| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 278, Issue 35, 32552-32560, August 29, 2003
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
B Activation and Chemokine Expression by Epithelial Cells*
¶
From the
University of Virginia Health System, Departments of Medicine,
Digestive Health Center of Excellence and
Microbiology, Charlottesville, Virginia
22908-0708
Received for publication, May 27, 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
B activation and the
induction of chemokine mRNA expression. By reverse transcriptase-PCR we
determined that MKN45 gastric epithelial cells express low but detectable
amounts of TLR2, -4, and -5 but no MD-2. To determine which, if any, TLRs may
play a role in the response of epithelial cells to H. pylori, HEK293
cells were cotransfected with the NF-B-Luc reporter, CD14 and MD2
expression plasmids, and expression plasmids for TLR2, TLR4, or TLR5.
Infection of the cultures with H. pylori (strain 26695) induced
NF-B activity in cells transfected with TLR2 and TLR5, but not TLR4.
Consistent with the HEK293 experiments, H. pylori-induced NF-B
activation was decreased in MKN45 gastric epithelial cells by transfection of
dominant-negative versions of TLR2 and TLR5 but not TLR4. Highly purified
lipopolysaccharide from H. pylori strain 26695 activated NF-B
in HEK293 via TLR2 but not TLR4. Partially purified flagellin from H.
pylori was also capable of inducing NF-B activation in HEK cells
transfected with TLR5. Additionally, chemokine gene expression was induced by
H. pylori in HEK293 cells following stable transfection with TLR2 or
TLR5 expression plasmids. These studies demonstrate that gastric epithelial
cells recognize and respond to H. pylori infection at least in part
via TLR2 and TLR5. Furthermore, the unique lipopolysaccharide of H.
pylori is a TLR2, not a TLR4 agonist. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
Gastric epithelial cells respond to infection with H. pylori by
activating numerous signal transduction cascades. To date, the responses that
have been best characterized are those which lead to expression and activation
of c-fos and c-jun via mitogen-activated protein kinase pathways
(1) and NF-B
(2–4).
One result of the activation of these pathways is the production of large
amounts of IL-81 by
the gastric epithelial cells
(2,
4,
5).
The mechanisms involved in the regulation of IL-8 production by H.
pylori-infected gastric epithelial cells have been extensively studied.
Aihara et al. (4) have
demonstrated that AP-1 and NF-B cooperate to regulate IL-8
gene expression in MKN45 cells. Mutation of both the AP-1 and
NF-B-binding sites in the IL-8 promoter resulted in a decrease in the
response to H. pylori infection. Mutation of the AP-1 site alone
resulted in a crippled promoter that was only partially responsive to H.
pylori. In contrast, mutation of the NF-B site resulted in a
promoter that was completely nonresponsive to H. pylori infection
demonstrating the critical importance of NF-B in the response. Maeda
et al. (6) further
demonstrated that H. pylori-induced NF-B activation in gastric
epithelial cells involved IB kinases, NIK, TRAF2, and TRAF6.
Cells of the innate immune system sense and respond to microbial products
via the Toll-like receptor family. Toll-like receptors (TLR) are an
evolutionarily conserved family of cell surface molecules that participate in
innate immune recognition of pathogen-associated molecular patterns
(7). Pathogen-associated
molecular patterns are generally unique, chemically diverse products with
conserved motifs that are produced by microorganisms. Pathogen-associated
molecular patterns often have an essential role in the structure of bacteria
and generally cannot be subtly modified as a result of mutation without
adversely affecting microbial growth or pathogencity. Examples include LPS
(specifically lipid A), peptidoglycan, lipoproteins, bacterial DNA, and
bacterial flagella. At least 10 different TLRs have been identified. In some
cases the bacterial ligand has also been identified. For example, TLR2
recognizes peptidoglycan (8)
and bacterial lipoproteins (9).
TLR4 recognizes LPS from most Gram-negative species
(10), TLR5 reacts with
flagellin (11), and TLR9 is a
receptor for bacterial CpG DNA
(12). The signaling events
occurring downstream of the TLRs are rapidly being elucidated and appear to
have many common features. In general, the cascade of events occurring
following ligation of the different TLRs involves the activation of a common
set of adapter proteins and protein kinases, the best characterized of which
leads to the activation of NF-B (reviewed in Refs.
13 and
14). All of the molecules
identified by Maeda et al.
(6) as being involved in H.
pylori-induced NF-B activation are also associated with responses
downstream of the Toll-like receptors. The identity of the cell surface
signaling molecule(s) responsible for the H. pylori-induced responses
is still unclear.
One potential signaling receptor for H. pylori that has been
examined in the literature is TLR4, originally identified as the signaling
molecule responsible for cellular responses to LPS
(10,
16). Recently, Kawahara et
al. (17) demonstrated
that H. pylori LPS stimulates TLR4 and induces mitogen oxidase 1 gene
expression in guinea pig gastric pit cells. However, other investigators have
provided evidence that H. pylori LPS does not stimulate human gastric
epithelial cell lines (3,
4). To better clarify the role
of Toll-like receptors in the responses of epithelial cells to H.
pylori infection, we examined the activation of NF-B in HEK293
cells transfected with specific TLR expression constructs and in the human
gastric cancer cell line MKN45 expressing dominant negative versions of TLR2,
TLR4, or TLR5. Our results indicated that live H. pylori induced
NF-B activation because of ligation of TLR2 and TLR5, but not TLR4. In
addition, we have determined that LPS derived from H. pylori strain
26695 is a TLR2 agonist and demonstrate that the inability of MKN45 cells to
respond to typical TLR4 agonists is because of a lack of the TLR4 accessory
molecule MD2.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
Plasmid Constructs—Human TLR2, TLR4, and MD-2 expression
plasmids were previously described
(19). The TLR5 cDNA
corresponding to the entire coding region was generated by reverse
transcriptase (RT)-PCR using primers corresponding to the published sequences
and cloned into pcDNA3.1Zeo (Invitrogen). Dominant-negative versions of TLR2,
TLR4, and TLR5 were created by generating truncated cDNAs that resulted in
deletion of the carboxyl-terminal portion of the molecule (TLR2 truncated at
amino acid 670, TLR4 at 700, and TLR5 at 685). Sequences of all clones were
confirmed by automated sequencing. The human CD14 expression vector,
pRc/RSVCD14, was a gift of R. Ulevitch (Scripps Research Institute, La Jolla,
CA); NF-B-Luc was from Clontech. All plasmid DNAs were isolated by
using endotoxin-free preparation kits from Qiagen (Valencia, CA).
Cell Culture and Transfection—HEK293 and AGS cells were from
ATCC, MKN45 cells were purchased from the JCRB Cell Bank (Japan), and CHO-CD14
cells were a gift of D. Golenbock (University of Massachusetts). HEK293 cells
were cultured in Dulbecco's modified Eagle's medium + 10% FBS (Hyclone), AGS
cells in Ham's F-12 + 10% FBS and MKN45 in RPMI 1640 + 10% FBS. All cells were
transfected in 24-well plates using LipofectAMINE (Invitrogen). Each
transfection contained 500 ng of NF-B-Luc (Clontech), 50 ng of TLR
expression vector or pcDNA3.1Zeo, 50 ng of pEF6-MD2, 100 ng of pRc/RSVCD14,
and 200 ng of pTK-Renilla (Promega) and 4 µl of LipofectAMINE.
Transfections were performed in triplicate, cultured for 24 h, and then
stimulated as indicated. Luciferase activities were determined using the dual
luciferase kit from Promega and all activities were normalized to the activity
of the cotransfected TK-Renilla plasmid.
RT-PCR of TLR mRNA—Total RNA was purified from cell lines using Trizol reagent (Invitrogen). Reverse transcription (RT) of 0.5 µg of total cellular RNA was performed in a final volume of 20 µl containing 5x first strand buffer (Invitrogen), 1 mM of each dNTP, 20 units of placental RNase inhibitor, 5 µM random hexamer, and 9 units of Moloney murine leukemia virus reverse transcriptase (Invitrogen). After incubation at 37 °C for 45 min, the samples were heated for 5 min at 92 °C to end the reaction and stored at –20 °C until PCR use. PCR was performed for 30 cycles under standard conditions using 2 µl of cDNA. Primers were generated such that the 5' and 3' primers were in separate exons. Sequences of primers used are shown in Table I.
|
Primary Culture of H. pylori—Gastric antrum specimens were taken from two patients recruited at the endoscopy unit at the University of Texas Medical Branch, Galveston, TX, according to the protocols outlined in a study approved by their Institutional Review Board. The specimens were placed in sterile normal saline and transported on ice. Homogenized tissues were cultured for <7 days at 37 °C on blood agar in an incubator containing 10% CO2. H. pylori growth was confirmed by Gram stain, and oxidase, catalase, and urease reactivity. The two isolates used in these studies were termed Hp98 and Hp99. H. pylori strain LC11 was a gift from P. Sherman (Hospital for Sick Children, Toronto, Ontario, Canada) (5) and strain 26695 was from K. Eaton (Ohio State University, Columbus, OH).
H. pylori in Vitro Infection—All H. pylori strains were grown for 24 h on Remel sheep blood agar plates, in 10% CO2 at 37 °C. One plate of bacteria was harvested with a sterile cotton swab and resuspended in 1 ml of Brucella broth. The bacteria were pelleted in a microcentrifuge at 6000 rpm for 5 min and resuspended in 1 ml of Brucella broth. This wash procedure was repeated and the bacteria were thoroughly resuspended in 1 ml of Brucella broth. Thirty microliters of the H. pylori suspension were plated to verify the multiplicity of infection. Twenty-four hours after transfection, the media was removed from the HEK293 or MKN45 cells and was replaced with fresh Dulbecco's modified Eagle's medium with 10% FBS or RPMI with 10% FBS, respectively. H. pylori then were added to the eukaryotic cells at a multiplicity of infection of 200–300:1. The infection proceeded for 24 h at 37 °C and 10% CO2.
H. pylori LPS Extraction—H. pylori strain 26695 was grown on
Remel sheep blood agar at 37 °C, in 10% CO2 for 72 h. The
confluent bacterial growth from 900 plates was harvested from each plate with
1 ml of sterile phosphate-buffered saline. The biomass was pooled and washed
twice with phosphate-buffered saline and subsequently freezedried, yielding
5 g of dried biomass. The LPS was extracted using the hot phenol-water
protocol of Moran et al.
(20), followed by protease
treatment. All equipment was treated with EtoxaClean to remove all endotoxin
contaminants. Briefly, the biomass was resuspended in endotoxin-free water at
a concentration of 1 g/ml and homogenized for 10 min at high speed in a
Sorvall Omni mixer. An equal volume of liquified phenol was added to the
bacterial suspension and heated at 65 °C for 1 h. The solution was then
blended for regular 2-min intervals, for a total of 10 min, at high speed in a
Sorvall Omni mixer, and then cooled on ice. The mixture was centrifuged at
5,000 x g at 4 °C for 15 min, and the aqueous layer was
collected. An equal volume of endotoxin-free water was added to the remaining
phenol layer, mixed, and the centrifugation and collection steps were repeated
for a total of 3 aqueous layer collections. The aqueous phases were combined
and dialyzed against 6 liters of endotoxin-free water for 72 h to remove all
residual phenol and freeze-dried. The crude LPS was resuspended at 10 mg/ml in
0.1 M Tris-HCl buffer, pH 7.5, containing 1 mg/10 ml of DNase II
and RNase A, and 0.1 mg/10 ml of proteinase K. The suspension was incubated at
37 °C for 16 h, and then at 65 °C for an additional 16 h. The
enzyme-treated LPS was dialyzed again against 6 liters of endotoxin-free water
over 72 h to remove degraded compounds. The resulting solution was
freeze-dried for 48 h. Prior to use, LPS was subjected to an additional
purification procedure as described by Hirschfeld et al.
(18).
H. pylori Flagellin Preparation—H. pylori 26695 were grown on Remel sheep blood agar at 37 °C, in 10% CO2 for 72 h. The bacteria were harvested from each plate with 1 ml of sterile Brucella broth. Bacteria from 100 plates were pooled and were blended for 1 min at high speed in a Sorvall Omni mixer. Bacterial cells were removed from solution by centrifugation at 15,000 x g at 4 °C for 25 min. The supernatant containing the sheared H. pylori flagella was ultracentrifuged at 100,000 x g at 4 °C for 90 min. The supernatant was discarded and the flagellar pellet was resuspended in 1 ml of sterile phosphate-buffered saline. Total protein concentration of the flagellin preparation was determined by the Bradford method using Pierce BCA reagents and protocol.
Quantitative RT-PCR—Total RNA was purified using the Trizol reagent (Invitrogen). RT of 0.5 µg of total cellular RNA was performed in a final volume of 20 µl containing 5x first strand buffer (Invitrogen), 1 mM of each dNTP, 20 units of placental RNase inhibitor, 5 µM random hexamer, and 9 units of Moloney murine leukemia virus reverse transcriptase (Invitrogen). After incubation at 37 °C for 45 min, the samples were heated for 5 min at 92 °C to end the reaction and stored at –20 °C until PCR use. Two µl of cDNA was subjected to real-time quantitative PCR using the Opticon system (MJ Research) with SYBR Green I (Molecular Probes, Eugene OR) as a fluorescent reporter. Chemokine and hypoxanthine phosphoribosyltransferase cDNAs were amplified in separate reactions. Threshold cycle number, of duplicate reactions, was determined using the Opticon software and levels of chemokine mRNA expression were normalized to hypoxanthine phsophoribosyltransferase levels using the formula 2(Rt–Et), where Rt is the mean threshold cycle for the reference gene (hypoxanthine phosphoribosyltransferase) and Et is the mean threshold cycle for the experimental gene. Data are thus expressed as arbitrary units. Primer sequences are provided in Table I.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
|
We next determined the capacity of MKN45 cells to activate NF-B in
response to TLR ligands. MKN45 cells were transiently transfected with a
NF-B/luciferase reporter plasmid and abilities of the synthetic
lipopetide PAM3CSK4 (a TLR2 agonist), highly purified
E. coli LPS (a TLR4 agonist), or live H. pylori, to induce
luciferase expression were determined. As shown in the transfection experiment
in Fig. 2A,
PAM3CSK4 and H. pylori were capable of
activating NF-B in MKN45 cells, whereas E. coli LPS could not.
Similar results were seen using another TLR2 agonist, peptidoglycan (data not
shown). Thus, MKN45 cells are in fact responsive to at least some TLR
agonists.
|
To determine whether the inability of MKN45 cells to respond to E.
coli LPS was because of the lack of MD2 expression, the experiment in
Fig. 2B was performed.
MKN45 cells were cotransfected with expression plasmids for CD14 and MD2 (or
empty vector control) along with the NF-B reporter. E. coli
LPS was capable of activating NF-B but only when the MD2 expression
plasmid was cotransfected. These results indicate that MKN45 cells do in fact
express a functional TLR4 protein but because of the lack of MD2 expression
they are incapable of responding to LPS. This then provides a potential
explanation for the lack of effect of H. pylori LPS on MKN45 cells
observed by Aihara et al.
(4).
H. pylori Activates NF-B in HEK293 Cells via TLR2 and
TLR5—To determine which, if any, TLRs may play a role in the
response of epithelial cells to H. pylori, we chose to examine the
H. pylori-induced NF-B response in HEK293 cells transfected
with TLR2, TLR4, or TLR5. In the representative transfection experiment (of 3
total) shown in Fig. 3, HEK293
cells were cotransfected with the NF-B-Luc reporter, CD14 and MD2
expression plasmids, and either empty pcDNA3.1 vector or expression plasmids
for TLR2, TLR4, or TLR5. Cultures were stimulated with either purified TLR
ligands (Pam3CSK4 for TLR2 or E. coli LPS for
TLR4) or live H. pylori for 24 h and assayed for luciferase activity.
The results indicated that cells transfected with TLR2 or TLR5 expression
plasmids, but not TLR4, were capable of activating NF-B in response to
H. pylori. The transfected TLR4 was indeed functional as demonstrated
by the ability of the cells to activate NF-B in response to E.
coli LPS.
|
This surprising result prompted us to examine other H. pylori
strains to determine whether the lack of TLR4 response observed for strain
26695 was a specific attribute of that particular laboratory-adapted strain.
To address this question, we performed infections of HEK293 cells transfected
with either TLR2 or TLR4 expression plasmids using three additional strains of
H. pylori: LC11 and two primary clinical isolates: HP98, and HP99. As
shown in Fig. 3B,
cotransfection of TLR2 but not TLR4 was able to confer the ability to activate
NF-Bin response to infection with all three strains. These results
additionally suggested that although H. pylori is a Gram-negative
bacterium, it has either developed a mechanism to avoid detection via TLR4 or
its unique LPS is not recognized by TLR4. Thus in HEK293 cells TLR2 and TLR5
can confer responsiveness to live H. pylori infection.
Role of Toll-like Receptors on Gastric Epithelial Cells in Response to
H. pylori Infection—The studies described above indicated that in a
reconstitution experiment, TLRs can participate in the response of epithelial
cells to H. pylori. To determine the role of endogenously expressed
TLRs in the response of gastric epithelial cells to H. pylori we
chose to inhibit TLR function through the expression of dominant-negative TLR
constructs. The TLRs function as dimers and expression of a mutated receptor,
deleted of the carboxyl-terminal cytoplasmic domain has been demonstrated to
act as a repressor of TLR signaling. MKN45 cells were cotransfected with
NF-B-Luc and either empty expression vector (pcDNA3.1) or expression
plasmids encoding truncated versions of TLR2, TLR4, or TLR5. Cultures were
stimulated by addition of live H. pylori for 24 h. The results of a
representative transfection experiment (of three) are shown in
Fig. 4. Consistent with the
results from the HEK293 experiments, cotransfection with dominant-negative
TLR2 and TLR5, but not TLR4, resulted in an inhibition of the response of
MKN45 cells to H. pylori infection. Thus, in a gastric epithelial
cell line, as in the HEK293 cells, H. pylori activated NF-B
primarily via TLR2 and TLR5 but not TLR4.
|
LPS from H. pylori Is a TLR2 Agonist—The data shown in Figs.
3 and
4 indicated that TLR2 and TLR5,
but not TLR4, participated in the response to live H. pylori
infection. This was a surprising result because H. pylori is a
Gram-negative bacterium. However, this bacterium can produce two types of
lipid A, a major tetraacyl and a minor hexaacyl variety with exceptionally
long fatty acid chains (20).
The structure of the tetraacyl lipid A moiety is clearly distinct from that
found in Enterobacteriaceae and more closely resembles that found in
Porphyromonas gingivalis, which is a TLR2 agonist
(20,
21). Notably, LPS from H.
pylori has been shown to be less toxic than that of E. coli. To
determine whether the H. pylori LPS is a TLR2 or a TLR4 agonist, we
purified the LPS from strain 26695 according to the procedure of Moran et
al. (20) with the
addition of three more extractions as described by Hirschfeld et al.
(18) to yield highly purified,
protein-free LPS. As shown in Fig.
5, HEK293 cells transfected with an NF-B/luciferase
reporter and TLR2, but not TLR4, responded to H. pylori 26695 LPS
indicating that the LPS derived from this organism is predominantly a TLR2
agonist. This result is in contrast to studies published by Kawahara et
al. (17,
22) that indicated that H.
pylori LPS activated guinea pig gastric pit cells via TLR4. In addition,
studies by Sakagami et al.
(23) suggested that the LPS of
Helicobacter felis was an important participant in
Helicobacter-mediated gastritis because no atrophy was seen in the infected
C3H/HeJ non-LPS responder animals, which have a non-functional TLR4. At
present the reasons for this discrepancy are unclear but our hypothesis is
that the lipid A composition of H. pylori LPS may differ between
bacterial strains and/or may be affected by differing culture conditions.
Studies to answer this question are underway. Alternatively, species
differences in the recognition of lipid A variants has been reported and could
potentially account for the discrepancies between our in vitro
studies with human TLRs and the in vivo mouse data
(24). Regardless, these
findings clearly indicated that the LPS from H. pylori strain 26695
when cultured under our routine conditions was a TLR2 and not a TLR4 agonist,
thus providing an explanation for the results shown in
Fig. 4.
|
CD14 Enhances the Response of Gastric Epithelial Cells to H.
pylori—The gastric epithelial cells examined in this study did not
express detectable levels of CD14 (data not shown). However, because CD14 can
be found in inflammatory exudates and serum as a soluble protein that can
enhance the response of cells to LPS
(25,
26), we wanted to examine the
potential role of this pattern recognition molecule in the response of gastric
epithelial cells to H. pylori infection. MKN45 cells were transiently
transfected with NF-B-Luc and either empty pRc/RSV vector or
pRc/RSV-CD14 and stimulated by infection with H. pylori as above. As
shown in Fig. 6A,
forced expression of CD14 in MKN45 cells was able to significantly enhance the
activation of NF-B following infection. To determine whether soluble
CD14 could likewise enhance the response of gastric epithelial cells to H.
pylori infection, we used culture supernatants from CHO cells stably
transfected with CD14 (a gift of D. Golenbock, University of Massachusetts) as
a source of soluble CD14. In the experiment shown in
Fig. 6B, MKN45 cells
were transfected with the NF-B-Luc reporter plasmid and infected with
H. pylori in the presence of a 10% volume of media from either CHO or
CHO-CD14 cells. In agreement with the previous experiment, soluble CD14 was
also capable of enhancing the response of MKN45 cells to H.
pylori.
|
To further characterize the effect of CD14 on the gastric epithelial
response to H. pylori, we examined the role of CD14 in enhancing the
response of MKN45 cells to LPS purified from H. pylori 26695. In the
experiment shown in Fig.
6C, MKN45 cells were transfected with the NF-B-Luc
reporter and either empty vector or pRc/RSV-CD14. Following transfection,
cells were stimulated with either PAM3CSK4 or H.
pylori LPS. In the absence of CD14, H. pylori LPS was incapable
of inducing NF-B activity, however, expression of CD14 rendered the
cells responsive to H. pylori LPS. Consistent with previous reports,
the synthetic bacterial lipopeptide PAM3CSK4 was capable
of inducing responses in the absence of CD14 but the response was enhanced by
CD14 expression (27).
Purified H. pylori Flagellin Activates NF-B via
TLR5— H. pylori is a flagellated bacterium that likely explains its
ability to stimulate HEK293 and MKN45 cells via TLR5. To confirm that the
responses observed above are in fact because of the interaction of H.
pylori flagellin with TLR5, we generated a partially purified flagellin
preparation from H. pylori cultures and assessed the ability of this
material to activate NF-B in both HEK293 and MKN45 cells. As shown in
Fig. 7A, stimulation
of HEK293 cells with this flagellin preparation resulted in NF-B
activation in cells transfected with either TLR5 or TLR2. Treatment of the
flagellin preparation with polymyxin B partially inhibited the response of
HEK-TLR2 cells but had no effect on HEK cells transfected with TLR5. Likewise,
as shown in Fig. 7B,
transfection of MKN45 cells with a dominant negative TLR5 construct partially
inhibited the response of the cells to the flagellin preparation. These
results therefore indicated that H. pylori is capable of activating
gastric epithelial cells as a consequence of flagella expression and as
predicted this response is mediated via TLR5.
|
H. pylori Induces Chemokine Gene Expression via TLR2 and
TLR5—The studies described above were limited to the analysis of
NF-B activation in response to H. pylori infection. To
determine whether Toll-like receptor expression is a prerequisite for the
other biological responses induced via H. pylori we chose to examine
chemokine gene expression in HEK293 cells that were stably transfected with
TLR2 and TLR5 expression plasmids. Following infection for 6 h with H.
pylori 26695, mRNA expression for IL-8, MIP-3
, and GRO was
analyzed by quantitative RT-PCR. As shown in
Fig. 8, in the absence of TLR
expression, H. pylori had a minimal effect on chemokine gene
expression: the expression of each gene was induced between 3- and 8-fold.
However, stable expression of TLR2 resulted in extremely enhanced expression
of all three chemokine genes. Likewise expression of TLR5 also enhanced the
induction of chemokine gene expression in response to H. pylori
infection although the effect was less dramatic than that observed for TLR2.
The differences in the abilities of TLR2 and TLR5 to enhance chemokine gene
expression may be reflective of differing concentrations/bioavaliabilities of
the respective ligands and/or different expression levels of TLR2 or TLR5 on
the surface of the HEK cell lines. Whether the low levels of chemokine
expression induced in response to infection of the wild-type cell line are
because of the low amount of TLR5 levels detected in HEK293 cells or are
because of other TLR-independent pathways remains to be determined.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
B cellular responses via TLR2 and TLR5,
but surprisingly not TLR4. Furthermore, our data using highly purified LPS
from H. pylori strain 26695 indicated that similar to LPS from P.
gingivalis (21) this LPS
was a TLR2 agonist. H. pylori has been demonstrated to require flagella for efficient colonization and maintenance of infection in the gastric epithelium (28, 29). An indication that H. pylori flagellin may play a significant role in disease progression was suggested by a report that demonstrated restriction fragment length polymorphisms in flagellin genes. Significant differences in IL-8 production were noted between groups with different restriction-fragment length polymorphisms and highly motile strains were associated with increased levels of IL-8 production (30). The response of cells of the innate immune system to flagellated bacteria was demonstrated by Hyashi et al. (11) to be mediated via interactions between bacterial flagellin and TLR5. The data presented herein indicate that gastric epithelial cells express TLR5 and respond to H. pylori at least in part through activation of this receptor.
Whereas the TLR5-mediated response to H. pylori was somewhat
predictable, the non-responsiveness of cells via TLR4 was surprising. While
these studies were in progress, Kawahara et al.
(17) demonstrated that H.
pylori LPS acted via TLR4 to stimulate
production from guinea pig gastric
pit cells. Guinea pig gastric pit cells were demonstrated to be deficient in
expression of TLR2 and thus all LPS-dependent responses were presumed to occur
via TLR4. Interestingly, in these studies only three of the six LPS
preparations tested were capable of activating Mox1 and inducing
p67phox expression. Previously, Moran et al.
(20) had demonstrated that the
lipid A component of H. pylori LPS was a mixture of two forms: a
predominant tetraacyl variety and a minor hexaacyl form. One important
implication of that study is that different strains of H. pylori, or
perhaps the same strain cultured under different conditions, may produce
lipopolysaccharides containing varying amounts of either hexaacyl or tetraacyl
lipid A. Of particular importance to the results presented here and the
studies of Kawahara et al.
(17,
22), is that the tetraacyl
lipid A structure of H. pylori is reminiscent of the lipid A
structure of P. gingivalis, which has been demonstrated to be a TLR2
agonist (21). Additionally,
consistent with this thought is the demonstration that LPS from H.
pylori and P. gingivalis bind poorly to LPS-binding protein and
are therefore transferred less efficiently to CD14
(31). Thus, we hypothesize
that the lack of responses observed for the three LPS preparations in the
Kawahara study may reflect differences in the lipid A structures produced by
the different strains of bacteria. Studies are in progress to define the lipid
A structure of the H. pylori LPS used in our studies and to determine
whether variations in culture conditions and/or strain of H. pylori
can influence the relative abilities to activate cells via TLR2 or TLR4.
The finding that H. pylori induces responses in epithelial cells
primarily via TLR2 and TLR5, and not TLR4, may provide some insight into the
reasons for the relatively low pathogenicity of this organism and the trend
toward the development of a chronic, low-level inflammation in the gastric
mucosa. Gene expression studies following stimulation of immune cells with
TLR2 or TLR4 agonists have been performed by numerous labs
(21,
32). The noteworthy finding of
these reports is that stimulation of cells via TLR4 results in the expression
of the same sets of proinflammatory mediators as does TLR2 with the exception
that TLR4 stimulation also results in the production of interferon-
and
concomitant downstream expression of interferon-inducible genes, including
IP-10, MCP-5, and iNOS (32).
Thus, the end result of TLR4-dependent responses is a more highly inflammatory
phenotype in response to infection. Because H. pylori may be more
inclined to activate cells via TLR2, the resulting inflammatory response will
be more muted perhaps allowing this organism to establish a chronic foothold
in the gastric epithelium. The gene expression profiles induced in response to
TLR5-dependent agonists, i.e. flagellin, have not as yet been
investigated. However, given the known similarities in downstream signaling
mechanisms for TLR2 and TLR5, i.e. both have an absolute requirement
for MyD88 (11,
33), it is likely that
stimulation via either receptor independently will result in similar gene
expression patterns and a moderate inflammatory response. The data presented
herein indicated that H. pylori induced IL-8, MIP-3, and
GRO in both HEK293-TLR2 and HEK293-TLR5 cells. Although in our studies
the absolute levels of chemokine gene expression were different between the
two cell types, perhaps as a result of different ligand and/or receptor
concentrations.
Finally, it is important to point out that our studies do not rule out a role for other H. pylori virulence factors in the responses of epithelial cells to infection. Clearly, numerous studies have pointed to the important role that the cag pathogenicity island plays in the development of gastritis and in responses induced in gastric epithelial cells. The CagA protein in particular appears to play a critical role in the regulation of gene expression in gastric epithelial cells (34). Although, the mechanism through which CagA effects cellular responses is still unclear, a report demonstrating the interaction of CagA with SHP2 is particularly intriguing (35). We would predict that stimulation of gastric epithelial cells with purified TLR2 and TLR5 agonists will result in the regulation of a subset of genes that are also regulated in response to H. pylori infection. It is highly likely that the response induced by H. pylori will be both qualitatively and quantitatively different from that which we observe with the purified TLR agonists, possibly due in part to the presence of cag pathogenicity island by H. pylori. Interestingly, Guillemin et al. (34) noted that the genes whose expressions were least effected by the Cag phenotype were those involved in the immune response thus suggesting that Toll-like receptor activation may be directly responsible for the regulation of that subset. Consistent with the studies reported here, a recent report by Bäckhed et al. (15) demonstrated that regulation of gastric mucosal responses to H. pylori is TLR4-independent but cag pathogenicity island-dependent. The role of other TLRs in regulating the response to H. pylori was not addressed in those studies, however. Importantly, those studies did demonstrate that both wild type and an isogenic CagA-deficient strain induced IL-8 production to the same extent, however, strains deleted in the entire cag pathogenicity island could not induce IL-8 production. Thus, when viewed in light of the data presented herein, which demonstrates a requirement for TLR expression to induce responses to H. pylori, the presence of an intact type IV secretion system is a necessary, but not sufficient, requirement for the induction of epithelial cell responses to H. pylori infection.
|
FOOTNOTES |
|---|
¶ To whom correspondence should be addressed: University of Virginia Health System, P.O. Box 800708, Charlottesville, VA 22908-0708. Tel.: 434-924-2573; Fax: 434-243-9645; E-mail: mfs3k{at}virginia.edu.
1 The abbreviations used are: IL-8, interleukin-8; TLR, Toll-like receptor;
LPS, lipopolysaccharide; RT, reverse transcriptase; FBS, fetal bovine serum;
CHO, Chinese hamster ovary.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  H. Wu, H. Wang, W. Xiong, S. Chen, H. Tang, and D. Han Expression Patterns and Functions of Toll-Like Receptors in Mouse Sertoli Cells Endocrinology, September 1, 2008; 149(9): 4402 - 4412. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Zorko and R. Jerala Alexidine and chlorhexidine bind to lipopolysaccharide and lipoteichoic acid and prevent cell activation by antibiotics J. Antimicrob. Chemother., July 16, 2008; (2008) dkn270v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Diya Zhang, Lili Chen, Shenglai Li, Zhiyuan Gu, and Jie Yan Lipopolysaccharide (LPS) of Porphyromonas gingivalis induces IL-1{beta}, TNF-{alpha} and IL-6 production by THP-1 cells in a way different from that of Escherichia coli LPS Innate Immunity, April 1, 2008; 14(2): 99 - 107. [Abstract] [PDF]
|
|
|
|
 |
|
 M. Mita, M. Satoh, A. Shimada, M. Okajima, S. Azuma, J. S. Suzuki, K. Sakabe, S. Hara, and S. Himeno Metallothionein is a crucial protective factor against Helicobacter pylori-induced gastric erosive lesions in a mouse model Am J Physiol Gastrointest Liver Physiol, April 1, 2008; 294(4): G877 - G884. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Adam, B. Schmausser, M. Gobeler-Kolve, H. K. Muller-Hermelink, and M. Eck Gastric extranodal marginal zone B-cell lymphomas of MALT type exclusively express toll-like receptor 4 in contrast to other lymphomas infiltrating the stomach Ann. Onc., March 1, 2008; 19(3): 566 - 569. [Abstract] [Full Text] [PDF]
|
|
|
|
