SaeR D53A mutant has a single chromosomal polymorphism (GAT→GCT) leading to substitution of the conserved aspartate residue D 53 by an alanine, thereby preventing SaeR phosphorylation by SaeS (Fig. 1) (13).The in vitro growth and morphology of the two mutants and the WT strain are similar in rich media (Fig. S1A and B).To assess the role of the SaeRS system during infection, we infected mice intravenously in a model of GBS meningoencephalitis.Less than half of the mice inoculated with ∆saeR or SaeR D53A mutants died, while all animals infected with WT GBS succumbed to infection (Fig. 2A).Quantification of bacteria in the organs of infected mice 24 h after challenge shows significantly lower bacterial counts in the blood and kidneys for both mutants compared to the WT strain (Fig. 2B through D).Similar lower bacterial loads are observed in blood at 48 h after infection, while kidney and brain invasion by saeR mutants is severely compromised (Fig. 2E through G), indicating that the SaeR regulator is necessary for systemic invasion.To test a second model of invasive infection, we intraperitoneally infected mice in a peritonitis-sepsis model of GBS infection.Both saeR mutants are less virulent than the WT strain (Fig. 3A) and are recovered in significantly lower numbers in the peritoneal cavity at 3 h post-infection (Fig. 3B).The saeR mutants do not efficiently spread systemically, as observed by reduced bacterial load in the blood at early time points and confirmed at 24 h post-infection (Fig. 3C through E), suggesting that SaeRS is critical for the initial phase of infection.

SaeRS regulates specialized virulence factors
To identify SaeRS-regulated genes, we performed RNA sequencing of the ΔsaeR and SaeR D53A mutants.Strikingly, no significant differences in gene expression were detected in either mutant compared with the parental strain after growing bacteria in THY (Fig. 4A  and B).This shows that the SaeRS system is not active in the WT strain in the tested condition.To overcome the requirement for the activating signal, we adopted a genetic approach called HK + (20).This approach relies on specific inactivation of the phosphatase activity of the histidine kinase to constitutively activate the signaling pathway independ ently of the signal (21,22).For SaeS, we introduced a single nucleotide polymorphism in the chromosome (ACT→GCT) leading to substitution of the catalytic threonine residue T 133 by an alanine.Transcriptome analysis of the SaeS T133A mutant revealed the highly specialized regulon of SaeRS (Fig. 4C).The regulon includes auto-regulation of the saeRS operon and activation of genes encoding for the virulence factors PbsP (15) and BvaP (12,23).We also observed moderate, but significant, activation of the operon down stream of the bvaP gene, which is likely due to terminator readthrough in the presence of massive bvaP transcription (Fig. 4C).Transcriptional hyperactivation of pbsP in the SaeS T133A mutant was confirmed by independent RT-qPCR (Fig. 4D).The overactivation of pbsP is corroborated by PbsP overexpression at the bacterial surface as shown by flow cytometry immunofluorescence analysis using PbsP polyclonal antibodies (Fig. 4E).Inactivation of SaeR by introducing the D 53 A mutation in the SaeS T133A mutant abol ishes pbsP and bvaP upregulation (Fig. 4F), confirming that the effects of the T 133 A mutation in SaeS depend on activation of the SaeR regulator.

SaeRS is activated during in vivo infection
To test SaeRS activation in vivo, we quantified pbsP and bvaP transcript levels by RT-qPCR in WT bacteria recovered from the peritoneal exudates of i.p. infected mice.The pbsP and bvaP genes are significantly upregulated 20-to 30-fold relative to in vitro grown bacteria (Fig. 5).To ascertain whether pbsP and bvaP expression is regulated in vivo by SaeRS, we infected mice with the ΔsaeR mutant.RT-qPCR analysis on bacterial RNA extracted from peritoneal lavage fluid (PLF) samples indicated that pbsP and bvaP expression is markedly reduced in the ΔsaeR mutant compared with the WT strain (Fig. 5), indicating that the SaeRS system is mainly responsible for the in vivo upregulation of pbsP and bvaP.
The pbsP gene was previously shown to be repressed by the CovR master regulator of virulence in the NEM316 strain (15).To compare SaeR-positive and CovR-negative regulations, we included as controls the cylE and bibA genes, which are directly regulated by CovR and encode for an enzyme required for the synthesis of the ß-hemolysin/ cytotoxin and the BibA adhesin, respectively (24,25).Increased transcription of the cylE and bibA genes in the WT and ΔsaeR mutant recovered from PLF samples confirms in vivo activation of the CovR regulon (i.e., the release of CovR repression) in the two strains compared to in vitro growth (Fig. 5).We next infected mice under the same conditions using a ΔcovR mutant and recovered total bacterial RNA from PLF samples.As expected, cylE and bibA transcription is up-regulated due to covR deletion and, interestingly, pbsP and bvaP transcription is also upregulated compared to the WT strain (Fig. 5).The moderate increase in pbsP and bvaP mRNA levels observed in the absence of CovR aligns with the slight but significant in vivo upregulation of these genes observed in the ∆saeR mutant in comparison with in vitro grown bacteria (Fig. 5).These findings demonstrate activation of both regulons during infection and confirm the co-regulation of PbsP and BvaP by primary SaeR activation and secondary CovR repression.

Dynamic SaeRS modulation is essential for infection
SaeRS is activated in vivo, necessary for virulence, and positively regulates only two virulence factors.To further characterize the role of SaeRS, we tested the activated SaeS T133A mutant for host-pathogen-related phenotypes.In vitro, the SaeS T133A mutant exhibits hyper-adhesion and hyper-invasion of A549 pulmonary alveolar epithelial cells and hCMEC/D3 brain endothelial cells compared to the WT strain (Fig. 6).To evaluate the role of the PbsP adhesin in these interactions, we repeated experiments with a ∆pbsP mutant constructed in the activated SaeS T133A mutant.Deletion of pbsP in the SaeS T133A mutant restores near WT levels of adhesion and invasion in both cellular models (Fig. 6).In agreement with SaeRS being inactive in vitro, deletion of saeR in the WT strain does not influence adhesion and invasion.By contrast, deletion of pbsP in the WT strain decreases adhesion and invasion by a twofold factor, in agreement with the presence of basal levels of PbsP expression in WT bacteria dependent on a secondary regulation.
Residual, low-level adherence of pbsP-deleted mutants is likely sustained by the activities of adhesins other than PbsP (15,23).Overall, these results demonstrate that the activa tion of SaeRS is a major determinant of host cell adhesion and invasion through positive regulation of expression of the PbsP adhesin.
Increased adhesion and invasion could enhance virulence by facilitating cellular translocation across crucial defensive barriers such as the blood-brain barrier (BBB).Previous studies have demonstrated that PbsP binds plasminogen (Plg), enabling GBS to migrate across endothelial cells following the conversion of Plg to plasmin by tissue plasminogen activator (tPa) (15,19).To investigate the role of SaeRS in GBS transmigra tion across endothelial cells, we utilized an in vitro BBB model involving hCMEC/D3 monolayers grown on transwell membrane inserts and bacteria pre-treated with Plg and tPa (16).Under these conditions, the activated SaeS T133A mutant crosses monolayers much more efficiently than the WT strain, a process dependent on the overexpression of PbsP (Fig. 7A).These findings suggest that the activation of SaeRS can increase virulence by promoting epithelial adhesion, invasion, and BBB crossing.However, the constitu tively activated SaeS T133A mutant is avirulent when directly injected into the blood stream (Fig. 7B).Accordingly, the SaeS T133A mutant is rapidly cleared from the circulating blood (Fig. 7C and D).Similarly, the SaeS T133A mutant is recovered in significantly lower numbers in the peritoneal cavity 1-3 h after intraperitoneal infection and is cleared after 24 h (Fig. 7E).The capsular polysaccharide enables GBS to evade host defenses and is, therefore, a major determinant of the ability of these bacteria to persist in vivo (3).However, it is unlikely that the impaired virulence observed in SaeS T133A GBS is linked to decreased capsule expression since the cps operon is not differentially expressed in this mutant compared to the WT strain (Fig. 4C).Collectively, our data indicate that constitu tive upregulation of SaeRS increases cellular invasion in vitro but decreases overall virulence, suggesting that dynamic regulation is necessary during in vivo infection.This underscores the importance of tightly regulating SaeRS activation, particularly the expression of PbsP, to promote interactions at epithelial and endothelial barriers while avoiding continuous over-activation that impedes bacterial virulence.

DISCUSSION
The present study establishes the SaeRS two-component system as a key regulator of host-GBS interactions.This system is specifically activated during in vivo infection and has a necessary role in several models of invasive GBS disease while being relatively silent during in vitro growth.However, non-physiological, constitutive activation of the SaeRS system decreases GBS virulence, highlighting the need for tight and dynamic regulation during different phases of pathogenesis.Originally, SaeRS was shown to be activated by a small molecule (<3 kDa, probably a heat-labile peptide) present in the vaginal fluid (12).The necessity for SaeRS regulation in several models of invasive infections, as shown here, suggests the presence of a widespread activating molecule that is not exclusive to the vaginal environment.The requirement for SaeRS at several stages of infection is consistent with the pronounced in vivo upregulation of the adhesin PbsP and its contribution to hematogenous dissemination (15), meningitis (16), and infection of diabetic wounds (17).Taken together, our study establishes SaeRS as a dynamic system that is specifically activated in vivo to rapidly upregulate virulence genes and promote invasive infection.
In addition to being highly dynamic, the SaeRS system displays a remarkable degree of specialization, being specifically dedicated to the positive regulation of the cell-wallanchored adhesin PbsP and the secreted protein BvaP.Previous in vitro transcriptomic analysis suggested that SaeRS regulates a large regulon of 301 to 466 genes depending on the growth media (12).However, absent or low-level in vitro activities severely limit functional analysis using SaeRS deletion mutants.For instance, we did not identify a single, significantly regulated gene when using in vitro inactivated saeR mutants of the NEM316 strain.To decipher SaeRS signaling, it is therefore necessary to activate the signaling pathway, either in the presence of the activating signal or by genetic engineer ing.By analyzing a ∆saeR mutant during in vivo growth in the mouse vaginal tract, the regulation of both pbsP and bvaP by SaeRS was previously demonstrated in the context of differential regulation of approximately one-third of the GBS genome (12).Our genetic approach bypasses the requirement for the signal and provides a focused, highresolution view of the regulon during in vitro growth in rich media.We have recently systematically applied the HK + approach in GBS to demonstrate the versatility of this gain-of-function strategy that works outstandingly well for SaeRS in deciphering TCS signaling (20).
The PbsP and BvaP virulence factors are positively regulated by SaeRS but are also repressed by the master regulator of virulence CovRS.One characteristic of the CovRS system is its plasticity, leading to strain-specific regulation (7).For example, CovR negatively regulated PbsP expression in the NEM316 strain (15), but this regulation was severely attenuated in the BM110 strain (16), a CC17 isolate representative of the hypervirulent lineage, although CovR binding to the pbsP promoter is conserved in both strains (7).In addition, overexpression of PbsP activated CovRS signaling in BM110, suggesting that PbsP acts as a signaling molecule connecting SaeRS and CovRS regulation in CC17 strains (20).We do not observe such activation of CovRS signaling in the SaeS T133A mutant in NEM316, likely due to an already significant activation of CovRS signaling in the NEM316 wild-type strain compared to BM110 (7).Overall, the SaeRS and CovRS regulatory pathways are connected through CovR repression of pbsP and activation of CovR signaling by PbsP.This is at variance with the inability of PbsP overexpression to trigger CovRS signaling in NEM316, as found here, which might be linked to differences in basal levels of CovR activity between the two strains (7).Overall, our results show that SaeRS is the main regulator of PbsP expression and acts as a strong activator, while CovRS is a secondary repressor that fine-tunes the expression of the adhesin.The SaeRS system appears necessary to enhance interactions with epithelial and endothelial cells during mucosal colonization and invasive infections, as observed here and in a previous study (12), while CovRS globally regulates bacterial pathogenicity.Interestingly, in vitro PbsP expression in WT strains is variable among isolates (15), raising the possibility that the equilibrium between SaeRS and CovRS influences the infectivity and colonization potential of each strain.It is interesting to note, in this respect, that the BvaP adhesin, whose SaeRS-dependent expression also contributes to vaginal coloniza tion, contains a variable number of repeated domains (23), suggesting selective pressure exerted by the host in altering the colonization potential of each strain.
In conclusion, we demonstrate the important role of the SaeRS pathway during systemic infection, the specialization of its restricted regulon in promoting adherence to and invasion of cellular barriers, and the presence of a dynamic regulatory network that involves, in addition to SaeRS, negative regulation by CovRS and contributes to strain specificity.This regulatory logic ensures that infection events occur efficiently, likely through initial overexpression of the PbsP adhesin and tight regulation of its production during subsequent phases of the infectious process.

Bacterial strains and mutagenesis
We used the GBS strain NEM316, a human prototype serotype III clinical strain belonging to clonal complex 23 (CC23), throughout the present study.NEM316 GBS and its mutants (Table S1) were cultured in Todd Hewitt Broth (Difco, BD) supplemented with 5 g/L of yeast extract (THY; BD) at 37°C.E. coli strains were cultured in Luria Bertani broth (BD) supplemented with erythromycin (150 µg/mL) at 37°C with shaking.Purification of GBS genomic DNA and E. coli plasmid DNA was carried out with, respectively, the DNeasy Blood and Tissue kit and the Quiaprep Spin Minipreps kit (both from Qiagen), following the manufacturer's instructions.The oligonucleotides used for genomic and transcriptomic analysis (provided by Eurofins MWG Operon) are listed in Table S2.PCRs for cloning and sequencing were performed using a high-fidelity polymerase (Phusion Plus DNA Polymerase; Thermo Scientific; cat.F630S).The pG1_ΔsaeR, pG1_SaeR D53A , pG1_ΔSaeS T133A , and pG1_ΔpbsP plasmids are listed in Table S3 and were constructed using splicing by overlap extension method with primers indicated in Table S4 (15).After transforming GBS with pG1_ΔsaeR, pG1_SaeR D53A , pG1_SaeS T133A , or pG1_ΔpbsP, integration, and de-recombination events were selected as described (15,20).The presence of the desired mutations was confirmed by whole-genome sequencing using the Illumina MiSeq platform.The ΔcovR strain was described in a previous study (26).

Animal models of GBS infection
Virulence of GBS was tested in 6-to 8-week-old CD1 female mice (Charles River) in accordance with the European Union guidelines for the use of laboratory animals.In the meningitis model, mice were infected i.v. with ~1×10 8 bacteria in a total volume of 0.1 mL of Dulbecco's PBS (DPBS, Sigma-Aldrich) and clinical signs were monitored every 12 h for 12 days.Animals with signs of irreversible disease or neurological signs, as assessed using a scoring system (27), were humanely euthanized.In a second group of experiments, i.v.infected mice were sacrificed at 24 or 48 h after infection to collect blood, brains, and kidneys.Organs were homogenized in the gentleMACS dissociation system (Miltenyi Biotec), as previously described (16,28).The number of colonies forming units (CFU) was measured in organ homogenates using previously described methods (15).In the sepsis model, CD1 mice were intraperitoneally (i.p.) injected with ~5×10 7 CFU in 0.2 mL of DPBS and monitored every 12 h for clinical signs as detailed above.GBS replication in the peritoneal cavity and its systemic spreading to other tissues was verified by plating blood, organ homogenates, and peritoneal lavage fluids (PLFs) at different time points.PLF samples were obtained by injecting 2 mL of phosphate-buffered saline (PBS) in the peritoneal cavity and subsequently aspirating a total of 1.7-1.9mL of fluid, as previously described (29)(30)(31)(32).

Quantitative RT-PCR and RNA sequencing
To measure the transcriptional levels of genes encoding for GBS virulence factors, bacterial RNA was extracted from PLF samples obtained at 1 h post-infection or from bacteria grown in vitro in THY, retro-transcribed, and analyzed using real-time PCR (RT-PCR), exactly as previously described (16).After PLF collection and centrifugation (12,000 × g for 10 min), eukaryotic cells were lysed by exposure to cold distilled water for 10 min.To collect bacteria, tissue debris and residual eukaryotic cells were removed by low-speed centrifugation (200 × g for 10 min) and, subsequently, superna tants were centrifuged at high speed (12,000 × g for 10 min) to obtain the bacterial pellet.Quantitative PCR (qPCR) was performed with the Taqman Gene Expression Master MIX (Applied Biosystem, cat.4369016) using probes (shown in Table S2) to detect the following transcripts of the following genes: pbsP, bvaP, cylE, bibA, and gyrA by the CFX Opus Real-time PCR System (Biorad).Relative gene expression levels were calculated with the ΔΔCT method, where expression values were normalized with the expression of the housekeeping gyrA gene.Each experiment was performed in triplicate.
Single-end strand-specific 75 bp reads were cleaned of adapter sequences and low-quality sequences (cutadapt version 1.15) and only sequences at least 25 nt in length were considered for further analysis.Bowtie v1.2.1.1 with default parameters was used for alignment on the NEM316 genome (NCBI: NC_004368).Genes were counted using featureCounts version v1.5.3 from a Subreads package (parameters: -t locus_tag -g ID -s 1).Count data were analyzed using R version 3.6.1 and the Bioconductor package DESeq2 version 1.26.0.The normalization and dispersion estimation were performed with DESeq2 using the default parameters but statistical tests for differential expression were performed by applying the independent filtering algorithm.A generalized linear model was set to test for the differential expression between the biological conditions.For each pairwise comparison, raw P-values were adjusted for multiple testing according to the Benjamini and Hochberg (BH) procedure and genes with an adjusted P-value lower than 0.05 were considered differentially.Raw sequencing reads and statistical analysis are publicly available (GEO accession number GSE269249).

Flow cytometry analysis
PbsP expression on the bacterial cell surface was visualized using flow cytometry immunofluorescence analysis using a mouse anti-PbsP serum and a normal serum control, as previously described (18).Briefly, GBS cells (~1×10 8 ) grown to the log phase in THY were washed in DPBS, fixed with 3.7% formaldehyde, and blocked using DPBS supplemented with 1% milk for 30 min at 22°C and gentle shaking.Bacteria were incubated with the mouse anti-PbsP serum diluted 1:50 for 1 h with gentle shaking.Subsequently, a phycoerythrin (PE-A)-conjugated goat anti-mouse IgG (ThermoFisher, cat.12-i10-82) diluted 1:50 in 1% milk was used to reveal primary antibody binding.Fluorescent bacteria were analyzed with a FACSCanto II flow cytometer using the FlowJo software (BD Biosciences).

Enzyme-linked immunosorbent assay
For testing PbsP expression on the GBS surface, 96 well Nunc MaxiSorp flat-bottom plates (Thermo Fisher Scientific; 44-2404-21) were coated at 4°C overnight with streptococci at a density of ~1×10 7 CFU/well in 0.05 M carbonate buffer (pH 9.5).After washing and blocking with 5% BSA in Tris-buffered saline pH 7.5 (TBS; 50 mM Tris-Cl; 150 mM NaCl), PbsP expression was evaluated using mouse anti-PbsP serum diluted 1:4,000 in TBS-1% BSA and incubated for 1 h at room temperature (RT) with gentle shaking.To detect antibody binding, anti-mouse IgG conjugated with horseradish peroxidase (HRP) diluted 1:1,000 in TBS-1% BSA was added and left for 45 min at room temperature.After the addition of o-phenylenediamine dihydrochloride (ODP; code 34006, Thermo Scientific), absorbance at 490 nm was determined in an enzyme-linked immunosorbent assay (ELISA) plate reader.

FIG 2 FIG 3
FIG 2 SaeR is required for virulence in a meningitis model.(A) Survival curves of intravenously (i.v.) infected mice.Adult female CD1 mice were infected with 10 8 CFU of NEM316 (WT), saeR deletion mutant (ΔsaeR), or SaeR-non-phosphorylable mutant with a D 53 A substitution in SaeR (SaeR D53A ).Animals with signs of irreversible disease were euthanized.***P < 0.001 by log-rank Mantel-Cox analysis.Shown are cumulative data from two experiments, each involving eight animals per group.(B-G) Effects of SaeR mutations on organ bacterial burden at 24 (B-D) and 48 h (E-G) post-infection.Shown are cumulative data from two experiments, each involving four animals per group.Horizontal red bars indicate mean values.The dashed lines indicate the limits of detection of the test.*P < 0.05; **P < 0.01; ns, non-significant as determined by the Wilcoxon test.

FIG 3 (FIG 4
FIG 3 (Continued) < 0.001 by log-rank Mantel-Cox analysis.Shown are cumulative data from two experiments, each involving eight animals per group.(B-E) Bacterial burden in peritoneal lavage fluid samples (PLFs) at 3 h (B), in the blood at 3 (C) and 24 (D) h, and in the kidneys at 24 h (E) post-infection.Mice were infected as indicated in panel A. Shown are cumulative data from two experiments, each involving four animals per group.Horizontal red bars indicate mean values.The dashed lines indicate the limits of detection of the test.*P < 0.05; **P < 0.01, ns, non-significant, as determined by the Wilcoxon test.

FIG 4 (FIG 5
FIG4 (Continued)    genes are color-coded (red: saeRS, pbsP, and bvaP; orange: bvaP adjacent operon).(D) Independent RT-qPCR validation of pbsP and bvaP overexpression in the genetically activated SaeS T133A mutant.Results are means ± SD from three independent experiments performed in triplicate.*P < 0.05, determined by Mann-Whitney statistical analysis.(E) Expression of PbsP on the GBS surface.Immunofluorescence flow cytometry analysis of PbsP expression on NEM316 WT (green area), SaeS T133A (red area), and SaeS T133A with a deletion of pbsP (pink area) using mouse polyclonal anti-PbsP serum.The gray area refers to the reactivity of each GBS strain with normal serum.(F) SaeR phosphorylation is required for pbsP overexpression.RT-qPCR analysis of pbsP mRNA levels in SaeS T133A and SaeS T133A SaeR D53A mutants.Results are means ± SD from three independent experiments performed in triplicate.*P < 0.05, as determined by Mann-Whitney statistical analysis.(G) SaeR phosphorylation is required for PbsP expression on the bacterial surface.Anti-PbsP mouse serum was used to measure PbsP surface expression using an enzyme-linked immunosorbent assay test on NEM316 WT, SaeS T133A , SaeS T133A ΔpbsP, SaeS T133A SaeR D53A , and ΔpbsP strains.

FIG 6 FIG 7
FIG 6 Activation of SaeRS signaling increases PbsP-dependent adhesion and invasion of epithelial and endothelial cells.(A and B) Adhesion and invasion of A549 epithelial cells by NEM316 WT strain (green), ∆pbsP (yellow), ∆saeR (blue), SaeS T133A (dark red), and SaeS T133A ∆pbsP (pink) mutants.Adherence and invasion are expressed as percentages of cell-associated bacteria relative to the total number of bacteria added to the monolayers.(C and D) Similar to panels A and B using hCMEC/D3 endothelial cells.Results are means ± SD from three independent experiments performed in triplicate.*P < 0.05; ns, non-significant, as determined by Mann-Whitney statistical analysis.

FIG 7 (
FIG7 (Continued)    of bacteria added to the monolayer in a transwell assay in the presence or not of plasminogen (Plg) and tissue Plg activator (tPA).Wild-type NEM316 (green), SaeS T133A (red), and SaeS T133A ΔpbsP (pink) strains were tested.Results are means ± SD from three independent experiments performed in triplicate.*P < 0.05, as determined by Mann-Whitney statistical analysis.(B) Survival curves of mice infected with WT (green) and SaeS T133A mutant (red) GBS.Adult female CD1 mice were infected i.v. with 1 × 10 8 CFU and clinical signs were monitored.Animals with signs of irreversible disease were euthanized.***P < 0.001 by log-rank Mantel-Cox analysis.(C and D) Bacterial burden in the blood 24 and 48 h post-infection.Mice were infected i.v. as in panel B. (E) Bacterial burden in PLF samples at the indicated times after i.p. challenge with 5 × 10 7 CFU.Shown are cumulative data from two experiments, each involving four animals per group.Horizontal red bars indicate mean values.The dashed lines indicate the limits of detection of the test.*P < 0.05; **P < 0.01, ns, non-significant, as determined by the Wilcoxon test.