Article

In vitro Intestinal Mucosal Epithelial Responses to Wild-Type Salmonella Typhi and Attenuated Typhoid Vaccines

Department of Pediatrics, Mucosal Biology Research Center, University of Maryland School of Medicine Baltimore, MD, USA.
Frontiers in Immunology 02/2013; 4:17. DOI: 10.3389/fimmu.2013.00017
Source: PubMed

ABSTRACT

Typhoid fever, caused by S. Typhi, is responsible for approximately 200,000 deaths per year worldwide. Little information is available regarding epithelium-bacterial interactions in S. Typhi infection. We have evaluated in vitro the effects of wild-type S. Typhi, the licensed Ty21a typhoid vaccine and the leading strains CVD 908-htrA and CVD 909 vaccine candidates on intestinal barrier function and immune response. Caco2 monolayers infected with wild-type S. Typhi exhibited alterations in the organization of tight junctions, increased paracellular permeability, and a rapid decrease in Trans-Epithelial Electrical Resistance as early as 4 h post-exposure. S. Typhi triggered the secretion of interleukin (IL)-8 and IL-6. Caco2 cells infected with the attenuated strains exhibited a milder pro-inflammatory response with minimal disruption of the barrier integrity. We conclude that wild-type S. Typhi causes marked transient alterations of the intestinal mucosa that are more pronounced than those observed with Ty21a or new generation attenuated typhoid vaccine candidates.

Full-text

Available from: Alessio Fasano, Mar 16, 2014
ORIGINAL RESEARCH ARTICLE
published: 12 February 2013
doi: 10.3389/fimmu.2013.00017
In vitro intestinal mucosal epithelial responses to wild-type
SalmonellaTyphi and attenuated typhoid vaccines
Maria Fiorentino
1
, Karen M. Lammers
1
, Myron M. Levine
2
, Marcelo B. Sztein
and Alessio Fasano
1
§
*
1
Department of Pediatrics, Mucosal Biology Research Center, University of Maryland School of Medicine, Baltimore, MD, USA
2
Department of Pediatrics, Center for Vaccine Development, University of Maryland School of Medicine, Baltimore, MD, USA
Edited by:
Eric Cox, Ghent University, Belgium
Reviewed by:
Diane Bimczok, University of Alabama
at Birmingham, USA
Hiroshi Ohno, RIKEN, Japan
*Correspondence:
Alessio Fasano, Mucosal Immunology
and Biology Research Center,
Massachusetts General Hospital East,
Building 114, 16th Street (Mail Stop
114-3503), Charlestown, MA
02129-4404, USA.
e-mail: afasano@partners.org
Current address:
Karen M. Lammers and Alessio
Fasano, Mucosal Immunology and
Biology Research Center,
Massachusetts General Hospital,
Charlestown, MA, USA.
§
Joint senior authorship
Typhoid fever, caused by S.Typhi, is responsible for approximately 200,000 deaths per year
worldwide. Little information is available regarding epithelium-bacterial interactions in S.
Typhi infection. We have evaluated in vitro the effects of wild-type S. Typhi, the licensed
Ty21a typhoid vaccine and the leading strains CVD 908-htrA and CVD 909 vaccine candi-
dates on intestinal barrier function and immune response. Caco2 monolayers infected with
wild-type S.Typhi exhibited alterations in the organization of tight junctions, increased para-
cellular permeability, and a rapid decrease in Trans-Epithelial Electrical Resistance as early
as 4 h post-exposure. S. Typhi triggered the secretion of interleukin (IL)-8 and IL-6. Caco2
cells infected with the attenuated strains exhibited a milder pro-inflammatory response
with minimal disruption of the barrier integrity. We conclude that wild-type S.Typhi causes
marked transient alterations of the intestinal mucosa that are more pronounced than those
observed with Ty21a or new generation attenuated typhoid vaccine candidates.
Keywords: intestinal mucosal barrier, epithelial permeability, mucosal immunity, Salmonella Typhi, typhoid
vaccines, cytokines
INTRODUCTION
Salmonella spp. are highly invasive pathogens. Salmonella enter-
ica serovar Typhi (S. Typhi) and S. Paratyphi A, B, and C are
the causing agents of enteric fevers. Typhoid fever is an acute,
life-threatening febrile illness caused by S. Typhi which results
in 200,000 deaths worldwide each year, largely in developing
nations (Crump et al., 2004). Typhoid fever also occurs among per-
sons living in industrialized countries, many belonging to known
risk groups, such as travelers to endemic regions, including mil-
itary personnel. Humans are the only reservoir for S. Typhi and
infection occurs through ingestion of contaminated food or water.
Following ingestion, the bacteria spread from the intestine via the
blood where they multiply to the intestinal lymph nodes, liver,
and spleen. Typhoid fever is characterized by fever and abdominal
symptoms. In 5–10% of infected people, neuropsychiatric man-
ifestations occur. Complications such as gastrointestinal bleed-
ing, intestinal perforation, and typhoid encephalopathy occur in
10–15% of patients (Fraser et al., 2007).
The first significant cellular contact enteric pathogens have with
the host, occurs at the level of the intestinal epithelium. Inva-
sive bacteria have evolved an array of mechanisms to breach the
integrity of the intestinal epithelial barrier, either by targeting tight
junction proteins directly or by altering transduction signals regu-
lating their assembly (Fasano et al., 1995; Aktories, 1997; Wu et al.,
1998; Pothoulakis, 2000; Simonovic et al., 2000). In most cases,
however it is unclear if these cellular events represent a direct
effect of bacterial mediators or compensatory responses of the
host epithelial cells.
Due to the high burden of typhoid fever and increasing antibi-
otic resistance, vaccine development remains a high priority.
Neither of the two vaccines currently available for the preven-
tion of typhoid fever, including the orally administered Ty21a, is
completely effective, with protection rates ranging between 60 and
80% (Levine et al., 1999, 2007; Guzman et al., 2006). Attenuated,
oral typhoid vaccine candidate strains harboring aro mutations
have been evaluated in volunteers (Hone et al., 1992) and some
of them, such as CVD 908-htrA and its derivative CVD 909, have
been shown to be well tolerated and highly immunogenic in clin-
ical trials (Tacket et al., 2000a,b, 2004; Salerno-Goncalves et al.,
2003, 2004; Sztein, 2007; Wahid et al., 2007, 2008, 2011).
Most of the studies aimed to evaluate the effect of Salmonella
spp. on epithelial cells have been focused on S. Typhimurium or
other non-typhoid spp. and demonstrated the ability of Salmonella
to disrupt barrier integrity by modulating epithelial permeability,
as indicated by a reduction in Trans-Epithelial Electric Resistance
(TEER) and alteration of tight junction expression and/or distrib-
ution (Finlay et al., 1988; Finlay and Falkow,1990;Jones et al., 1994;
Clark et al., 1998; Jepson et al., 2000; Bertelsen et al., 2004; Otte and
Podolsky, 2004; Kohler et al., 2007). Very limited information is
available regarding S. Typhi’s ability to disrupt the epithelial barrier
and no studies so far have described the interaction of aro mutants
with human enterocytes. To our knowledge, only one study
demonstrated that S. Typhi is able to disrupt the epithelial barrier
in vitro by decreasing TEER in Hep-2 and Caco2 cell monolayers
(Solano et al., 2001). Despite the wealth of available information
on the pathogenesis of Salmonella spp., a systematic study of the
www.frontiersin.org February 2013 | Volume 4 | Article 17 | 1
Page 1
Fiorentino et al. Epithelial cell responses to S. Typhi
effects of S. Typhi or attenuated S. Typhi vaccines on the integrity
and function of human epithelial cells has never been performed.
To this end, we have exploited a human cellular intestinal model
system to evaluate the epithelial events occurring at the host-
bacterial interface following infection with wild-type S. Typhi, the
licensed attenuated oral Ty21a typhoid vaccine and the leading
strains CVD 908-htrA and CVD 909 typhoid vaccine candidates.
MATERIALS AND METHODS
CELL CULTURE
Human Caco2 intestinal epithelial cells [HTB-37, American Type
Culture Collection (ATCC), Rockville, MD, USA] were grown in
Dulbeccos Modified Eagle Medium (DMEM) supplemented with
fetal bovine serum, glutamine, and antibiotics. Caco2 cells are
polarized epithelial cells that form apical junctional complexes,
resulting in high electrical resistance, useful for studying effects
of bacteria on permeability. Monolayers (passages 22–30) were
grown on 1.12 cm
2
permeable polyester filters with 0.4 µm pore
size (Corning, Lowell,MA,USA) and utilized after 14–21 days until
having reached a confluent, polarized, and differentiated state. For
some immunofluorescence staining experiments (IFL),Caco2 cells
were grown on eight-well slide culture chambers (Lab Tek II, Nunc,
IL, USA) and were used 2 days after confluence.
BACTERIA STRAINS AND GROWTH CONDITIONS
Wild-type S. Typhi strain Ty2 (Deng et al., 2003) was used in this
study. Ty21a, a commercially available licensed live oral typhoid
vaccine, is an attenuated strain developed by chemical mutage-
nesis of S. Typhi strain Ty2 (Germanier and Fuer, 1975). CVD
908-htrA and CVD 909 are aro attenuated strains of S. Typhi
(Tacket et al., 2000b, 2004). Bacteria were routinely pre-cultured
at 37˚C overnight in Luria-Bertani (LB) broth. Aliquots of the pre-
cultures were inoculated into 5 ml DMEM and grown for 2–3 h at
37˚C in a shaking incubator (200 rpm) until the cultures reached
the exponential phase (OD
600nm
= 0.5). The aro mutant strains
were grown on aro agar, as previously described (Hone et al., 1991;
Tacket et al., 2000b, 2004), pre-cultured in LB broth and cultured
in DMEM with the addition of Dihydroxybenzoate (DHB, 0.1%).
GENERATION OF CONDITIONED MEDIA AND HEAT-KILLED CULTURES
Aliquots of overnight pre-cultured bacteria were grown in DMEM
for 2–3 h to a final OD
600
of 0.5. Cells were pelleted and
supernatants filter sterilized by passing through a 0.22 µm pore
size filter. Supernatants were used immediately upon filtration.
These supernatants are hereafter referred to as conditioned
media (CM).
Bacteria grown in DMEM for 2–3 h as described above were
heat-killed (HK) by boiling at 100˚C for 30 min. In both cases, the
effectiveness of filtration and killing by heat was confirmed by lack
of bacterial growth from 100 µl of these media plated onto agar
plates and then incubated overnight at 37˚C.
MEASUREMENT OF TEER
Trans-Epithelial Electric Resistance was used to monitor the
integrity of the epithelial monolayer using a Millicel ERS Volt-ohm
meter (World Precision Instruments, New Haven, CT, USA). Only
those monolayers that exhibited a TEER of 1100–1700 .cm
2
were considered to have an appropriate barrier function and were
used in the study. The average number of cells/monolayer was
approximately about 1–1.5 × 10
5
at confluency.
Cell monolayers were drained of media, gently washed with PBS
and then incubated with DMEM without antibiotics and serum at
37˚C for 2 h before bacterial infection.
The bacteria suspension, HK bacteria, or bacterial supernatants
were added apically at an inoculation ratio [Multiplicity Of Infec-
tion (MOI)] of 40:1, 400:1, and 4000:1 bacteria:epithelial cell
ratios, corresponding to 4–6 × 10
6
, 4–6 × 10
7
, and 4–6 × 10
8
CFUs, respectively and incubated at 37˚C. After 4 h infection,
cells were washed with PBS to remove non-adherent bacteria
and treated with gentamicin (480 µg/ml). Monolayers were incu-
bated at 37˚C overnight. TEER was measured at 2, 4, and 22 h
post-infection.
CELL VIABILITY
The viability of the Caco2 cells after bacterial infection with wild-
type S. Typhi Ty2, Ty21a, CVD 908-htrA, and CVD 909, was
assessed by a lactate dehydrogenase (LDH) secretion assay. The
LDH secretion was measured from the cellular supernatant by a
commercially available LDH assay kit (Cytotox 96, Promega, WI,
USA) and carried as described by the manufacturer’s instructions.
Lysis of the cells with 1% Triton X-100 served as positive control.
The absorbance was measured at 490 nm using a multifunctional
microplate reader. Cell viability was also assessed by Propidium
Iodide staining by flow cytometry following a standard technique
(Diebel et al., 2005; Riccardi and Nicoletti, 2006).
ASSESSMENT OF Caco2 CELL MONOLAYER PARACELLULAR
PERMEABILITY
The permeability of the Caco2 cell monolayers was evaluated
by measuring the influx of Fluorescein isothiocyanate (FITC)-
dextran and FITC-Bovine Serum Albumin (BSA) [molecular
weights of 4.0 and 40 kDa (Sigma), respectively].
Fluorescein isothiocyanate-dextran and -Bovine Serum Albu-
min were dissolved in P buffer (10 mM HEPES, pH 7.4, 1 mM
sodium pyruvate, 10 mM glucose, 3 mM CaCl
2
, 145 mM NaCl)
or P/EGTA buffer [10 mM HEPES, pH 7.4, 1 mM sodium pyru-
vate, 10 mM glucose, 145 mM NaCl, 2 mM ethylene glycol-bis(ß-
aminoethyl ether)-N,N,N
0
,N
0
-tetraacetic acid (EGTA)].
Briefly, the apical surface of Caco2 cell monolayers was infected
with bacteria (MOI 400:1) for 4 h, washed, treated with gentam-
icin, and incubated at 37˚C overnight. To measure the paracellular
flux, the apical, and basolateral cell culture media were replaced
with P buffer containing FITC-dextran (10 mg/ml) or FITC-BSA
(10 mg/ml) and P buffer alone, respectively. P/EGTA buffer con-
taining FITC-dextran (10 mg/ml) or FITC-BSA (10 mg/ml) and
P/EGTA buffer were used as positive controls. After incubation for
4 h, the amounts of FITC-dextran and FITC-BSA in the basolateral
media were measured with a fluorometer(excitation at 492 nm and
emission at 520 nm). Data are expressed as fluorescent intensity.
CYTOKINE ASSAYS
To determine the cytokine response of Caco2 cell monolayers to
bacterial infection, cells were incubated with increasing amounts
of bacteria (MOIs of 40:1, 400:1, and 4000:1 bacteria:cell ratios).
Frontiers in Immunology | Mucosal Immunity February 2013 | Volume 4 | Article 17 | 2
Page 2
Fiorentino et al. Epithelial cell responses to S. Typhi
After 4 h of incubation, bacteria were removed and monolayers
incubated with fresh culture medium at 37˚C overnight. After
22 h media were collected from either upper or lower cham-
bers of each transwell. Samples were centrifuged at 2,000 rpm
for 10 min to remove any residual cells or debris. Supernatants
were stored at -80˚ and then assayed for interleukin (IL)-8, IL-6,
tumor necrosis factor (TNF)-α, IL-17, IL-1β, interferon (IFN)-
γ, and transforming growth factor (TGF)-β using a Luminex 100
platform (Luminex,Austin, TX, USA) and the Milliplex MAP High
Sensitivity Human Cytokine kit. Kits were run according to the
manufacturer’s instructions, with the exception of sample collec-
tion and processing as described above. Incubation of beads and
test samples for all kits was performed overnight at 4˚C for max-
imum sensitivity. Samples were run in duplicate. Media collected
from Ty21a infected cells at a MOI of 400:1, were processed for
detection of IL-8 and IL-6, only.
IMMUNFLUORESCENCE
Monolayers on chamber slides (4 h post-infection staining) or
transwell filter inserts (22 h post-infection staining) following
infection were washed three times with PBS and fixed in PFA
4%/PBS for 20 min at room temperature. Cells were then blocked
with 2% PBS-diluted normal goat serum (blocking solution) for
30 min and incubated with blocking solution-diluted primary
antibody overnight at 4˚C [anti zonula occludens (ZO)-1, 1:100].
After three washes with PBS, monolayers were incubated with
TRITC-conjugated secondary antibody (1:5000 in blocking solu-
tion) at room temperature for 1 h in the dark. Phalloidin staining
(1:1000) of actin filaments was carried out at the same time as
the secondary antibody. Monolayers were washed with PBS and
nuclei stained with DAPI (1:1000 in PBS) solution for 2 min at
room temperature. Tissue culture filters housing the epithelial
cell monolayers were carefully detached from their support and
mounted on coverslips. Monolayers were analyzed with a Nikon
Eclipse TE2000-E fluorescent microscope.
WESTERN BLOT ANALYSIS
Triton X-100-soluble and -insoluble fractions
Triton X-100-soluble and -insoluble fractions are working defi-
nitions applied to biochemically define the localization of tight
junction proteins. Several studies have made use of this method
to fractionate tight junctional protein in their cytosolic and mem-
brane bound components (Youakim and Ahdieh, 1999; Nusrat
et al., 2000, 2001; Andreeva et al., 2001; Chen et al., 2002). Proteins
found in the Triton X-100-insoluble fraction have been associated
with the tight junction complex.
Bacteria were added to the apical surface of Caco2 cell mono-
layers at a MOI of 400:1 for 4 h at 37˚C. Bacteria were removed,
cells were washed with PBS and incubated with DMEM sup-
plemented with gentamicin. Cells were harvested at 4 and 22 h
post-infection and Triton X-100-soluble and -insoluble protein
fractions were prepared. Monolayers were harvested on ice in lysis
buffer [1% Triton X-100, 100 mM NaCl, 10 mM HEPES, 2 mM
EDTA, 4 mM Na3VO4, 40 mM NaF 200 mM PMSF, and a protease
inhibitor cocktail (Complete Mini, Roche Molecular Biochemi-
cals, Mannheim, Germany) and phosphatase inhibitors (Sigma, St.
Louis, MO, USA)]. Lysates were rotated at 4˚C, 30 min, centrifuged
(14000 g for 30 min at 4˚C) and the supernatant suspension, repre-
senting the Triton X-soluble fraction, was collected. The remaining
pellet was re-suspended in lysis buffer supplemented with 1% SDS
and sonicated (5W, 5 s) two–three times on ice. The resulting
suspension was centrifuged (14000 g for 5 min at 4˚C) and the
supernatant, representing the Triton X-insoluble fraction, was col-
lected. Samples were used immediately or stored at 80˚C. Protein
concentration was quantified by the Bradford method (Bio-Rad,
Hercules, CA, USA). Samples were electrophoresed through a
10–20% gradient SDS polyacrylamide gel and transferred onto
polyvinylidene difluoride membranes (Millipore, Bedford, MA,
USA). Membranes were blocked in blocking buffer (Tris-buffered
saline, 0.1% Tween 20, 5% BSA), for 1 h at room temperature. The
blots were incubated overnight at 4˚C with mouse anti-occludin
diluted in blocking buffer. After washing, membranes were incu-
bated for 1 h at room temperature with the appropriate secondary
antibody diluted in blocking buffer. The hybridized band was
detected by chemiluminescence using an ECL kit (Amersham)
according to the manufacturer’s instructions. Membranes were
stripped and reprobed (Blot restore solution, Millipore) for the
detection of phosphothreonine followed by actin that served as
loading control. Band intensity was normalized to actin and quan-
titated by densitometry using Image J software (National Institutes
of Health). Data represent the average of two separate experiments.
ANTIBODIES
Mouse anti-occludin (OC-3F10, cat # 33–1500), mouse anti-ZO-1
(1A12, cat # 339100), and rabbit anti-phosphothreonine (ZPT1cat
# 718200) antibodies were purchased from Invitrogen (Camarillo,
CA, USA). Mouse anti-actin protein antibody (5C, cat # 82353)
was purchased by Thermo Fisher Scientific (IL, USA). TRITC-
conjugated anti-mouse and anti-rabbit secondary antibodies and
FITC-conjugated phalloidin were obtained from Sigma (S. Louis,
MO, USA).
STATISTICS
Data are expressed as means ± SEM. Data were analyzed by using
GraphPad (San Diego, CA, USA) software. Two-way (TEER) and
one-way ANOVA were used to compare the data among groups.
Differences were considered to be statistically significant if P
values were <0.05.
RESULTS
SALMONELLA ENTERICA SEROVAR TYPHI AFFECTS EPITHELIAL
BARRIER FUNCTION BY DECREASING TEER IN A DOSE-DEPENDENT
FASHION
To investigate the effects of S. Typhi on mucosal barrier integrity,
we infected Caco2 monolayers with S. Typhi at different MOIs.
Since modulation and/or disruption of epithelial barrier function
can be measured by changes in TEER, we used this technique
to monitor alterations in mucosal permeability caused by the
bacteria. As shown in Figure 1A we found that wild-type S.
Typhi induced a decrease in TEER in the monolayer in a dose-
dependent manner. All inoculation ratios used (MOIs of 40:1,
400:1, and 4000:1 bacteria:cell ratios, respectively) caused a sig-
nificant drop in TEER [from 1147.0 ± 21.2 .cm
2
at baseline to
804.1 ± 37.5 .cm
2
, 318.7 ± 38.2 .cm
2
, and 252.6 ± 8.1 .cm
2
,
www.frontiersin.org February 2013 | Volume 4 | Article 17 | 3
Page 3
Fiorentino et al. Epithelial cell responses to S. Typhi
FIGURE 1 | Trans-epithelial electrical resistance (TEER) responses of
CaCo2 cell monolayers to wild-type S.Typhi, HK bacteria, and culture
supernatants. (A) TEER changes upon infection with wild-type S. Typhi at
different MOIs. (B) TEER in HK bacteria and culture supernatants (Cond.
Media) treated monolayers. Data are expressed as means ± SEM for
triplicate samples for all conditions tested. These results are representative
of three experiments with similar results. #Denotes p < 0.001 of S. Typhi
strains over t = 0 (ANOVA).
respectively] as early as 2 h post-infection. The decrease in TEER
was even more dramatic at 4 h post-infection with TEER values
of 527.6 ± 34.9 .cm
2
, 206 ± 15.5 .cm
2
, and 170.1 ± 1.6 .cm
2
,
respectively compared to baseline (1185.2 ± 40.0 .cm
2
). At 4 h,
bacteria were removed and monolayers were treated with gentam-
icin in order to eliminate non-adherent bacteria and incubated
in medium overnight at 37˚C. At 22 h post-infection, TEER val-
ues were still low for the monolayers infected at higher titers
(298.8 ± 17.5 .cm
2
and 146.3 ± 7.4 .cm
2
for MOIs of 400:1
and 4000:1, respectively). In contrast, at a bacterial MOI of 40:1
we observed a substantial recovery in TEER, although still signif-
icantly lower than baseline (1235.3 ± 71.4 .cm
2
compared to a
baseline value of 1487 ± 75.0 .cm
2
) (Figure 1A). Taken together,
these data demonstrate that S. Typhi alters Caco2 monolayer
barrier function in a dose-dependent manner. Interestingly, the
removal of bacteria allows the cells to slowly recover and counter-
act the adverse effect caused by the pathogen. This process appar-
ently starts earlier for lower bacterial loads, presumably because
the damage to the barrier function is less severe.
We then investigated whether the effects seen with wild-type S.
Typhi on Caco2 cell TEER required the interaction of viable bac-
teria with the target enterocyte in order to elicit the host response.
We thus compared wild-type with HK and filtered wild-type bac-
teria (CM). As shown in Figure 1B both the HK bacteria and the
CM from S. Typhi, applied at the same titer as the wild-type strain,
failed to significantly decrease the TEER of Caco2 cell monolayer.
These results demonstrate that the effects of S. Typhi on barrier
function depend on direct interaction of viable bacteria with target
enterocytes and are not mediated by secretion of toxins or other
mediators.
AMELIORATION OF EPITHELIAL BARRIER CHANGES EXHIBITED BY
ATTENUATED S. TYPHI VACCINE STRAINS
To determine the effect of attenuated mutant strains of S. Typhi
on mucosal permeability in vitro, Caco2 cell monolayers were
infected with S. Typhi Ty21a, CVD 908-htrA, and CVD 909 atten-
uated strains (Figure 2). Wild-type S. Typhi was used as a positive
control. Monolayers reached confluence in about 2 weeks, with a
baseline TEER between 1200 and 1700 .cm
2
(t = 0). As described
in our first series of experiments, wild-type S. Typhi induced
a significant decline in TEER at all MOIs as early as 2 h post-
infection (Figure 2). In contrast, both CVD 908-htrA and CVD
909 mutant strains failed to induce TEER changes at their low-
est infection titers (MOI 40:1; Figure 2A). At a MOI of 400:1
(Figure 2B) we observed a decrease in TEER as early as 2 h post-
infection that became more pronounced at 4 h when we registered
a drop to 842 ± 43.5 .cm
2
and 586.75 ± 13.3 .cm
2
for CVD
909 and CVD 908-htrA, respectively compared to baseline values
(1185.2 ± 40.0 .cm
2
). Interestingly, 22 h after exposure to these
attenuated strains we observed a recovery in TEER values,although
the difference with the uninfected monolayer was still significant
(Figure 2B).
Of note, at the highest bacterial load (MOI 4000:1) the resulting
effects of the two S. Typhi vaccine candidates became similar to
TEER decrease caused by the wild-type strain. At 22 h the TEER for
all strains was very low, reaching its nadir at 293.5 ± 57.4 .cm
2
for the attenuated CVD 909 strain (Figure 2C). The only statis-
tically significant difference we observed with the wild-type was
for CVD 909 at 2 h post-infection. The vaccine strain Ty21a was
tested at MOIs of 4000 and 400 in parallel with wild-type S. Typhi,
as positive control. As shown in Figure 2D, at a MOI of 4000:1
Ty21a induced a drop in TEER (283.3 ± 37.1 .cm
2
) similar to
the wild-type (139.5 ± 5.3 .cm
2
) at 4 h post-infection. At 22 h,
we observed a slight recovery of TEER values for the mutant
infected cells (440.1 ± 114.9 .cm
2
) suggesting a less disruptive
effect of this strain on Caco2 monolayer barrier function than
wild-type S. Typhi (126.7 ± 5.1 .cm
2
). At a bacterial infection
load of 400:1 the effects of Ty21a on TEER were again similar
to those observed with the wild-type strain with values at 22 h
of 349.9 ± 56.3 .cm
2
(p = ns) compared to 194.7 ± 15.3 .cm
2
with wild-type Salmonella (data not shown).
These data demonstrate that the new generation mutant strains
tested, although capable of affecting the integrity of the epithelial
cell monolayers at high inoculation ratios, have a considerably
milder effect on the intestinal barrier integrity than wild-type S.
Typhi at lower inoculation ratios. Ty21a effect on Caco2 cell TEER
was overall similar to that of wild-type Salmonella.
EFFECTS OF S. TYPHI INFECTION ON CELLULAR VIABILITY
In order to evaluate if the disruption of the epithelial barrier func-
tion caused by S. Typhi is mediated by enhanced cell death, the
viability of the monolayer was assessed by measuring the levels of
LDH released in the cell medium at 22 h post-infection. As shown
Frontiers in Immunology | Mucosal Immunity February 2013 | Volume 4 | Article 17 | 4
Page 4
Fiorentino et al. Epithelial cell responses to S. Typhi
FIGURE 2 | The effect of S.Typhi attenuated strain on theTEER of
polarized Caco2 monolayers. Wild-type S. Typhi served as control. (A)
Aro mutants-infected monolayers (MOI of 40:1); (B) Aro mutants-
infected monolayers (MOI of 400:1). (C) Aro mutants-infected
monolayers (MOI of 4000:1). (D) TEER in Caco2 cells infected with
Ty21a applied apically at a MOI of 4000:1. Data are expressed as
means ± SEM for triplicate samples for all conditions tested. These
results are representative of three experiments with similar results.
Statistical comparisons over wild-type S. Typhi at the same time point;
*p < 0.05 (ANOVA).
in Figure 3A, all strains induced a significantly higher release of
LDH compared to uninfected cells, at the highest MOI (4000:1).
At a MOI of 400:1 only wild-type S. Typhi induced a significant
higher release of LDH. No differences with the uninfected cells
were observed at a MOI of 40:1 for any of the strains studied.
Similarly, as shown in Figure 3B, a significantly higher level of
LDH was observed only for the wild-type Ty2 S. Typhi at a MOI
of 4000:1 compared to uninfected controls. These results were
confirmed by staining with Propidium Iodide and flow cytometry
(data not shown). The range of dead cells varied between 4.3% in
the uninfected monolayers and 13.7% in cultures with wild-type
S. Typhi at a MOI of 4000.
These data show that, although all strains can affect cell via-
bility when applied at high doses, the overall monolayer viability
is mostly preserved, with the greatest cytotoxicity level observed
being less than 20% of the positive control. Taken together with
the data discussed above showing that at a bacterial load of 40:1
there is a recovery of TEER over time and even at a MOI of 400:1
we observed the complete recovery of TEER after 3 days (data not
shown), these data suggest that the decrease observed is likely to
be largely due to modulation of cellular permeability, with only a
minor component attributable to cell loss or toxicity resulting in
cell death.
S. TYPHI INFECTION INCREASES PARACELLULAR FLUX IN CACO2 CELL
MONOLAYERS
We next sought to determine whether alterations induced by S.
Typhi with regard to TEER correlated with changes in epithelial
paracellular permeability. We evaluated alterations in barrier func-
tion in response to the pathogen, by measuring the trans-epithelial
flux of fluorescently labeled Dextran (FITC-dextran, 4 kDa) and
BSA (FITC-BSA, 40 kDa). As shown in Figure 4, Caco2 cells
exposed to wild-type S. Typhi showed a significantly increased
transport of both FITC-BSA (Figure 4A) and FITC-Dextran
(Figures 4B,C) from the apical chamber to the basolateral side
compared to uninfected monolayers. None of the mutant strains
caused increased paracellular permeability to these markers.
(Figures 4A–C). At a MOI 40:1, wild-type S. Typhi failed to
increase the paracellular flux of labeled markers (Figure 5). EGTA
was used as a maximum permeability positive control for its abil-
ity to completely open tight junctions. As expected, EGTA-treated
monolayers showed a remarkably higher degree of transport of
both tracers confirming the validity of the assays. Although we
can’t exclude that to some extent the paracellular flux of labeled
markers might be due to the minimal cell death we observed
when infecting monolayers with wild-type S. Typhi, these results
likely represent a S. Typhi-mediated increase in epithelial per-
meability resulting from modulation of intercellular tight junc-
tions function. In light of this, the mutant vaccine strains appear
to have either an attenuated or no effect on the tight junction
barrier.
S. TYPHI AFFECTS PARACELLULAR PERMEABILITY BY ALTERING
TIGHT-JUNCTION PROTEINS DISTRIBUTION AND SCAFFOLDING IN
POLARIZED CELLULAR MONOLAYERS
We have determined that S. Typhi affects epithelial barrier integrity
largely by increasing Caco2 monolayer paracellular permeabil-
ity. The maintenance of TEER and the relative impermeability
of polarized epithelium to macromolecules require the correct
functioning and integrity of intercellular tight junctions and their
www.frontiersin.org February 2013 | Volume 4 | Article 17 | 5
Page 5
Fiorentino et al. Epithelial cell responses to S. Typhi
FIGURE 3 | Effect of S. Typhi infection on cell viability. LDH levels were
measured in the apical media as an index of cell death. Values represent
LDH release as a percentage of total LDH (positive control, Triton X-100).
(A) LDH levels after infection with wild-type S. Typhi and its aro mutants at
different MOI’s. (B) LDH release after infection with wild-type S. Typhi and
Ty21a applied at a MOI of 4000:1. Data are expressed as means ± SEM for
triplicate samples for all conditions tested. These results are representative
of three experiments with similar results. ***p< 0.001, **p < 0.01,
*p < 0.05 compared to uninfected cells (ANOVA).
association with cytoskeletal proteins at the apico-lateral cell sur-
face (Balda and Matter, 1998). In order to investigate the possible
disruption of the tight junctional complex by the pathogen, we
examined the tight junction-associated protein ZO-1 in infected
monolayers by immunofluorescence microscopy.
As shown in Figure 6A, in uninfected monolayers ZO-1 is
localized at the cell-cell boundary in a typical chicken wire-likepat-
tern throughout the monolayer, indicating intact tight-junction
complexes. In contrast, following infection with wild-type S. Typhi
for 4 h (Figure 6B; MOI 400:1), ZO-1 distribution appears altered:
while still observing the protein at the cell boundaries, ZO-1 also
appeared clustered into the cytoplasm and co-localized with aggre-
gates of actin fibers of the disrupted cytoskeleton (Figures 6D,F).
In contrast, as expected, in uninfected monolayers the actin
cytoskeleton was organized in a network of filaments normally dis-
tributed beneath the plasma membrane and throughout the cyto-
plasm (Figures 6C,E). As a plausible effect of this tight-junction
and cytoskeletal rearrangement we observed the detachment of
adjacent cells from each other (Figure 6B, arrows). Similar to ZO-
1, in infected monolayers we observed the clustering of claudin-
1into the cytoplasm around the nucleus and the detachment of
adjacent cells (data not shown). Consistent with TEER data, in
monolayers infected with CVD 909 we did not observe major
alterations of ZO-1 localization. As shown in Figures 6G,H, ZO-1
is localized at the cell-cell boundary with minimal internalization
and also minimal alteration of the actin cytoskeleton. At 22 h post-
infection, we still observed a severely altered distribution of ZO-1
FIGURE 4 | Paracellular permeability in Caco2 cell monolayers after
infection with wild-type S. Typhi and its mutant strains. (A) FITC-BSA
(40 kDa) net transport after infection with wild-type S. Typhi at a MOI of
400:1. (B) FITC-Dextran (4 kDa) net transport after infection with wild-type
S. Typhi or the attenuated aro mutants CVD 908-htrA and CVD 909, applied
at a MOI of 400:1. (C) Permeability to FITC-Dextran (4 kDa) after infection
with wild-type S. Typhi and the vaccine strain Ty21a at a MOI of 4000:1.
Calcium-free medium supplemented with EGTA to disrupt TJs served as
positive control. Results are expressed as mean ± SEM of triplicate samples
for each condition and are representative of three experiments with similar
results. ***p < 0.001 compared to uninfected cultures (ANOVA).
in wild-type S. Typhi treated cells (Figure 7D) whereas the effect
appears largely attenuated in Ty21a infected cells, for which only a
few areas of cell–cell detachment and some ZO-1 internalization
were detected (Figure 7C, arrows). No effect on the distribution of
ZO-1 was observed with CVD 909 (Figure 7B) or CVD 908 (not
shown). These results strongly support the notion that changes in
epithelial permeability induced by bacterial infection are caused
by alterations in the paracellular pathway due to disruption and/or
modulation of the sealing function of tight junctions.
PHOSPHORYLATION OF OCCLUDIN PLAYS A ROLE IN THE REGULATION
OF TIGHT JUNCTION FUNCTION UPON S. TYPHI INFECTION
Occludin has been widely described to play an important role in
the regulation of tight-junction integrity (Rao, 2009). It has been
shown that its overexpression results in TEER elevation (Balda
et al., 1996; McCarthy et al., 1996). We have analyzed the sol-
ubility of occludin in the non-ionic detergent Triton X-100 as
an indicator of its association with the tight-junction complex,
Frontiers in Immunology | Mucosal Immunity February 2013 | Volume 4 | Article 17 | 6
Page 6
Fiorentino et al. Epithelial cell responses to S. Typhi
FIGURE 5 | Paracellular permeability in Caco2 cell monolayers after
infection with wild-type S. Typhi and its mutant strains (MOI 40:1).
(A) FITC-BSA (40 kDa) net transport after infection with wild-type S. Typhi
and attenuated strains. (B) FITC-Dextran (4 kDa) net transport after infection
with wild-type S. Typhi or the attenuated mutants. Calcium-free medium
supplemented with EGTA to disrupt TJs served as positive control. Results
are expressed as mean ± SEM of triplicate samples for each condition and
are representative of three experiments with similar results. p = ns
compared to uninfected controls (ANOVA).
in uninfected and infected monolayers in order to determine its
implication in S. Typhi induced increased permeability of Caco2
monolayers.
After infection, equal protein amounts of cell lysates from the
Triton X-100 soluble and insoluble fractions were resolved by SDS-
PAGE and then analyzed by immunoblotting in parallel with unin-
fected controls (Figure 8). In the uninfected monolayer occludin
was mainly localized in the insoluble fraction and is visible on west-
ern blots as a strong band of about 65 kDa [Low Molecular Weight
(LMW); Figure 8A, upper panel]. Upon infection with wild-type
S. Typhi, an additional band of about 72–79 kDa was detected in
the insoluble membrane fraction at 4 h post-infection. This High
Molecular Weight (HMW) band was also observed in both CVD
mutants-treated cell lysates, although weaker than that observed
with the wild-type S. Typhi (Figure 8A, upper panel; Figure 8B).
At 22 h, we observed an increase in LMW occludin in the soluble
fraction paralleled by a decrease of the HMW species in all bac-
teria samples (Figure 8D, upper panel; Figure 8E). The HMW
species (72–79 kDa) has been previously shown to represent a
hyperphosphorylated form of occludin and represents a sub-pool
of this protein specifically associated with the functional seal-
ing components of tight-junction (Sakakibara et al., 1997; Wong,
1997; Seth et al., 2007). Specifically it has been demonstrated that
occludin undergoes dephosphorylation on Ser/Thr residues dur-
ing the disruption of tight junctions by various insults. Analysis
of the threonine phosphorylation status of our samples showed
that occludin appears to be phosphorylated on Thr in the rest-
ing epithelium (uninfected controls, LMW band, Figures 8A,D,
middle panels). Conversely, in 4 h wild-type S. Typhi-infected
cell lysates, only the HMW band appeared phosphorylated and
this represents the hyperphosphorylated form of occludin. We
observed a similar, albeit milder, shift of the occludin phospho-
rylation status toward the HMW species in both CVD 909 and
CVD 908-htrA S. Typhi mutants (Figure 8C). Analysis of lysates
at 22 h post-infection revealed that S. Typhi induced both a loss
of occludin and hyperphosphorylated occludin from the insoluble
fraction and a shift of the signal to the soluble fraction (Figure 8D,
middle panel; Figure 8F) in all Salmonella strain-treated samples.
As in the uninfected control, the phosphorylation status of the
65 kDa band of the insoluble membrane fraction was apparently
not affected.
Taken together, these data clearly show that wild-type S. Typhi
induces changes in the phosphorylation status of occludin and this
is likely to result in the disruption of tight junctions.
S. TYPHI INTERACTION WITH EPITHELIAL CELLS TRIGGERS THE
RELEASE OF PRO-INFLAMMATORY CYTOKINES
Cytokines play a central role in regulating immune and
inflammatory responses during infection with pathogens. Pro-
inflammatory cytokines are crucial components of the host
response to pathogenic microbes; they are produced early in infec-
tion and contribute to various steps of the host inflammatory and
immune response. A few studies have shown that S. Typhi induces
the release of IL-8 and/or IL-6 in epithelial cells (Weinstein et al.,
1997, 1998; Sharma and Qadri, 2004; Raffatellu et al., 2005; Win-
ter et al., 2008) however, no studies to date have evaluated the
immune response elicited by the interaction of attenuated strains
of S. Typhi with the host epithelial cells. Thus, we analyzed the
early immune response of intestinal epithelial cells to infection
with the aro mutants CVD 908-htrA and CVD 909 and the licensed
vaccine Ty21a.
Supernatants from both the monolayer apical and basolateral
compartments were evaluated by Luminex assay for the detection
of the pro-inflammatory cytokines IL-1β, IL-8, IL-6, TNF-α, IL-17,
www.frontiersin.org February 2013 | Volume 4 | Article 17 | 7
Page 7
Fiorentino et al. Epithelial cell responses to S. Typhi
FIGURE 6 | Fluorescence microscopy of Caco2 monolayers labeled
with ZO-1 or fluorescein-phalloidin (actin) show disruption of the
tight-junction complex and the cytoskeleton at 4 h post-infection.
Caco2 polarized monolayers were infected with wild-type S. Typhi (MOI
400:1) for 4 h, washed with PBS, fixed and stained along with
uninfected controls. (A) ZO-1 staining in uninfected monolayers. Note
the characteristic chicken wire patterning. (B) Disrupted
ZO-1organization in wild-type S. Typhi -infected monolayers. Areas of
cell-cell detachment are marked by arrows. (C) Actin staining of the
cytoskeleton in uninfected controls. (D) Actin fibers stained in the S.
Typhi -infected cells. (E) Merge of (A,C). (F) Merge of (B,D). (G) ZO-1
distribution in CVD 909 infected monolayers. (H) ZO-1 (red), actin
(green) and DAPI for the nuclei (blu) merged staining of CVD 909
infected cells. Bar, 25 µm.
Frontiers in Immunology | Mucosal Immunity February 2013 | Volume 4 | Article 17 | 8
Page 8
Fiorentino et al. Epithelial cell responses to S. Typhi
FIGURE 7 | ZO-1 staining of S. Typhi and mutant strains infected monolayers at 22 h post-infection. (A) Uninfected control. (B) CVD 909. (C) Ty21a. (D)
Wild-type S. Typhi. ZO-1 staining (red), DAPI for the nuclei (blu). Areas of severe ZO-1 disruption are marked by arrows. Bar, 30 µm.
and IFN-γ . Moreover, we assayed supernatants for the measure-
ment of IL-10 and TGF-β. Media collected from Ty21a infected
cells were assessed for levels of IL-8 and IL-6 only. Cytokines were
measured at 4 and 22 h post-infection. The only cytokines detected
in these studies were IL-8 and IL-6. Levels of all other cytokines
were very low or below detection. Figure 9A shows the results
for IL-8 in media collected from the apical and basolateral sides
of uninfected, live wild-type S. Typhi as positive control, filtrate,
heat-inactivated S. Typhi, and S. Typhi aro mutants-infected Caco2
monolayers at 22 h post-infection. As expected, we detected a sig-
nificant cytokine secretion in the basolateral compartment but
interestingly a remarkable release of IL-8 was also measured on
the apical side. IL-8 secretion induced by wild-type S. Typhi was
significantly higher than that observed in uninfected cells on both
sides and at all bacterial loads applied except for CVD909, apical
secretion (Figure 9A). The largest IL-8 secretion was measured at
a MOI of 400:1 for which we detected amounts of 2239 ± 573 and
838 ± 197 pg/ml compared to the uninfected monolayer, on the
basolateral and apical sides, respectively. Even at the lowest MOI
of 40:1 the fold increase was highly significant compared to unin-
fected, both on the basolateral (918 ± 163) and apical (340 ± 585;
Figure 9B) side. IL-8 secretion in the basolateral side induced
by filtered and heat-inactivated wild-type bacteria (Figure 9A) is
remarkably lower than live, wild-type S. Typhi, suggesting that a
physical interaction with the live pathogen is needed to trigger a
strong cytokine response. S. Typhi aro mutants elicit remarkable
levels of cytokine secretion, although the overall IL-8 amounts
are lower than those secreted by wild-type bacteria-infected cells,
in both the apical and basolateral compartments (Figures 9A,B).
Similar to the wild-type strain, we measured the highest secre-
tion of IL-8 at a MOI of 400:1 CFU/cell for CVD 908-htrA:
336.5 ± 66.5 and 832 ± 175 pg/ml in the apical and basolateral
sides, respectively. IL-8 secretion elicited by the mutant strain CVD
909 was higher than untreated cells at all MOIs, with the greatest
difference being observed at the MOI of 4000:1, with IL-8 lev-
els 210 ± 40.1 and 814 ± 179 pg/ml compared to 7.49 ± 2.03 and
15.0 ± 2.94 pg/ml of uninfected cells, in the apical and basolateral
compartments, respectively.
Although wild-type S. Typhi induces more IL-8 secretion than
the mutant strains added at the same titers, data show that the
differences are consistently statistically significant for the cytokine
release in the apical compartment (Figures 9A,B). IL-8 amounts
in the basolateral side were significantly different from wild-
type strain only for both CVD 908-htrA and CVD 909-treated
monolayers at the MOIs of 400:1 and for CVD 909 at 40:1.
Between the two aro mutants, data show no significant differ-
ences at any concentration. We observed a slight although not
significant increase in IL-8 secretion upon treatment of cells with
filtered and heat-inactivated bacteria, compared to uninfected
monolayer (Figure 9A). Interestingly, IL-8 levels measured at 4 h
www.frontiersin.org February 2013 | Volume 4 | Article 17 | 9
Page 9
Fiorentino et al. Epithelial cell responses to S. Typhi
FIGURE 8 | Occludin is hyperphosphorylated on threonine and
translocates into the cytoplasm. We analyzed the solubility of occludin
as an indication of its association with tight junctions. Protein lysates in the
form of X-100 Triton soluble (S) and insoluble (I) fractions were obtained
from Caco2 cell monolayers after infection with wild-type S. Typhi, CVD
908-htrA, and CVD 909 for 4 and 22 h. Equal amounts were loaded on the
gel, electrophoresed, transferred onto a membrane and blotted with
antibodies. (A) Western blot of protein samples blotted with anti-occludin
(upper panel), anti-phosphothreonine (middle panel), and anti-actin (lower
panel) as loading control, 4 h following exposure to bacteria. (B,C)
Quantification of the data shown in (A). (D) Western blot of protein
samples blotted with anti-occludin (upper panel), anti-phosphothreonine
(middle panel) and anti-actin (lower panel) 22 h post-infection with S. Typhi
strains. (E,F). Quantification of the data shown in (D). Data have been
normalized to the actin loading control. These results represent the
average of two individual experiments.
post-infection, were significantly higher in the apical side than the
basolateral for both wild-type S. Typhi and CVD 908-htrA applied
at a MOI of 400:1 (Figure 9C).
IL-8 secreted by Caco2 monolayers upon infection with Ty21a
strain (MOI 400:1), was 50 and 100-fold greater than the unin-
fected control in the apical and basolateral sides, respectively
(Figure 9D). Ty21a elicited a reduced IL-8 release compared to
wild-type S. Typhi, with a significant difference in the basal com-
partment. At a higher MOI, significant difference was observed
between wild-type and Ty21a only for IL-8 amounts released api-
cally. Of note, in our in vitro system we did not observe any
substantial difference in the ability of Ty21a to elicit IL-8 secretion
when Ty21a was grown in the absence or presence of 0.05% galac-
tose, a concentration which has been previously shown to allow
complete LPS O-antigen synthesis without reducing the strain
growth rate (Shi et al., 2010). As shown in Figure 10, IL-6 was
secreted by all strains and at all conditions applied. IL-6 levels
measured apically were significantly higher than uninfected cells
for most of the strains and conditions except CM, CVD 908-htrA
at a MOI of 40:1 and CVD 909 at MOIs of 40:1 and 400:1. On the
basolateral surface IL-6 secretion levels appeared to be statistically
higher than uninfected controls for wild-type S. Typhi and CVD
908-htrA applied at a MOI of 400:1 and S. Typhi applied at a MOI
of 40:1. At the highest bacterial loads (MOIs 4000:1 and 400:1)
we did not observe any significant difference in the level of IL-6
secretion between the wild-type strain and either vaccine candi-
date in both compartments; conversely, both CVD 908-htrA and
CVD 909 applied at a MOI of 40:1 induced a basolateral secretion
of IL-6 significantly lower than wild-type Salmonella. No statis-
tical differences were observed between CVD 908-htrA and CVD
909 strains.
Taken together, these data suggest that S. Typhi induces a sig-
nificant immune response by epithelial cells through the release
of the pro-inflammatory cytokines IL-8 and IL-6. Of importance,
these results provide additional information regarding the attenu-
ation of the S. Typhi mutant strains CVD 908-htrA, CVD 909 and
Ty21a as evidenced by their induction of measurable, but reduced
responses compared to the wild-type strain, particularly at lower
titers.
DISCUSSION
Bacterial enteric infections exact a heavy toll on human popula-
tion, particularly among children. Despite the explosion of knowl-
edge on the pathogenesis of enteric diseases experienced during
the past decade, the number of diarrheal episodes and human
deaths reported worldwide remains remarkably high. Typhoid
fever is a serious, life-threatening disease widespread in devel-
oping countries, affecting about 21.5 million persons each year.
About 400 individuals in the United States each year, 75% of which
get infected during international travels develop typhoid fever. In
addition to risk of death or persistent infection, other potential
complications of S. Typhi infection include toxemia, myocarditis,
liver damage and intestinal lesions. Due to increased resistance to
antibiotics, the development of effective and safe vaccines against
Frontiers in Immunology | Mucosal Immunity February 2013 | Volume 4 | Article 17 | 10
Page 10
Fiorentino et al. Epithelial cell responses to S. Typhi
FIGURE 9 | S. Typhi attenuated strains induce IL-8 secretion by Caco2
cells. (A) IL-8 secreted by Caco2 cells infected with vaccine candidates CVD
908-htrA and CVD 909 applied at MOIs of 4000:1 and 400:1, HK S. Typhi and
conditioned media (22 h post-infection). (B) IL-8 released by Caco2 cells
infected with the aro mutants at a MOI of 40:1at 22 h post-infection; (C) IL-8
secretion measured at 4 h post-infection at a MOI of 400:1 (statistical analysis
between the apical and basolateral compartments for the same strain); (D) IL-8
measured at 22 h after infection with the vaccine strain Ty21a applied at MOIs
of 4000 and 400. Wild-type S. Typhi served as positive control. Values shown
represent the mean ± SEM of three independent assays. #p < 0.05 over
uninfected; ***p < 0.001, **p < 0.01, *p < 0.05 for comparisons between
wild-type S. Typhi and mutant strains applied at the same titer (ANOVA).
S. Typhi infection is a public health priority. The scientific effort to
study the cross talk between enteric pathogens and intestinal host
has been mainly focused on studying bacterial pathogenesis and
how microorganisms can cause diarrhea. In this manuscript, we
believe for the first time, we have shifted our attention from how
bacteria can induce diarrhea and inflammation to the integrated
response of the intestinal epithelial milieu following exposure to
S. Typhi.
The mammalian gastrointestinal tract is lined by a dynamic
epithelium exquisitely responsive to stimuli of innumerable vari-
ety, and is populated by a complex community of microbial
partners, far more numerous than the cells of the intestine itself.
In normal homeostasis, the GI epithelial layer forms a tight, but
selective barrier: microbes and most antigens are held at bay, but
nutrients from the essential to the trivial are absorbed efficiently.
This selectivity is based on the capability of intestinal epithelial
cells to perceive the luminal presence of possible danger signals.”
The interaction between the epithelium and the luminal microbial
population is mainly based on modulation of intestinal permeabil-
ity and intestinal mucosal defenses. The tightness of the epithelial
barrier is itself dynamic, though the mechanisms governing and
effecting dynamic permeability are only partially understood.
Our data demonstrate that exposure to S. Typhi causes a time-
and dose-dependent impairment of the gut epithelial barrier func-
tion without overt damage of the intestinal monolayer. The doses
of bacteria we used in this study range from 4 to 600 times lower
than a single dose of the Ty21a typhoid vaccine. Although these
bacterial amounts applied to the single monolayer on a mem-
brane insert area of 1.1 cm
2
might appear too high compared
to a single dose of the vaccine, we have to consider that physi-
ologically in the intestinal tract the distribution of the ingested
vaccine won’t be homogeneous. There are sections of the GI tract,
including the duodenum and the jejunum that will be exposed
to a higher bacterial load than the ileum. Moreover, the num-
bers of bacteria present in particular GI segments are likely to
be affected by many factors, including the preference of S. Typhi
to colonize the small intestine, bacterial replication in the local
microenvironment, numbers of typhoid bacilli present in acute or
chronic infection, etc. Our study shows that at a lower S. Typhi
MOI (40:1), Caco2 monolayers have an almost complete barrier
function recovery (Figure 1A), while at higher concentrations
there is a slight (MOI 400:1) or absent (MOI 4000:1) recovery
of monolayer’s tightness at 22 h post-infection. After removing
the bacteria, the monolayer starts slowly to recover and counteract
www.frontiersin.org February 2013 | Volume 4 | Article 17 | 11
Page 11
Fiorentino et al. Epithelial cell responses to S. Typhi
FIGURE 10 | S. Typhi attenuated strains induce secretion of IL-6.
IL-6 secreted by Caco2 cells infected with vaccine candidates CVD
908-htrA and CVD 909 applied at different bacterial loads, conditioned
media, and the wild-type strain, as positive control. Values shown
represent the mean ± SEM of three independent assays. #p < 0.05
over uninfected; ***p < 0.001, **p < 0.01, *p < 0.05 for comparisons
between wild-type S. Typhi and mutant strains applied at the same
titer (ANOVA).
the adverse effect caused by the pathogen. This process apparently
begins earlier for lower bacterial loads, presumably because the
damage to the barrier function is less severe. We have shown that
although occurring, cell death is low and limited mostly to the
higher bacteria infection ratios. Based on the results presented, we
believe that cell death accounts only marginally for the increased
permeability. In support of this, we were able to provide the mol-
ecular bases for the changes in barrier integrity by showing that
intercellular tight junctions of Caco2 monolayers exposed to wild-
type S. Typhi were disassembled (Figures 68) and by demon-
strating that TJ disassembly was due to S. Typhi induced early
(4 h) hyperphosphorylated occludin (Figures 8A–C), followed by
a late (22 h) shift of the occludin signal to the soluble fraction
(Figures 8D–F). These changes in occludin phosphorylation and
localization were paralleled by similar departure of the scaf-
fold proteins ZO-1 (Figures 6B and 7D) and claudin-1 (not
shown) from cell boundaries. These data suggest that exposure
of epithelial cells to S. Typhi causes phosphorylation of the trans-
membrane TJ component occludin, followed by translocation of
occludin, ZO-1 and claudin-1 from cell boundary to cytoplasm,
leading to increased paracellular permeability as demonstrated
by the increased passage of both paracellular markers dextran
and BSA from the apical to the basolateral monolayer compart-
ments. Interestingly, S. Typhi vaccine candidates CVD 908-htrA
and CVD 909 caused either no changes in TEER at lower doses
(MOI 40:1) or less dramatic changes at higher doses compared
to wild-type Salmonella (see Figure 2), paralleled by a decreased
passage of paracellular markers dextran and BSA (Figures 4
and 5). Also it was interesting to note that the licensed and
currently clinically used S. Typhi Ty21a typhoid vaccine caused
TEER changes similar, although not as severe, as those observed
with wild-type S. Typhi (Figure 2D). This was further sup-
ported by immunofluorescence staining of ZO-1 at 22 h, showing
an attenuated effect on tight-junction barrier compared to the
wild-type strain, although more severe than the CVD mutants
(Figure 7C).
Ty21a is an attenuated mutant strain of S. Typhi Ty2, iso-
lated in the early 1970s by random chemical mutagenesis. It has
a prominent GalE- and Vi-negative phenotype (Germanier and
Fuer, 1975) that, associated with further spontaneous mutations
(Germanier and Fuer, 1975; Coynault et al., 1996), resulted in a
vaccine exhibiting remarkable safety and good efficacy (Levine
et al., 1999). The Ty21a is the active constituent of Vivotif® (Berna
Biotech Ltd., Switzerland), currently the only licensed live oral vac-
cine against typhoid fever. Because of the multiple doses needed
to achieve protection, novel attenuated S. Typhi strains that may
serve as single dose oral attenuated vaccines have been developed.
CVD 908-htrA and CVD 909 are mutant strains carrying muta-
tions in the aromatic amino acid synthesis pathway. In particular,
deletions in the aro genes (Ty2 aroC and aroD) make these strains
Frontiers in Immunology | Mucosal Immunity February 2013 | Volume 4 | Article 17 | 12
Page 12
Fiorentino et al. Epithelial cell responses to S. Typhi
auxotrophic for some aromatic compounds. The mutant bacte-
ria become attenuated because they are unable to scavenge these
molecules in mammalian cells. A further deletion of the htrA gene
impairs their response to stress and the ability to survive inside
macrophages (Lowe et al., 1999). CVD 909 constitutively expresses
the Vi antigen. Its invasiveness might be reduced compared to its
isogenic parent, CVD 908-htrA and the wild-type since it has been
shown in vitro that the expression of the Vi capsule is negatively
associated with the expression of the genes regulating invasion
such as flagellin and T3SS-1 (Arricau et al., 1998). It has been
shown that Vi expression in S. Typhi is repressed in high osmo-
larity environments (such as the intestinal lumen) and viceversa
is induced in conditions of low osmolarity [i.e., blood and tissue
(Arricau et al., 1998)]. We have grown all our strains in LB and/or
DMEM media, mimicking the physiological environment of the
intestinal lumen and our invasion assays confirm that CVD 909
is less invasive than its isogenic mutant CVD 908-htrA (data not
shown), most likely due to the constitutive expression of the Vi
antigen. While to a lesser extent than CVD 909, the CVD 908-htrA
and Ty21a mutants showed a less invasive phenotype as compared
to the wild-type, as well (data not shown), reflecting their overall
attenuated properties. Although it might be argued that our results
are affected by decreased bacterial growth and/or invasion of the
mutant strains as compared to the wild-type during infection, it
is important to emphasize that the aim of our study was to eval-
uate the outcome of the host-bacteria interaction at the mucosal
level and hence the effects of these mutants and their attenuated
properties on the intestinal epithelial barrier in as close to phys-
iological conditions as possible. Future studies will be directed
to evaluate the molecular determinants of the remarkable effects
on barrier function and cytokine production described in this
manuscript.
Overall we were able to demonstrate that S. Typhi impairment
of the epithelial barrier function occurs by modulating the seal-
ing function of tight junctions with the consequent increase of
the monolayer paracellular permeability. The mutant strains show
either an attenuated or no effect on tight junction disassembly
and paracellular permeability compared to the more aggressive
phenotype detected in cells exposed to the wild-type strain. It is
reasonable to hypothesize that the observed differences between
attenuated and wild-type S. Typhi strains might be essential for the
lack of reactogenicity and remarkable immunogenicity observed
when these vaccine strains were fed to volunteers in Phase 1 and
2 clinical trials (Tacket et al., 2000b, 2004; Levine et al., 2001;
Salerno-Goncalves et al., 2003, 2004; Sztein, 2007; Wahid et al.,
2007, 2008, 2011, 2012; McArthur and Sztein, 2012). Another
aspect of importance in intestinal mucosal defense is the capa-
bility of intestinal epithelial cells to “prime” the gut associated
lymphoid tissue to possible danger caused by enteric pathogens by
releasing pro-inflammatory cytokines and chemokines, including
IL-8. We have demonstrated that Caco2 monolayers exposure to
luminal S. Typhi secrete IL-8 in a polarized manner, with higher
release in the basolateral side than in the luminal side at 22 h post-
infection. Both HK bacteria and CM caused the release of much
lower IL-8 levels, suggesting that live bacteria are more efficient
in triggering danger signals to be released by enterocytes, e.g., the
release of large amounts of IL-8, which in turn causes neutrophil
recruitment into the lamina propria to properly face a possible
enteric infection. S. Typhi attenuated vaccines triggered a milder
IL-8 release, in keeping with their milder effects seen regarding
barrier function. Noteworthy is the apical secretion of both IL-8
and IL-6.
Several studies have described the polarized secretion of these
acute phase pro-inflammatory molecules (Zeillemaker et al., 1995;
Carolan et al., 1997; Nasreen et al., 2001; Fahey et al., 2005; Sun
et al., 2008). Their release on the “luminal” side can have signifi-
cant implications. Epithelial cells represent the first site of contact
bacteria have with the host and this interaction is likely to trig-
ger a series of early events that ultimately will prevent harmful
microorganisms from invading and damaging the host. The apical
secretion of IL-8 and IL-6 may create a chemotactic gradient and
facilitate neutrophil trans-epithelial migration to the luminal side
where neutrophils can constitute a defense barrier and perform
their essential function of eliminating invading microorganisms.
This possibility is supported by our data showing that in the first
few hours following infection IL-8 release is significantly polar-
ized toward the apical side (Figure 9C). The basolateral release of
cytokines at later times will result in the recruitment of immune
cells to the lamina propria, thus playing a role in supporting and
amplifying the early epithelial inflammatory response. In explor-
ing these interesting possibilities in future studies we should take
into account that expression of the Vi antigen might affect the
secretion of IL-8. It has been shown that deletion of the genes
associated with the regulation, biosynthesis and export of the Vi-
capsular antigen (viaB locus) increases IL-8 expression elicited by
S. Typhi in HEK293 cell line (Raffatellu et al., 2005). Similarly,
infection of Caco2 cells with Vi
+
S. Typhi produced significantly
lower levels of IL-8 as compared with Vi
S. Typhi (Sharma and
Qadri, 2004). Thus, the constitutive expression of the Vi antigen
might have played a role in our results with the mutant CVD 909.
Future studies will address this possibility.
Overall, a better understanding of the interactions between
enteric bacteria and intestinal epithelial cells is the basis for the
development and improving of preventive interventions. Our
results show that the intestinal mucosa is more than just a merely
physical barrier against infection. Intestinal epithelial cells operate
as an active extension of our innate immune system, performing
a surveillance function that can specifically identify enemies and
activate an offensive response to block infection.
In conclusion, the application of a multidisciplinary approach
to study bacterial pathogenesis, along with the recent sequenc-
ing of entire microbial genomes have made possible discoveries
that are changing the way scientists view bacterium-host inter-
actions. Currently, research on the molecular basis of the patho-
genesis of infective enteric diseases of necessity transcends estab-
lished boundaries between microbiology, cell biology, intestinal
pathophysiology, and immunology. Our contribution outline the
need to integrate studies on bacterial pathogenesis with the host
response in order to better understand the clinical outcome of
bacterial enteric diseases and develop proper preventive inter-
ventions, including the development of attenuated, protective
vaccines.
www.frontiersin.org February 2013 | Volume 4 | Article 17 | 13
Page 13
Fiorentino et al. Epithelial cell responses to S. Typhi
ACKNOWLEDGMENTS
These studies were funded by NIAID, NIH, DHHS grants R01-
AI036525 and U19 AI082655 (Cooperative Center for Transla-
tional Research in Human Immunology and Biodefense; CCHI)
to Marcelo B. Sztein and Alessio Fasano. The content is solely the
responsibility of the authors and does not necessarily represent the
official views of the National Institute of Allergy and Infectious
Diseases or the National Institutes of Health.
REFERENCES
Aktories, K. (1997). Bacterial toxins that
target Rho proteins. J. Clin. Invest.
99, 827–829.
Andreeva,A. Y., Krause, E., Muller, E. C.,
Blasig, I. E., and Utepbergenov, D.
I. (2001). Protein kinase C regulates
the phosphorylation and cellular
localization of occludin. J. Biol.
Chem. 276, 38480–38486.
Arricau, N., Hermant, D., Waxin, H.,
Ecobichon, C., Duffey, P. S., and
Popoff, M. Y. (1998). The RcsB-
RcsC regulatory system of Salmo-
nella Typhi differentially modulates
the expression of invasion proteins,
flagellin and Vi antigen in response
to osmolarity. Mol. Microbiol. 29,
835–850.
Balda, M. S., and Matter, K. (1998).
Tight junctions. J. Cell. Sci. 111(Pt.
5), 541–547.
Balda, M. S., Whitney, J. A., Flo-
res, C., Gonzalez, S., Cereijido,
M., and Matter, K. (1996). Func-
tional dissociation of paracellu-
lar permeability and transepithelial
electrical resistance and disruption
of the apical-basolateral intramem-
brane diffusion barrier by expres-
sion of a mutant tight junction
membrane protein. J. Cell Biol. 134,
1031–1049.
Bertelsen, L. S., Paesold, G., Mar-
cus, S. L., Finlay, B. B., Eck-
mann, L., and Barrett, K. E. (2004).
Modulation of chloride secretory
responses and barrier function of
intestinal epithelial cells by the
Salmonella effector protein SigD.
Am. J. Physiol. Cell. Physiol. 287,
C939–C948.
Carolan, E. J., Mower, D. A., and Casale,
T. B. (1997). Cytokine-induced neu-
trophil transepithelial migration is
dependent upon epithelial orienta-
tion. Am. J. Respir. Cell Mol. Biol. 17,
727–732.
Chen, Y. H., Lu, Q., Goodenough, D.
A., and Jeansonne, B. (2002). Non-
receptor tyrosine kinase c-Yes inter-
acts with occludin during tight junc-
tion formation in canine kidney
epithelial cells. Mol. Biol. Cell 13,
1227–1237.
Clark, M. A., Hirst, B. H., and Jep-
son,M. A. (1998). Inoculum compo-
sition and Salmonella pathogenic-
ity island 1 regulate M-cell invasion
and epithelial destruction by Salmo-
nella typhimurium. Infect. Immun.
66, 724–731.
Coynault, C., Robbe-Saule, V., and
Norel, F. (1996). Virulence and
vaccine potential of Salmonella
typhimurium mutants deficient in
the expression of the RpoS (sigma
S) regulon. Mol. Microbiol. 22,
149–160.
Crump, J. A., Luby, S. P., and Mintz,
E. D. (2004). The global burden of
typhoid fever. Bull. World Health
Organ. 82, 346–353.
Deng, W., Liou, S. R., Plunkett, G. III,
Mayhew, G. F., Rose, D. J., Bur-
land, V., et al. (2003). Compara-
tive genomics of Salmonella enterica
serovar Typhi strains Ty2 and CT18.
J. Bacteriol. 185, 2330–2337.
Diebel, L. N., Liberati, D. M., Baylor,
A. E., III, Brown, W. J., and Diglio,
C. A. (2005). The pivotal role of
tumor necrosis factor-alpha in sig-
naling apoptosis in intestinal epithe-
lial cells under shock conditions. J.
Trauma 58, 995–1001.
Fahey, J. V., Schaefer, T. M., Chan-
non, J. Y., and Wira, C. R.
(2005). Secretion of cytokines and
chemokines by polarized human
epithelial cells from the female
reproductive tract. Hum. Reprod. 20,
1439–1446.
Fasano, A., Fiorentini, C., Donelli, G.,
Uzzau, S., Kaper, J. B., Margaretten,
K., et al. (1995). Zonula occludens
toxin modulates tight junctions
through proteinkinase C-dependent
actin reorganization, in vitro. J. Clin.
Invest. 96, 710–720.
Finlay, B. B., and Falkow, S. (1990).
Salmonella interactions with polar-
ized human intestinal Caco-2
epithelial cells. J. Infect. Dis. 162,
1096–1106.
Finlay, B. B., Gumbiner, B., and Falkow,
S. (1988). Penetration of Salmo-
nella through a polarized Madin-
Darby canine kidney epithelial cell
monolayer. J. Cell Biol. 107, 221–230.
Fraser, A., Goldberg, E., Acosta, C. J.,
Paul, M., and Leibovici, L. (2007).
Vaccines for preventing typhoid
fever. Cochrane Database Syst. Rev.
3, CD001261.
Germanier, R., and Fuer, E. (1975). Iso-
lation and characterization of Gal E
mutant Ty 21a of Salmonella typhi:
a candidate strain for a live, oral
typhoid vaccine. J. Infect. Dis. 131,
553–558.
Guzman, C. A., Borsutzky, S., Griot-
Wenk, M., Metcalfe, I. C., Pearman,
J.,Collioud,A.,et al. (2006). Vaccines
against typhoid fever. Vaccine 24,
3804–3811.
Hone, D. M., Harris, A. M., Chatfield,
S., Dougan, G., and Levine, M. M.
(1991). Construction of genetically
defined double aro mutants of Sal-
monella typhi. Vaccine 9, 810–816.
Hone,D. M., Tacket,C. O., Harris, A. M.,
Kay, B., Losonsky, G., and Levine, M.
M. (1992). Evaluation in volunteers
of a candidate live oral attenuated
Salmonella typhi vector vaccine. J.
Clin. Invest. 90, 412–420.
Jepson, M. A., Schlecht, H. B.,
and Collares-Buzato, C. B. (2000).
Localization of dysfunctional tight
junctions in Salmonella enter-
ica serovar typhimurium-infected
epithelial layers. Infect. Immun. 68,
7202–7208.
Jones, B. D., Ghori, N., and Falkow,
S. (1994). Salmonella typhimurium
initiates murine infection by pen-
etrating and destroying the spe-
cialized epithelial M cells of the
Peyer’s patches. J. Exp. Med. 180,
15–23.
Kohler, H., Sakaguchi, T., Hurley, B. P.,
Kase, B. A., Reinecker, H. C., and
McCormick, B. A. (2007). Salmo-
nella enterica serovar Typhimurium
regulates intercellular junction pro-
teins and facilitates transepithelial
neutrophil and bacterial passage.
Am. J. Physiol. Gastrointest. Liver
Physiol. 293, G178–G187.
Levine, M. M., Ferreccio, C., Abrego, P.,
Martin, O. S., Ortiz, E., and Cryz,
S. (1999). Duration of efficacy of
Ty21a, attenuated Salmonella typhi
live oral vaccine. Vaccine 17(Suppl.
2), S22–S27.
Levine, M. M., Ferreccio, C., Black, R.
E., Lagos, R., San Martin, O., and
Blackwelder, W. C. (2007). Ty21a live
oral typhoid vaccine and preven-
tion of paratyphoid fever caused by
Salmonella enterica Serovar Paraty-
phi B. Clin. Infect. Dis. 45(Suppl. 1),
S24–S28.
Levine, M. M., Tacket, C. O., and Sztein,
M. B. (2001). Host-Salmonella inter-
action: human trials. Microbes Infect.
3, 1271–1279.
Lowe, D. C., Savidge, T. C., Pickard,
D., Eckmann, L., Kagnoff, M. F.,
Dougan, G., et al. (1999). Charac-
terization of candidate live oral Sal-
monella typhi vaccine strains har-
boring defined mutations in aroA,
aroC, and htrA. Infect. Immun. 67,
700–707.
McArthur, M. A., and Sztein, M. B.
(2012). Heterogeneity of multifunc-
tional IL-17A producing S. Typhi-
specific CD8+ T cells in vol-
unteers following Ty21a typhoid
immunization. PLoS ONE 7:e38408.
doi:10.1371/journal.pone.0038408
McCarthy, K. M., Skare, I. B.,
Stankewich, M. C., Furuse, M.,
Tsukita, S., Rogers, R. A., et al.
(1996). Occludin is a functional
component of the tight junction. J.
Cell. Sci. 109(Pt. 9), 2287–2298.
Nasreen, N., Mohammed, K. A., Hard-
wick, J.,Van Horn, R. D., Sanders, K.
L., Doerschuk, C. M., et al. (2001).
Polar production of interleukin-8
by mesothelial cells promotes the
transmesothelial migration of neu-
trophils: role of intercellular adhe-
sion molecule-1. J. Infect. Dis. 183,
1638–1645.
Nusrat, A., Parkos, C. A., Verkade,
P., Foley, C. S., Liang, T. W.,
Innis-Whitehouse, W., et al. (2000).
Tight junctions are membrane
microdomains. J. Cell. Sci. 113(Pt.
10), 1771–1781.
Nusrat, A., von Eichel-Streiber, C.,
Turner, J. R., Verkade, P., Madara,
J. L., and Parkos, C. A. (2001).
Clostridium difficile toxins disrupt
epithelial barrier function by alter-
ing membrane microdomain local-
ization of tight junction proteins.
Infect. Immun. 69, 1329–1336.
Otte, J. M., and Podolsky, D. K. (2004).
Functional modulation of entero-
cytes by gram-positive and gram-
negative microorganisms. Am. J.
Physiol. Gastrointest. Liver Physiol.
286, G613–G626.
Pothoulakis, C. (2000). Effects of
Clostridium difficile toxins on
epithelial cell barrier. Ann. N. Y.
Acad. Sci. 915, 347–356.
Raffatellu, M., Chessa, D., Wilson, R. P.,
Dusold, R., Rubino, S., and Baumler,
A. J. (2005). The Vi capsular anti-
gen of Salmonella enterica serotype
Typhi reduces Toll-like receptor-
dependent interleukin-8 expression
in the intestinal mucosa. Infect.
Immun. 73, 3367–3374.
Rao, R. (2009). Occludin phosphoryla-
tion in regulation of epithelial tight
junctions. Ann. N. Y. Acad. Sci. 1165,
62–68.
Riccardi, C., and Nicoletti, I. (2006).
Analysis of apoptosis by propidium
iodide staining, and flow cytometry.
Nat. Protoc. 1, 1458–1461.
Frontiers in Immunology | Mucosal Immunity February 2013 | Volume 4 | Article 17 | 14
Page 14
Fiorentino et al. Epithelial cell responses to S. Typhi
Sakakibara, A., Furuse, M., Saitou,
M., Ando-Akatsuka, Y., and Tsukita,
S. (1997). Possible involvement of
phosphorylation of occludin in tight
junction formation. J. Cell Biol. 137,
1393–1401.
Salerno-Goncalves, R., Fernandez-
Vina, M., Lewinsohn, D. M., and
Sztein, M. B. (2004). Identification
of a human HLA-E-restricted
CD8+ T cell subset in volunteers
immunized with Salmonella enter-
ica serovar Typhi strain Ty21a
typhoid vaccine. J. Immunol. 173,
5852–5862.
Salerno-Goncalves, R., Wyant, T. L.,
Pasetti, M. F., Fernandez-Vina, M.,
Tacket, C. O., Levine, M. M., et al.
(2003). Concomitant induction of
CD4+ and CD8+ T cell responses
in volunteers immunized with Sal-
monella enterica serovar typhi strain
CVD 908-htrA. J. Immunol. 170,
2734–2741.
Seth, A., Sheth, P., Elias, B. C., and
Rao,R. (2007). Protein phosphatases
2A and 1 interact with occludin
and negatively regulate the assem-
bly of tight junctions in the CACO-2
cell monolayer. J. Biol. Chem. 282,
11487–11498.
Sharma, A., and Qadri, A. (2004).
Vi polysaccharide of Salmonella
typhi targets the prohibitin fam-
ily of molecules in intestinal
epithelial cells and suppresses
early inflammatory responses.
Proc. Natl. Acad. Sci. U.S.A. 101,
17492–17497.
Shi, H., Santander, J., Brenneman, K.
E., Wanda, S. Y., Wang, S., Senechal,
P., et al. (2010). Live recombi-
nant Salmonella typhi vaccines con-
structed to investigate the role of
rpoS in eliciting immunity to a
heterologous antigen. PLoS ONE
5:e11142. doi:10.1371/journal.pone.
0011142
Simonovic, I., Rosenberg, J., Kout-
souris, A., and Hecht, G. (2000).
Enteropathogenic Escherichia coli
dephosphorylates and dissociates
occludin from intestinal epithelial
tight junctions. Cell. Microbiol. 2,
305–315.
Solano, C., Sesma, B., Alvarez, M.,
Urdaneta, E., Garcia-Ros, D., Calvo,
A., et al. (2001). Virulent strains
of Salmonella enteritidis disrupt the
epithelial barrier of Caco-2 and
HEp-2 cells. Arch. Microbiol. 175,
46–51.
Sun, Y., Wu, F., Sun, F., and Huang,
P. (2008). Adenosine promotes IL-
6 release in airway epithelia. J.
Immunol. 180, 4173–4181.
Sztein, M. B. (2007). Cell-mediated
immunity and antibody responses
elicited by attenuated Salmonella
enterica Serovar Typhi strains used
as live oral vaccines in humans.
Clin. Infect. Dis. 45(Suppl. 1),
S15–S19.
Tacket, C. O., Galen, J., Sztein, M. B.,
Losonsky, G., Wyant, T. L., Nataro,
J., et al. (2000a). Safety and immune
responses to attenuated Salmonella
enterica serovar typhi oral live vec-
tor vaccines expressing tetanus toxin
fragment C. Clin. Immunol. 97,
146–153.
Tacket, C. O., Sztein, M. B., Wasserman,
S. S., Losonsky, G., Kotloff, K. L.,
Wyant, T. L., et al. (2000b). Phase
2 clinical trial of attenuated Sal-
monella enterica serovar typhi oral
live vector vaccine CVD 908-htrA in
U.S. volunteers. Infect. Immun. 68,
1196–1201.
Tacket, C. O., Pasetti, M. F., Sztein,
M. B., Livio, S., and Levine, M. M.
(2004). Immune responses to an oral
typhoid vaccine strain that is modi-
fied to constitutively express Vi cap-
sular polysaccharide. J. Infect. Dis.
190, 565–570.
Wahid, R., Pasetti, M. F., Maciel, M. Jr.,
Simon, J. K., Tacket,C. O., Levine, M.
M., et al. (2011). Oral priming with
Salmonella typhi vaccine strain CVD
909 followed by parenteral boost
with the S. Typhi Vi capsular poly-
saccharide vaccine induces CD27+
IgD S. Typhi-specific IgA and IgG
B memory cells in humans. Clin.
Immunol. 138, 187–200.
Wahid, R., Salerno-Goncalves, R.,
Tacket, C. O., Levine, M. M., and
Sztein, M. B. (2007). Cell-mediated
immune responses in humans
after immunization with one or
two doses of oral live attenuated
typhoid vaccine CVD 909. Vaccine
25, 1416–1425.
Wahid, R., Salerno-Goncalves, R.,
Tacket, C. O., Levine, M. M., and
Sztein, M. B. (2008). Generation of
specific effector and memory T cells
with gut- and secondary lymphoid
tissue-homing potential by oral
attenuated CVD 909 typhoid vac-
cine in humans. Mucosal Immunol.
1, 389–398.
Wahid, R., Simon, R., Zafar, S. J.,
Levine, M. M., and Sztein, M.
B. (2012). Live oral typhoid vac-
cine Ty21a induces cross-reactive
humoral immune responses against
Salmonella enterica serovar Paraty-
phi A and S. Paratyphi B in
humans. Clin. Vaccine Immunol. 19,
825–834.
Weinstein, D. L., O’Neill, B. L., Hone, D.
M., and Metcalf, E. S. (1998). Dif-
ferential early interactions between
Salmonella enterica serovar Typhi
and two other pathogenic Sal-
monella serovars with intestinal
epithelial cells. Infect. Immun. 66,
2310–2318.
Weinstein, D. L., O’Neill, B. L., and
Metcalf, E. S. (1997). Salmonella
typhi stimulation of human intesti-
nal epithelial cells induces secretion
of epithelial cell-derived interleukin-
6. Infect. Immun. 65, 395–404.
Winter, S. E., Raffatellu, M., Wilson,
R. P., Russmann, H., and Baumler,
A. J. (2008). The Salmonella enter-
ica serotype Typhi regulator TviA
reduces interleukin-8 production in
intestinal epithelial cells by repress-
ing flagellin secretion. Cell. Micro-
biol. 10, 247–261.
Wong, V. (1997). Phosphorylation of
occludin correlates with occludin
localization and function at the tight
junction. Am. J. Physiol. 273, C1859–
C1867.
Wu, Z., Nybom, P., Sundqvist, T.,
and Magnusson, K. E. (1998).
Endogenous nitric oxide in MDCK-I
cells modulates the Vibrio cholerae
haemagglutinin/protease (HA/P)-
mediated cytotoxicity. Microb.
Pathog. 24, 321–326.
Youakim, A., and Ahdieh, M. (1999).
Interferon-gamma decreases barrier
function in T84 cells by reducing
ZO-1 levels and disrupting apical
actin. Am. J. Physiol. 276, G1279–
G1288.
Zeillemaker, A. M., Mul, F. P., Hoynck
van Papendrecht, A. A., Kuijpers, T.
W., Roos, D., Leguit, P., et al. (1995).
Polarized secretion of interleukin-8
by human mesothelial cells: a role in
neutrophil migration. Immunology
84, 227–232.
Conflict of Interest Statement: Drs. M.
M. Levine and M. B. Sztein are co-
inventors in a patent for the develop-
ment of S. Typhoid and S. Paratyphoid
vaccines licensed to Bharat Biotech
International Limited. The remaining
authors declare no conflicting financial
interests.
Received: 17 October 2012; accepted: 09
January 2013; published online: 12 Feb-
ruary 2013.
Citation: Fiorentino M, Lammers KM,
Levine MM, Sztein MB and Fasano
A (2013) In vitro intestinal mucosal
epithelial responses to wild-type Sal-
monella Typhi and attenuated typhoid
vaccines. Front. Immun. 4:17. doi:
10.3389/fimmu.2013.00017
This article was submitted to Frontiers in
Mucosal Immunity, a specialty of Fron-
tiers in Immunology.
Copyright © 2013 Fiorentino, Lammers,
Levine, Sztein and Fasano. This is an
open-access article distributed under the
terms of the Creative Commons Attribu-
tion License, which permits use, distrib-
ution and reproduction in other forums,
provided the original authors and source
are credited and subject to any copy-
right notices concerning any third-party
graphics etc.
www.frontiersin.org February 2013 | Volume 4 | Article 17 | 15
Page 15
  • Source
    • "After evaluation of different combination of S1K3 and Salmonella in vitro, the pathogenic strain disrupted the monolayer integrity while the probiotic maintained it, even in competition with Salmonella (Fig. 2A–2F). Previous in vitro studies investigating mechanisms of Salmonella traversal have been done mostly using epithelial cell monolayers grown on permeable filters, for example, Caco-2 cells[6,28,48,60,68], and some have concluded that an alteration of epithelial tight junctions is involved. Tight junctions contain integral membrane proteins (occludin, claudins, JAM, CAR and TAMPs) and peripheral membrane proteins that directly interact with the cytoskeleton[25,36,62]. "
    [Show abstract] [Hide abstract] ABSTRACT: We investigated the mechanism by which an Indian indigenous probiotic culture, Lactobacillus rhamnosus S1K3, could overcome the pathogenic strain Salmonella enterica with an emphasis on the response at the intestinal mucosal level after long-term (30 days) consumption. S1K3 was able to produce antimicrobial compounds against the pathogens. The probiotic adhered strongly to intestinal epithelium and maintained its integrity in presence of Salmonella through stimulation of tight junction and antimicrobial peptide genes in vitro. Mice pre-fed for 30 days with S1K3-fermented milk exhibited low incidence of pathogenic Salmonella at mucosal and systemic levels. The probiotic induced TLRs transcripts at the Peyer's patches, followed by an increase in the Secretory-IgA in intestinal fluid, the IgA-secreting cells in lamina propria of small intestine and the IgA level in serum. Moreover, S1K3 maintained the protein level of IL-12, increased the IL-4 and reduced the TGF-β level in intestinal fluid/serum at the later stage of infection. All these actions concurred to lower the count of Salmonella in feces, its invasion in spleen, liver and intestine tissues and improved the health status of probiotic fed group. In view of this performance, S1K3 appears to be a suitable candidate for the development of nutraceutical food.
    Full-text · Article · Jan 2016 · The Journal of Nutritional Biochemistry
  • Source
    • "Interestingly, and unlike our observations with Salmonella Typhi [40], we detected an increase in TEER in the initial few hours of infection (6 hours), when the monolayers were treated with bacteria at 105 and 106 CFU/monolayer. The greatest increase for both Shigella strains was observed when a 105 bacterial load was applied, for which we recorded TEER values of 1121±59.9 "
    [Show abstract] [Hide abstract] ABSTRACT: Bacterial dysentery due to Shigella species is a major cause of morbidity and mortality worldwide. The pathogenesis of Shigella is based on the bacteria's ability to invade and replicate within the colonic epithelium, resulting in severe intestinal inflammatory response and epithelial destruction. Although the mechanisms of pathogenesis of Shigella in the colon have been extensively studied, little is known on the effect of wild-type Shigella on the small intestine and the role of the host response in the development of the disease. Moreover, to the best of our knowledge no studies have described the effects of apically administered Shigella flexneri 2a and S. dysenteriae 1 vaccine strains on human small intestinal enterocytes. The aim of this study was to assess the coordinated functional and immunological human epithelial responses evoked by strains of Shigella and candidate vaccines on small intestinal enterocytes. To model the interactions of Shigella with the intestinal mucosa, we apically exposed monolayers of human intestinal Caco2 cells to increasing bacterial inocula. We monitored changes in paracellular permeability, examined the organization of tight-junctions and the pro-inflammatory response of epithelial cells. Shigella infection of Caco2 monolayers caused severe mucosal damage, apparent as a drastic increase in paracellular permeability and disruption of tight junctions at the cell-cell boundary. Secretion of pro-inflammatory IL-8 was independent of epithelial barrier dysfunction. Shigella vaccine strains elicited a pro-inflammatory response without affecting the intestinal barrier integrity. Our data show that wild-type Shigella infection causes a severe alteration of the barrier function of a small intestinal cell monolayer (a proxy for mucosa) and might contribute (along with enterotoxins) to the induction of watery diarrhea. Diarrhea may be a mechanism by which the host attempts to eliminate harmful bacteria and transport them from the small to the large intestine where they invade colonocytes inducing a strong inflammatory response.
    Full-text · Article · Jan 2014 · PLoS ONE