Regulation of Staphylococcus aureus α-Toxin Gene (hla) Expression by agr, sarA and sae In Vitro and in Experimental Infective Endocarditis
Regulation of Staphylococcus aureus a-Toxin Gene (hla) Expression by agr, sarA, and sae In Vitro and in Experimental Infective Endocarditis
Yan Q. Xiong 1 2
Julie Willard 2
Michael R. Yeaman 1 2
Ambrose L. Cheung () 0
Arnold S. Bayer 1 2
0 Department of Microbiology, Dartmouth Medical School , Hanover, New Hampshire
1 David Geffen School of Medicine at University of California Los Angeles , Los Angeles
2 Department of Medicine, Division of Infectious Diseases, Los Angeles Biomedical Research Institute at Harbor-University of California Los Angeles Medical Center , Torrance
Background. The Staphylococcus aureus global regulators-agr, sarA, and sae-coordinately control a-toxin gene (hla) expression in vitro. However, their relative in vivo contributions to hla expression, particularly in endovascular infections, have not been defined Methods. A plasmid-based hla-promoter:green fluo escent protein reporter system was constructed in 2 genetically related S. aureus strains: RN6390 (a natural sigma factor B [sigB]-deficien mutant), SH1000 (a sigBrepaired variant of RN6390 lineage), and their respective agr, sarA, agr/sarA, and sae mutants. These strain sets were used to quantify hla expression in vitro and in an experimental infective endocarditis (IE) model using flo cytometry. Results. In vitro, hla expression was positively modulated by all 3 regulons (sae 1 agr/sarA 1 agr and sarA) in both RN6390 and SH1000 backgrounds. In the IE model, hla expression in cardiac vegetations was lower in all single mutants than in the respective parental strains (P ! .05 for sae mutant) but was maintained at nearparental levels in the agr/sarA double mutant in both backgrounds. A similar findin was also observed in kidneys and spleens. Conclusions. These results indicate that, although several distinct regulatory circuits can affect hla expression in vitro and in vivo, sae appears to play a crucial role in this context.
Staphylococcus aureus a-toxin (a 33.2-kDa polypeptide
encoded by hla) is a pore-forming cytotoxin that is
produced by the majority of S. aureus strains  and
targets a broad range of host cell types [1, 2]. Like most
staphylococcal extracellular proteins, a-toxin is not
expressed constitutively but is tightly regulated by an array
of extracellular and intracellular signals. At least 3 global
regulatory loci—the accessory gene regulator (agr), the
staphylococcal accessory gene regulator (sarA), and the
staphylococcal accessory protein effector (sae)—appear
to coordinately control hla expression in S. aureus in
vitro [3–7]. In addition, 2 sarA homologues, rot and
sarT, repress hla expression [8–10]. The agr locus exerts
a direct positive impact on hla expression, whereas sarA
positively affects hla expression by both agr-dependent
and agr-independent pathways [3, 5–7, 11]. The sae
locus consists of a 2-component signal-transduction
system encoded by saeS (sensor) and saeR (response
regulator) that positively regulates the expression of hla
at the transcriptional level [4, 12, 13]. In addition, the
complex transcriptional profil of sae activation is
influence by agr, as well as by certain environmental
stimuli, and it may also be influence by sigma factor
B (sigB) .
Although the regulation of hla has been studied
extensively in vitro, its in vivo regulation remains
incompletely defined especially that involved in
staphylococcal endovascular infections. Recently, we examined the
in vitro versus in vivo expression patterns of a number
of key regulatory and structural virulence genes in S.
aureus in an experimental infective endocarditis (IE) model.
These investigations demonstrated that their in vitro expression
and regulation profile were often not precisely mirrored in
vivo [15–18]. Moreover, we define target tissue-specifi
differences in gene expression, as well as unexpected regulatory
pathways that were active exclusively in the IE model.
Therefore, these data strongly implicated that host environmental
cues and/or additional regulatory circuits play a major role in
the in vivo activation of S. aureus virulence genes.
The present study was designed to compare hla activation
profile in vitro versus in vivo and to examine the effects of
the global regulators (agr, sarA, and sae) on such activation
profile in 2 genetically related S. aureus strains that differ
principally in sigB functionality: RN6390 (a natural sigB-deficien
mutant) and SH1000 (a sigB-repaired variant of the RN6390
lineage). A well-characterized S. aureus IE model was used in
these investigations. The model mimics its human counterpart
microbiologically, immunologically, histopathologically,
pathogenetically, and anatomically . At an inoculum of ID95,
sequential stages of pathogenesis ensue, including the
colonization of damaged cardiac endothelium, microbial persistence
and proliferation, eventual hematogenous seeding of distant
organs, and reseeding of the endocardial lesions.
MATERIALS AND METHODS
Microorganisms and plasmids. The S. aureus strains and
plasmids used in the study are listed in table 1. S. aureus RN6390
is a a-toxin–producing strain that has been used in numerous
studies [5, 15, 20]. It contains a natural 11-bp deletion in the
rsbU sensing locus of its sigB operon that renders it functionally
sigB deficien . SH1000 is a sigB-repaired variant (rsbU+)
in the RN6390 lineage strain, 8325-4, that renders it sigB
functional. For construction of the hla promoter:reporter plasmid,
pALC1740 in the N315 genome, the promoter fragment of hla
(nt 1,141,972–1,140,534) was amplifie by polymerase chain
reaction (PCR) on basis of the hla sequence of strain N315
. This PCR fragment was cloned into pCR2.1 and then
ligated into pALC1484 upstream of the red-shifted green fluo
rescent protein (GFP) gene, gfpuvr, in Escherichia coli [16, 17].
Correct insertion of the hla promoter fragment was confi med
by restriction digestion and DNA sequencing. The recombinant
plasmid pALC1740, which contains the hla promoter, was
electroporated firs into RN4220 to obtain
chloramphenicol-resistant transformants. The plasmid was then purifie from
RN4220derived transformants and electroporated into RN6390 and
SH1000 and their isogenic agr, sarA, agr/sarA, or sae mutants.
Of note, the number of copies of this plasmid has been shown
to be ∼15 copies/cell . For verificatio of plasmid
maintenance and for routine storage, S. aureus strains were grown
with appropriate antibiotic selection markers: chloramphenicol
(10 mg/mL) for the selection of pALC1740 or its derivatives,
tetracycline (5 mg/mL) for Dagr::tetM, and erythromycin (5 mg/
mL) for DsarA::ermC and saeR::Tn551 mutants [4, 24, 25].
Importantly, the pALC1740 plasmid was retained in all
constructs after in vitro and in vivo animal model passage, as
evidenced by restriction analyses and retention of
Laboratory strain related to 8325-4 harboring an 11-bp deletion in rsbU
RN6390 with pALC1484
RN6390 with pALC1740
agr mutant of RN6390 with pALC1740
sarA mutant of RN6390 with pALC1740
agr/sarA double mutant of RN6390 with pALC1740
sae mutant of RN6390 with pALC1740
Repaired sigB stress response operon in 8325-4 (RN6390 lineage)
SH1000 with pALC1484
SH1000 with pALC1740
agr mutant of SH1000 with pALC1740
sarA mutant of SH1000 with pALC1740
agr/sarA double mutant of SH1000 with pALC1740
sae mutant of SH1000 with pALC1740
RN6390 with Dspa::Etbr mutation
Shuttle vector containing pUC19 cloned into the HindIII site of pC194
pSK236 containing gfpuvr preceded by the hla promoter fragment
Derivative of pSK236 containing the promoterless gfpuvr gene
phenicol resistance . All S. aureus strains were stored at
70 C until they were thawed for in vitro and in vivo use.
Expression of hla in vitro as determined by flo cytometry.
Overnight S. aureus cultures were diluted into fresh tryptic soy
broth (TSB) to an initial inoculum of 1 107 cfu/mL and then
incubated at 37 C on a rotary shaker (at 200 rpm) for 24 h
[17, 18]. At selected time points, 0.5-mL aliquots were obtained
for the assessment of hla promoter expression by flo
cytometry (FACScalibur; Becton-Dickinson), as described elsewhere
[17, 18]. For each sample, 10,000 cells were acquired and
analyzed for quantitative GFP fluo escence [16–18]. RN6390 and
SH1000 that contained a promoterless gfpuvr plasmid were used
as GFP-negative controls. Data were expressed as the mean
SD percentage of acquired cells expressing GFP above the
baseline (set at 10 fluo escent units in the FL-1 channel, as described
elsewhere) [17, 18].
Northern-blot analysis of hla transcription. To verify the
in vitro patterns of hla transcription revealed by the GFP
reporter methods described above, we compared gene expression
in the parental strains with that in selected mutants, using
Northern analyses. In brief, S. aureus cells in the early stationary
phase were processed by use of a Trizol isolation kit (Gibco
BRL) in combination with 0.1-mm sirconia-silica beads in a
Biospec reciprocating shaker to yield RNA, as described
elsewhere . Then, 15 mg of each sample was electrophoresed
in a 1.5% agarose–0.66 mol/L formaldehyde gel in
morpholine propanesulfonic acid (MOPS) running buffer (20 mmol/L
MOPS, 10 mmol/L sodium acetate, and 2 mmol/L EDTA [pH
7.0]). Blotting of RNA onto Hybond N+ membranes (Amersham)
was performed using the Turboblotter alkaline transfer system
(Schleicher & Schuell). For detection of the hla transcript,
gelpurifie DNA probes were radiolabeled with a-32P dCTP using
the random-primed DNA labeling method (Roche Diagnostics)
and hybridized under aqueous-phase conditions at 65 C. The
blots were subsequently washed and autoradiographed.
Chemiluminescent Western-blot analysis. To delineate the
correlation between hla transcription and a-toxin production,
we tested the specifi production of this protein in both RN6390
and SH1000 backgrounds. S. aureus culture supernatants from
equal numbers of cells in the early stationary phase of growth
were applied to a 12% SDS-polyacrylamide gel. The proteins
were transferred to nitrocellulose membranes (Millipore). After
transfer, the membranes were blocked overnight with 1%
nonfat dry milk in Tris-buffered saline that contained 0.05% Tween
20 and then incubated for 5 h with a sheep anti–staphylococcal
a–toxin IgG:horseradish peroxidase conjugate (1:600 dilution;
Toxin Technology) that contained 0.05% Tween 20. The blot
was then developed with supersignal West Dura substrate
(Pierce Biotechnology). The chemiluminescent emissions from
each construct were quantitatively compared using the
Fluorchem 8900 Imaging System (Alpha Innotech). Purifie a-toxin
(20 mg) was used as a positive control (Toxin Technology).
Hemolytic activity assay. We evaluated the relative
hemolytic activities of the parental strains and sae mutants using
rabbit erythrocytes, as described elsewhere . Briefl ,
overnight cultures of S. aureus were equalized to the optical density
at 600 nm and then centrifuged at 2500 g to pellet bacteria.
Aliquots of serial dilutions of culture supernatants were added
to a 1% suspension of washed rabbit erythrocytes in 0.01 mol/
L PBS (pH 7.2) that contained 0.1% bovine serum albumin
(Sigma) and were incubated for 30 min at 37 C . Triton
(10%) was used as a positive control. The hemolytic activity
was expressed as mean units of hemolytic activity ( SD) per
milliliter. Hemolytic units (HU) were define as the reciprocal
of the highest dilution of the culture supernatant that caused
50% hemolysis .
Rabbit IE model. A well-characterized rabbit IE model was
used to evaluate regulation of hla expression in vivo. In brief,
female New Zealand white rabbits (Irish Farms Products and
Services) underwent transcarotid-transaortic valve
catheterization [15–18]. At 24 h after catheterization, rabbits were
challenged intravenously with each S. aureus construct at the
previously define ID95 inoculum (1 107 cfu/rabbit) [17, 18].
Rabbits were euthanized by a rapid intravenous injection of
sodium pentobarbital (100 mg/kg; Abbott Laboratories) at 48
h after infection. At death, vegetations, kidneys, and spleen were
removed for the determination of bacterial densities, and these
were also processed in parallel for the detection of hla
expression (see below). Only samples from rabbits with proper
catheter placement and macroscopic vegetations were analyzed in
the present study.
Microbiologic evaluation. Homogenized tissue samples
were quantitatively cultured onto TSB agar plates and incubated
at 37 C for 24 h. S. aureus densities were expressed as mean
SD log10 colony-forming units per gram of tissue.
hla regulation in vivo. Tissue samples were prepared for
flo cytometric-analysis by a series of processing steps to
remove tissue debris, as detailed elsewhere [17, 18]. In addition,
to enhance the specificit of these analyses, an
immunodetection system featuring a labeled anti–staphylococcal protein A
antibody was used to distinguish S. aureus cells from
hosttissue debris [17, 18]. This dual fluo ophore system features
nonoverlapping emission spectra for GFP and the labeled
protein A probe, which causes no quenching of the GFP signal
[17, 18]. A protein A knockout mutant of S. aureus (ALC1645)
was used to confi m the specificit of this immunolabeling, as
described elsewhere [17, 18]. GFP quantitation was limited to
protein A–positive S. aureus cells.
We recognize that the 3 regulators (agr, sarA, and sae) each
modulate the production of protein A in S. aureus. For instance,
both agr and sarA repress the transcription of spa (the protein
A gene), whereas sae up-regulates the transcription of protein
A. However, the flow-cytometri analyses were highly sensitive,
and we limited the detection system to only a baseline level of
protein A expression in vivo among the 10,000 events captured,
which circumvented potential selection bias due to different
levels of protein A expression.
Statistical analysis. To statistically compare GFP fluo
escence and target tissue densities among the various S. aureus
strains, the Kruskal-Wallis analysis of variance test was used
with Tukey post hoc adjustments for multiple comparison
where indicated. P ! .05 was considered to be significant
hla regulation in vitro. Figure 1 shows the results of in vitro
hla expression and regulation in both RN6390 and SH1000
backgrounds over the course of a 24-h growth cycle. As
expected, a growth phase–dependent hla activation profil was
observed in both RN6390 and SH1000 parental strains that
reached its maximum during the late log/early stationary phases
of growth. In RN6390, hla expression decreased by ∼10% in
the agr single mutant, compared with the parental strain. For
the sarA and agr/sarA mutants, hla activation was substantially
decreased throughout the growth cycle, compared with that of
the parental strain (∼50% reduction of hla expression during
8–16 h incubation). Of interest, during 16–24 h of incubation,
hla activation increased substantially in the agr/sarA mutants
(i.e., the percentage of hla-expressing S. aureus cells increased
from 16% at 16 h to 56% at 24 h). In SH1000, hla expression
was lower than that in the RN6390 parental strain (e.g., 56%
vs. 72% hla-expressing cells, respectively, at 24 h). Similar to
the RN6390 data, hla activation increased substantially in the
agr/sarA double mutant, compared with the parental strain in
SH1000, after 16 h of incubation. Minimal GFP expression was
observed in the sae mutants in both RN6390 and SH1000
backgrounds (figu e 1). In addition, the promoterless gfpuvr construct
in both backgrounds demonstrated !2% GFP expression during
24 h of incubation. Of importance, the above percentage of
GFP expression in temporal profile paralleled the GFP fluo
rescence intensity per cell in study constructs (data not shown).
Northern-blot analysis of hla transcription. RNA
transcript analysis confi med several important aspects of the hla
expression profile as define by the GFP reporter described
above in both the RN6390 and SH1000 backgrounds (figu e
2). In particular, the extent of hla activation was lower in the
SH1000 than in the RN6390 parental strains. Additionally, the
expression of hla in the agr/sarA double mutants was
maintained but was lower than that of the parent strains in both
backgrounds. Furthermore, hla transcription was undetectable
in both sae mutants.
Chemiluminescent Western-blot analysis. Paralleling the
hla transcription data above, the RN6390 parental strain
produced higher amounts of a-toxin than the SH1000 parental
strain (figu e 3). In addition, the hierarchy of a-toxin
production paralleled the transcriptional data by GFP reporter and
Northern-blot analyses (i.e., parent 1 agr/sarA 1 sae mutants)
(figu e 3).
Hemolytic activity assay. As expected, the RN6390 parental
strain had greater hemolytic activity (221 27 HU/mL) than
that of SH1000 parental strain (17 1 HU/mL), which paralleled
the higher a-toxin secretion by RN6390 than that by SH1000
that was determined by Western-blot analysis. The sae mutants
exhibited no detectable hemolytic activity (!10 HU/mL).
Microbiologic evaluation in the rabbit IE model. S. aureus
RN6390 densities achieved in vegetations were significantl
higher than those in kidneys and spleen for each individual
study construct (P ! .05) (table 2). However, only the sarA
mutant exhibited a significan reduction in bacterial counts in
kidneys, compared with the parental strain (P ! .05) (table 2).
Similar to RN6390, S. aureus counts in vegetations in all
SH1000 strains were significantl higher than those in the
respective kidneys and spleen (P ! .05) (table 3). The sae mutant
Table 2. Staphylococcus aureus RN6390 densities in 48-h target tissues from
rabbits with endocarditis.
(no. of rabbits)
SD, log10 cfu/g tissue
Spleen 7.05 6.51 5.92
NOTE. P ! .05, vegetation counts vs. those in kidneys or spleen of each individual strain.
a P ! .05 vs. control.
Table 3. Staphylococcus aureus SH1000 densities in 48-h target tissues from
rabbits with endocarditis.
(no. of rabbits)
SD, log10 cfu/g tissue
Kidneys Spleen 7.96 7.57
NOTE: P ! .05, vegetation counts vs. those in kidneys or spleen of each individual strain.
a P ! .05 vs. the parental strain.
of SH1000 achieved significantl lower S. aureus densities in
all 3 target tissues, compared with the parental strain (P !
.05). When RN6390 was compared with SH1000, RN6390
strains had somewhat lower S. aureus counts than SH1000
strains in vegetations. However, these differences did not reach
hla regulation in vivo in the rabbit IE model. Figure 4
depicts the in vivo hla expression profile by the isogenic strain
sets within vegetations, kidneys, and spleen. Data (mean, 6–10
target tissue assays) are presented that compare the various
mutants to the respective parental strains (set at 100%). In the
RN6390 strain set, the percentage of hla-expressing cells in
cardiac vegetations was modestly (but not significantly lower
in the agr and sarA single mutants than in the parental controls.
Importantly, hla expression in all target tissues was significantl
lower in the sae mutant (P ! .05 vs. the parental strain), whereas
the agr/sarA double mutant retained hla expression at or above
the parental level. Similarly, in SH1000 strains, hla expression
was maintained at or above the parental level in the agr/sarA
double mutant in all 3 target tissues. In addition, hla expression
was substantially lower in the sae mutant, compared with the
other constructs in all 3 target tissues, although this did not
reach statistical significanc (P p .08). Moreover, the extent of
GFP expression per cell for the various constructs paralleled
the percentage of positive GFP expression in both RN6390 and
SH1000 backgrounds in all target tissues. For example, for the
sae mutants, the mean GFP fluo escence was substantially lower
in all target tissues (42%, 41%, and 59% mean GFP fluo escence
vs. the parental strain [set at 100%] in vegetations, kidneys,
and spleen in the RN6390 strain set, respectively; and 48%,
53%, and 42% mean GFP fluo escence vs. the parental strain
[set at 100%] in vegetations, kidneys, and spleen in the SH1000
strain set, respectively). Importantly, 2 promoterless gfp strains
(in RN6390 and SH1000) had minimal GFP expression in any
target tissue (i.e., !2% GFP expression in vegetations in both
The staphylococcal a-toxin gene, hla, is positively controlled
by agr, sarA, and sae [3–5]. It appears that agr activates hla at
both the transcriptional and translational levels , whereas
sarA exerts a complex positive impact on hla expression by
both agr-dependent and agr-independent pathways [5, 8, 11].
In addition, sae appears to positively activate hla via an
agrdependent pathway in vitro [4, 14]. Moreover, the stress
response regulon, sigB, has been demonstrated to be involved
indirectly in hla expression by its composite effects on sarA,
agr, and sae [1, 14, 25, 29]. These latter interactions can be
complex, because sigB can enhance sarA expression but inhibit
agr expression during standard in vitro growth conditions [26,
29–31]. Of note, Cheung et al.  previously demonstrated
enhanced hla expression and a-toxin production in complete
sigB operon mutants in RN6390. By contrast, the influenc of
sigB on sae is still a matter of debate. For instance, Novick and
Jiang  demonstrated that sigB activation represses sae
expression, whereas other investigators could not confi m these
data in different S. aureus genetic background strains [30, 33].
In addition to the in vitro regulatory pathways, there has
been recent evidence that the expression of hla and/or
production of a-toxin are modifie after interactions with host
cells. For example, neutrophils may induce stress responses in
the organism that result in the differential activation of key
virulence regulons. Alternatively, neutrophils can release
elastases and serine proteases that may cleave a-toxin .
Moreover, Rothfork et al.  documented the capacity of
neutrophils to attenuate agr expression (and, presumably, hla
activation) by the generation of reactive oxygen and nitrogen
intermediates (i.e., HOCL and ONOO , respectively). Such
reactive species appear to oxidize the C-terminal methionine and,
hence, to inactivate the agr autoinducing peptide of S. aureus
cells within neutrophils, thus preventing the full activation of
the agr operon .
Despite the above in vitro and ex vivo observations, little
has been published about the in vivo expression and regulation
of hla, particularly in endovascular infections. Thus, the present
investigation was designed to defin the regulatory paradigms
controlling hla expression in an S. aureus IE model. Several
interesting observations emerged from the study. Consistent
with the results of prior in vitro studies, hla expression was
positively affected by agr, sarA, and especially sae [1, 4, 5, 36].
Interestingly, the expression of hla was increased in agr/sarA
double mutants during late periods of stationary growth, albeit
to a level lower than that in their respective parental strains.
By contrast, minimal hla expression was observed in sae
mutants over the course of the same growth periods. These in
vitro data suggest that additional regulatory factor(s) other than
agr and sarA are responsible for maintaining hla expression,
and sae appears to play a key role in this regard. This in vitro
paradigm was seen, in part, in vivo. Thus, hla expression in
cardiac vegetations was moderately reduced in the agr and sarA
single mutants; however, hla activation was markedly reduced
in the sae mutants. These finding suggest that in vivo hla
activation is more dependent on sae than on agr and/or sarA.
This apparent agr/sarA-independent but sae-dependent in vivo
regulation of hla expression has recently been confi med in a
different animal model: Goerke et al.  used transcript
analyses in a soft-tissue guinea pig infection model to show retained
hla expression in vivo in 2 agr-deficien clinical isolates and in
2 laboratory agr mutants of strains Newman and RN6390 and
to show markedly attenuated hla expression in a sae knockout
for all study strains discussed above. Collectively, the present
data in the IE model and those in the model of Goerke et al.,
in an entirely different model of pathogenesis, corroborate the
role of sae as an important regulatory system controlling hla
expression both in vitro and in vivo. Finally, it should be noted
that, despite nearly absent hla expression in vitro in each sae
mutant in our study, measurable but reduced hla expression
was observed in vivo. Such in vitro–in vivo disparities suggest
complex influence of host factors (e.g., blood proteins and
cells or vascular tissue) and/or in vivo microenvironmental cues
(e.g., pH, oxygen tension, and nutrient concentrations) on
staphylococcal gene expression. In addition, regulatory
elements other than agr, sarA, and sae are likely involved.
The overall impact of sae on the virulence of S. aureus
remains to be elucidated. Because hla expression has been
previously documented to be a virulence factor in several animal
models, including IE , it is reasonable to determine that
sae is also a key virulence factor in this latter model. This
assumption was supported by bacterial counts within target
tissues in the IE model being substantially lower in the sae
mutant than in the respective parental strains.
Another important findin is that the expression of hla is
substantially and negatively influence by sigB both in vitro
and in vivo. For instance, higher hla activation in RN6390 than
in SH1000 was observed both in vitro (figu e 1) and in all target
tissues in the IE model (∼20% higher in RN6390 vs. SH1000;
data not shown). These in vitro finding are consistent with those
of Horsburgh et al. , who found that restoration of jB activity
in SH1000 resulted in a marked decrease in the levels of a-toxin.
To our knowledge, the confi mation of this in vitro paradigm in
an experimental model is reported here for the firs time. These
results could be explained by (1) the lack of the repression of
sae transcription by sigB in RN6390  and/or (2) a large
reduction in expression of agrRNAIII in SH1000, compared with
8325-4 lineage strains (e.g., RN6390) .
In summary, the present results suggest that sae plays a
critical role in regulating hla expression in vitro and in the
experimental IE model. It appears that although agr and sarA are
both involved in hla expression, in their absence, sae can
directly support hla expression.
We thank Yin Li for her excellent technical assistance.
1. Bhakdi S , Tranum-Jensen J. Alpha-toxin of Staphylococcus aureus . Microbiol Rev 1991 ; 55 : 733 - 51 .
2. Walev I , Martin E , Jonas D , et al. Staphylococcal alpha-toxin kills human keratinocytes by permeabilizing the plasma membrane for monovalent ions . Infect Immun 1993 ; 61 : 4972 - 9 .
3. Arvidson S , Janzon L , Lofdahl S. The role of the d-lysin gene (hld) in the agr-dependent regulation of exoprotein synthesis in Staphylococcus aureus . In Novick RP, ed. Molecular biology of the staphylococci . New York : VCH, 1990 : 419 - 31 .
4. Giraudo AT , Cheung AL , Nagel R. The sae locus of Staphylococcus aureus controls exoprotein synthesis at the transcriptional level . Arch Microbiol 1997 ; 168 : 53 - 8 .
5. Goerke C , Flucklger U , Steinhuber A , Zimmerli W , Wolz C. Impact of the regulatory loci agr, sarA and sae of Staphylococcus aureus on the induction of a-toxin during device-related infection resolved by direct quantitative transcript analysis . Mol Microbiol 2001 ; 40 : 1439 - 47 .
6. Morfeldt E , Taylor D , von Gabain A , Arvidson S. Activation of a-toxin translation in Staphylococcus aureus by the trans-encoded antisense RNA , RNAIII. EMBO J 1995 ; 14 : 4569 - 77 .
7. Vandenesch F , Kornblum J , Novick RP. A temporal signal, independent of agr, is required for hla but not spa transcription in Staphylococcus aureus . J Bacteriol 1991 ; 173 : 6313 - 20 .
8. Cheung AL , Bayer AS , Zhang G , Gresham H , Xiong YQ. Regulation of virulence determinants in vitro and in vivo in Staphylococcus aureus . FEMS Immunol Med Microbiol 2004 ; 40 : 1 - 9 .
9. McNamara PJ , Milligan-Monroe KC , Khalili S , Proctor RA . Identifi cation, cloning, and initial characterization of rot, a locus encoding a regulator or virulence factor expression in Staphylococcus aureus . J Bacteriol 2000 ; 182 : 3197 - 203 .
10. Schmidt KA , Manna AC , Gill S , Cheung AL. SarT, a repressor of alphahemolysin in Staphylococcus aureus . Infect Immun 2001 ; 69 : 4749 - 58 .
11. Cheung AL , Projan SJ . Cloning and sequencing of sarA of Staphylococcus aureus, a gene required for the expression of agr . J Bacteriol 1994 ; 176 : 4168 - 72 .
12. Giraudo AT , Calzolari A , Cataldi AA , Bogni C , Nagel R. The sae locus of Staphylococcus aureus encodes a two-component regulatory system . FEMS Microbiol Lett 1999 ; 177 : 15 - 22 .
13. Novick RP . Pathogenicity factors and their regulation . In: Fischetti VA, Novick RP , Ferretti JJ , Portnoy DA , Rood JI, eds. Gram-positive pathogens. Washington, DC: ASM Press, 2000 : 392 - 407 .
14. Novick RP , Jiang D. The staphylococcal saeRS system coordinates environmental signals with agr quorum sensing . Microbiology 2003 ; 149 : 2709 - 17 .
15. Cheung AL , Nast CC , Bayer AS . Selective activation of sar promoters with the use of green fluo escent protein transcriptional fusions as the detection system in the rabbit endocarditis model . Infect Immun 1998 ; 66 : 5988 - 93 .
16. van Wamel W , Xiong YQ , Bayer AS , Yeaman MR , Nast CC , Cheung AL . Regulation of Staphylococcus aureus type 5 capsular polysaccharides by agr and sarA in vitro and in an experimental endocarditis model . Microb Pathog 2002 ; 33 : 73 - 9 .
17. Xiong Y-Q , Van Wamel W , Nast CC , Yeaman MR , Cheung AL , Bayer AS . Activation and transcriptional interaction between agr RNAII and RNAIII in Staphylococcus aureus in vitro and in an experimental endocarditis model . J Infect Dis 2002 ; 186 : 668 - 77 .
18. Xiong YQ , Bayer AS , Yeaman MR , van Wamel W , Manna AC , Cheung AS . Impacts of sarA and agr in Staphylococcus aureus strain Newman on fib onectin-binding protein A gene expression and fib onectin adherence capacity in vitro and in experimental infective endocarditis . Infect Immun 2004 ; 72 : 1832 - 6 .
19. Moreillon P , Que YA , Bayer AS . Pathogenesis of streptococcal and staphylococcal endocarditis . Infect Dis Clin North Am 2002 ; 16 : 297 - 318 .
20. Blevins JS , Beenken KE , Elasri MO , Hurlburt BK , Smeltzer MS . Straindependent differences in the regulatory roles of sarA and agr in Staphylococcus aureus . Infect Immun 2002 ; 70 : 470 - 80 .
21. Horsburgh MJ , Aish JL , White IL , Shaw L , Lithgow JK , Foster SJ. jB modulates virulence determinant expression and stress resistance: characterization of a functional rsbU strain derived from Staphylococcus aureus 8325-4 . J Bacteriol 2002 ; 184 : 5457 - 67 .
22. Kuroda M , Ohta T , Uchiyama I , et al. Whole genome sequencing of methicillin-resistant Staphylococcus aureus . Lancet 2001 ; 357 : 1218 - 9 .
23. Novick RP . The Staphylococcus as a molecular genetic system . In: Novick RP, ed. Molecular biology of the staphylococci . New York : VCH, 1990 ; 1 - 37 .
24. Giraudo AT , Mansilla C , Chan A , Raspanti C , Nagel R. Studies on the expression of regulatory locus sae in Staphylococcus aureus . Curr Microbiol 2003 ; 46 : 246 - 50 .
25. Kupferwasser LI , Yeaman MR , Nast CC , et al. Salicylic acid attenuates virulence in endovascular infections by targeting global regulatory pathways in Staphylococcus aureus . J Clin Invest 2003 ; 112 : 222 - 33 .
26. Ziebandt AK , Weber H , Rudolph J , et al. Extracellular proteins of Staphylococcus aureus and the role of SarA and jB . Proteomics 2001 ; 1 : 480 - 93 .
27. Cheung AL , Eberhardt K , Fischetti VA . A method to isolate RNA from gram-positive bacteria and mycobacteria . Anal Biochem 1994 ; 222 : 511 - 4 .
28. Bayer AS , Ramos MD , Menzies BE , Yeaman MR , Shen AJ , Cheung AL. Hyperproduction of alpha-toxin by Staphylococcus aureus results in paradoxically reduced virulence in experimental endocarditis: a host defense role for platelet microbicidal proteins . Infect Immun 1997 ; 65 : 4652 - 60 .
29. Bischoff M , Entenza JM , Giachino P. Influenc of a functional sigB operon on the global regulators sar and agr in Staphylococcus aureus . J Bacteriol 2001 ; 183 : 5171 - 79 .
30. Bischoff M , Dunman P , Kormanec J , et al. Microarray-based analysis of the Staphylococcus aureus jB regulon . J Bacteriol 2004 ; 186 : 4085 - 99 .
31. Gertz S , Engelmann S , Schmid R , et al. Characterization of the jB regulon in Staphylococcus aureus . J Bacteriol 2000 ; 182 : 6983 - 91 .
32. Cheung AL , Chien YT , Bayer AS . Hyperproduction of alpha-hemolysin in a sigB mutant is associated with elevated SarA expression in Staphylococcus aureus . Infect Immun 1999 ; 67 : 1331 - 7 .
33. Goerke C , Fluckiger U , Steinhuber A , et al. Role of Staphylococcus aureus global regulators sae and sigmaB in virulence gene expression during device-related infection . Infect Immun 2005 ; 73 : 3415 - 21 .
34. Worlitzsch D , Kaygin H , Steinhuber A , Dalhoff A , Botzenhart K , Doring G . Effects of amoxicillin, gentamicin, and moxifloxaci on the hemolytic activity of Staphylococcus aureus in vitro and in vivo. Antimicrob Agents Chemother 2001 ; 45 : 196 - 202 .
35. Rothfork JM , Timmins GS , Harris MN , et al. Inactivation of a bacterial virulence pheromone by phagocyte-derived oxidants: new role for the NADPH oxidase in host defense . Proc Natl Acad Sci USA 2004 ; 101 : 13867 - 72 .
36. Cheung AL , Koomey JM , Butler CA , Projan SJ , Fischetti VA . Regulation of exoprotein expression in Staphylococcus aureus by a locus (sar) distinct from agr . Proc Natl Acad Sci USA 1992 ; 89 : 6462 - 6 .