| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
*
Institute of Virology and Departments of
Pathology and
Dermatology, University of Cologne, Cologne, Germany
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-inducible protein-10 was almost undetectable in all cell
lines tested. CD40L was able to induce MCP-1 production; however,
despite much higher CD40 expression in malignant cells, MCP-1 induction
was significantly lower compared with nontumorigenic cells. After
sensitization with IFN-, another T cell-derived cytokine showing
minimal effects on CD40 expression levels, CD40 ligation led to a more
than 20-fold MCP-1 induction in carcinoma cell lines. An even stronger
effect was observed for IFN--inducible protein-10. Our study
highlights the synergism of T cell-derived mediators such as CD40L and
IFN- for chemokine responses in cervical carcinoma cells, helping to
understand the chemokine expression patterns observed in
vivo. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Chemotactic cytokines are classified as CXC chemokines (
-chemokines)
or CC chemokines (ß-chemokines), depending on whether or not the
first two cysteines are separated by one amino acid. CXC chemokine
family members, i.e., the prototype IL-8 (2, 3) or growth-stimulatory
activity-, -ß, and -, are primarily responsible for the
recruitment of neutrophils. Exceptions are CXC chemokines such as
IP-103 (IFN--inducible protein) and MIG
(monokine induced by IFN-), which may act on T lymphocytes. CC
chemokines, i.e., the prototype MCP-1 (4, 5), MCP-2 to -4, RANTES, and
MIP-1 and -ß, preferentially attract monocytes, T lymphocytes,
eosinophils, and NK cells (reviewed in 6). Synthesis of chemokines
can occur in many cell types after stimulation with proinflammatory
cytokines, particularly IFN-, IL-1, and members of the TNF family,
including CD40L (7). However, which of these proinflammatory mediators
are primarily responsible for the induction of chemokines in cervical
carcinomas is as yet unresolved.
CD40, the receptor for CD40L, is a member of the TNFR family. It is expressed on B cells, monocytes (8, 9), dendritic cells (10), and a variety of nonhemopoietic cells, including normal keratinocytes (11), tumor cells, and many in vitro transformed and carcinoma-derived cell lines (12, 13, 14, 15). CD40L (gp39), a type II transmembrane protein with homology to TNF, has been shown to be transiently induced on CD4+ T cells upon activation (16). The interaction between CD40L and CD40 is not only crucial in B cell differentiation and activation (reviewed in 17), but is also important for different inflammatory responses of nonhemopoietic cell types (18). Examples are the induction of chemokines such as IL-8 in skin keratinocytes and MCP-1 in renal cells (7, 19) and the production of IL-6 (20), the latter being a growth factor for a variety of cell types including cervical carcinomas (21).
As a central signaling mechanism, CD40 induces the translocation of
NF-
B (22), a transcription factor regulating many genes such as
IL-6, TNF, and MHC class I and MHC class II controlling inflammatory
responses (23). Three intracellular proteins, TNFRassociated
factor (TRAF) 2, TRAF5, and TRAF6, that associate with the
cytoplasmic domain of CD40, were shown to mediate NF-B activation
(24, 25, 26). At least TRAF2 and TRAF6 involve the mitogen-activated
protein 3 kinase-related kinase NIK (27), a pathway
that is shared with the TNF receptor. In fact, NF-B is also involved
in the regulation of both prototypes of chemokine families, namely IL-8
(28) and MCP-1 (29, 30).
In vivo, MCP-1 is markedly expressed in cervical squamous cell carcinomas (SCCs) with an inflammatory reaction. In situ hybridization and immunohistochemical studies revealed that MCP-1 transcripts and protein predominated in neoplastic epithelial cells (31). This contrasts with the in vitro situation, in which cervical carcinoma cell lines, e.g., HeLa, were reported to express only extremely low MCP-1 mRNA levels (32, 33). It therefore was anticipated that additional environmental factors might exist that up-regulate chemokines in cervical carcinoma cells in vivo (31).
Our initial studies have shown high expression of CD40 in cervical
carcinoma and of CD40L in adjacent mononuclear cells in vivo. Hence, we
asked the question as to whether CD40 activation might be involved in
the regulation of chemokine production in HPV-positive epithelial
cells. We provide evidence that two synergistic stimuli, CD40L- and
IFN--dependent signals, are sufficient for efficient chemokine
induction in cervical carcinoma cells in vitro. These results might
explain the expression pattern of chemokines in cervical SCC observed
in vivo.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The HPV16- or HPV18-positive cervical carcinoma cell lines SiHa (ATCC HTB-35), CaSki (ATCC CRL-1550), and SW756 (kindly provided by Dr. M. von Knebel-Doeberitz, Heidelberg, Germany), and the HPV16-transformed foreskin keratinocyte cell line HPK IA (34) and baby hamster kidney (BHK) cells were cultured in DMEM (Life Technologies, Eggenstein, Germany). The routine culture medium for BHK cells transfected with CD40L cDNA (BHKCD40L) (35) contained 200 nM methotrexate (Sigma, Deisenhofen, Germany) and 100 µg/ml G418 (Life Technologies). All media were supplemented with 10% heat-inactivated FCS, 100 U/ml penicillin, 0.1 mg/ml streptomycin, 1 mM sodium pyruvate, and 2 mM L-alanyl-L-glutamine (all from Life Technologies). The in vitro HPV18 E6/7-transformed cell strains K51 and I56 (kindly provided by Dr. L. A. Laimins, Chicago, IL) were maintained in DMEM containing 25% Ham’s F12 medium, 10% FCS, 50 µg/ml gentamicin, 0.4 µg/ml hydrocortisone, 10-10 M cholera toxin, 5 µg/ml transferrin, 2 x 10-11 M triiodthyronine, 1.8 x 10-4 M adenine, 5 µg/ml insulin (all from Sigma), and 10 ng/ml epidermal growth factor (Life Technologies).
Immunohistochemistry
Ten tissue specimens with cervical intraepithelial neoplasia (CIN) representing all grades of dysplasia (CIN I-III) and six specimens with SCC of the cervix derived from the files of the Institute of Pathology, University of Cologne (Cologne, Germany), routinely fixed in 5% Formalin and paraffin embedded, were investigated immunohistochemically. After deparaffinization, the slides were subjected to microwave treatment for 3 x 5 min at 750W in 0.1 M citrate buffer for enhanced Ag retrieval. Subsequently, the ImmunoMax technique (36), modifying the catalyzed reporter deposition technique (37), was applied. The primary mAbs anti-CD40 mAb G28-5 (38) and anti-CD40L mAb (PharMingen, Hamburg, Germany) were diluted in 10% normal mouse serum and incubated overnight at 4°C. Biotinylated rabbit anti-mouse Ab (E354; Dako, Carpinteria, CA) was added for 30 min at room temperature, followed by a peroxidase-coupled streptavidin-biotin complex (k355; Dako) for 30 min at room temperature. In the next step, biotinylated tyramine solution (20 mg of N-hydroxysuccinimido- sulfo-LC-biotin (Pierce, Seattle, WA) dissolved in 0.5 ml of DMSO allowed to react with 6.4 mg of tyramine (Sigma) in 10 ml of 0.1 M borate buffer, pH 8) was applied for 10 min at room temperature. This solution had been preincubated for 72 h at 4°C, readjusted to pH 8, and stored at -80°C. Before application, it was diluted 1/50 (v/v) in Tris-buffered saline (TBS)/0.05% H2O2. StreptAB-alkaline phosphatase complex (K391; Dako) was incubated for 30 min at room temperature. Between all steps, threefold washing with TBS was performed. Finally, the reaction products were visualized using naphthol-AS-biphosphate and new fuchsin as chromogens. Nuclei were counterstained with hematoxylin.
The subpopulations of inflammatory cells that could be recognized in all specimens, i.e., T lymphocytes as well as myelomonocytic cells, were characterized by routine immunohistochemistry (ABC alkaline phosphatase) using the mAbs directed against CD8 (DK25; Dako), CD15 (LeuM1; Becton Dickinson, Heidelberg, Germany), and CD68 (PG-M1; Dako). Only the mAb OPD4 (anti-CD45R0; Dako), detecting a helper/inducer phenotype of T cells, was applied using the above-described ImmunoMax technique.
FACS analysis
Cells either unstimulated or after stimulation with 1000 U/ml
IFN- (Life Technologies) for 48 h were detached with PBS
containing 5 mM EDTA (Sigma). Following blocking with 2% BSA (Sigma)
in PBS, cells were incubated with 5 µg/ml anti-CD40 mAb (IgG1;
PharMingen) or MOPC-21 (Sigma) as an isotype-matched control. Cells
were then stained with FITC-conjugated goat anti-mouse
F(ab')2 (Dianova, Hamburg, Germany), and CD40 expression
was determined by flow cytometry (FACScan; Becton Dickinson, Mountain
View, CA).
Induction of cytokine production
Cells were either used unstimulated or prestimulated with 1000
U/ml IFN- (Life Technologies) for 24 h. They were then seeded
in 24-well plates at a density of 1.5 x 105
cells/well again in the absence or presence of 1000 U/ml IFN-,
respectively. After 24 h, they were stimulated with BHK cells
expressing CD40L, CD40L-negative BHK wild-type cells, or medium as a
control in a 300-µl volume. BHK cells had been freshly detached with
PBS containing 5 mM EDTA preserving CD40L expression on their surface.
After 16 h, cellular supernatants (SN) were collected,
centrifuged, and stored at -20°C.
Determination of cytokines by ELISA
Cytokine ELISAs were essentially done as described (20). Maxisorp plates (Nunc, Wiesbaden, Germany) were coated with 1 µg/ml anti-IL-6, anti-MCP-1, or anti-IL-8 mAb (PharMingen) or 2 µg/ml anti-RANTES mAb (Biosource, Fleurus, Belgium) overnight. After blocking of the plates for 1 h with PBS containing 0.5% BSA, 0.05% Tween 20 (Serva, Heidelberg, Germany), and 0.02% NaN3, SN or serial dilutions of the respective recombinant human cytokines (Tebu, Frankfurt, Germany) as standards were added for 6 h. In the case of RANTES, 5% skim milk was used as blocking reagent. Plates were then incubated with anti-IL-6, anti-MCP-1 pAb, or anti-IL-8 pAb at 0.5 µg/ml or anti-RANTES pAb at 1 µg/ml overnight (all pAbs were purchased from Tebu). For the detection of MCP-3 and IP-10, concentrations of the capture Abs were 5 and 2 µg/ml, and for the biotinylated detection Abs 3 µg/ml and 500 ng/ml, respectively (PharMingen and R&D, Wiesbaden, Germany). After 2-h incubation with peroxidase-labeled goat anti-rabbit F(ab')2, or streptavidin-peroxidase (Dianova) in the case of MCP-3 and IP-10, the substrate was applied and the extinction was measured with a SLT ELISA reader at 405 nm.
Electrophoretic mobility shift assay (EMSA)
Cervical carcinoma cell lines were grown in 10-cm culture
plates overnight to subconfluency. For CD40 stimulation,
paraformaldehyde-fixed BHKCD40L or BHKwt cells
were used. These cells were prepared as described (39). Briefly, BHK
cells were detached with 5 mM EDTA in PBS, fixed with 3%
paraformaldehyde, and washed six times with PBS. The cervical cell
lines were then stimulated either with medium or 2.4 x
107 BHK cells. After 1.5-h incubation at 37°C, fixed BHK
cells were removed, the carcinoma cells were harvested, and nuclear
extracts were prepared according to Dignam et al. (40). After
preincubation of the nuclear extracts containing 5 µg protein with
poly(dI-dC), NF-B-binding activity was examined. The extracts were
incubated for 20 min at room temperature with a 32P-labeled
double-stranded 30-bp probe from the mouse light chain enhancer
containing one NF-B site (41, 42). In supershift experiments, the
nuclear extracts were preincubated with the appropriate Abs
(anti-p50 and anti-p65; Santa Cruz Biotechnology, Santa
Cruz, CA) for 40 min before adding the probe.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In normal squamous epithelium of the cervix, CD40 was found only
in cells of the basal layer (Fig. 1
a).
Specimens with different stages of dysplasia showed a similar pattern
of reactivity regardless of the CIN grade (Fig. 1b). On the
other hand, CD40 expression was strongly up-regulated in SCCs. An
intense membrane-associated reactivity was observed, which was
accompanied by a slighter, diffuse cytoplasmic Ag distribution in all
parts of the tumors (Fig. 1c). However, neighboring
intraepithelial neoplasia behaved as described for CIN above. Five of
the six SCCs studied displayed a moderate to strong inflammatory
reaction; the sixth tumor exhibited a strong one. Numerous macrophages
were detected in the infiltrate, as revealed with mAb PG-M1, which
binds a macrophage-restricted epitope of the CD68 Ag (Fig. 1d). Infiltrates of the SCCs also contained granulocytes and
monocytes, as revealed by staining of the CD15 Ag. Moreover, both
CD45R0- and CD8-expressing lymphocyte populations were present in the
infiltrates surrounding tumor growth (not shown). CD40L was detected on
a subpopulation of tumor-infiltrating lymphocytes (Fig. 1e),
providing the basis for a paracrine stimulation mechanism via CD40L and
CD40 interactions in cervical SCC.
|
, lower
panel). Nontumorigenic keratinocytes that were transformed with
HPV in vitro (I56, K51, HPK IA) showed only low basal CD40 expression
(Fig. 2, upper panel) and, as with other cell types, CD40
expression was up-regulated by IFN- treatment (Fig. 2). Of these
cells, HPK IA was the one that displayed the greatest up-regulation of
CD40 expression in response to IFN- treatment.
|
HPK IA cells representing nontumorigenic in vitro
transformed keratinocytes were compared with the carcinoma cell lines
SiHa and SW756 with respect to MCP-1 production. All three cell lines
constitutively secreted only low levels of MCP-1 (less than 120 pg/ml)
into the culture medium (Fig. 3, upper
panel). The low MCP-1 production of cervical carcinoma cell lines
in vitro contrasted to the up-regulation of MCP-1 in cervical SCC
observed in vivo. This discrepancy might either be explained by
suppressive factors present in culture, but not in vivo, by the lack of
MCP-1 regulating intracellular signals in cultured cell lines or by
additional stimulatory factors present in vivo, but not in vitro. Our
analyses showed that TGF-ß1, which suppresses MCP-1
production at least in macrophages (43), was ruled out as a negative
regulator of MCP-1 in our carcinoma cells, as the investigated cell
lines did not produce any detectable amounts of TGF-ß1
(data not shown). Moreover, the transcriptional regulator SP-1, which
is relevant for constitutive human MCP-1 gene expression (29), was
prevalent in nuclear extracts of the cervical carcinoma cell lines
according to gel-shift experiments with a SP-1 binding site-specific
oligonucleotide (data not shown). We therefore speculated that in vivo
environmental factors were present, leading to the up-regulation of
MCP-1 in cervical carcinoma cells.
|
As CD40 was found to be highly expressed in cervical SCC, while
CD40L was detected in the tumor infiltrate, functional experiments were
performed to further analyze whether the CD40L-CD40 interaction could
account for MCP-1 expression in cervical carcinoma. Previous work has
shown that membrane-bound CD40L resulted in stronger stimulation of
CD40 compared with soluble, proteolytically released CD40L (20).
Therefore, CD40 was activated with CD40L-bearing BHK cells that did not
interfere with any of the cytokine assays (data not shown). Stimulation
of HPKIA cells with CD40L-bearing cells resulted in a more
than ninefold induction of MCP-1 (up to 922 pg/ml) (Fig. 3
A,
left panel). BHKwt cells did not alter MCP-1
production. Up-regulation of MCP-1 was also seen after stimulation with
paraformaldehyde-fixed BHK cells, although their stimulatory capacity
was weaker, as already described earlier (20). In contrast, SiHa and
SW756 cells were significantly less efficient at producing MCP-1 after
CD40 ligation. Indeed, SiHa and SW756 cell MCP-1 production reached
maximums of 278 and 322 pg/ml, respectively (Fig. 3A,
middle and right panel), this despite their
higher CD40 expression levels.
CD40 induces NF-B-binding activity in cervical carcinoma cell
lines
To obtain information on the mechanism of CD40-mediated MCP-1
induction, we investigated whether CD40 was able to activate the
NF-B transcription factor in cervical carcinoma cell lines. NF-B
had previously been shown to be crucially involved in MCP-1 regulation.
Nuclear extracts prepared from SW756 showed weak constitutive
NF-B-like binding activity (Fig. 4A,
lane 4). Membrane-bound CD40L strongly induced
NF-B-binding activity in both SW756 and SiHa (Fig. 4A,
lanes 2 and 5), while BHKwt cells did
not (Fig. 4A, lanes 3 and 6). In both
cell lines, two complexes (complex I and II) resulted in EMSA
experiments. Supershift analysis using specific Abs recognizing the p50
or p65 subunits of NF-B, respectively, revealed that complex I,
which was strongly mobilized by CD40L, mainly contained the p65 and
also p50 subunits of NF-B. The much weaker complex II only
contained p50/p50 homodimers (Fig. 4B). These results
indicated that binding activity of NF-B complexes is strongly
induced by CD40L in cervical carcinoma cells, with a predominance of
complex I probably representing functional p65/p65 homodimers and
p50/p65 heterodimers.
|
B-binding activity,
and the weak MCP-1 response after CD40 activation, one could speculate
on a selectively reduced inducibility of MCP-1 in these cell lines. To
put this hypothesis to the test, CD40-mediated induction of IL-6, a
growth factor for cervical carcinoma, was investigated. This cytokine
does not belong to the chemokine family, but the promoter of the IL-6
gene shares several regulatory elements, including binding sites for
NF-B responsible for transcriptional activation. We have reported
recently that IL-6 may be induced by CD40 in non-B cells and that its
induction correlates to the CD40 expression level within a cell line
(20). When HPK IA cells were stimulated with CD40L, no IL-6 production
was measurable (Fig. 3B, left panel), which could
therefore not be compared with MCP-1 induction by CD40. SiHa cells
produced up to 9.5 ng/ml of IL-6 after CD40 ligation, and IL-6
production after CD40 stimulation was even higher in SW756 cells
reaching 23.6 ng/ml (Fig. 3B, middle and
right panels). However, both of these cell lines also
constitutively expressed high amounts of IL-6. Thus, the CD40-mediated
inducibility of IL-6 in cervical carcinoma cells (1.3–2.4-fold) was of
the same order as that of MCP-1 (2.8–3.5-fold), but took place on a
much higher niveau. These experiments strongly argued against a
selective unresponsiveness of the MCP-1 gene to inflammatory stimuli.
IFN- induces MCP-1 production in HPV-positive cervical
keratinocytes
In vivo, CD40L may not represent the only stimulus responsible for
MCP-1 induction in cervical carcinoma. Besides CD40L, activated
CD4+ tumor-infiltrating T cells may also produce high
amounts of IFN-, a cytokine that is known to exert a regulatory
action on some chemokines, including MCP-1 (44, 45). IFN-
stimulation of SiHa cells raised the production of MCP-1 into the
culture supernatants only to low levels (258 pg/ml). However, a
stronger response was seen with the tumorigenic cell line SW756. The
latter showed an induction of MCP-1 to even higher levels than in HPK
IA cells (1200 and 819 pg/ml, respectively) (Fig. 5).
|
sensitizes cervical carcinomas for CD40-mediated MCP-1
production
Stimulation of IFN--preactivated cells with CD40L-bearing cells
led to a strong increase in MCP-1 production not only in HPK IA cells,
but also most notably in the tumorigenic cell lines SiHa and SW756. In
the case of SiHa, up to 1600 pg/ml MCP-1, and in the case of
SW756, up to 2800 pg/ml MCP-1 were produced (Fig. 5). Compared
with stimulation with medium alone, the combined effect of IFN-
preactivation and CD40L stimulation resulted in a more than 20-fold
induction of MCP-1 production by SiHa cells, a 25-fold increase for
SW756, and a 39-fold increase for HPK IA cells.
As shown in Fig. 2, IFN- stimulation also up-regulated CD40
expression in HPK IA cells and to a lesser extent in SiHa and SW756. To
examine whether IFN--preactivated carcinoma cells generally
responded better to CD40 ligation, IL-6 was again taken as a
"reference cytokine" and its production was measured. While in
IFN--preactivated HPK IA cells IL-6 could be induced at least to
measurable amounts (72 pg/ml) after CD40 activation, IL-6 levels of
CD40-stimulated SiHa and SW756 were not altered by IFN-
preactivation (data not shown). These findings suggested that
synergistic induction of MCP-1 after IFN- and CD40L stimulation is
not simply due to the up-regulation of CD40, which should have
influenced also the induction of other cytokines, i.e., IL-6, but takes
place on the level of intracellular signaling.
IFN- sensitizes cervical carcinomas for CD40L-mediated
CC-chemokine and IP-10, but not IL-8 production
To investigate whether the sensitization by IFN- was a
specificity of MCP-1 regulation or a more general phenomenon, we
measured the CD40L-mediated induction of other chemokines of the CC
(RANTES and MCP-3) and CXC chemokine family (IL-8 and IP-10) combined
with or without IFN- preactivation (Fig. 6).
|
). However, IFN- stimulation
clearly sensitized not only HPK IA, but also SW756 and to a lesser
extent SiHa cells for the induction of both chemokines by CD40L (Fig. 6, upper panel).
|
, lower
panel). The basal quantities of IP-10 in the supernatants of all
three cell lines were at the detection limit of the IP-10 ELISA
(between 0 and 75 pg/ml), and CD40L stimulation could increase the
IP-10 production only to low amounts (or not at all in the case of
SW756 cells). However, high IP-10 levels were measured already after
IFN- stimulation. Concentrations ranged from 5 ng/ml in SiHa cells,
30 ng/ml in HPK IA, to 38 ng/ml in SW756 cells. When these
IFN--sensitized cells were stimulated with CD40L, a strong synergism
was measured for IP-10 production, reaching up to 106 ng/ml in HPK IA,
66 ng/ml in SiHa, and 65 ng/ml in SW756 cells.
A regulation pattern different from that seen for the above-mentioned
chemokines was observed for IL-8, a chemokine that predominantly
attracts neutrophils to sites of inflammation (Fig. 6, lower
panel). IL-8 was induced more than 4.5-fold in HPK IA as well as
in SiHa cells after CD40 ligation and to a lower degree in SW756 cells.
IFN-, rather, had a negative than stimulatory effect. Thus, IL-8
production after CD40L stimulation was reduced in IFN--pretreated
cells.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Recently, it has been shown that cervical carcinoma cell lines lack MCP-1 mRNA expression in vitro (32). This chemokine attracts not only monocytes, but also CD4- and CD8-positive memory T cells (47). Studies on MCP-1 in vivo confirmed a low expression in high grade CIN. However, SCC of the uterine cervix showed a mild to strong expression of MCP-1 localized to the cancer cells themselves (31). Interestingly, MCP-1 expression was pronounced at the epithelial-mesenchymal interface and correlated with the inflammatory reaction within the tumor. Those differences in MCP-1 in vivo and in vitro led to the postulate that additional environmental factors were influencing this effect.
In the present study, we demonstrate that the interaction between CD40L
and CD40 in synergism with IFN- leads to a strong up-regulation of
MCP-1 and functionally related cytokines in cervical carcinoma cell
lines in vitro, a mechanism that could possibly explain the phenomena
observed in vivo.
CD40 was found to be strongly expressed on SCC of the uterine cervix. The ligand for CD40 was detected on a subset of infiltrating mononuclear cells. In vitro, CD40 was strongly expressed on cervical carcinoma cell lines, e.g., SiHa, SW756, and CaSki, corresponding well to the in vivo situation. The only exception were HeLa cells, which were therefore excluded from the study. Foreskin keratinocytes transformed in vitro with HPV16 or 18, i.e., HPK IA, I56, and K51 cells, showed only low CD40 expression levels.
In contrast to the in vivo situation, basal MCP-1 production was low in all cell lines tested in vitro. This low MCP-1 expression could neither be attributed to autocrine production of TGF-ß1, which can suppress MCP-1 production at least in macrophages (43), nor to the lack of the SP-1 transcription factor, which is required for constitutive MCP-1 expression (30). SP-1-binding activity was in fact strong in cervical carcinoma cells (data not shown). Therefore, our analyses suggested that in cultured cervical carcinoma cells, the MCP-1 gene might not be repressed in vitro, but might lack stimuli that act on the cells in vivo. As we had detected, both CD40L and CD40 being highly expressed in cervical carcinoma tissue, the consequence of the interaction between both molecules was investigated in vitro. Stimulation of the cervical carcinoma cells (SiHa and SW756) with CD40L induced MCP-1 production. However, when compared with the nontumorigenic cells (HPK IA), MCP-1 production was rather low, despite their much higher CD40 expression levels. Additional experiments revealed that the inducibility of MCP-1 was in a similar range as the inducibility of IL-6 in the carcinoma cells, indicating that there was no selective defect of regulation within the MCP-1 gene itself. As basal IL-6 production was already high in the carcinoma cells, IL-6 was strongly secreted after CD40 ligation.
For efficient gene induction, often two synergistic signals are
required. In the case of the IL-6 gene, at least one crucial signal
might be constitutively active in cervical carcinoma cells. CD40
activation may provide a second stimulus, probably via NF-B, which
is a regulatory factor for both IL-6 and MCP-1 (30, 48, 49) and is
inducible in these cells, as revealed by gel- and supershift analysis.
In the case of MCP-1, CD40L provides a stimulatory signal via NF-B,
which, however, seems not to be sufficient to explain the strong
up-regulation seen in vivo. As an intriguing hypothesis, T cells might
provide a second stimulus not via the same cytokine, but via different
effectors. In fact, IFN-, another T cell-derived cytokine that
induces MCP-1 in endothelial and some epithelial cells (44, 45), also
raised MCP-1 production in the cell lines investigated in this study.
Stimulation of IFN--primed cells with membrane-bound CD40L led to a
strong synergistic response (more than 20-fold), and MCP-1 was produced
at nanogram levels. These results suggest that NF-B activation might
be necessary for MCP-1 induction, but other signaling factors, e.g.,
those induced by IFN-, are additionally required to induce MCP-1
production in cervical carcinoma cells.
A similar effect, but even much stronger synergistic induction, was
observed for another T cell-attracting chemokine, IP-10, and to a
lesser extent for MCP-3 and RANTES. In contrast, neutrophil-attracting
IL-8 was induced in both cervical carcinoma cell lines and in vitro
transformed cells and reached comparable levels of expression.
Furthermore, IL-8 did not respond to IFN- pretreatment indicating
different regulatory mechanisms of IL-8.
We conclude from our results that production of MCP-1 is not
irreversibly repressed in cervical carcinoma cell lines in vitro. The
combined signals of CD40L and IFN- seem to be adequate stimuli to
up-regulate MCP-1 production in these cells. In vivo, both CD40 and
CD40L are expressed in close vicinity in cervical SCC. IFN- produced
by activated infiltrating T cells may sensitize the tumor cells to
produce MCP-1 after CD40L contact. This paracrine mechanism may account
for the high expression levels of MCP-1 at the mesenchymal-epithelial
interface in cervical carcinomas observed in vivo, contributing to the
control of disease progression. In cervical carcinomas showing no
intense inflammatory reaction, therapeutic application of IFN- may
be considered as a strong sensitizer for endogenous defense mechanisms
via CD40L-CD40-mediated responses.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Sigrun Hess, Institute of Virology, University of Cologne, Fürst-Pückler-Str. 56, 50935 Cologne, Germany. E-mail address:
3 Abbreviations used in this paper: IP-10, IFN--inducible protein; BHK, baby hamster kidney; CD40L, CD40 ligand; CIN, cervical intraepithelial neoplasia; EMSA, electrophoretic mobility shift assay; HPV, human papilloma virus; MCP-1, monocyte-chemoattractant protein-1; NF-B, nuclear factor-B; pAb, polyclonal Ab; SCC, squamous cell carcinoma; SN, supernatant; TRAF, TNFR -associated factor wt, wild type.
Received for publication September 28, 1998. Accepted for publication January 8, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
B and induction of IL-6 production. J. Immunol. 155:4588.[Abstract]
B. J. Immunol. 153:4357.[Abstract]
B in the immune system. Annu. Rev. Immunol. 12:141.[Medline]
B by TNF receptor 2 and CD40. Science 269:1424.B induction by TNF, CD95 and IL-1. Nature 385:540.[Medline]
B- and cis-regulatory enhancer binding protein-like factor binding elements in activating the interleukin-8 gene by pro-inflammatory cytokines. J. Biol. Chem. 265:21128.B and SP-1 regulate transcription of the human monocyte chemoattractant protein-1. J. Immunol. 153:2052.[Abstract]
B sites and NF-B/Rel subunit specificity. J. Biol. Chem. 272:31092.B: affinity, specificity, Zn2+ dependence, and differential half-site recognition. J. Biol. Chem. 266:252. regulation of human renal cortical epithelial cell-derived monocyte chemotactic peptide-1. Kidney Int. 44:43.[Medline]
B transcription factor. Mol. Cell. Biol. 10:2327.B-like transcription factor in the activation of the interleukin-6 gene by inflammatory lymphokines. Mol. Cell. Biol. 10:561.This article has been cited by other articles:
|
|
 |
|
  K. S. Hill, F. Errington, L. P. Steele, A. Merrick, R. Morgan, P. J. Selby, N. T. Georgopoulos, D. M. O'Donnell, and A. A. Melcher OK432-Activated Human Dendritic Cells Kill Tumor Cells via CD40/CD40 Ligand Interactions J. Immunol., September 1, 2008; 181(5): 3108 - 3115. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Usui, Y. Okunuki, T. Hattori, M. Takeuchi, T. Kezuka, H. Goto, and M. Usui Expression of Costimulatory Molecules on Human Retinoblastoma Cells Y-79: Functional Expression of CD40 and B7H1. Invest. Ophthalmol. Vis. Sci., October 1, 2006; 47(10): 4607 - 4613. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. Bugajska, N. T. Georgopoulos, J. Southgate, P. W. M. Johnson, P. Graber, J. Gordon, P. J. Selby, and L. K. Trejdosiewicz The Effects of Malignant Transformation on Susceptibility of Human Urothelial Cells to CD40-Mediated Apoptosis J Natl Cancer Inst, September 18, 2002; 94(18): 1381 - 1395. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Batrla, M. Linnebacher, W. Rudy, S. Stumm, D. Wallwiener, and B. Guckel CD40-expressing Carcinoma Cells Induce Down-Regulation of CD40 Ligand (CD154) and Impair T-Cell Functions Cancer Res., April 1, 2002; 62(7): 2052 - 2057. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Smola-Hess, U. Sandaradura de Silva, D. Hadaschik, and H. J. Pfister Soluble interleukin-6 receptor activates the human papillomavirus type 18 long control region in SW756 cervical carcinoma cells in a STAT3-dependent manner J. Gen. Virol., October 1, 2001; 82(10): 2335 - 2339. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. BASOK, A. SHNAIDER, L. MAN, C. CHAIMOVITZ, and A. DOUVDEVANI CD40 Is Expressed on Human Peritoneal Mesothelial Cells and Upregulates the Production of Interleukin-15 and RANTES J. Am. Soc. Nephrol., April 1, 2001; 12(4): 695 - 702. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. E. King, R. W. Kelly, H. O. D. Critchley, A. Malmstrom, M. Sennstrom, and R. P. Phipps CD40 Expression in Uterine Tissues: A Key Regulator of Cytokine Expression by Fibroblasts J. Clin. Endocrinol. Metab., January 1, 2001; 86(1): 405 - 412. [Abstract] [Full Text]
|
|
|
|
|
|
S. Hess, H. Smola, U. Sandaradura de Silva, D. Hadaschik, D. Kube, S. E. Baldus, U. Flucke, and H. Pfister Loss of IL-6 Receptor Expression in Cervical Carcinoma Cells Inhibits Autocrine IL-6 Stimulation: Abrogation of Constitutive Monocyte Chemoattractant Protein-1 Production J. Immunol., August 15, 2000; 165(4): 1939 - 1948. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. V. Jalukar, B. S. Hostager, and G. A. Bishop Characterization of the Roles of TNF Receptor-Associated Factor 6 in CD40-Mediated B Lymphocyte Effector Functions J. Immunol., January 15, 2000; 164(2): 623 - 630. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) or the isotype-matched control mAb MOPC-21 (
), in vitro
transformed cell lines (top) or cervical carcinoma cell
lines (bottom) were either left untreated
