JOURNAL OF BACTERIOLOGY, Apr. 2008, p. 2726–2738
Copyright © 2008, American Society for Microbiology. All Rights Reserved.
Vol. 190, No. 8
ExoS Controls the Cell Contact-Mediated Switch to Effector Secretion
in Pseudomonas aeruginosa?†
Michelle Cisz, Pei-Chung Lee, and Arne Rietsch*
Department of Molecular Biology and Microbiology, Case Western Reserve University, 10900 Euclid Ave.,
Cleveland, Ohio 44106-4960
Received 26 September 2007/Accepted 3 November 2007
Type III secretion is used by many gram-negative bacterial pathogens to directly deliver protein toxins
(effectors) into targeted host cells. In all cases, secretion of effectors is triggered by host cell contact, although
the mechanism is unclear. In Pseudomonas aeruginosa, expression of all type III secretion-related genes is
up-regulated when secretion is triggered. We were able to visualize this process using a green fluorescent
protein reporter system and to use it to monitor the ability of bacteria to trigger effector secretion on cell
contact. Surprisingly, the action of one of the major type III secreted effectors, ExoS, prevented triggering of
type III secretion by bacteria that subsequently attached to cells, suggesting that triggering of secretion is
feedback regulated. Evidence is presented that translocation (secretion of effectors across the host cell plasma
membrane) of ExoS is indeed self-regulated and that this inhibition of translocation can be achieved by either
of its two enzymatic activities. The translocator proteins PopB, PopD, and PcrV are secreted via the type III
secretion system and are required for pore formation and translocation of effectors across the host cell plasma
membrane. Here we present data that secretion of translocators is in fact not controlled by calcium, implying
that triggering of effector secretion on cell contact represents a switch in secretion specificity, rather than a
triggering of secretion per se. The requirement for a host cell cofactor to control effector secretion may help
explain the recently observed phenomenon of target cell specificity in both the Yersinia and P. aeruginosa type
III secretion systems.
Pseudomonas aeruginosa is a frequent cause of hospital-
acquired infections, including ventilator-associated pneumonia
(9) and catheter infections in immunocompromised patients
(45), as well as burn wound infections (31, 49). It is also the
leading cause of the chronic lung infections afflicting patients
with cystic fibrosis and is the leading cause of death in this
patient group (37). Type III secretion is one of the most im-
portant virulence factors that this bacterial pathogen has at its
disposal. The ability to secrete effector proteins is associated
with higher rates of relapse and mortality in patients with
ventilator-associated pneumonia (24, 29). It also has an impor-
tant role in virulence in a wide variety of animal models of
infection, including mouse models of acute lung infection and
burn wound infection, as well as in hosts as distantly removed
from humans as moth larvae and the slime mold Dictyostelium
discoideum (43, 50, 52).
Four effector proteins (exoenzymes) in P. aeruginosa have
been described to date: ExoS, ExoT, ExoU, and ExoY. ExoS
and ExoT are highly homologous, bifunctional enzymes that
(GAP) domains with similar target specificities (both can stim-
ulate GTPase activity of RhoA, Rac1, and CDC42) and car-
boxy-terminal ADP-ribosylation domains (3). The substrate
specificities of the ADP-ribosylation domains differ markedly,
however, with ExoS targeting a wide variety of proteins, in-
cluding small Ras-like GTPases (23, 41, 53), ezrin-radixin-
moesin proteins (38), vimentin (10), and cyclophilin (14), and
ExoT affecting CrkI and CrkII (63). ExoU is a phospholipase
(59) and ExoY an adenlyate-cyclase (76). Strains of P. aerugi-
nosa differ in their complement of effectors. Interestingly,
ExoS and ExoU appear to be mutually exclusive, since almost
no strains which code for both effectors have been isolated
(18). There appears to be some correlation of effectors with
specific diseases. In particular, ExoS-producing strains appear
to be more common in cystic fibrosis isolates (19), whereas
ExoU-producing strains appear to be more common in isolates
from keratitis patients (20).
Transcription of the type III secretion genes is induced by
triggering of secretion (21). Secretion can be triggered in vitro
by removing calcium from the medium or by contact with host
cells (21, 33, 67, 69). Induction of gene expression is linked to
secretion by virtue of a regulatory system comprised of four
proteins. ExsA, the master regulator of the type III secretion
genes, is an AraC-type transcription regulator that is required
for the expression of the type III secretion genes (22). Its
activity is controlled by the antiactivator ExsD (40). ExsD
activity, in turn, is regulated by the type III secretion chaper-
one ExsC (12). The final regulator, ExsE, is a small secreted
protein which interacts with ExsC (55, 68). When effector se-
cretion is off, ExsE is thought to bind to and sequester ExsC.
This in turn allows ExsD to bind to and inactivate ExsA. Upon
triggering of secretion, ExsE is exported via the type III secre-
tion machinery, freeing ExsC to bind to ExsD and activating
ExsA (78). In this study we have used a green fluorescent
protein (GFP) reporter to monitor triggering of exoS expres-
sion on cell contact.
* Corresponding author. Mailing address: Department of Molecular
Biology and Microbiology, Case Western Reserve University School of
Medicine, 10900 Euclid Ave., Cleveland, OH 44106-4960. Phone:
(216) 368-2249. Fax: (216) 368-3055. E-mail: email@example.com.
† Supplemental material for this article may be found at http://jb
?Published ahead of print on 26 November 2007.
While the exact steps involved in the triggering of effector
secretion on cell contact have not been well defined, we do
know of several events that have to occur. First, the bacterium
has to make contact with the cell, a process mediated by spe-
cific adhesins. In P. aeruginosa the type IV pili are important
for this initial attachment step; however, they are not specifi-
cally required for type III secretion, since they can be replaced
by the unrelated Yersinia ph6 adhesin (65). Subsequently, the
type III secretion system (T3SS) has to be brought close to the
plasma membrane. The translocator proteins PopB and PopD
have to be inserted into the host membrane to form the trans-
location pore to which the type III secretion apparatus is
docked (7, 11). Insertion of PopD, but not PopB, requires the
needle tip protein PcrV (27, 28). In the closely related Yersinia
T3SS, the PcrV homolog LcrV appears to be absolutely re-
quired for the insertion of the PopB homolog YopB but only
partially required for YopD insertion (8, 27). Evidence has
been presented that Yersinia secretes the translocator proteins
(but not the effectors) into the culture medium during the
infection process, an event that may be triggered by serum
proteins in tissue culture medium (35, 36). However, at least in
the case of YopB, it has been explicitly demonstrated that
these secreted translocators cannot cross-complement a yopB
null mutant, suggesting that the translocators that form the
pore, to which a given apparatus is attached, have to be se-
creted and retained in close proximity to the needle (58). This
is consistent with a recent observation that only minimal levels
of secreted YopB and YopD are required for successful intox-
ication of cells (15). After the formation of the translocation
pore and docking of the needle to the pore, effector secretion
is triggered. In the closely related Yersinia T3SS, triggering of
effector secretion has been proposed to be a function of sens-
ing the lowered calcium concentration in the eukaryotic cell
cytosol (35). Sensing of the host cell is thought to involve a
conformational shift in the needle protein that is perpetuated
from the site of contact down the needle to the base of the
secretion apparatus (71). A somewhat different model has
been proposed for triggering of type III secretion in Shigella.
Here it has been proposed that the PopB homolog IpaB is, in
fact, part of the needle tip complex (71). Insertion of IpaB into
the host cell membrane is thought to result in a conformational
shift in the needle tip which results in secretion of the remain-
ing translocator, as well as the effector proteins. In support of
this model, it has been reported that IpaB is surface displayed
in Shigella (47, 71). To date, however, no one has been able to
detect YopB or YopD at the tips of Yersinia T3SS needles (8,
44) or PopB or PopD at the surface of P. aeruginosa (unpub-
lished observation), suggesting that the insertion of the trans-
locon into the target membrane may proceed via a slightly
different mechanism in these two classes of T3SSs.
Recent reports suggest that translocation of effectors by the
Yersinia T3SS is controlled by a eukaryotic factor and modu-
lated by effectors, YopE and YopK. Deletion of either gene
results in an increase in effector translocation (1, 2, 32). YopK
has been proposed to modulate the size of the translocation
pore formed by YopB and YopD (32). YopE is a secreted
toxin that exhibits Rho-GAP activity and displays some homol-
ogy with the amino-terminal portions of ExoS and ExoT. YopE
displays a somewhat different target specificity, however (77).
Interestingly, the Rho-GAP domain of ExoS was able to sub-
stitute for YopE and reestablish Yop translocation control.
The ADP-ribosylation domain of ExoS, however, had no effect
in this assay (1).
In this study we have used the fact that expression of type III
secretion-related genes is controlled by secretion of ExsE to
study triggering of effector secretion on cell contact. Interest-
ingly, we found that intoxication of cells by ExoS blocked
subsequent induction of our reporter constructs, suggesting
that triggering of effector secretion is controlled by a host
factor that is inactivated by ExoS. Similar to the inhibition of
Yop translocation by YopE in Yersinia, intoxication of cells by
ExoS limits its own translocation into host cells. In fact, ExoS
can also limit effector secretion when expressed in trans in a
Yersinia yopE null mutant (1). Unlike in the Yersinia system,
however, where only the Rho-GAP activity of ExoS was able to
limit effector translocation when expressed in a yopE mutant,
either enzymatic activity of ExoS could serve to limit translo-
cation in P. aeruginosa. Triggering of exoS expression on cell
contact depends on the presence of both PopB and PopD.
These results suggest a hierarchy of secretion, in which trans-
locators are secreted before effector proteins. In agreement
with this hypothesis, we present evidence that secretion of
translocator proteins is apparently not controlled by calcium.
Interestingly, intracellular calcium levels in the host cell do not
appear to control effector secretion either, suggesting that a
signal other than calcium has to be responsible for the switch
to effector secretion on cell contact. Taken together, these data
allow us to propose a step-by-step model for triggering of
effector secretion on cell contact. Triggering involves translo-
con insertion, docking, and subsequent switch of secretion
specificity to allow secretion of effectors, a process that is
controlled by a yet-to-be-identified eukaryotic factor.
MATERIALS AND METHODS
Bacterial strains, cells, and growth conditions. The bacterial strains and plas-
mids used in this study are listed in Table 1. P. aeruginosa was grown in LB with
200 mM NaCl, 10 mM MgCl2, and 0.5 mM CaCl2unless noted otherwise. Where
indicated, calcium was removed from the medium by adding EGTA to a final
concentration of 5 mM. Tissue culture cells were grown in RPMI supplemented
with 2 mM glutamine and 10% fetal bovine serum (FBS) (RP10) in the presence
of 5% CO2at 37°C. In some instances medium without FBS was used. A549 cells
(lung epithelial cell line, ATCC CCL-185) were obtained from the laboratory of
Susann Brady-Kalnay. Inhibitors were added 30 min before infection after re-
placing the medium with prewarmed RP10. Cycloheximide was added at a final
concentration of 200 ?g/ml, cytochalasin D at 10 ?M, streptolysin O at 5 U/?l,
alpha-hemolysin at 10 U/?l, and calcimycin at 10 ?M.
Plasmid and strain construction. To construct pCTX2-groE-mCherry, a frag-
ment containing the groE promoter and first two codons of groES was amplified
(PgroE5, 5?-AAAAAactagtACGACCTGAACGCCCGCTACGGA-3?; PgroE3,
5?-AAAAAgaattcCTTCATAGTCGTAACTCTCCCAAA-3? [restriction sites
are lowercase]), as well as mCherry (CFP-5R, 5?-AAAAAgaattcATGGTGAGC
AAGGGCGAGGA-3?; Cherry-3H, 5?AAAAAaagcttTTACTTGTACAGCTCG
TCC-3?) The PCR fragments were digested with SpeI/EcoRI and EcoRI/
HindIII, respectively, and cloned into SpeI/HindIII-digested pMiniCTX2 (30).
The groE-mCherry reporter construct was then crossed onto the chromosome of
PAO1F, and extraneous plasmid sequences were removed by FLP recombination
using plasmid pFLP2 as described previously (30). Deletion of the three effector
genes, as well as exsE and pscC, was performed using published constructs (55,
70, 74). Mutant constructs to introduce active-site mutations into the exoS gene
on the PAO1F chromosome were generated by amplifying two flanking se-
quences using two outside primers (exoS5-2, 5?-AAAAAggtaccGCTTGCAAGGG
TCCTGGCTGAACA-3?; exoS3-1, 5?-AAAAAaagcttCCGTACCCTGCCGCTAC
TGAACT-3?) as well as the corresponding internal primers harboring the relevant
mutations (lowercase) (SADR3-1, 5?-ATATCGAACTACAAGAATGAtAAAGAt
ATTCTCTATAACAAAGAAACCGA-3?; SADR5-2, 5?-TCGGTTTCTTTGTTA
VOL. 190, 2008HOST FACTOR CONTROLS CELL CONTACT-MEDIATED EFFECTOR SECRETION2727
TAGAGAATaTCTTTaTCATTCTTGTAGTTCGATAT-3?; SGAP3-1, 5?-TGGC
3?; SGAP5-2, 5?-TGCCGGCCAAGGCGGTGCTCAGCGAtttCAGCGCCCCAT
CTCCGCTGGCCA-3?). Flanks were joined by splicing by overlap extension PCR
(73) and cloned into the allelic exchange vector pEXG2 (55). After sucrose
selection, the presence of the appropriate mutation was confirmed using test
SGAPtest [5?-TGCCGGCCAAGGCGGTGCTCAGCGAttt-3?]) in conjunction
with the appropriate flanking primer. pcrV was cloned into plasmid pPSV35 as an
EcoRI/HindIII fragment (pcrV2-5R, 5?-AAAAgaattcTGGCTTGTTGATCTGA
GGAATCACGA-3?; pcrV2-3H, 5?-AAAAAaagctTCGGCTGGTTCATGGAT
ACCTCTA-3?). popB was cloned into plasmid pPSV35 as an EcoRI/HindIII
fragment (popB5R, 5? AAAAAgaattcTTAGGAGGCGCCCCCATGAATCCG
ATAACGCTTGAA-3?; popBEX3, 5?-AAAAAaagcttGACGTCTCCTCAGAT
CGCTGCCGGT-3?). popD was cloned into plasmid pPSV35 as an EcoRI/
HindIII fragment (popD5R, 5?-AAAAAgaattcTTAGGAGGCGCCCCCATGA
TCGACACGCAATATTCCCT-3?; popDEX3, 5?-AAAAAaagcttCGCGCGGA
GACGGCTCAGACCACT-3?). The popD deletion construct, pEXG2-?popD,
was created by amplifying the two flanking regions (popD5-1, 5?-AAAAAgaatt
cGTGGTCAGCTTCGGCGGCTCAGCGGT-3?; popD5-2, 5?-AACTCGAGCC
GAGGAA-3?; popD3-2, 5?-AAAAAaagcttAGGGTCAGTTGCGCTGCGAGA
AT-3?), which were subsequently stitched together by splicing by overlap exten-
sion PCR, digested with EcoRI and HindIII, and cloned into pEXG2.
Microscopy. A549 cells (105) were seeded 2 days ahead of time on glass
coverslips placed in six-well plates and allowed to attach tightly to the substrate
before the experiment was performed. On the day of the experiment, the cells
were washed twice with phosphate-buffered saline (PBS), and fresh medium was
added to the wells. If indicated, cells were pretreated with toxins and inhibitors
30 min prior to infection, which was subsequently carried out in the continued
presence of the given toxin or inhibitor. Overnight cultures of the appropriate
strains of P. aeruginosa were used to infect cells for the indicated amounts of
time. Subsequently, cells were gently washed twice with PBS (to prevent loss of
rounded cells) and fixed with 4% paraformaldehyde in a 200 mM phosphate
buffer (pH 7.4) (Poly Scientific) for 15 min at room temperature. Fixed slides
were washed once with PBS, dabbed dry, and mounted on slides using Prolong
mounting medium (Invitrogen). Slides were imaged using a Zeiss Axioplan2
microscope. While cell rounding was evident in cells intoxicated with ExoS
and/or ExoT, cell lifting was minimized by allowing the cells to attach for 2 days
and washing the cells gently after the infection (PBS was added to the wall of the
well and allowed to wash over the cells, rather than added onto the cells directly).
The similarity of cell numbers/field in slides with rounded and flat cells suggests
that loss of rounded cells was minimal in the reported experiments.
For video microscopy 8 ? 104cells were seeded 2 days ahead of time in
glass-bottom culture dishes. On the day of the experiment, the cells were washed
twice with PBS, the medium was exchanged for fresh medium, and the cells were
infected at a multiplicity of infection (MOI) of 25 (assuming 2 ? 105cells). Image
collection was started 1.5 h postinfection. Images were collected at a rate of one
image every 2 min using a 60? objective on a Deltavision RT epifluorescence
microscope with an automated stage (Applied Precision, Inc.) and captured
using a charge-coupled device camera. Phase and fluorescent images were over-
laid and assembled into movies using Softworx analysis software. Still images
were cropped using Adobe Photoshop CS3 (Adobe, Inc.).
Translocation assay. The translocation assay was based on an established
protocol (46). Two days prior to the experiment, 1.5 ? 106A549 cells were
TABLE 1. Strains and plasmids
Strain or plasmidDescriptionReference
PAO1F (wild-type PAO1)
PAO1F ?exoS ?exoT ?exoY (?3TOX)
PAO1F ?exoS ?exoT ?exoY attB::groE-mCherry
PAO1F ?pscC attB::groE-mCherry
PAO1F ?exoS::GL3 ?exoT ?exoY attB::groE-mCherry
PAO1F ?exoS::GL3 ?exoT ?exoY ?popB attB::groE-mCherry
PAO1F ?exoS::GL3 ?exoT ?exoY ?popD attB::groE-mCherry
PAO1F ?exsE ?exoS ?exoT ?exoY
PAO1F ?exsE ?exoT ?exoY (“exoS?”)
PAO1F ?exsE ?exoS ?exoY (“exoT?”)
PAO1F ?exsE ?exoS ?exoT (“exoY?”)
PAO1F ?exsE ?pscC
PAO1F ?exsE ?exoT ?exoY exoS-R146K (“exoSGap?”)
PAO1F ?exsE ?exoT ?exoY exoS-E379D/E381D (“exoSADPR?”)
PAO1F ?exsE ?exoT ?exoY exoS-R146K/E379D/E381D (“exoSGAP?/ADPR?”)
PAO1F ?exsE ?pcrV2
pPSV35 Expression plasmid with lacIq, lacUV5 promoter, gentamycin resistance, colE1 and
Pseudomonas origins of replication, oriT
pcrV cloned under control of lacUV5 promoter
popB cloned under control of lacUV5 promoter
popD cloned under control of lacUV5 promoter
pexoSpromoter controlling GFP expression, gentamycin resistance
pgroEcontrolling mCherry expression
Allelic exchange vector, colE1 origin, oriT, gentamycin resistance, sacB
pEXG2 with exoSR146K allele
pEXG2 with exoSE379D/E381D allele
pEXG2 with exoSR146K/E379D/E381D allele
pEXG2 with pcrV deletion allele (?4–180)
pEXG2 with popD deletion allele (?3–268)
pEXGW with popB deletion allele
2728CISZ ET AL.J. BACTERIOL.
seeded in 100-mm-diameter tissue culture dishes. On the day of the experiment,
the cells were washed twice with PBS, and RPMI without FBS was added to each
dish. For each strain three plates of cells were infected. Cells were infected at an
MOI of 50 for 2 h and then washed twice with PBS and treated as follows. The
first plate was not treated with protease (PBS control), and the second and third
plates were treated with 1 ml PBS (Invitrogen; supplemented with Mg and Ca)
plus 250 ?g/ml proteinase K for 20 min at room temperature. Cells were col-
lected using a cell scraper, pelleted in a microcentrifuge, and resuspended in 75
?l PBS with 2 mM phenylmethylsulfonyl fluoride. To the cells from the first and
second plates, 75 ?l PBS with 1% Triton X-100 was added. To the cells derived
from the third plate, 75 ?l of PBS with 1% sodium dodecyl sulfate (SDS) was
added. The lysates were incubated at room temperature for 20 min, at which
point the protein concentration was determined by the Bradford method (Bio-
Rad). Equivalent amounts of protein were loaded on a 12.5% gel, separated by
SDS-polyacrylamide gel electrophoresis (SDS-PAGE), and transferred to nitro-
cellulose using a semidry transfer apparatus. Indicated proteins were detected by
Western blotting. The anti-actin antibody (hybridoma supernatant, catalog no.
JLA20) developed by J. J.-C. Lin was obtained from the Developmental Studies
Hybridoma Bank developed under the auspices of the NICHD and maintained
by The University of Iowa, Department of Biological Sciences, Iowa City.
Mouse-anti RpoB was purchased from NeoClone. Antisera against ExoS were a
gift from Stephen Lory (Harvard Medical School). Antisera directed against
PopB and PopD were a gift from Ina Attree (CEA Grenoble). All other antisera
were raised against purified His-tagged proteins (Covance).
Secretion assay. Bacteria were diluted 1:300 into 3-ml cultures of “high-salt”
LB (NaCl adjusted to 200 mM and medium supplemented with 0.5 mM CaCl2
and 10 mM MgCl2[final concentrations]) and grown for 2 h, at which point
secretion was induced by the addition of EGTA (5 mM final concentration). The
cultures were grown for another 30 min, at which point the bacteria were pelleted
by centrifugation. Supernatant proteins were precipitated with 10% (final con-
centration) trichloroacetic acid (TCA). Pellets were washed once with acetone,
resuspended in sample buffer, and normalized according to the optical density at
600 nm (OD600) of the culture. Samples were separated by SDS-PAGE, trans-
ferred to nitrocellulose membranes, and probed with the appropriate antibodies.
In the case of the assay to reestablish calcium control, overnight cultures were
diluted 1:300 into 5 ml “high-salt” LB with and without 5 mM EGTA. Bacteria
were again cultured for 2 hours, at which point bacteria from 2-ml aliquots of
each original culture were pelleted and resuspended in medium with or without
5 mM EGTA. The cultures were incubated for another 30 min, at which point the
bacteria were pelleted and resuspended in sample buffer (cell pellet fraction).
Supernatant protein was precipitated with TCA as noted above. Both cell pellet
and supernatant fractions were normalized according to the OD600of the cul-
While secretion of effectors via the T3SS can be triggered in
vitro by removing calcium from the medium, the more relevant
cue for triggering type III secretion in the context of an infec-
tion is most likely host cell contact. We set out to study trig-
gering of type III secretion on host cell contact by utilizing a
convenient property of Pseudomonas aeruginosa, the fact that
triggering of effector secretion via the T3SS also results in
up-regulation of all type III secretion-associated genes (both
toxin and structural genes) (74). To this end we used one of
two previously described reporter constructs: a plasmid-based
reporter, in which GFP expression is controlled by the exoS
promoter, and a chromosomal reporter, in which GFP and a
translationally coupled copy of lacZ were integrated in the
exoS locus, replacing the exoS open reading frame (54, 56).
In order to visualize all cell-associated bacteria, we further
labeled our bacteria by creating a constitutively expressed re-
porter encoding an amino-terminal translational fusion of the
first two amino acids of groES to the red fluorescent mCherry
protein. The reporter was expressed from the groES promoter
and inserted in the neutral CTX phage attachment site (see
Fig. S1 in the supplemental material). All reporters were neu-
tral with regard to T3SS function as assayed by the ability of
bacteria carrying the reporters to intoxicate epithelial cells (see
Fig. S1 in the supplemental material). Our readout for cell
contact-mediated triggering of type III secretion therefore is
the percentage of cell-associated bacteria (red fluorescent)
that are expressing GFP.
Initial experiments using a wild-type strain of P. aeruginosa
failed to detect significant numbers of cell-associated bacteria
expressing GFP. This was surprising to us, since induction of
type III secretion gene expression on cell contact had been
described previously (33, 69). However, triggering of exoS ex-
pression on cell contact was readily detectable when the ex-
periment was performed with a P. aeruginosa strain lacking the
genes for all three of the known toxin open reading frames,
exoS, exoT, and exoY (?3TOX) (Fig. 1A). We next performed
time-lapse video microscopy to study triggering of exoS expres-
sion in real time. Interestingly, in these experiments both the
wild-type and the ?3TOX bacteria readily turned on GFP
expression on cell contact (Fig. 1B and C).
This apparent discrepancy is most easily explained by pos-
tulating that intoxication of cells by one or all of the known
effector proteins prevents subsequent triggering of type III
secretion. If this is the case, the first wild-type bacterium to
contact a cell would still trigger type III secretion, inject its
toxins, and up-regulate T3SS gene expression. Intoxication,
however, would prevent subsequent triggering of exoS expres-
sion on cell contact by other bacteria, thereby accounting for
the very low number of cell-associated GFP-positive bacteria
in the fixed-cell experiments. Since the ?3TOX strain does not
intoxicate the targeted host cell, the cell remains permissive for
triggering type III secretion by bacteria that attach after the
initial contact, and the percentage of GFP-positive bacteria
would be expected to increase as we observed.
Intoxication of cells by ExoS prevents subsequent triggering
of type III secretion on cell contact. To examine the hypothesis
that intoxication of cells prevents subsequent triggering of ef-
fector secretion, we engineered three strains expressing only
ExoS, ExoT, or ExoY. Expression was deregulated by deleting
exsE, and all strains were tested to ensure that they expressed
and secreted the appropriate toxin (Fig. 2C). Epithelial cells
were preintoxicated with the indicated strains for 1 h, at which
point the cells were washed twice with PBS and infected with
the ?3TOX reporter strain bearing the pexoS-GFP reporter
plasmid. After 4 h of infection, the cells were washed and the
ability of the reporter bacteria (readily distinguishable by vir-
tue of the constitutively expressed mCherry reporter) to induce
exoS expression was assayed. While preinfection of the cells
with strains bearing mutation in a T3SS structural gene
(?pscC) or a strain lacking the three known effectors (?3TOX)
had no apparent effect on the ability of the reporter to trigger
exoS expression, preintoxication with a strain expressing all the
toxins or just ExoS was able to completely inhibit subsequent
triggering of exoS expression (Fig. 2A). These results could not
be explained by an inability of the reporter bacteria to bind the
epithelial cells (Fig. 2B). ExoT had an intermediate effect in
these experiments that was not statistically significant, while
ExoY did not prevent triggering of exoS expression by the
reporter strain (Fig. 2A).
We next dissected which of the two enzymatic functions of
ExoS is responsible for the negative effect on subsequent T3SS
triggering. To this end, we engineered strains that constitu-
VOL. 190, 2008HOST FACTOR CONTROLS CELL CONTACT-MEDIATED EFFECTOR SECRETION 2729
tively expressed either wild-type ExoS or mutant forms of the
protein in which Rho-GAP (R146K) (25), ADP-ribosylation
(E379D/E381D) (51), or both activities had been inactivated.
All proteins were readily secreted and expressed at similar
levels (Fig. 3C). Interestingly, the Rho-GAP mutant and ADP-
ribosylation mutant proteins were still able to prevent subse-
quent triggering of exoS expression in the reporter (Fig. 3A).
Mutation of both active sites, however, resulted in a toxin that
was unable to prevent induction of GFP expression in the
reporter strain. This indicates that the repressive activity of
ExoS is tied to its enzymatic activities and that at least one of
the signaling pathways affected by ExoS is required for cell
contact-mediated triggering of effector secretion.
Triggering of exoS expression is tied to host cell integrity but
not to the low-calcium environment of the host cell cytosol. We
further analyzed the cellular requirements for triggering of
type III secretion by preintoxicating cells and monitoring the
effect of intoxication on the ability of a reporter strain to
induce exoS expression on cell contact.
Several of the enzymatic activities of P. aeruginosa exoen-
zymes affect the actin cytoskeleton, so we determined whether
cytochalasin D, which prevents elongation of F-actin, leading
to a net disassembly of the actin cytoskeleton, prevents trig-
gering of the T3SS. Cytochalasin D did not prevent induction
of exoS expression on cell contact. If anything, induction of
exoS expression was enhanced (Table 2). Cycloheximide pre-
treatment also did not prevent triggering of exoS expression,
suggesting that the defect in triggering after ExoS intoxication
is not due to the loss of a labile receptor protein.
On the other hand, intoxication of cells with either of two
different pore-forming toxins, streptolysin O (cholesterol de-
pendent, up to 30-nm-diameter pores) or alpha-toxin (not cho-
lesterol dependent, 0.6- to 1-nm pores) (4), prevented trigger-
ing of the T3SS (Table 2), indicating that cells have to have an
intact plasma membrane to allow triggering of the T3SS.
It has been proposed in the case of Yersinia that triggering of
effector secretion on cell contact depends on sensing of the low
calcium concentration in the host cell cytosol (35). We decided
to test this hypothesis directly by using the ionophore calcimy-
cin to shuttle calcium into the tissue culture cells and assay its
effect on triggering of exoS expression (Fig. 4). Calcium was
clearly shuttled into the cells as visualized using the calcium-
FIG. 1. Cell contact-dependent induction of exoS expression. (A) The attB::groE-mCherry chromosomal marker, as well as pP33-pexoS-GFP,
were introduced into wild-type PAO1F, a ?pscC derivative, or a strain lacking all three exoenzymes (?3TOX) and used to monitor induction of
exoS expression over time. The cells were infected at an MOI of 25 for 1 h and washed twice with PBS, and the medium was replaced with fresh
tissue culture medium. Subsequently, slides were removed at 1-h intervals, washed, and fixed with paraformaldehyde before being mounted for
microscopic examination. Generally four fields, comprising 60 to 80 A549 cells and ?100 to 300 cell-associated bacteria in total, were evaluated
per strain and time point. The standard deviation represents the variation between fields. (B and C) Still images captured by live video microscopy
of PAO1F/pP33-pexoS-GFP (wild type) (B) and PAO1F ?3TOX/pP33-pexoS-GFP (C) infecting A549 cells. Bar, 10 ?m. Arrowheads indicate
bacteria that have made contact and induce GFP expression. The time postinfection in minutes is indicated above each panel.
2730CISZ ET AL.J. BACTERIOL.
sensitive dye rhodamine 5N (Kd, 320 ?M; Invitrogen), but
triggering of exoS expression on cell contact was unaffected.
The translocator proteins PopB and PopD are required for
cell contact-mediated up-regulation of T3SS expression. Two
models have been formulated with regard to triggering of type
III secretion on cell contact. On the one hand, interaction of
the needle tip or a needle tip-associated translocator protein
with the plasma membrane or a putative receptor could result
in triggering of the T3SS (as proposed for Shigella ). This in
turn results in secretion of the remaining translocators, which
are inserted into the plasma membrane, forming a pore to
which the needle is subsequently docked. On the other hand,
the translocators could be secreted prior to cell contact and
FIG. 2. Effect of preintoxication on subsequent triggering of exoS
expression. Cells were either mock infected or preinfected at an MOI
of 50 for 1 h with PAO1F ?pscC, PAO1F ?exsE, PAO1F ?exsE in
which all three exoenzymes had been deleted (?3TOX), or PAO1F
?exsE in which two of the exoenzymes had been deleted so that only
one exo gene remained (i.e., “?exsE exoS?” is ?exsE ?exoT ?exoY).
After the preinfection period, the cells were washed twice with PBS
and the medium was replaced with fresh medium carrying the reporter
strain (PAO1F ?3TOX attB::groE-mCherry/pP33-pexoS-GFP) at an
MOI of 25. The infection was continued for another 4 h, at which point
the cells were washed twice with PBS and fixed with paraformaldehyde
before being mounted for microscopic examination. Four to seven
fields totaling ?40 to 70 A549 cells and ?120 to 250 cell-associated
bacteria were analyzed for each condition. The error bars indicate
standard deviations between fields. (A) Percentage of GFP-positive,
cell-associated reporter bacteria (mCherry positive). The preintoxica-
tion strains are noted on the x axis. Statistical significance was deter-
mined using Student’s t test (*, P ? 0.05). (B) Attachment of the
reporter bacteria to cells (mCherry-positive bacteria/cell). The prein-
toxication strains are noted on the x axis. (C) Secretion of effectors.
The indicated strains were grown in LB. Secretion of effectors was
triggered by removing calcium from the medium. Culture supernatants
were collected after removing the bacteria by centrifugation, TCA
precipitated, and resuspended in sample buffer. Samples were normal-
ized based on the OD600of the culture, separated by SDS-PAGE,
blotted, and probed with a mix of antibodies directed against ExoS,
ExoT, ExoY, and PopN.
FIG. 3. Influence of ExoS active-site mutations on the block in
subsequent triggering of exoS expression by the reporter bacteria. The
experiment was conducted as described for Fig. 2. Preintoxication
strains were derived from PAO1F ?exsE and expressed only the indi-
cated exoenzymes. Ten to 12 fields totaling ?150 to 200 A549 cells and
?300 to 700 cell-associated bacteria were analyzed for each condition.
The error bars indicate standard deviations between fields. (A) Per-
centage of GFP-positive, cell-associated reporter bacteria (mCherry
positive). The preintoxication strains are noted on the x axis. Statistical
significance was determined using Student’s t test (**, P ? 0.01).
(B) Attachment of the reporter bacteria to cells (mCherry-positive
bacteria/cell). The preintoxication strains are noted on the x axis.
(C) Secretion of effectors. Secretion was monitored as described for
Fig. 2. The blot was probed with an anti-ExoS antiserum. The band
above the ExoS band in the PAO1F ?exsE supernatant is ExoT, which
weakly cross-reacts with the anti-ExoS antiserum.
VOL. 190, 2008 HOST FACTOR CONTROLS CELL CONTACT-MEDIATED EFFECTOR SECRETION 2731
inserted into the plasma membrane, and docking of the needle
to the translocation pore results in a specificity switch and
secretion of effector proteins (35). If the former is true, then
induction of exoS expression on cell contact should be inde-
pendent of at least one of the translocator proteins; if the latter
is the case, then induction of the reporter should depend on
both PopB and PopD. It was recently reported that triggering
of exoS expression upon cell contact depends on PopD (67).
This does not, however, fully distinguish the two proposed
Mutants lacking either popB or popD were not able to trig-
ger exoS expression on cell contact (Fig. 5A). The defect in
exoS induction could be complemented by supplying the de-
leted gene on a plasmid. Loss of either translocator protein did
not significantly affect attachment (Fig. 5B). Triggering of exoS
expression on cell contact therefore depends on a functional
TABLE 2. Effect of toxins and inhibitors on cell contact-dependent exoS induction
Mean ? SD
% GFP-positive cell-
No. of bacteria/cell
% GFP-positive bacteria
No cells ? EGTA
21.7 ? 4.4
0.1 ? 0.2
0.4 ? 0.4
3.7 ? 2.1
30.9 ? 5.4
17 ? 3.4
0.1 ? 0.1
0.1 ? 0.2
0.1 ? 0.3
55.5 ? 11.2
62 ? 11.9
34.5 ? 2
6.2 ? 4.5
33.5 ? 7.9
15.2 ? 8.1
13.5 ? 3.8
8 ? 1.1
6.4 ? 1.9
aThe experiments were performed in RPMI with 10% FBS (RP10) or without FBS (RPMI).
b“No cells” indicates bacteria grown under tissue culture conditions in the indicated medium and with the corresponding toxin/inhibitor but in the absence of A549
cND, not determined.
FIG. 4. Effect of the calcium ionophore calcimycin on triggering of exoS expression on cell contact. A549 cells were pretreated for 30 min with
dimethyl sulfoxide (DMSO) or 10 ?M calcimycin and subsequently infected with the exoS reporter strain PAO1F ?3TOX attB::groE-mCherry
?exoS::GL3 at an MOI of 10 for 4 h. Cells were then washed twice with PBS lacking calcium and incubated for 10 min with 5 ?M rhodamine 5N
in PBS (no calcium). Cells were washed again twice with PBS and fixed with paraformaldehyde before being mounted for microscopic examination.
Four (DMSO) or five (calcimycin) fields totaling 78 (DMSO) or 90 (calcimycin) A549 cells and 298 (DMSO) or 465 (calcimycin) cell-associated
bacteria were analyzed. (A) representative differential interference contrast (DIC), rhodamine channel, and GFP channel fluorescent images.
(B) Percent GFP-positive cell-associated bacteria. (C) Average cell-associated (mCherry-positive) bacteria/cell. Error bars indicate standard
deviations between fields.
2732CISZ ET AL. J. BACTERIOL.
translocase. This observation differentiates induction on cell
contact from induction under low-calcium conditions, which is
independent of PopB and PopD (Fig. 5C) (67). As had been
demonstrated for YopB in the Yersinia system (58), neither a
popB mutant nor a popD mutant could be trans-complemented
in a coinfection experiment (see Fig. S2 in the supplemental
material), suggesting that both translocators have to be se-
creted in cis.
ExoS feedback regulates effector secretion. To confirm that
ExoS controls effector secretion, we monitored translocation
of wild-type ExoS into A549 cells. All experiments were per-
formed with a deregulated strain lacking exsE to rule out the
effect of changes in expression on the result of the transloca-
tion assay. Cells were infected at a relatively high MOI (50) for
2 h. The cells were then washed and in some cases treated with
proteinase K to digest protein associated with the outside of
the cells. Translocated proteins were liberated by lysing the
host cells with the detergent Triton X-100, which does not lyse
P. aeruginosa. As a control, cells and cell-associated bacteria
were lysed using SDS.
While a strain expressing the wild-type ExoS (exoS?) trans-
located only very little ExoS into the A549 cells (Fig. 6, lanes
1 and 2), a mutant expressing equal amounts of the GAP/ADP-
ribosylation double mutant ExoS (G/A) translocated signifi-
cantly more protein into the targeted epithelial cells (Fig. 6,
lanes 4 and 5). Consistent with the preintoxication experiment,
therefore, these data indicate that translocation of wild-type
ExoS is self-inhibited in a manner that depends on its enzy-
matic activity. As expected (64), translocation depended on a
functional needle tip protein (PcrV) (Fig. 6, lanes 10 and 11).
ExoT exhibited an intermediate phenotype in this assay (Fig. 6,
lanes 7 and 8), consistent with the intermediate phenotype
observed in the preintoxication experiment (Fig. 2).
Calcium controls effector secretion but not translocator se-
cretion in vitro. The requirement of PopB and PopD for trig-
gering of exoS expression on cell contact suggests that trans-
locators are secreted before effectors. In vitro, however,
secretion of the effectors and translocators appears to be si-
multaneous upon depletion of calcium from the medium (75).
To test whether calcium in fact controls secretion of trans-
locator proteins we performed experiments in which we rees-
tablished calcium control of secretion after depleting calcium
from the medium using EGTA. Wild-type PAO1 was grown in
LB either in the presence of 0.5 mM Ca2?or in the same
medium supplemented with 5 mM EGTA. Once the bacteria
had reached an OD600of ?0.2, the cultures were split in half,
the cells were pelleted, and the bacteria from each culture were
resuspended in medium with calcium or depleted of calcium
with EGTA. The cultures were then incubated for another 30
FIG. 5. PopB and PopD are required for triggering of exoS expres-
sion on cell contact. (A) A549 cells were infected for 4 h (MOI, 10)
with the reporterstrain PAO1F
?exoS::GL3 or a ?popB or ?popD derivative thereof harboring either
the vector plasmid (pPSV35) or the complementation plasmid (pP35-
popB or pP35-popD, respectively). The infection was carried out in
RPMI with 10% FBS and 10 ?M IPTG (isopropyl-?-D-thiogalactopy-
ranoside). GFP expression was assayed by microscopy. Five to eight
fields totaling 100 to 130 A549 cells and 500 to 1,000 cell-associated
bacteria were analyzed for each strain. The error bars indicate stan-
dard deviations between fields. Statistical significance was determined
using Student’s t test (*, P ? 0.05). (B) Attachment of the reporter
bacteria to cells(mCherry-positive
?exoS::GL3, a ?popB derivative, and a ?popD derivative were grown
in “high-salt” LB. exoS expression was induced by removing calcium
from the medium through the addition of EGTA (5 mM final concen-
tration) and assayed by beta-galactosidase assay. Activity (averages
from three assays) is depicted in Miller units, and error bars denote the
standard deviations between replicates. w.t., wild type.
bacteria/cell). (C) PAO1F
FIG. 6. Translocation of effectors into A549 cells. A549 cells were
infected in medium lacking FBS for 2 h (MOI, 50). All strains were
?exsE. exoS?and exoS(G/A?) were deleted for exoY and exoT, exoT?
was deleted for exoS and exoY, and ?pcrV expressed all effectors. Cells
were washed with PBS and treated with proteinase K where indicated,
and proteins were extracted with Triton X-100 or SDS as indicated.
Proteins were detected by Western blotting, and the relevant bands are
identified on the right of each blot. Individual lanes are numbered
below the blot.
VOL. 190, 2008HOST FACTOR CONTROLS CELL CONTACT-MEDIATED EFFECTOR SECRETION 2733
min at 37°C before pelleting the bacteria and precipitating
supernatant proteins with TCA.
While secretion of “effector” class proteins (ExoS, ExoT,
ExoY, ExsE, and PopN) was blocked immediately by resus-
pending bacteria that had been induced with EGTA in me-
dium containing calcium, secretion of the translocator proteins
(PopB, PopD, and PcrV) continued unabated (Fig. 7). Some
secretion of PopD was apparent even in the culture that had
not been treated with EGTA at any point. These data suggest
that calcium does not, in fact, control secretion of the translo-
cator proteins and that the apparent triggering of translocator
secretion witnessed in the standard experiment is likely a com-
bination of up-regulation of translocator expression and an
increase in the number of apparatuses. Additionally, degrada-
tion of secreted proteins by chelator-sensitive proteases (55)
may also mask the secretion of translocators in the presence of
calcium. In the case of Yersinia, it has been postulated that
secretion of translocator proteins is stimulated by serum pro-
teins, such as serum albumin (35). In our hands, however,
translocators were secreted even in minimal medium lacking
any protein (see Fig. S3 in the supplemental material), sug-
gesting that secretion of translocator proteins is simply consti-
Type III secretion is a common virulence mechanism among
gram-negative pathogens. One of the hallmarks of type III
secretion is that toxins are delivered vectorially into the host
cell cytoplasm in a contact-dependent manner. While elegant
experiments with Salmonella have demonstrated that trigger-
ing of secretion is a very rapid process (60), the precise order
of steps involved or, indeed, the involvement of host factors in
controlling type III secretion is poorly understood.
Here we present data that support a model for cell contact-
mediated triggering of effector secretion by P. aeruginosa (Fig.
8). We have presented data that secretion of translocator pro-
teins (PopB, PopD, and PcrV) proceeds even in the presence
of calcium. Cell contact, therefore, rather than triggering se-
cretion per se, results in a switch in the secretion specificity,
allowing secretion of effector class proteins (ExoS, ExoT,
ExoY, PopN, and ExsE). While the data presented suggest that
secretion of translocators is calcium independent, a closer look
at the data, in particular for the deregulated exsE mutant (see
Fig. S3 in the supplemental material), suggests that secretion
of translocators may be more efficient in the absence of cal-
cium. There could be several explanations for this observation.
For one, it is known that removing calcium from the medium
FIG. 7. Calcium control of effector and translocator secretion.
Wild-type PAO1 was grown to an OD600of ?0.2 in either the presence
or absence of calcium, and bacteria were subsequently spun down and
resuspended in medium with calcium or lacking calcium and incubated
for an additional 30 min before the cells were pelleted. The cell pellet
and TCA-precipitated supernatant protein were separated by SDS-
PAGE gel, transferred to nitrocellulose, and probed with the indicated
antiserum. The four conditions are indicated above each lane: growth
with calcium, spun down, and resuspended in medium with calcium
(?/?); growth with calcium, spun down, and resuspended in medium
without calcium (5 mM EGTA) (?/?); growth without calcium, spun
down, and resuspended in medium with calcium (?/?); and growth
without calcium, spun down, and resuspended in medium without
FIG. 8. Model of translocon insertion and triggering of effector secretion on cell contact. The individual steps are depicted from left to right.
V, PcrV; B, PopB; D, PopD. The question mark indicates a yet-unidentified eukaryotic factor (or factors) required for triggering of effector
secretion, which is negatively controlled by translocated ExoS. The changes in the translocon and needle that are proposed to signal effector
secretion are depicted as a thick border (PopB and PopD) or black shading of the needle.
2734 CISZ ET AL.J. BACTERIOL.
stimulates cyclic AMP production, which in turn up-regulates
type III secretion gene expression (74). In other words, even
though the exsE mutant is deregulated, it may still receive a
slight boost in expression by the cyclic AMP/Vfr pathway when
grown in the absence of calcium. Consistent with this observa-
tion, exoS expression is slightly increased when P. aeruginosa is
grown in the absence of calcium, even when exsE is deleted (55,
68). Alternatively, degradation of secreted protein by a chela-
tor-sensitive protease could account for the apparent reduction
of secreted PopB. An interesting third possibility is that the
rate of secretion increases when effector secretion is triggered.
There is some evidence to support this hypothesis, since the
activity of the translocation ATPase in Yersinia, YscN (PscN in
P. aeruginosa), is negatively regulated by the associated protein
YscL (PscL) (5). YscL is required for type III secretion, and its
level has to be carefully controlled since overexpression also
negatively affects secretion (5). Interestingly, a 5-amino-acid
deletion of the flagellar YscL homolog, FliH, was found to
stimulate the ATPase activity of the flagellar ATPase, FliI,
raising the possibility that YscN (PscN) ATPase activity could
be activated as well as repressed (26).
These data also suggest that the current model of how YopN
controls effector secretion in the closely related Yersinia T3SS
may not apply to P. aeruginosa. A yopN mutant, just like a P.
aeruginosa popN mutant, secretes effectors constitutively (13,
54, 64). It was proposed that YopN enters the secretion chan-
nel and blocks it by virtue of its C-terminal interaction with
TyeA, which in turn is bound to a yet-to-be-identified compo-
nent of the secretion machinery (19). Here we demonstrated
that translocators are clearly secreted under conditions where
secretion of PopN (the P. aeruginosa YopN homolog) is shut
down, suggesting that PopN cannot be blocking the secretion
channel (Fig. 7). This is likely also the case for Yersinia, since
the proposed model is also not consistent with the finding that
translocators but not effectors are secreted into the culture
supernatant by Yersinia infecting tissue culture cells (34, 58). A
simple model to explain these results would be that the PopN-
Pcr1-Pcr2-PscB complex (YopN-TyeA-SycN-YscB in Yersinia
spp.) occludes a component of the type III secretion machinery
that allows access of effector-chaperone complexes to the se-
cretion ATPase, while translocator-chaperone complexes gain
access via a different route. Consistent with this hypothesis, it
was recently demonstrated that certain yscU mutants fail to
secrete translocators while retaining the ability to secrete ef-
fectors (62). Our assay to reestablish calcium control of secre-
tion is a simple method to discriminate proteins secreted as
“translocator” and “effector” class secretion substrates. It will
be interesting to determine the molecular basis of the signals
that allow the T3SS to differentially recognize translocators
and effectors. In this context, it is tempting to speculate that
the cognate chaperones may be involved in targeting translo-
cators for secretion, since translocator and effector class chap-
erones differ structurally (48).
It was recently demonstrated that PopD is required for trig-
gering of T3SS gene expression on cell contact (67). We have
found that the switch to effector class secretion requires a
functional translocation pore. Mutants lacking PopB or PopD
fail to induce exoS expression on cell contact (which relies on
the secretion of ExsE [55, 67, 68]). Our findings imply that the
PcrV-dependent insertion of PopD into the host cell plasma
membrane itself cannot be the signal for triggering effector
secretion, since PopD is still inserted in a popB mutant (28). As
has been demonstrated for YopB in the case of Yersinia (58),
neither popB nor popD mutants can be cross-complemented,
arguing that insertion of the translocators has to occur in close
proximity to the secretion needle. This is consistent with the
previous finding that insertion of PopD requires the needle tip
protein PcrV (27, 28). The simplest model for secretion con-
trol, therefore, is that after attachment to the cell, when the
needle tip is brought into close proximity to the host cell, the
translocators are inserted into the plasma membrane, forming
the translocation pore. A yet-unknown host factor acts at the
stage of either translocator insertion, docking of the needle to
the translocation pore, or stabilization of the translocase. In
our estimation the latter two possibilities are more likely, since
PopB and PopD can form pores in model membranes in the
absence of other cofactors (17). Interestingly, PcrV cannot
bind to these pores (61). This could of course also be due to an
inherently weak affinity of PcrV for the translocation pore,
requiring either other cofactors or multiple simultaneous in-
teractions as found in the assembled needle tip. The needle tip
most likely harbors multiple PcrV molecules (8) that can in-
teract with the translocation pore (purified PcrV, on the other
hand, is a monomer ). We propose that it is docking of the
needle tip to the translocation pore that results in a confor-
mational change in PcrV, which is propagated down the needle
and results in the activation of effector secretion. In this con-
text, it should be noted that it has been proposed that removing
calcium from the medium in vitro results in triggering effector
secretion by causing a conformational change in the needle
(66). It is therefore easy to envision how removing calcium
from the needle in vitro essentially bypasses the need for a
host-derived signal, which would originate at the assembled
translocation apparatus, by switching the organization of the
needle to a conformation that signals effector secretion.
Our findings are analogous to data from the Yersinia system,
where it was found that translocation of effector proteins can
be modulated by the action of the effector protein YopE (2).
The amino-terminal portion of ExoS and ExoT (comprising
the secretion signal, chaperone binding site, membrane local-
ization domain, and Rho-GAP domain) are homologous to
YopE. Interestingly, ExoS and ExoT can substitute for YopE
with regard to control of effector translocation, suggesting that
these two instances of translocation control are related (1).
There are some differences between these systems, however,
since a second protein, YopK, which has no homolog in P.
aeruginosa, also contributes to translocation control in Yersinia.
In our hands, both the Rho-GAP activity of ExoS and the
ADP-ribosylation activity can serve to control effector translo-
cation. This observation could be interpreted to mean that the
signaling pathways affected by either activity of ExoS converge
to control the activity of the host factor required for triggering
effector secretion on cell contact. Alternatively, either activity
could affect a separate process, both of which are required for
the host factor to be active (e.g., down-regulating a step in
phosphoplipid biogenesis and affecting phospholipid traffick-
ing, either of which results in a change in plasma membrane
phospholipid composition). Clearly, a better understanding of
the nature of the host factor will be required to understand the
role of ExoS in controlling the “effector switch.”
VOL. 190, 2008HOST FACTOR CONTROLS CELL CONTACT-MEDIATED EFFECTOR SECRETION2735
The observation that either activity can control the switch to
effector secretion is in contrast to the case for Yersinia, where
only the Rho-GAP domain of ExoS was able to control effector
translocation when expressed in a yopE null mutant. This dis-
crepancy is perhaps due to differences in the cells used in these
experiments (HeLa versus A549 cells). In the case of Yersinia,
it has been suggested that injection of YopE generates a neg-
ative signal that shuts down effector translocation (1). Our
results are more consistent with the interpretation that the
switch to effector secretion requires a preexisting host factor,
which is inactivated by ExoS. Permeabilization of cells with
either streptolysin O or alpha-hemolysin, two pore-forming
toxins with differing pore sizes and membrane requirements,
abolishes triggering of effector secretion. Since these toxins kill
the cell and allow release of small metabolites and nucleotides,
it is unlikely that activation of a signaling cascade is responsible
for the lack of exoS induction in this case. It has been described
for both the Yersinia and P. aeruginosa systems that type III
secretion can display target cell specificity (39, 42, 57). The
absence of a cellular function required for activation of effector
secretion offers a simple explanation for this phenomenon.
The nature of the signal or cellular function required for
triggering of effector secretion is unclear. It has been proposed
that the signal for triggering type III secretion is in fact the
low-calcium environment of the host cell cytosol, which could
conceivably be sensed once the translocation pore has formed
and docked to the needle (35). Raising the intracellular cal-
cium pool by using the ionophore calcimycin, however, had no
effect on the ability of P. aeruginosa to induce exoS expression,
suggesting that calcium plays no role in triggering of effector
secretion on cell contact. In the Yersinia system YopE was
demonstrated to prevent accumulation of actin at the site of
bacterial attachment. In a yopE mutant, accumulation of actin
was correlated with an increase in pore formation, and it was
hypothesized that localized membrane ruffling resulted in the
translocation pore being jarred loose from the needle (72).
Accumulation of actin, however, is apparently not involved in
controlling effector secretion, since addition of cytochalasin D
did not prevent triggering of exoS expression.
Clearly, determining the molecular basis of triggering of
effector secretion on cell contact will add an important facet to
our understanding of type III secretion. As mentioned above,
controlling the switch to effector secretion may underlie the
observed target cell specificity. It is easy to imagine how being
able to target specific cell types, or perhaps more importantly
to avoid injecting toxins into innocuous cells, can help shape
the course of an infection. A clearer understanding of the
molecular principles governing the specificity of effector secre-
tion on cell contact will allow us to design experiments to
directly test its role in infection.
We thank Piet DeBoer and Patrick Viollier for critical reading of the
manuscript, Stephen Lory (Harvard Medical School) for the ExoS
antiserum, and Ina Attree (CEA Grenoble) for the PopB and PopD
antisera. We also thank Charles Stopford for technical assistance, as
well as David McDonald for allowing us access to the Deltavision
microscope and for instruction in its use.
This work was supported in part by pilot and feasibility award
RIETSC06I0 from the Cystic Fibrosis Foundation to A.R. Preliminary
data were collected in the laboratory of John J. Mekalanos, and that
work was supported by NIH grant AI26289.
1. Aili, M., E. L. Isaksson, S. E. Carlsson, H. Wolf-Watz, R. Rosqvist, and M. S.
Francis. Regulation of Yersinia Yop-effector delivery by translocated YopE.
Int J. Med. Microbiol., in press.
2. Aili, M., E. L. Isaksson, B. Hallberg, H. Wolf-Watz, and R. Rosqvist. 2006.
Functional analysis of the YopE GTPase-activating protein (GAP) activity of
Yersinia pseudotuberculosis. Cell. Microbiol. 8:1020–1033.
3. Barbieri, J. T., and J. Sun. 2004. Pseudomonas aeruginosa ExoS and ExoT.
Rev. Physiol. Biochem. Pharmacol. 152:79–92.
4. Bhakdi, S., H. Bayley, A. Valeva, I. Walev, B. Walker, M. Kehoe, and M.
Palmer. 1996. Staphylococcal alpha-toxin, streptolysin-O, and Escherichia
coli hemolysin: prototypes of pore-forming bacterial cytolysins. Arch. Micro-
5. Blaylock, B., K. E. Riordan, D. M. Missiakas, and O. Schneewind. 2006.
Characterization of the Yersinia enterocolitica type III secretion ATPase
YscN and its regulator, YscL. J. Bacteriol. 188:3525–3534.
6. Bleves, S., C. Soscia, P. Nogueira-Orlandi, A. Lazdunski, and A. Filloux.
2005. Quorum sensing negatively controls type III secretion regulon expres-
sion in Pseudomonas aeruginosa PAO1. J. Bacteriol. 187:3898–3902.
7. Broms, J. E., A. L. Forslund, A. Forsberg, and M. S. Francis. 2003. PcrH of
Pseudomonas aeruginosa is essential for secretion and assembly of the type
III translocon. J. Infect. Dis. 188:1909–1921.
8. Broz, P., C. A. Mueller, S. A. Muller, A. Philippsen, I. Sorg, A. Engel, and
G. R. Cornelis. 2007. Function and molecular architecture of the Yersinia
injectisome tip complex. Mol. Microbiol. 65:1311–1320.
9. Chastre, J., and J. Y. Fagon. 2002. Ventilator-associated pneumonia. Am. J.
Respir. Crit. Care Med. 165:867–903.
10. Coburn, J., S. T. Dillon, B. H. Iglewski, and D. M. Gill. 1989. Exoenzyme S
of Pseudomonas aeruginosa ADP-ribosylates the intermediate filament pro-
tein vimentin. Infect. Immun. 57:996–998.
11. Dacheux, D., J. Goure, J. Chabert, Y. Usson, and I. Attree. 2001. Pore-
forming activity of type III system-secreted proteins leads to oncosis of
Pseudomonas aeruginosa-infected macrophages. Mol. Microbiol. 40:76–85.
12. Dasgupta, N., G. L. Lykken, M. C. Wolfgang, and T. L. Yahr. 2004. A novel
anti-anti-activator mechanism regulates expression of the Pseudomonas
aeruginosa type III secretion system. Mol. Microbiol. 53:297–308.
13. Day, J. B., and G. V. Plano. 1998. A complex composed of SycN and YscB
functions as a specific chaperone for YopN in Yersinia pestis. Mol. Micro-
14. DiNovo, A. A., K. L. Schey, W. S. Vachon, E. M. McGuffie, J. C. Olson, and
T. S. Vincent. 2006. ADP-ribosylation of cyclophilin A by Pseudomonas
aeruginosa exoenzyme S. Biochemistry 45:4664–4673.
15. Edqvist, P. J., M. Aili, J. Liu, and M. S. Francis. 2007. Minimal YopB and
YopD translocator secretion by Yersinia is sufficient for Yop-effector deliv-
ery into target cells. Microbes Infect. 9:224–233.
16. Espina, M., S. F. Ausar, C. R. Middaugh, M. A. Baxter, W. D. Picking, and
W. L. Picking. 2007. Conformational stability and differential structural
analysis of LcrV, PcrV, BipD, and SipD from type III secretion systems.
Protein Sci. 16:704–714.
17. Faudry, E., G. Vernier, E. Neumann, V. Forge, and I. Attree. 2006. Syner-
gistic pore formation by type III toxin translocators of Pseudomonas aerugi-
nosa. Biochemistry 45:8117–8123.
18. Feltman, H., G. Schulert, S. Khan, M. Jain, L. Peterson, and A. R. Hauser.
2001. Prevalence of type III secretion genes in clinical and environmental
isolates of Pseudomonas aeruginosa. Microbiology 147:2659–2669.
19. Ferracci, F., F. D. Schubot, D. S. Waugh, and G. V. Plano. 2005. Selection
and characterization of Yersinia pestis YopN mutants that constitutively
block Yop secretion. Mol. Microbiol. 57:970–987.
20. Fleiszig, S. M., T. S. Zaidi, M. J. Preston, M. Grout, D. J. Evans, and G. B.
Pier. 1996. Relationship between cytotoxicity and corneal epithelial cell
invasion by clinical isolates of Pseudomonas aeruginosa. Infect. Immun. 64:
21. Frank, D. W. 1997. The exoenzyme S regulon of Pseudomonas aeruginosa.
Mol. Microbiol. 26:621–629.
22. Frank, D. W., and B. H. Iglewski. 1991. Cloning and sequence analysis of a
trans-regulatory locus required for exoenzyme S synthesis in Pseudomonas
aeruginosa. J. Bacteriol. 173:6460–6468.
23. Fraylick, J. E., M. J. Riese, T. S. Vincent, J. T. Barbieri, and J. C. Olson.
2002. ADP-ribosylation and functional effects of Pseudomonas exoenzyme S
on cellular RalA. Biochemistry 41:9680–9687.
24. Garau, J., and L. Gomez. 2003. Pseudomonas aeruginosa pneumonia. Curr.
Opin. Infect. Dis. 16:135–143.
25. Goehring, U. M., G. Schmidt, K. J. Pederson, K. Aktories, and J. T. Barbieri.
1999. The N-terminal domain of Pseudomonas aeruginosa exoenzyme S is a
GTPase-activating protein for Rho GTPases. J. Biol. Chem. 274:36369–
26. Gonzalez-Pedrajo, B., G. M. Fraser, T. Minamino, and R. M. Macnab. 2002.
Molecular dissection of Salmonella FliH, a regulator of the ATPase FliI and
the type III flagellar protein export pathway. Mol. Microbiol. 45:967–982.
2736CISZ ET AL.J. BACTERIOL.
27. Goure, J., P. Broz, O. Attree, G. R. Cornelis, and I. Attree. 2005. Protective
anti-V antibodies inhibit Pseudomonas and Yersinia translocon assembly
within host membranes. J. Infect. Dis. 192:218–225.
28. Goure, J., A. Pastor, E. Faudry, J. Chabert, A. Dessen, and I. Attree. 2004.
The V antigen of Pseudomonas aeruginosa is required for assembly of the
functional PopB/PopD translocation pore in host cell membranes. Infect.
29. Hauser, A. R., E. Cobb, M. Bodi, D. Mariscal, J. Valles, J. N. Engel,
and J. Rello. 2002. Type III protein secretion is associated with poor clinical
outcomes in patients with ventilator-associated pneumonia caused by
Pseudomonas aeruginosa. Crit. Care Med. 30:521–528.
30. Hoang, T. T., A. J. Kutchma, A. Becher, and H. P. Schweizer. 2000. Integra-
tion-proficient plasmids for Pseudomonas aeruginosa: site-specific integra-
tion and use for engineering of reporter and expression strains. Plasmid
31. Holder, I. A., A. N. Neely, and D. W. Frank. 2001. Type III secretion/
intoxication system important in virulence of Pseudomonas aeruginosa in-
fections in burns. Burns 27:129–130.
32. Holmstrom, A., J. Petterson, R. Rosqvist, S. Hakansson, F. Tafazoli, M.
Fallman, K. E. Magnusson, H. Wolf-Watz, and A. Forsberg. 1997. YopK of
Yersinia pseudotuberculosis controls translocation of Yop effectors across
the eukaryotic cell membrane. Mol. Microbiol. 24:73–91.
33. Hornef, M. W., A. Roggenkamp, A. M. Geiger, M. Hogardt, C. A. Jacobi, and
J. Heesemann. 2000. Triggering the ExoS regulon of Pseudomonas aerugi-
nosa: a GFP-reporter analysis of exoenzyme (Exo) S, ExoT and ExoU syn-
thesis. Microb. Pathog. 29:329–343.
34. Lee, V. T., D. M. Anderson, and O. Schneewind. 1998. Targeting of Yersinia
Yop proteins into the cytosol of HeLa cells: one-step translocation of YopE
across bacterial and eukaryotic membranes is dependent on SycE chaperone.
Mol. Microbiol. 28:593–601.
35. Lee, V. T., S. K. Mazmanian, and O. Schneewind. 2001. A program of
Yersinia enterocolitica type III secretion reactions is activated by specific
signals. J. Bacteriol. 183:4970–4978.
36. Lee, V. T., and O. Schneewind. 1999. Type III machines of pathogenic
yersiniae secrete virulence factors into the extracellular milieu. Mol. Micro-
37. Lyczak, J. B., C. L. Cannon, and G. B. Pier. 2002. Lung infections associated
with cystic fibrosis. Clin. Microbiol. Rev. 15:194–222.
38. Maresso, A. W., M. R. Baldwin, and J. T. Barbieri. 2004. Ezrin/radixin/
moesin proteins are high affinity targets for ADP-ribosylation by Pseudomo-
nas aeruginosa ExoS. J. Biol. Chem. 279:38402–38408.
39. Marketon, M. M., R. W. DePaolo, K. L. DeBord, B. Jabri, and O. Schnee-
wind. 2005. Plague bacteria target immune cells during infection. Science
40. McCaw, M. L., G. L. Lykken, P. K. Singh, and T. L. Yahr. 2002. ExsD is a
negative regulator of the Pseudomonas aeruginosa type III secretion regu-
lon. Mol. Microbiol. 46:1123–1133.
41. McGuffie, E. M., D. W. Frank, T. S. Vincent, and J. C. Olson. 1998. Modi-
fication of Ras in eukaryotic cells by Pseudomonas aeruginosa exoenzyme S.
Infect. Immun. 66:2607–2613.
42. McGuffie, E. M., J. E. Fraylick, D. J. Hazen-Martin, T. S. Vincent, and J. C.
Olson. 1999. Differential sensitivity of human epithelial cells to Pseudomonas
aeruginosa exoenzyme S. Infect. Immun. 67:3494–3503.
43. Miyata, S., M. Casey, D. W. Frank, F. M. Ausubel, and E. Drenkard. 2003.
Use of the Galleria mellonella caterpillar as a model host to study the role of
the type III secretion system in Pseudomonas aeruginosa pathogenesis. In-
fect. Immun. 71:2404–2413.
44. Mueller, C. A., P. Broz, S. A. Muller, P. Ringler, F. Erne-Brand, I. Sorg, M.
Kuhn, A. Engel, and G. R. Cornelis. 2005. The V-antigen of Yersinia forms
a distinct structure at the tip of injectisome needles. Science 310:674–676.
45. National Nosocomial Infections Surveillance System. 2004. National Noso-
comial Infections Surveillance (NNIS) System Report, data summary from
January 1992 through June 2004, issued October 2004. Am. J. Infect Control
46. Nordfelth, R., and H. Wolf-Watz. 2001. YopB of Yersinia enterocolitica is
essential for YopE translocation. Infect. Immun. 69:3516–3518.
47. Olive, A. J., R. Kenjale, M. Espina, D. S. Moore, W. L. Picking, and W. D.
Picking. 2007. Bile salts stimulate recruitment of IpaB to the Shigella flexneri
surface, where it colocalizes with IpaD at the tip of the type III secretion
needle. Infect. Immun. 75:2626–2629.
48. Parsot, C., C. Hamiaux, and A. L. Page. 2003. The various and varying roles
of specific chaperones in type III secretion systems. Curr. Opin. Microbiol.
49. Pirnay, J. P., D. De Vos, C. Cochez, F. Bilocq, J. Pirson, M. Struelens, L.
Duinslaeger, P. Cornelis, M. Zizi, and A. Vanderkelen. 2003. Molecular
epidemiology of Pseudomonas aeruginosa colonization in a burn unit: per-
sistence of a multidrug-resistant clone and a silver sulfadiazine-resistant
clone. J. Clin. Microbiol. 41:1192–1202.
50. Pukatzki, S., R. H. Kessin, and J. J. Mekalanos. 2002. The human pathogen
Pseudomonas aeruginosa utilizes conserved virulence pathways to infect the
social amoeba Dictyostelium discoideum. Proc. Natl. Acad. Sci. USA 99:
51. Radke, J., K. J. Pederson, and J. T. Barbieri. 1999. Pseudomonas aeruginosa
exoenzyme S is a biglutamic acid ADP-ribosyltransferase. Infect. Immun.
52. Rahme, L. G., E. J. Stevens, S. F. Wolfort, J. Shao, R. G. Tompkins, and
F. M. Ausubel. 1995. Common virulence factors for bacterial pathogenicity in
plants and animals. Science 268:1899–1902.
53. Riese, M. J., A. Wittinghofer, and J. T. Barbieri. 2001. ADP ribosylation of
Arg41 of Rap by ExoS inhibits the ability of Rap to interact with its guanine
nucleotide exchange factor, C3G. Biochemistry. 40:3289–3294.
54. Rietsch, A., and J. J. Mekalanos. 2006. Metabolic regulation of type III
secretion gene expression in Pseudomonas aeruginosa. Mol. Microbiol. 59:
55. Rietsch, A., I. Vallet-Gely, S. L. Dove, and J. J. Mekalanos. 2005. ExsE, a
secreted regulator of type III secretion genes in Pseudomonas aeruginosa.
Proc. Natl. Acad. Sci. USA 102:8006–8011.
56. Rietsch, A., M. C. Wolfgang, and J. J. Mekalanos. 2004. Effect of metabolic
imbalance on expression of type III secretion genes in Pseudomonas aerugi-
nosa. Infect. Immun. 72:1383–1390.
57. Rucks, E. A., and J. C. Olson. 2005. Characterization of an ExoS type III
translocation-resistant cell line. Infect. Immun. 73:638–643.
58. Ryndak, M. B., H. Chung, E. London, and J. B. Bliska. 2005. Role of
predicted transmembrane domains for type III translocation, pore forma-
tion, and signaling by the Yersinia pseudotuberculosis YopB protein. Infect.
59. Sato, H., D. W. Frank, C. J. Hillard, J. B. Feix, R. R. Pankhaniya, K.
Moriyama, V. Finck-Barbancon, A. Buchaklian, M. Lei, R. M. Long, J.
Wiener-Kronish, and T. Sawa. 2003. The mechanism of action of the
Pseudomonas aeruginosa-encoded type III cytotoxin, ExoU. EMBO J. 22:
60. Schlumberger, M. C., A. J. Muller, K. Ehrbar, B. Winnen, I. Duss, B.
Stecher, and W. D. Hardt. 2005. Real-time imaging of type III secretion:
Salmonella SipA injection into host cells. Proc. Natl. Acad. Sci. USA 102:
61. Schoehn, G., A. M. Di Guilmi, D. Lemaire, I. Attree, W. Weissenhorn, and
A. Dessen. 2003. Oligomerization of type III secretion proteins PopB and
PopD precedes pore formation in Pseudomonas. EMBO J. 22:4957–4967.
62. Sorg, I., S. Wagner, M. Amstutz, S. A. Muller, P. Broz, Y. Lussi, A. Engel,
and G. R. Cornelis. 2007. YscU recognizes translocators as export substrates
of the Yersinia injectisome. EMBO J. 26:3015–3024.
63. Sun, J., and J. T. Barbieri. 2003. Pseudomonas aeruginosa ExoT ADP-
ribosylates CT10 regulator of kinase (Crk) proteins. J. Biol. Chem. 278:
64. Sundin, C., J. Thelaus, J. E. Broms, and K. Forsberg. 2004. Polarisation of
type III translocation by Pseudomonas aeruginosa requires PcrG, PcrV and
PopN. Microb. Pathog. 37:313–322.
65. Sundin, C., M. C. Wolfgang, S. Lory, A. Forsberg, and E. Frithz-Lindsten.
2002. Type IV pili are not specifically required for contact dependent trans-
location of exoenzymes by Pseudomonas aeruginosa. Microb. Pathog. 33:
66. Torruellas, J., M. W. Jackson, J. W. Pennock, and G. V. Plano. 2005. The
Yersinia pestis type III secretion needle plays a role in the regulation of Yop
secretion. Mol. Microbiol. 57:1719–1733.
67. Urbanowski, M. L., E. D. Brutinel, and T. L. Yahr. 2007. Translocation of
ExsE into Chinese hamster ovary cells is required for transcriptional induc-
tion of the Pseudomonas aeruginosa type III secretion system. Infect. Immun.
68. Urbanowski, M. L., G. L. Lykken, and T. L. Yahr. 2005. A secreted regula-
tory protein couples transcription to the secretory activity of the Pseudomo-
nas aeruginosa type III secretion system. Proc. Natl. Acad. Sci. USA 102:
69. Vallis, A. J., T. L. Yahr, J. T. Barbieri, and D. W. Frank. 1999. Regulation
of ExoS production and secretion by Pseudomonas aeruginosa in response to
tissue culture conditions. Infect. Immun. 67:914–920.
70. Vance, R. E., A. Rietsch, and J. J. Mekalanos. 2005. Role of the type III
secreted exoenzymes S, T, and Y in systemic spread of Pseudomonas aerugi-
nosa PAO1 in vivo. Infect. Immun. 73:1706–1713.
71. Veenendaal, A. K., J. L. Hodgkinson, L. Schwarzer, D. Stabat, S. F. Zenk,
and A. J. Blocker. 2007. The type III secretion system needle tip complex
mediates host cell sensing and translocon insertion. Mol. Microbiol. 63:1719–
72. Viboud, G. I., and J. B. Bliska. 2001. A bacterial type III secretion system
inhibits actin polymerization to prevent pore formation in host cell mem-
branes. EMBO J. 20:5373–5382.
73. Warrens, A. N., M. D. Jones, and R. I. Lechler. 1997. Splicing by overlap
extension by PCR using asymmetric amplification: an improved technique
for the generation of hybrid proteins of immunological interest. Gene 186:
74. Wolfgang, M. C., V. T. Lee, M. E. Gilmore, and S. Lory. 2003. Coordinate
regulation of bacterial virulence genes by a novel adenylate cyclase-depen-
dent signaling pathway. Dev. Cell 4:253–263.
75. Yahr, T. L., L. M. Mende-Mueller, M. B. Friese, and D. W. Frank. 1997.
VOL. 190, 2008 HOST FACTOR CONTROLS CELL CONTACT-MEDIATED EFFECTOR SECRETION2737
Identification of type III secreted products of the Pseudomonas aeruginosa
exoenzyme S regulon. J. Bacteriol. 179:7165–7168.
76. Yahr, T. L., A. J. Vallis, M. K. Hancock, J. T. Barbieri, and D. W. Frank.
1998. ExoY, an adenylate cyclase secreted by the Pseudomonas aeruginosa
type III system. Proc. Natl. Acad. Sci. USA 95:13899–13904.
77. Zhang, Y., Q. Deng, J. A. Porath, C. L. Williams, K. J. Pederson-Gulrud, and
J. T. Barbieri. 2007. Plasma membrane localization affects the RhoGAP
specificity of Pseudomonas ExoS. Cell. Microbiol. 9:2192–2201.
78. Zheng, Z., G. Chen, S. Joshi, E. D. Brutinel, T. L. Yahr, and L. Chen. 2007.
Biochemical characterization of a regulatory cascade controlling transcrip-
tion of the Pseudomonas aeruginosa type III secretion system. J. Biol. Chem.
2738CISZ ET AL.J. BACTERIOL.