| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Max von Pettenkofer-Institut für Hygiene und Medizinische Mikrobiologie, Ludwig-Maximilians-Universität München, D-80336 München, Germany;
Medizinische Poliklinik, Ludwig-Maximilians-Universität München, D-80336 München, Germany; and
Klinik für Dermatologie und Allergologie, Ludwig-Maximilians-Universität München, D-80337 München
3Correspondence: Max von Pettenkofer-Institut, Ludwig-Maximilians-Universität München, Pettenkoferstrasse 9a, D-80336 München, Germany. E-mail: Autenrieth{at}m3401.mpk.med.uni-muenchen.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
B element.
This NF-B binding site preferentially binds Rel p65-p65 homodimers
as well as some p50-p65 heterodimers in response to stimulation by
invasin. Invasin-induced NF-B activation correlated with degradation
of IB
and the inhibition of NF-B by specific inhibitors of
IB activation blocked invasin-induced IL-8 secretion.
Invasin-triggered IL-8 production does not depend on invasin-triggered
uptake of bacteria, and is independent of a functional PI3-kinase. This
report is the first to demonstrate the molecular basis of IL-8
production triggered by enteropathogenic bacteria. Together, these data
elucidate the possible early pathomechanisms operating in
Yersinia infection and may have implications for the
design of novel therapeutics directed against this
enteropathogen.—Schulte, R., Grassl, G. A., Preger, S., Fessele,
S., Jacobi, C. A., Schaller, M., Nelson, P. J., Autenrieth,
I. B. Yersinia enterocolitica invasin protein
triggers IL-8 production in epithelial cells via activation of Rel
p65-p65 homodimers.
Key Words: bacteria • chemokine • transcription • NF-B
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(TNF-) (1
2
3
4
5
6)
.
Recent reports using animal models and in vitro systems have
identified IL-8 as an important chemoattractant in PMN migration to
sites of infection or tissue injury (3
, 7)
.
Yersinia enterocolitica causes a broad range of
gastrointestinal syndromes ranging from acute enteritis and
enterocolitis to mesenteric lymphadenitis (8
, 9)
. It has
been suggested that the triggering of IL-8 production by
Yersinia may be part of its pathogenic strategy.
Yersinia that is resistant to phagocytosis by PMN may take
advantage of recruited PMN and/or other cells in order to induce tissue
pathology and thus gain access to, or disseminate within, host tissue
(10
, 11)
.
The virulence of Y. enterocolitica is controlled by
chromosomal (yst, inv) (12
13
14
15)
and
plasmid-encoded genes (16
, 17)
. The pYV virulence plasmid
directs production of the outer membrane protein YadA and secretion of
at least 11 antihost proteins called Yops (17
18
19)
.
Adherence to or internalization of Yersinia by epithelial
cells (20
, 21)
depends on Yersinia surface
proteins including invasin (22
23
24)
, attachment-invasion
locus (14
, 25)
, and Yersinia adhesin A
(24
, 26
27
28
29
30)
.
The outer membrane protein invasin plays an important role in the early
phase of intestinal infection (31)
. Invasion of Y.
enterocolitica into epithelial cells depends on the interaction
between Yersinia invasin and ß1
integrins on the surface of the eukaryotic cell (22)
.
Infection of HeLa or T84 cells with Y. enterocolitica
induces IL-8 transcription and subsequent secretion (32)
.
The invasin protein of Y. enterocolitica plays an essential
role in Yersinia-triggered IL-8 production. Inv
mutant Y. enterocolitica pYV- strain
does not induce IL-8 production, whereas a recombinant E.
coli strain-expressing Yersinia inv can
(33)
. Although invasin is required for IL-8 gene
activation, host cell invasion appears not to be essential for this
process as blocking of Yersinia cell invasion by
cytochalasin D or wortmannin does not affect invasin-triggered IL-8
expression and secretion (33)
.
The promoter region of the human IL-8 gene contains consensus sequences
for several transcription factors including the nuclear factors (NF)
IL-6, NF-B, AP-1, AP-3, and octamer binding proteins
(34)
. Although cooperation with other transcription
factors including NF-IL-6 and AP-1 (34
35
36
37)
, is thought to
be necessary, the NF-B element has been shown to be crucial for IL-8
transcription (34
35
36
37)
.
Previous studies have demonstrated that NF-B is induced by a
pleiotropic array of agents including various cytokines,
double-stranded RNA, phorbol esters, and several viruses
(38)
. More recent investigations have shown that infection
of host cells with pathogenic bacteria is associated with activation of
NF-B and IL-8 expression in a variety of cell types, including
cultured enterocytes. Thus, infection of epithelial cells with
bacterial pathogens such as Helicobacter pylori, S.
typhimurium, or enteropathogenic E. coli can lead to
the activation of NF-B (5
, 39
40
41
42
43)
. To date, the
mechanism by which enteric bacterial pathogens induce NF-B
activation in epithelial cells is unknown.
The purpose of the present study was to clarify whether and how the
outer membrane protein invasin of Y. enterocolitica
contributes to the induction of IL-8 expression. We show that bacteria
or inert latex particles coated with a truncated Y.
enterocolitica invasin protein comprising the carboxyl-terminal
195 amino acids trigger degradation of IB and subsequent
activation of Rel p65-p65 homodimers, which leads to IL-8 expression
and production by epithelial cells. These events depend on
invasin-mediated attachment of bacteria or beads to epithelial cells,
but are independent of invasin-triggered internalization, which depends
on functional PI3-kinase activity.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
,
(pYV-) inv mutant Y.
enterocolitica WA314 (45)
, noninvasive E.
coli HB101, and the E. coli HB101 (pINV1914) strain
expressing the Y. enterocolica inv (33)
were
grown in Luria-Bertani broth (LB). For infection experiments, overnight
cultures were diluted to an optical density (OD) at 600 nm (OD600) of
0.2 in LB and incubated for 3 h at 27°C or 37°C, respectively.
Cell culture and infection protocol
Human HeLa cervical epithelial cells (ATCC CCL-2.1) were grown
in RPMI 1640 (Biochrom KG, Berlin, Germany) and 10% fetal bovine serum
(Gibco BRL, Paisley, Scotland), supplemented with 2 mM L-glutamine
(Biochrom KG), penicillin (100 U/ml), and streptomycin (100 µg/ml)
(Biochrom KG). The cells were grown in a humidified 5%
CO2 atmosphere at 37°C. For infection, bacteria
grown for 3 h in LB at 27°C or 37°C, respectively, were
collected by centrifugation and washed twice in sterile
phosphate-buffered-saline (PBS) (pH 7.4). After determination of the
OD, appropriate dilutions of the bacteria in PBS were performed before
infection. Cells were infected with a bacterium-to-cell ratio of 100:1
or as indicated. The actual number of bacteria administered was
determined by plating 0.1 ml of 1:10 serial dilutions on Mueller-Hinton
(MH) agar and counting of colony-forming units. Monolayers of HeLa
cells (60–70% confluent) were washed twice with PBS and incubated in
medium containing heat-inactivated fetal bovine serum without
antibiotics. After 1–2 h of equilibration, bacterial samples were
added. Monolayers and bacteria were incubated for 2 h or as
indicated to allow bacterial adherence and entry. After removal of the
medium, cultures were washed three times with PBS to remove
extracellular bacteria and further incubated for 6 h or as
indicated in the presence of 100 µg of gentamicin/ml to kill
remaining extracellular bacteria. Then culture supernatants were
removed and centrifuged for 20 min to pellet residual bacteria and
cells before IL-8 determination. Cells were lysed with 1% Triton X-100
in PBS. The number of released viable bacteria was determined by
plating serial 10-fold dilutions on MH agar. TNF- was a gift from G.
Adolf from Bender (Vienna, Austria) and was used as a positive control
for IL-8 stimulation. Alternatively, invasin-coated beads (see below)
or anti-ß1 integrin monocloncal antibodies (clone TS2/16 at 5 µg/ml
induced submaximal responses; Endogen, Woburn, Mass.), followed by goat
anti-mouse IgG antibodies (20 µg/ml; Sigma, St. Louis, Mo.) were
used. MG-132 (Biomol, Hamburg, Germany) and curcumin (Biomol) were
used to inhibit NF-B activation by incubating HeLa cells 30 min
before stimulation or infection as indicated. HeLa cells were treated
with 100 nM of wortmannin (Sigma), a PI3-kinase inhibitor, 20 min prior
to infection to prevent bacterial internalization.
DNA constructs
Vectors containing fusions between the 5'-flanking region
sequences of the human IL-8 gene (34
, 35)
or sequentially
deleted fragments of the IL-8 gene 5'-flanking region and a Luciferase
reporter gene were kindly provided by Ron Crystal (Cornell Medical
Center, New York). Vectors were constructed from a pUC8-derived vector
(pCMV-luciferase) as described by Nakamura et al. (46)
.
The pIL-8B chloramphenicol acetyltransferase (CAT; wild-type) and
pIL-8Bm CAT (mutant) were provided by Charles Kunsch (Atherogenics
Inc., Norcross, Ga.) and have been described previously
(47)
. In these reporter constructs, CAT expression is
under the control of the human IL-8 genomic sequence from -420 to +101
bp. Specific substitution mutation was introduced into the IL-8
promoter region to disrupt the NF-B binding site (designated
pIL-8B CAT) (47)
. ‘Full-length’ IL-8 luciferase (Luc)
reporter constructs together with mutant constructs that lack either
the NF-B, NF-IL-6, or AP-1 binding sites were obtained from Andrew
Keates (Beth Israel Deaconess Medical Center, Boston, Mass.). The
full-length reporter construct contains a 1512 bp fragment (nucleotides
-1481 to +40) of the promoter region of the IL-8 gene (34
, 35)
cloned into the pGL2-Basic Luciferase expression vector
(Promega, Madison, Wis.). The mutants are based on the full-length
construct in which nucleotides -71 to -82 (for NF-B mutant),
nucleotides -84 to -91 (for NF-IL-6 mutant), or nucleotides -120 to
-126 (for AP-1) have been deleted. To normalize for transfection
efficiency, cotransfections were performed using pCMV-ß-galactosidase
(Clontech, Palo Alto, Calif.) in which ß-galactosidase is
constitutively expressed from the CMV promoter. pNF-B-Luciferase
(Clontech) was used as an NF-B transcription reporter vector. Part
of the inv gene encoding the 397 or 195 carboxyl-terminal
amino acids was amplified from Y. enterocolitica serotype
O:8 by polymerase chain reaction (PCR) using the forward primer
5'-ACGTGAATTCCACGTTGACCGTTATTGTGC-3' (EcoRI
restriction site underlined) for Inv397,
ACGTGAATTCCTACCCAGTACCGAAGATAA (EcoRI-site
italicized) for Inv195, and the reverse primer
5'-GCCGCTCGAGCTATTGCGGCTCCGCAC-3' (XhoI
restriction site italicized) according to the published sequence
(48)
. The amplified DNA fragments were digested with
EcoRI and XhoI and cloned into pGEX-4T-3
expression vector (Amersham-Pharmacia, Little Chalfont, U.K.) resulting
in pINV397 and pINV195, respectively.
Expression and purification of GST-Inv397 and GST-Inv195 fusion
protein
E. coli BL21 harboring pGEX-4T-3, pINV397, or pINV195
was grown at 24°C in LB to an OD of 0.7. Expression of the GST and
GST-Inv fusion protein was induced with IPTG
(isopropyl-ß-D-thiogalactopyranoside) at a final concentration of 0.1
mM. Cells were grown for 2 additional hours before being harvested by
centrifugation and frozen at -20°C. Frozen cells were resuspended in
PBS containing 1 mM phenylmethylsulfonylfluoride (PMSF) and complete
protease inhibitor mixture (Boehringer Mannheim). After disrupting
bacterial cells by French press, lysates were cleared for cellular
debris by centrifugation. GST-Inv397, GST-Inv195 and GST were purified
from supernatants using glutathione Sepharose 4B (Pharmacia Biotech,
Brussels, Belgium) and Superdex 200 gel filtration. Purity and identity
of GST-Inv397 or GST-Inv195 fusion protein were analyzed by sodium
dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and
Western blot using anti-GST and anti-Inv antisera. Protein
concentration was determined by the BCA protein assay (Pierce,
Rockford, Ill.).
Coating of proteins to latex beads
For noncovalent coating of beads, purified protein was dialyzed
against PBS pH 7.0. About 109 latex beads (1 µm
diameter, sulfate-modified; Molecular Probes, Eugene, Oreg.) were
washed with 1 ml of PBS and resuspended in 500 µl of PBS. Purified
GST, GST-Inv397, or GST-Inv195 fusion protein (0–2 mg) was added and
allowed to adsorb to the beads for 3 h at room temperature (RT).
After adding 500 µl of 20 mg/ml bovine serum albumin (BSA), the
solution was incubated at RT for another hour. Then beads were washed
in PBS containing 1 mg/ml BSA and stored at 4°C in 500 µl of PBS
containing 0.2 mg/ml BSA. To determine the coupling efficiency, the
protein concentration of the starting solution and of the supernatant
before adding BSA was determined. Integrity of coated GST-Inv397 or
GST-Inv195 fusion protein was checked by Western blot analysis.
IL-8 ELISA
The amount of IL-8 secreted into the supernatant was determined
as described previously (32)
using an enzyme-linked
immunosorbent assay (ELISA) with optimal concentrations of a mouse
anti-human IL-8 monoclonal antibody (mAb) (G265–5; PharMingen, San
Diego, Calif.) and a biotinylated mouse anti-human IL-8 mAb (G265–8;
PharMingen) as detecting antibody. IL-8 concentrations were calculated
from the straight-line portion of a standard curve using recombinant
human IL-8 (PharMingen).
Reverse transcription-PCR analysis
Total RNA of infected HeLa cells in 6-well plates was extracted
using 1 ml of TRIzol reagent (Gibco BRL). Five micrograms of RNA was
reverse transcribed as described by Schulte et al. (33)
.
cDNA products were amplified by PCR in 50 µl (10 mM Tris pH 8.3, 50
mM KCl, 2.5 mM MgCl2), 200 µM each of dATP,
dCTP, dGTP, and dTTP in the presence of 25 pmol each of 5' and 3'
primer (4)
and 2.0 U of AmpliTaq Gold DNA polymerase
(Perkin Elmer, Überlingen, Germany). The temperature profile of
the amplification consisted of 25 cycles of 1 min denaturation at
95°C and 2.5 min annealing and extension at 60°C (IL-8) or 72°C
(ß-actin) (4)
. Negative controls were performed by
omitting RNA from the cDNA synthesis and specific PCR amplifications.
PCR products were separated in 2% agarose gels. Quantification was
performed using a FluoroS Imager (Bio-Rad, München, Germany).
Transient transfection
For transient transfection assays, 1 x
105 HeLa cells were seeded in 12-well plates, and
cotransfected 24 h later with 1 µg of CAT or Luc reporter
constructs using ExGen (Euromedex; Souffelweyersheim, France)
transfection reagent according to the manufacturer’s instructions, and
incubated for 12 h at 37°C. The constructs were cotransfected
with 0.25 µg of ß-galactosidase construct (pCMV-ß-gal) to
determine transfection efficiency. For stimulation, transfected cells
were washed twice with PBS and incubated in antibiotic-free medium.
HeLa cells were infected with bacteria at a MOI (multiplicity of
infection) of 100 and incubated at 37°C. After 2 h, cells were
washed twice with PBS to remove extracellular bacteria and incubated
for an additional 6 h in the presence of 100 µg gentamicin/ml.
Then supernatants were removed to determine IL-8 secretion. HeLa cells
were washed twice with PBS and lysed. Lysates were centrifuged and
supernatants were removed for protein determination, measurement of
ß-galactosidase, and determination of CAT or Luc activity.
Reporter gene activity
To determine CAT activity, cells were washed twice with PBS and
once with TEN (40 mM Tris/HCl pH 7.8; 1 mM EDTA pH 8.0; 150 mM NaCl).
Cells were incubated with 200 µl TEN for 5 min on ice and finally
resuspended. After centrifugation at 12,000 rpm for 1 min at 4°C,
cells were resuspended in 100 µl 0.25 M Tris/HCl pH 8.0. Finally,
cells were lysed by three successive rounds of freezing and thawing.
Luc activity was measured as follows. Cell lysates were added to a
solution containing 25 µg of n-butyryl coenzyme A, 0.2 µCi
[C14] chloramphenicol in a final volume of 50
µl 0.25 M Tris/HCl (pH 8.0). The reaction was allowed to proceed for
1 h at 37°C and then stopped by adding 200 µl of mixed xylene
(Sigma-Aldrich, Steinheim). The organic phase was then extracted twice
with 100 µl of 0.25 M Tri/HCl, pH 8.0, and the 100 µl of the
organic phase was counted in a scintillation counter MicroBeta TriLux
(Wallac, Turku, Finland). To evaluate Luc activity, cells were washed
with PBS and lysed with 0.1% Triton X-100 in 25 mM Tris-phosphate pH
7.8, 10% glycerol, and 1 mM DTT. Lysates were centrifuged at 12.000
rpm for 5 min at 4°C. CAT activity in supernatants were assayed in
Luc buffer containing 20 mM tricine; 1.07 mM
(MgCO3)4
Mg(OH)2 x 5H2O; 0.1 mM
EDTA; 33.3 mM D/L-DTT; 270 µM Li3-coenzyme A;
470 µM D(-)-luciferin; 530 µM Mg-ATP. Luminescence was measured
with a MicroBeta TriLux (Wallac). The protein concentration of
supernatants was determined by the Bradford method (Bio-Rad protein
assay). ß-Galactosidase activity was determined according to a
standard protocol (49)
. Levels of CAT and Luc expression
were normalized by ß-galactosidase activity and total protein
concentration.
IB immunoblotting
To examine the presence of IB or IBß at different
times after infection, 5 x 105 HeLa cells
in 6-well plates were infected at a MOI of 100. Bacterial infection was
synchronized by centrifugation for 5 min at 500 g at 20°C.
At indicated time points, cells were washed twice with PBS and 400 µl
lysis buffer (10 mM Tris, pH 8.0, 60 mM KCl, 1 mM EDTA, 0.5% Nonidet
P-40, 1 mM DTT, 1 mM PMSF, 1 mM benzamidine, 20 mM
ß-glycerophosphate, 5 mM p-nitrophenyl phosphate, and 0.1 mM
Na3VO4) was added. Cells
were scraped off, transferred into an Eppendorf tube, and centrifuged
to pellet cell debris. Proteins were separated by 10% SDS-PAGE and
electrotransferred to polyvinylidene difluoride membrane.
Immunostaining for I-B was performed with polyclonal anti-I-B
or I-Bß antibodies (Santa Cruz Biotechnology, Santa Cruz, Calif.).
Immunoreactive bands were visualized by incubation with donkey
anti-rabbit antibodies conjugated to horseradish peroxidase using
enhanced chemiluminescence reagents (Amersham).
Electrophoretic mobility shift assay (EMSA)
HeLa cells (5x106) were infected as
described above or incubated with GSTINV397 beads at a MOI of 2000. At
various intervals after infection or stimulation, nuclear extracts were
prepared according to the protocol described by Schreiber et al.
(50)
. Aliquots of the supernatant containing nuclear
proteins were stored at -70°C. Protein concentrations were
determined by the Bradford assay. The oligonucleotide probes described
below were labeled with [
32P]ATP (NEN)
by using the T4-Polynucleotide kinase (Boehringer Mannheim,
Mannheim, Germany) and then purified on a NucTrap probe
purification column (Stratagene, San Diego, Calif.). The following
oligonucleotides were used: NF-B consensus (NF-Bc):
5'-AGTTGAGGGGACTTTCCCAGGC-3' (Santa Cruz); NF-B mutant
(NF-Bm): 5'-AGTTGAGGCGACTTTCCCAGGC-3' (Santa Cruz); IL-8-B
consensus (IL-8Bc): 5'-ATCGTGGAATTTCCTCTGA-3' (Metabion, Munich,
Germany); IL-8-B mutant (IL-8Bm): 5'-ATCcTGcAATgTCgTCTGA-3'
(Metabion); hIL8:
5'GGGCCATCAGTTGCAAATCGTGGAATTTCCTCTGACATAATGAAAAGAT-3';
C/EBP consensus (C/EBPc): 5'-TGCAGATTGCGCAATCTGCA-3'; C/EBP mutant
(C/EBPm): 5'-TGCAGAGACTAGTCTCTGCA-3'; RANTES
E:5'-CTTTTCCGTTTTGTGCAATTTCACTTATG-3'. Promoter sequence
analysis for transcription factor binding sites was performed by using
professional MatInspector software (Genomatix, Munich, Germany) as
described by Quandt et al. (51)
. Nuclear extracts (3 to 6
µg) were incubated with 30,000 cpm of the
32P-labeled oligonucleotide probe for 45 min on
ice in a buffer containing 5% glycerol, 90 mM NaCl, 1 mM
dithiothreitol 1 mM EDTA pH 8.0, 10 mM Tris-HCl pH 7.2, and 1 µg
dIdC. For supershift analysis, antibodies against p50, p52 (UBI), p53,
p65, c-Rel, RelB, and C/EBP (Santa Cruz) were included in the binding
reaction. Samples were resolved on a 5% nondenaturating polyacrylamide
gel using 0.5 x TBE (25 mM Tris-HCl, 25 mM boric acid, 0.5 mM
EDTA) as running buffer. Gels were transferred to Whatman 3M paper and
dried under vacuum. Protein binding was assessed via autoradiography.
Transmission electron microscopy
One hour after infection, cells were washed three times with
PBS, harvested by trypsin/EDTA treatment, washed twice with medium
containing 10% fetal calf serum, and once with PBS. After
centrifugation at a speed of 150 g for 10 min, the resulting
pellets were fixed for 2 h in 2% formaldehyde/2.5%
glutaraldehyde in 0.1 M phosphate-buffer (pH 7.3) for 2 h at room
temperature. Postfixation was based on 1% osmium tetroxide containing
1% K-dichromate in 0.85% NaCl at pH 7.3 at room temperature for 45
min. After several washings and dehydration procedures, the specimen
were embedded in glycide ether. The small blocks of cells were cut
using an ultra microtome (Ultracut, Reichert, Vienna, Austria).
Semi-thin sections (1 µm) were studied with a light microscope after
staining with 1% toluidine blue and 1% pyronine G (Merck, Darmstadt,
Germany). The sections were viewed at a magnification of x400.
Ultra-thin sections (80 nm) were stained with 0.5% uranyl acetate for
10 min at 30°C and 2.7% lead citrate for 5 min (Ultrastainer, LKB,
Sweden) at 20°C. Grids were examined using a Zeiss EM 902
transmission electron microscope (Zeiss, Oberkochen, Germany) operating
at 80 kV at magnifications between x2000 and x30,000.
Immunofluorescence staining and confocal laser scan microscopy
Immunofluorescence microscopy (CLSM) was performed as recently
described by V. Kempf et al. (unpublished results). In brief, after 3%
paraformaldehyde (Sigma) fixation of the monolayers, extracellular
bacteria or beads were stained with polyclonal rabbit
anti-Yersinia invasin antibodies, followed by
FITC-conjugated goat anti-rabbit antibodies (Dianova, Hamburg,
Germany). After three washings, cells were permeabilized by 2% Triton
X-100 in PBS, washed, and intracellular bacteria or beads were stained
by anti-Yersinia invasin antibodies, followed by
Cy-5-conjugated goat anti-rabbit antibodies (Dianova). F-actin was
stained by TRITC-conjugated phalloidin (Sigma). The fluorescence images
were obtained with a confocal laser scan microscope (Leica TCS 4D).
Statistics
Data shown in the figures are from representative
experiments. Comparable results were obtained in additional
experiments. Differences between mean values were analyzed using the
Student’s t test. P<0.05 was considered
statistically significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). In contrast, infection with Yersinia inv mutant or
inv-negative E. coli strain did not induce
significant IL-8 mRNA expression (Fig. 1)
. Gentamicin-killed E.
coli pINV1914 induced comparable IL-8 mRNA expression levels (not
shown). These results suggest that Yersinia invasin is
essential for induction of the IL-8 gene.
|
Yersinia invasin-coated latex beads are internalized
and trigger IL-8 production in HeLa cells
The aforementioned findings suggest that particles such as killed
bacterial cells carrying invasin may induce IL-8 activation in HeLa
cells. To test this hypothesis, we cloned a 397 and a 195 amino acid
carboxyl-terminal fragment of Yersinia invasin, known to be
sufficient to bind to ß1 integrins on the host cell surface
(52)
, onto a GST expression vector. Inert latex beads
coated with the purified GST-Inv397 or GST-Inv195 fusion protein were
first tested in cell internalization assays as described by Dersch and
Isberg (53)
. For this purpose, HeLa cells were incubated
with beads coated with either purified GST-Inv397, GST-Inv195, or GST
for 60 min and further processed for transmission electron microscopy.
Electron microscopic analysis revealed that beads coated with GST only
occasionally adhered to and were internalized by HeLa cells (Fig. 2A
). In contrast, beads coated with GST-Inv397 or GST-Inv195
efficiently adhered to and were internalized by HeLa cells in a way
resembling the zipper mechanism by which Yersinia bacteria
are internalized into HeLa cells (24)
(Fig. 2B
, C
, D
).
|
Next, HeLa cells were incubated with GST, GST-Inv195, or
GST-Inv397-coated beads and IL-8 production was determined in culture
supernatants by ELISA. The results show that GST-Inv397 or GST-Inv195,
but not GST-coated beads, induced IL-8 secretion by HeLa cells in a
dose-dependent manner (Fig. 2E
, F
, G
, H
). Submaximal
IL-8 production was achieved with beads coated with 500 µg
GST-Inv397/ml at a bead-to-cell ratio of 2000 to 4000 whereas
GST-Inv195 were less efficient in stimulating IL-8 (Fig. 2F
, G
, H
). The addition of soluble GST-Inv397 or GST-Inv195 to
HeLa cells did not induce IL-8 production (not shown). GST-Inv195 or
GST-Inv397-coated beads induced ~20% less IL-8 levels than TNF-,
but twofold higher IL-8 levels than stimulation with cross-linked
anti-ß1 integrin antibodies (Fig. 2E
, F
, G
, H
).
Identification of the minimal IL-8 promoter fragment required for
invasin-mediated activation of the IL-8 gene
To localize the minimal promoter fragment of the human IL-8 gene
necessary and sufficient to mediate Yersinia-induced IL-8
expression, we analyzed the Luc activity of HeLa cells transfected with
IL-8 promoter/luciferase fusion constructs after bacterial infection.
HeLa cells were transiently transfected with either full-length IL-8
promoter constructs (pN1481L, which spans the region from -1481 to
+44) or 5'-truncated promoter constructs (pN130L or pN112L, which span
the regions from -130 to +44, or -112 to +44, respectively).
Transfectants were stimulated either with TNF- (positive control) or
infected with Y. enterocolitica pYV-
or E. coli strains.
Analysis of Luc activity of HeLa cells transfected with the full-length
(pN1481L) construct revealed a threefold increase in response to
TNF- activation or infection by Y. enterocolitica or
E. coli pINV1914 (Fig. 3A
). In contrast, Y. enterocolitica inv mutant or
the parental E. coli strain were not able to stimulate Luc
activity (Fig. 3A
). Consistent with previous results
(6)
, Y. enterocolitica
pYV+, which suppresses IL-8 secretion in
epithelial cells, did not induce Luc activity in transfected HeLa cells
(Fig. 3A
), which suggests that inv-expressing
Y. enterocolitica and E. coli can influence IL-8
expression in HeLa cells at the level of transcription.
|
HeLa cells transfected with pN130L or pN112L reporter constructs showed
comparable results indicating that the promoter fragment from
nucleotide -112 to +44 contains the essential regulatory elements for
invasin-mediated induction of IL-8 transcription. These sequences were
subjected to analysis with MatInspector software (51)
. The
pN112L reporter construct contains one putative NF-B and two
putative NF-IL-6 binding sites, suggesting a potential role for these
elements in invasin-mediated IL-8 transcriptional activation. The
NF-B binding site does not show a strong consensus for the binding
of p50 but does identify the region as a ‘good’ c-rel and p65 binding
site (Fig. 3B
).
Activation of IL-8 transcription by invasin-expressing bacteria is
mediated by Rel proteins
To assess whether NF-B-like and/or NF-IL-6 binding sequences
within the IL-8 promoter contribute to the stimulatory activity by
invasin-producing bacteria, transient transfections with IL-8
promoter/Luc reporter constructs with mutations in either the NF-B
(pIL-8-B), NF-IL-6 (pIL-8-NF-IL-6), or AP-1 (pIL-8-AP-1) binding
site were performed. Mutation of the AP-1 or NF-IL-6 site did not
significantly change stimulation of Luc activity after infection of
transfected HeLa cells with Y. enterocolitica
pYV- (data not shown). In contrast, HeLa cells
transfected with a construct in which the NF-B binding site is
mutated did not respond to infection by production of Luc activity (not
shown). Together, these results suggest that Yersinia
invasin-mediated induction of the IL-8 gene requires at least a
functional NF-B binding site.
Activation of NF-B in HeLa cells infected with Y.
enterocolitica, invasin-expressing E. coli, or
invasin-coated beads
The experiments described above suggest that the NF-B
site is essential for invasin-mediated activation of the IL-8 promoter
in HeLa cells. To determine which Rel factors are actually induced in
HeLa cells after infection with invasin-producing bacteria, nuclear
extracts from invasin-stimulated HeLa cells were prepared and EMSA was
performed. These experiments revealed that infection with Y.
enterocolitica pYV- and E. coli
pINV1914 or incubation with beads coated with invasin induces NF-B
binding activity in the nuclei, which was absent in unstimulated cells
or in cells treated with Yersinia inv mutant strain,
E. coli, or GST-coated beads (Fig. 4A
and data not shown). NF-B binding activity in HeLa cells
after Yersinia infection was detected in a range of 45 to 90
min after infection, with maximum activity at 60 min (data not shown).
Moreover, activation of nuclear B-DNA binding complex was observed
in the presence of wortmannin (Fig. 4A
), an inhibitor of the
PI3-kinase, which blocks invasion of E. coli pINV1914 (Fig. 4B
) or Y. enterocolitica (33)
into
epithelial cells, suggesting that bacterial invasion and PI3-kinase
activity are not required for NF-B activation.
|
To characterize the Rel proteins that mediate this effect, HeLa cells
were stimulated with Inv397-coated beads and the nuclear extracts were
incubated with oligonucleotide probe hIL-8, representing bp -101 to
-53 (Fig. 3B
). The EMSA data depicted in Fig. 5
indicate that stimulation of HeLa cells with Inv397-coated beads
induced two complexes that bound to the IL-8 promoter sequence.
Competition experiments with cold oligonucleotides revealed that
NF-Bc but not C/EBP consensus nucleotides could compete for binding
to these complexes. In addition, supershift analysis with antibody
reagent specific for anti-p65, anti-p50, and anti-C/EBPpan antibodies
revealed that only the Rel protein-specific anti-p65 and anti-p50
antibodies induced supershifts identifying the complexes as p50-p65
heterodimers and p65-p65 homodimers (Fig. 5)
. Thus, no specific binding
of C/EBP family members could be demonstrated in these cells.
|
Binding of invasin-induced Rel proteins to IL-8 B consensus
sequence
We then focused on the role of Rel proteins in invasin-induced
control of IL-8 expression. To establish the specificity of Rel
proteins binding to the IL-8B consensus sequence, nuclear extracts
of invasin-stimulated HeLa cells were incubated with IL-8B consensus
or IL-8B mutant oligonucleotide probe. As evident from Fig. 6A
, specific complexes bound to IL-8B consensus but not to
the mutated sequence of the IL-8Bm oligonucleotide probe.
|
To confirm the functional relevance of these data, an IL-8 promoter/CAT
reporter construct representing bp -410 to +101 in which the NF-B
binding site has been mutated by nucleotide exchange (pIL-8mCAT) to the
sequence of the aforementioned IL-8 Bm (47
, 54)
was
used. Mutation of the NF-B binding site significantly reduced
invasin-mediated CAT activity in pIL-8Bm CAT-transfected HeLa cells
(Fig. 6B
). These results confirm that activation of the IL-8
promoter in HeLa cells in response to infection with either Y.
enterocolitica or E. coli-expressing inv
requires an intact binding site for NF-B.
To better characterize the Rel proteins binding to the IL-8 B
element, EMSA supershift experiments were performed. Antibody reagent
specific for p65 completely shifted both complexes whereas anti-p50
antibodies shifted only the lower complex (Fig. 6C
). In
contrast, anti-cRel and anti-RelB antibodies caused slight supershifts
but anti-p52 had no effect. These results suggest that the
invasin-induced upper band is comprised of p65-p65 homodimers and the
lower complex is primary classical p50-p65 heterodimers.
NF-B consensus sequence binds p50/p50 and p50/p65 complexes
while IL-8-B consensus sequence binds p50/p65 and p65/p65 complexes
According to the computer-based analysis, the ‘NF-B’ binding
site of the IL-8 promoter does not show a strong consensus for the
binding of p50; our results suggest a prominent role for p65-p65
homodimers in invasin-mediated activation of IL-8 expression (Fig. 6)
.
To provide direct evidence for this hypothesis, EMSA competition and
supershift experiments (including comparison binding of labeled NF-B
consensus and IL-8B consensus oligonucleotide probes) were
performed. The data presented in Fig. 7
Aclearly demonstrate that nuclear extracts of invasin-stimulated HeLa
cells comprise various complexes, some of which bind predominantly to
NF-B consensus probe whereas others bind predominantly to IL-8B
consensus oligonucleotide probe.
|
Subsequently, EMSA supershift experiments (including the NF-B
consensus oligonucleotide probe) were performed and revealed that the
addition of anti-p65 antibodies completely shifted the p50-p65
heterodimers, whereas anti-p50 antibodies shifted both p50-p50 and
p50-p65 complexes (Fig. 7B
). Moreover, anti-p52 antibodies
caused a significant supershift and shifted complexes, including both
p50-p50 as well as p50-p65 complexes. In contrast, anti-cRel and
anti-RelB antibodies caused slight supershifts.
In addition, we used an NF-B/luciferase reporter vector
pNF-B-Luc. When HeLa cells transfected with pNF-B-Luc were
infected with bacteria, a two- to threefold stimulation of Luc activity
after infection with Y. enterocolitica
pYV- and E. coli pINV1914 was
observed whereas Y. enterocolitica
pYV- inv or E. coli did not stimulate
significant levels of Luc activity (not shown). These results further
confirm that infection of HeLa cells with Y. enterocolitica
pYV- or E. coli pINV1914 leads to
activation of NF-B.
Inhibition of NF-B activation blocks invasin-induced IL-8
secretion
To show a link between NF-B activation and IL-8 secretion,
inhibitors of NF-B activation were used to modulate
invasin-triggered IL-8 production. MG-132 blocks NF-B activation by
specifically interfering with IB degradation (55)
.
Curcumin is an antioxidant that inhibits NF-B activation (56
, 57)
. HeLa cells were pretreated with various amounts of
inhibitors before infection with Y. enterocolitica
pYV-. As shown in Fig. 7C
, MG-132 and
curcumin treatment markedly inhibited Y.
enterocolitica-induced IL-8 secretion (by ~ 90% using 20
µM MG-132 and ~ 80% using 40 µM curcumin) (Fig. 7C
). These results indicate that activation of NF-B by
inv-expressing bacteria leads to IL-8 expression and
subsequent secretion.
IB degradation in HeLa cells infected with invasin-producing
E. coli
In the next step, we investigated whether bacteria expressing the
inv gene were capable of inducing IB degradation in HeLa
cells. HeLa cells were infected with E. coli or E.
coli pINV1914 and the level of IB was determined by Western
blot analysis. Infection of HeLa cells with Y.
enterocolitica pYV- or E. coli
pINV1914 showed only a weak degradation of IB within 45 to 60 min
after infection (data not shown). However, when synthesis of IB
was blocked by treating the cells with cycloheximide prior to the
infection, a significant degradation of IB after stimulation with
E. coli pINV1914 or Y. enterocolitica
pYV-, but not upon infection with E.
coli or Y. enterocolitica pYV-
inv, was found (Fig. 8
). Comparable results were obtained after stimulation with Inv397-coated
beads. Significant degradation of IBß was not observed (not
shown). These results show that invasin-producing bacteria induce
degradation of IB, an essential early event in the activation of
NF-B.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
B and IL-8 production and secretion
(39
40
41
42
43)
. The mechanism by which these different bacterial
pathogens induce IL-8 expression and secretion is not yet clear.
However, a close contact between the pathogen and the host cells as
well as a type III protein secretion machinery is essential for NF-B
activation by the aforementioned microorganisms (39
40
41
42)
.
In contrast, Yersinia-induced IL-8 production does not
require a type III protein secretion machinery that is encoded by the
virulence plasmid pYV of Yersinia. In fact, Y.
enterocolitica pYV- bacteria that lack the
type III protein secretion system induce IL-8 production, whereas the
presence of the pYV plasmid-encoded type III secretion machinery in
Y. enterocolitica wild-type strains may even suppress IL-8
secretion (6)
. In the light of the present study and in
keeping with previous published data (45
, 58)
, it is
conceivable that suppression of IL-8 secretion by Y.
enterocolitica pYV+ may rely on the
inhibition of NF-B activation by YopP/YopJ.
The most salient finding of this study is that exposure of human
epithelial cells to inert bead particles coated with the
carboxyl-terminal 195 amino acids of Y. enterocolitica
invasin results in degradation of IB, activation of the
transcription factor NF-B, and expression of the human IL-8 gene.
The production of IL-8 is largely controlled on the transcriptional
level and activated NF-B is a prime regulator of IL-8 gene
expression in response to different stimuli (34
, 35
, 36
, 51)
. In addition to NF-B, the human IL-8 promoter contains
binding motifs for other transcription factors, i.e., AP-1 and NF-IL-6
(34
, 35)
. NF-B and NF-IL-6 have been described to
synergistically activate IL-8 gene transcription (34
35
36
37
, 59)
. In Jurkat T lymphocytes, it has been shown that NF-B and
members of the NF-IL-6 family functionally cooperate in transcriptional
activation of the IL-8 gene (58)
. Analysis of IL-8
promoter reporter constructs revealed that the promoter fragment
spanning from nucleotide -112 to +44 contains the essential
cis-acting elements required to mediate invasin-triggered
up-regulation in HeLa cells. This essential IL-8 promoter fragment
contains one NF-B and two putative NF-IL-6 binding sites. In the
experiments presented here, we were not able to demonstrate a
contribution of C/EBP family members to control IL-8 expression in HeLa
cells by either reporter gene analysis or EMSA analysis. By contrast,
NF-B is absolutely required for invasin-mediated activation of the
IL-8 gene.
By means of electrophoretic mobility shift assays, we showed that
NF-B p50-p65 heterodimer, p65-p65, and p50-p50 homodimers are
activated in HeLa cells after infection with invasin-producing bacteria
or exposure to GST-Inv397-coated beads. The IL-8 B element appears
to preferentially bind p65-p65 homodimers. Exposure of HeLa cells to
E. coli pINV1914 or GST-Inv-coated beads induced degradation
of IB, suggesting a mechanisms for activation of NF-B
(38)
. However, in the experimental conditions used here,
degradation of IB in HeLa cells was observed only in the presence
of an inhibitor of protein translation, suggesting the induction of
rapidly operating feed back loops or a selective induction of IB,
but not IBß, degradation as observed in Listeria
monocytogenes-infected cells (60)
.
Blocking of invasin-triggered IL-8 secretion by MG132 or curcumin, both
of which inhibit NF-B, suggests that the induction of IL-8 secretion
by invasin in HeLa cells relies significantly on invasin-mediated
NF-B activation. Similar observations have been reported in H.
pylori infection. Thus, inhibition of IL-8 transcription and
subsequent secretion by NF-B inhibitors have been shown in gastric
epithelial cells after Helicobacter pylori infection
(40)
.
The question that arises now is whether, during Peyer’s patch
infection, the proinflammatory action of invasin via activation of
NF-B or the antiinflammatory action of YopP/YopJ via inhibition of
NF-B predominates. Preliminary data from our laboratory show that in
the early phase of intestinal Yersinia infection, a rapid
increase of cytokine and chemokine mRNA expression can be observed,
suggesting an activation of NF-B. Moreover, previous data from other
groups suggest that invasin plays an important role during colonization
of Peyer’s patches (31
, 61)
. In keeping with these
observations, we found that invasin is crucial for translocation of
Y. enterocolitica through M cells and that
Yersinia bacilli-colonizing Peyer’s patches at least
partially express invasin (unpublished results). On the other hand,
YopJ plays an important role during intestinal Yersinia
infection (58)
. Therefore, it is possible that both
activation and inhibition of NF-B may occur in infected Peyer’s
patches and that the resulting ‘netto effect’ may be decisive for the
pathogenesis of yersiniosis.
Killed invasin-exposing E. coli pINV1914 could induce
nuclear B DNA binding activity and IL-8 production in HeLa cells
(data not shown), suggesting that the binding of invasin-expressing
bacteria to the host cell surface is sufficient to induce signal
transduction pathways that lead to NF-B activation. More strikingly,
inert latex beads coated with purified recombinant GST-Inv protein
induced NF-B activation as well as elevated levels of IL-8 mRNA and
IL-8 secretion. This finding suggests that Y. enterocolitica
invasin, a known ligand of host cell ß1 integrins, induces activation
of host cell NF-B, thereby up-regulating transcription of the IL-8
gene. Nevertheless, at present we cannot exclude that additional
bacterial components may be involved in NF-B activation by
Yersinia. LPS, however, a common bacterial surface
component, does not appear to play a role in this process as addition
of LPS in the presence or absence of bacteria did not demonstrate any
effect on IL-8 transcription or secretion by epithelial cells (data not
shown).
In a previous study (33)
we showed that inhibition of
Y. enterocolitica internalization into epithelial cells by
wortmannin, an inhibitor of the PI3-kinase, does not interfere with
Yersinia-induced IL-8 secretion. This finding suggests that
bacterial adherence rather than bacterial invasion is the trigger for
IL-8 secretion by epithelial cells. In keeping with this observation,
we found that invasin-mediated activation of NF-B in HeLa cells
proceeds in the presence of wortmannin. These data are supported by a
recent report by Eaves-Pyles and colleagues (42)
, which
showed that activation of NF-B in Caco-2 cells by Salmonella
ssp. occurs in the absence of invasion. From these data we can
conclude that invasin triggers at least two pathways: one leading to
activation to NF-B, followed by IL-8 transcription; the other
leading to activation of PI3-kinase, followed by internalization of
bacteria. Preliminary data from our laboratory suggest that
invasin-induced signaling in epithelial cells involves p38
mitogen-activated protein (MAP) kinase, but not MAPKK (G. A.
Grassl et al., unpublished results).
The invasin proteins of Y. enterocolitica and Y.
pseudotuberculosis are outer membrane proteins that are involved
in the penetration of these bacteria into mammalian cells
(22)
. Invasin binds to ß1 integrins with high affinity.
The integrin binding domain of invasin has been mapped to the
carboxyl-terminal 192 amino acids of the molecule (62)
.
Expression of this fragment alone on the cell surface of noninvasive
bacteria is sufficient to confer the invasive phenotype of Y.
pseudotuberculosis (52)
. Binding of invasin to ß1
integrins on epithelial host cells induces tyrosine phosphorylation,
cytoskeletal rearrangement, and subsequent internalization of the
bacterium into the host cell (24
, 63)
. In B cells invasin
induces expression of several activation markers and proliferation
(64)
. In human T cells, invasin provides costimulatory
activity through interaction with the ß1 integrins (65)
.
The mechanisms by which binding of invasin to ß1 integrins leads to
these different types of cellular responses are unknown. Considering
the findings reported here, it is conceivable that activation of
NF-B by invasin might be a key step in these cellular responses.
In summary, our data suggest that invasin-coated particles such as
bacteria or beads bind to ß1 integrins on the surface of HeLa cells.
This leads to activation of NF-B, thereby inducing transcription of
the IL-8 gene. In human monocytic cell lines, ligand binding of
integrin or ligation of ß1 integrins with antibodies causes a rapid
tyrosine phosphorylation of proteins (66)
. In addition,
integrin ligation leads to nuclear translocation of the p50 and p65
subunits of NF-B and to increased levels of mRNAs for
immediate-early genes, including IL-1ß (67
, 68)
. The
actual signal transduction pathways leading to activation of NF-B in
epithelial cells on invasin ligations are a matter of ongoing research.
The specific interaction of Yersinia invasin with ß1
integrins resembles the action principle of pattern recognition
molecules/receptors (69)
and triggers effector mechanisms
of the innate immune system. Hence, Yersinia-induced IL-8
production may account for recruitment of PMN, one of the earliest host
responses to infection. However, Yersinia may subvert this
host defense mechanisms by switching on plasmid-encoded genes, which
render them resistant against phagocytosis by these cells. Invasin
binds more efficient to ß1 integrins than the natural ligand,
fibronectin, suggesting that invasin represents an optimized surface
for integrin binding in comparison with host substrates, possibly as a
result of convergent evolution (70)
. In keeping with these
findings, we found that invasin induces IL-8 responses in epithelial
cells to levels nearly as strong as, e.g., the agonist TNF-.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Present address: Plantamed GmbH, Kerschensteinerstr. 11–15, D-92318 Neumarkt OPf, Germany.
Received for publication September 20, 1999.
Revision received January 25, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
B in intestinal epithelial cells by enteropathogenic Escherichia coli. Am. J. Physiol. 273,C1160-C1167
