| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Articles: Hormones and Reproductive Health |
Department of Obstetrics and Gynecology, University of Missouri, Columbia, Missouri 65212
Address all correspondence and requests for reprints to: Dr. Kathy L. Sharpe-Timms, Department of Obstetrics and Gynecology, University of Missouri, 1 Hospital Drive, Columbia, Missouri 65212.
Abstract
Based on our previous observation that peritoneal endometriotic (PE) lesions synthesize in vivo substantially more haptoglobin (Hp) than related eutopic tissues, we hypothesized that this increase in Hp production was due to endometrial-peritoneal interactions. As interleukin-6 (IL-6) stimulates Hp in other tissues and is produced by endometrial cells, we tested our hypothesis by evaluating the effects of IL-6 on Hp production by PE cells, normal peritoneal (P) cells, and eutopic endometrial cells from women with (UE-E) and without endometriosis (UE-C) using semiquantitative RT-PCR and enzyme-linked immunoabsorbent assay. Endogenous production of IL-6 was also assessed. Treatment with human recombinant IL-6 and dexamethasone significantly increased Hp production by P or PE cells in a dose- and time-dependent manner (P < 0.05). Hp messenger ribonucleic acid was not detected in UE-E and UE-C cells in the absence or presence of IL-6 and dexamethasone. PE and UE-E cells expressed significantly more IL-6 messenger ribonucleic acid than P and UE-C cells (P < 0.05). Moreover, UE-E cells secreted 6 times more IL-6 protein than UE-C cells (P < 0.05). These findings support our hypothesis and suggest a novel endometrial-peritoneal interaction whereby locally synthesized IL-6 and Hp may participate in the establishment and persistence of peritoneal endometriosis.
HAPTOGLOBIN (Hp) is an acute phase response plasma protein whose synthesis by the liver is increased severalfold by inflammation or injury (1). A variety of tissues and cells of other than liver origin have also been shown to synthesize Hp (2, 3, 4, 5, 6, 7, 8, 9, 10). Regulation of Hp expression is complex and both tissue and species specific (11). In human hepatic cells, maximal Hp expression requires glucocorticoids and interleukin-6 (IL-6) (12), whereas in rat hepatic cells, maximal expression requires glucocorticoids and both IL-6 and IL-1ß (13). Hp expression by rat intestinal epithelial (3) and by rat Sertoli cells (6) is not, however, influenced by IL-6. In vivo, inflammation increases Hp expression in rat liver (6) and murine liver, skin, spleen, kidney (4), adipose tissue (9), and lung (4, 10, 14). Yet, Hp expression can be reduced 4-fold by the same inflammatory stimuli in the rat testis (6).
Inflammatory cytokines, including IL-6, have been thought to contribute to the maintenance of peritoneal (15) and ovarian (16) endometriosis. Endometrial stromal cells from women with endometriosis have been reported to secrete significantly more IL-6 than endometrial stromal cells from normal women (17). In addition, stromal cells from ovarian endometriomata have been shown to produce significantly larger amounts of IL-6 than endometrial stromal cells from women without endometriosis (17, 18). No information is currently available regarding IL-6 production by peritoneal endometriotic (PE) and peritoneal (P) cells in comparison with eutopic endometrial cells from women with endometriosis (UE-E) and eutopic endometrial cells from women without endometriosis (UE-C) cells.
We previously reported that in vivo Hp messenger ribonucleic acid (mRNA) levels are significantly higher in peritoneal endometriotic lesions than in eutopic endometrial and peritoneal tissues. Individually, the two major components of the peritoneal endometriotic lesion, endometrium and peritoneum, express very low levels of Hp mRNA (2, 19, 20). These findings suggest that some interaction between peritoneal and endometrial tissues is needed to increase Hp gene expression. Thus, it is possible that endometrial IL-6 stimulates Hp production by peritoneum. Alternatively, immune cells that infiltrate endometriotic lesions (21, 22) produce cytokines that may induce the Hp gene. Therefore, we sought to determine whether IL-6 stimulates Hp synthesis by PE, P, UE-E, and UE-C cells and to assess the possible origin of this inflammatory cytokine.
Experimental Subjects
Tissues were obtained from informed volunteers (n = 26) routinely presenting to the physicians at the Department of Obstetrics and Gynecology at the University of Missouri School of Medicine, as approved by the institutional review board, Health Sciences Section. Women undergoing laparoscopy or laparotomy for evaluation of infertility, pelvic pain, tubal anastomosis, or elective tubal sterilization were recruited. Patients had a history of regular menses, and 24 of 26 women were not taking any sex steroids or steroid-modulating medication at the time of the surgery. Two patients, who contributed only peritoneal endometriotic lesions, were taking oral contraceptives when the surgical procedure was performed. When data from these two patients were excluded from the analyses, the statistical conclusions regarding the production of Hp and IL-6 were not altered.
Endometrial tissues were collected at the time of hysterectomy or with
an endometrial cell sampler (Endocell, Wallach, Orange, CT) from women
with endometriosis and without pelvic disease. Endometrial control
subjects underwent laparoscopy for tubal sterilization and reported no
symptoms of chronic pelvic pain or infertility. The absence of
endometriosis in control subjects was assessed by systematic
observation of the pelvis during laparoscopy in accordance with
American Society of Reproductive Medicine standards. Peritoneal
endometriotic lesions and peritoneum were obtained at the time of
laparoscopy or laparotomy as previously described (23).
Endometriosis was confirmed histologically by the presence of
endometrial glands and stroma. Endometria were dated histologically
according to Noyes’ criteria (24). Biopsy source,
patients’ age, menstrual cycle date, and stage of endometriosis were
also recorded (Table 1
).
|
Cell preparation and culture
P, PE, and UE cell cultures were established to study interactions between the endometrial tissue and peritoneum as an in vitro model for the establishment of peritoneal endometriosis. Our hypothesis stated that 1) P and PE cells produced an acute phase response protein (Hp) when stimulated an inflammatory cytokine (IL-6); and 2) only endometrial stromal cells from women with endometriosis synthesized and secreted enough amount of IL-6 to achieve this stimulation. To test our hypothesis, we evaluated basal Hp and IL-6 production by P, PE, and UE cell cultures as well as the effect of exogenous IL-6 on Hp synthesis and secretion by these cultures. The effect of UE-E-conditioned media on Hp production by P cells was also studied. P, PE, and UE tissue specimens for cell cultures weighed between 40–100 mg. Specimens were transported to the laboratory in saline solution. Tissues were minced into small pieces and dissociated with collagenase (Life Technologies, Inc., Grand Island, NY) and deoxyribonuclease (Sigma, St. Louis, MO) as previously described (23). No cell isolation procedures were performed. Cells from each individual specimen were plated at a density of 0.5–2 x 105 cells/mL in 35-mm diameter culture dishes (Becton Dickinson and Co., Franklin Lakes, NJ). After 16 h of incubation, nonadherent cells were washed away. This technique usually yields more than 90% stromal cells (25, 26). Cells were grown in 1 mL DMEM/Ham’s F-12 (Sigma) containing 10% heat-inactivated FBS (Life Technologies, Inc.), 100 IU/mL penicillin, and 100 µg/mL streptomycin (Sigma) in a humidified incubator at 37 C with 5% CO2. Adherent cells were identified immunocytochemically with specific monoclonal antibodies to cytokeratin (Biodesign, Kennebunk, ME) and vimentin (Roche, Indianapolis, IN). The proportions of cytokeratin-positive (endometrial epithelial or peritoneal mesothelial) cells and vimentin-positive but cytokeratin-negative cells (endometrial stromal or peritoneal subserosal) were assessed in each cell culture as described previously (27). Results (mean ± SEM) are presented as the percentage of cells that stained for vimentin, but not for cytokeratin, intermediate filaments. Normal mouse serum (Sigma) was used instead of the primary antibody as negative control. The presence of leukocytes was determined by immunostaining with a monoclonal antibody to human CD45 (PharMingen, San Diego, CA).
Cultures reached 90–95% confluence within 7 ± 1 days (mean ± SEM) of initial isolation. To evaluate basal Hp gene expression and the effect of exogenous IL-6 on Hp transcript levels, PE, P, UE-E, or UE-C cells were washed twice with PBS and then incubated in serum-free medium [phenol red free-DMEM/Ham’s F-12 containing 1% ITS+ (Collaborative Biomedical Products, Bedford, MA), 100 IU/mL penicillin, and 100 µg/mL streptomycin] for 24 h. Nontreated or basal cultures were then switched to serum-free medium containing 0.1% (vol/vol) ethanol. Treated cultures were incubated in the same serum-free medium containing 1 µmol/L dexamethasone (Dex; Sigma) in ethanol (0.1%, vol/vol, final concentration) for 14 h to increase the number of high affinity IL-6 receptors (28). This Dex concentration is equivalent to serum levels of endogenous glucocorticoids under inflammatory conditions (29, 30, 31). To evaluate the effect of Dex alone on Hp gene expression, a second set of controls was added to selected experiments; P or PE cells were incubated for 14 h in serum-free medium containing Dex. After this preincubation period, serum-free medium containing 0.1% ethanol was added to nontreated cultures. Serum free-medium containing either Dex alone or Dex and 100 ng/mL human recombinant (hr) IL-6 (Sigma) was added to the Dex control or treated cultures, respectively.
The effects of different doses of IL-6 on Hp production were studied by incubating P or PE cells for 14 h in serum-free medium containing 1 µmol/L Dex and then treating them with serum-free medium containing 1 µmol/L Dex and 0, 10, or 100 ng/mL hrIL-6 for 24 h. Accumulation of Hp in culture medium over time was evaluated by incubating P cells for 14 h in serum-free medium containing 1 µmol/L Dex and then treating them with serum-free medium containing 1 µmol/L Dex and 100 ng/mL hrIL-6 for 6, 12, or 24 h.
As UE-E cell culture medium contained high levels of IL-6 (see Results), the effect of UE-E cell-conditioned medium on Hp production by P cells was also evaluated. P cells were incubated for 14 h in serum-free medium containing 1 µmol/L Dex and then treated for 24 h with 1 µmol/L Dex and nontreated UE-E cell-conditioned medium, which contained 3 ng endogenous IL-6/mL. Basal Hp gene expression by these P cultures was determined by incubation with serum-free medium containing vehicle only.
Endometrial explant culture
The effect of hypoxia on Hp gene expression was tested by assessing the relative levels of Hp mRNA in a nontreated UE-E explant culture at different incubation times. The UE-E tissue specimen for explant culture weighed 800 mg (wet weight). A fraction was immediately frozen in liquid nitrogen for RNA isolation (time zero, representing in vivo mRNA levels). The rest was divided into three equal portions and incubated as described previously (23). After 12, 24, and 48 h of incubation, the culture medium was discarded, and the tissue was homogenized for RNA isolation.
Semiquantitative RT-PCR procedure
Basal Hp and IL-6 relative mRNA levels and the effect of
exogenous IL-6 and cell-conditioned medium on Hp gene expression were
evaluated by RT-PCR. Briefly, total RNA was isolated from the cultured
cells or the explants using the RNeasy kit (QIAGEN,
Valencia, CA), following the manufacturer’s instructions. The yield
was approximately 6 µg total RNA/1 x 106
cells. mRNAs in a given sample were reverse transcribed with an adapter
primer (Life Technologies, Inc.) as described previously
(2). For RT, 1 µg total RNA was used. Hp gene expression
by the cells with the various treatments was analyzed by
semiquantitative RT-PCR using glyceraldehyde-3-phosphate dehydrogenase
(GAPDH) as internal standard. Basal IL-6 mRNA and protein levels were
determined in nontreated cell cultures only. The composition of PCR
reactions were as follows: 1) Hp: 2 µL original complementary DNA
(cDNA) solution, 20 mmol/L Tris-HCl (pH 8.9 at 25 C), 50 mmol/L KCl, 2
mmol/L MgCl2, 0.2 mmol/L NTPs, 0.1 µmol/L
forward and reverse primers, and 1.25 U/µL recombinant Taq
DNA polymerase (Life Technologies, Inc.); 2) IL-6: 4 µL
original cDNA solution, 20 mmol/L Tris-HCl (pH 8.9 at 25 C), 50 mmol/L
KCl, 1.5 mmol/L MgCl2, 0.2 mmol/L NTPs, 0.2
µmol/L forward and reverse primers, and 1.25 U/µL recombinant
Taq DNA polymerase; and 3) GAPDH: 1 µL original cDNA
solution, 20 mmol/L Tris-HCl (pH 8.9 at 25 C), 50 mmol/L KCl, 1.5
mmol/L MgCl2, 0.2 mmol/L NTPs, 0.1 µmol/L
forward and reverse primers, and 1.25 U/µL recombinant Taq
DNA polymerase. Primer sequences and the sizes of PCR products from
cDNA and genomic DNA are shown in Table 2. The numbers of cycles needed to
perform the PCR in linear phase for target and reference amplicons were
determined as previously indicated (19). Cycling
conditions were as follows: 1) Hp: 96 C for 30 s, then 27, 30, 33,
or 36 cycles of 96 C for 30 s, 62 to 56 C for 30 s, and 72 C
for 90 s plus a 5-s autoextension, followed by a 10-min final
extension at 72 C; 2) IL-6: 96 C for 30 s, then 19, 22, 25, 28, or
31 cycles of 96 C for 30 s, 54 to 49 C for 30 s, and 72 C for
90 s plus a 5-s autoextension, followed by a 10-min final
extension at 72 C; and 3) GAPDH: 96 C for 30 s, then 22, 25, 28,
or 31 cycles of 96 C for 30 s, 55 to 50 C for 30 s, and 72 C
for 90 s plus a 5-s autoextension, followed by a 10-min final
extension at 72 C.
|
- and two
ß-subunits encoded by the same mRNA (20, 32). The
three human Hp phenotypes arise from two variant -chains. Our Hp PCR
conditions allowed us to distinguish among different Hp phenotypes
(20). Ethidium bromide-stained agarose gel electrophoresis
of PCR reactions and Southern blot analysis of Hp PCR products were
carried out as previously described (19). All PCR
amplification products originated from their respective cDNAs (Table 2
). No genomic contamination was detected in any of the
experiments. The ethidium bromide-stained PCR products were quantified by densitometry using the ImageQuant program (Molecular Dynamics, Inc., Sunnyvale, CA). Results were expressed as the ratio between the amount of the Hp or IL-6 amplification product over the GAPDH amplification product for each sample. These ratios were called Hp or IL-6 relative transcript or mRNA levels. Each densitometric value was calculated as an average of two independent PCRs. Experiments were conducted under identical conditions, or an internal control (cDNA from PE1 cells) was included to normalize for PCR efficiencies and photographic exposure. Results are presented as the mean ± SEM.
Hp enzyme-linked immunoabsorbent assay (ELISA)
The ability of P or PE cells to secrete Hp protein upon IL-6 and
Dex stimulation was assessed by ELISA. First, the specificity of the Hp
ELISA capture antibody was tested against PE cell culture medium by
Western blot analysis. Cells from the specimen PE1 were incubated for
72 h in serum-free medium without Dex or IL-6. The culture medium
was then concentrated using Ultrafree-0.5 centrifugal filter units
(Millipore Corp., Bedford, MA). An equal amount of
concentrated nonconditioned culture medium was used as a negative
control. Human hepatic Hp (Calbiochem, La Jolla, CA) was
used as a positive control. Concentrated samples were electrophoresed
through a 12% SDS-PAGE and transferred to nitrocellulose membrane
(Micron Separations, Inc., Westborough, MA) using established
protocols. The membrane was immunostained with a rabbit polyclonal
antibody to human Hp (DAKO Corp., Carpenteria, CA) as
previously described (2). To ensure that the signals on
the immunoblot corresponded to Hp - and ß-chains, a second
negative control was included, where the antibody to human Hp was
replaced by an equivalent amount of normal rabbit serum.
Second, the Hp concentration in the cell culture medium was determined as previously indicated (33) with minor modifications. Briefly, 96-well microtiter plates (Costar, Cambridge, MA) were coated overnight at room temperature, with 4 µg/mL rabbit antibody to human Hp (DAKO Corp.) as capture antibody. After three washes with PBS (pH 7.2)-0.05% Tween 20, 200 µL Hp standards (made in culture medium) or samples were added in duplicate and incubated for 1 h at 37 C. Wells were washed before adding a 1:3600 dilution of goat antibody to human Hp (Sigma) and incubated for 1 h at 37 C. After washing, a final incubation was performed with a 1:1000 dilution of alkaline phosphatase-conjugated rabbit antibody to goat IgG (Sigma) for 1 h at 37 C. Unbound conjugate was then washed away, and 1 mg/mL p-nitrophenol phosphate (Sigma) in 10% diethanolamine, 0.5 mol/L MgCl2, and 0.02% sodium azide was added. After 10 min at room temperature, the reaction was stopped with 3 mol/L NaOH, and the absorbance values were read at 410 mm in a Dynatech Corp. MR5000 ELISA reader (Dynex, Chantilly, VA). The assay detected 0.05 ng Hp/200 µL in a linear fashion up to 0.50 ng/200 µL, with intra- and interassay coefficients of variation of less than 7% and 6%, respectively. The number of cells was determined with the vital dye 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (Sigma) as previously indicated (25). The coefficient of variation for the 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide assay was 9%.
IL-6 ELISA
To determine the potential source of inflammatory cytokine (UE-E, UE-C, or P cells) in the endometrial tissue-peritoneum interaction, the IL-6 protein concentration in nontreated cell culture medium was quantified by ELISA. A monoclonal antibody to human IL-6 was used as the capture antibody (R\|[amp ]\|D Systems, Inc., Minneapolis, MN), and a polyclonal antibody to human IL-6 as the detection antibody (R\|[amp ]\|D Systems, Inc.). The capture antibody shows no cross-reactivity to human IL-1, IL-2, IL-3, IL-4, IL-5, IL-7, granulocyte-macrophage colony-stimulating factor (CSF), granulocyte CSF, or macrophage CSF (R\|[amp ]\|D Systems, Inc., specifications sheet). The detection antibody does not react with human IL-2, IL-4, IL-5, CSF-1, mouse IL-3, or mouse granulocyte-macrophage CSF. Briefly, 96-well microtiter plates (Costar) were coated overnight at room temperature with 2 µg/mL capture antibody. After three washes with PBS-0.05% Tween 20, 300 µL PBS containing 1% BSA (Sigma), 5% sucrose (Sigma), and 0.05% sodium azide were added per well and incubated at 37 C for 1 h. After washing three times with PBS-0.05% Tween 20, 200 µL IL-6 standards (made in culture medium) or samples were added in duplicate and incubated for 1 h at 37 C. Wells were washed again before adding 200 µL of a 1:1000 dilution of detection antibody and incubated for 1 h at 37 C. After the washings, 200 µL of a 1:2000 dilution of biotinylated antirabbit IgG (Vector Laboratories, Inc., Burlingame, CA) were added and incubated for 1 h at 37 C. After washing as described above, a final incubation was performed with a 1:1000 dilution of alkaline phosphatase-streptavidin conjugate (Vector Laboratories, Inc.) for 1 h at 37 C, and the assay was developed as indicated for the Hp ELISA. The assay detected 0.05 ng IL-6/200 µL in a linear fashion up to 0.75 ng/200 µL, with intra- and interassay coefficients of variation of less than 9% and 8%, respectively.
Statistical analysis
When needed, logarithmic transformation of the data was performed to homogenize variances and normalize the distribution (34). One-way ANOVA was used to detect differences in the proportions of different cell types in the mixed cultures (SigmaStat statistical software, Jandel, San Rafael, CA). Hp relative transcript levels among cell types for nontreated cultures were also analyzed by one-way ANOVA. When significant P values were determined, post-hoc analyses were performed using the Student-Newman-Keuls test. Model variables for two-way ANOVA included cell type (P vs. PE cells), treatment (nontreated vs. IL-6/Dex-treated cultures), and interaction (cell type x treatment). Treated and nontreated UE-E and UE-C cells were not included in this analysis because they did not express the Hp gene (see Results). No significant difference in Hp relative transcript levels was detected between P and PE cells (P = 0.233); therefore, data from cells were pooled, and the effect IL-6 on Hp production was analyzed by paired t test. No significant interaction was detected between cell type and treatment (P = 0.261). Hp dose-response data, IL-6 relative transcript levels, and IL-6 protein concentrations were examined by one-way ANOVA and Student-Newman -Keuls tests. The effects of Dex and Dex/IL-6 on Hp production by selected P and PE cell cultures and the accumulation of Hp protein in cell culture medium over time were analyzed by one-way repeated measures ANOVA due to the heterogeneity of the specimens.
Results
Cell cultures
Overall, more than 95% of the P, PE, UE-E, and UE-C cells
strongly reacted with the antivimentin antibody, but did not stain for
cytokeratin. Most P, PE, and UE cells were fuse-shaped and elongated
and displayed a pattern of single cell monolayer growth. A few clusters
of polygonal-shaped cells (100% minus percentage of mesenchymal cells)
stained for cytokeratin intermediate filaments and were, thus, of
epithelial/mesothelial origin. No significant difference in the
relative proportion of these cells was detected among the different
populations (Table 3; F3,
17 = 2.4; P = 0.14). These results
indicated that all cultures consisted basically of cells of mesenchymal
origin, with only a small proportion (<5%, overall) of epithelial or
mesothelial cells. In all cultures, less than 1% of the cells stained
with the antibody to human CD45.
|
All PCRs yielded products of the expected size. No contaminating
genomic DNA and/or primers-dimers were detected in any of the samples
tested (Table 2 and Fig. 1). In addition,
no amplification was obtained when reverse transcriptase was omitted
from the reactions (data not shown). A series of PCRs was carried out
to determine the linear amplification range as a function of the number
of cycles. Amplification was linear between 27 and 36 cycles for Hp, 22
and 31 cycles for GAPDH, and 19 and 25 cycles for IL-6 (Fig. 1).
Correlation coefficients were 1.000 for Hp and IL-6 and 0.997 for
GAPDH. The coefficient of variation among PCR efficiencies was 6%.
|
Immunoblot analysis of PE1 cell culture medium revealed a major
immunoreactive protein at approximately 42–47 kDa (Fig. 2, lane 1, arrow), similar to
the apparent molecular mass of human hepatic Hp ß-chain (lane 3). The
arrowhead in lane 1 indicates a minor band of about 52 kDa,
which is possibly uncleaved Hp -ß polypeptide precursor
(35). Human hepatic Hp -chain was visualized at the
bottom of the blot (lane 3). No signal was detected in an equivalent
amount of nonconditioned culture medium (lane 2) or when the primary
antibody was replaced by an equivalent amount of normal rabbit serum
(data not shown).
|
Under nontreated conditions, Hp relative mRNA levels were
undetectable in UE-E and UE-C (n = 9). Therefore, data from UE-E
and UE-C were pooled and are referred to as UE throughout the section.
However, in vivo, endometrial tissues express very low
levels of Hp mRNA that are detectable by RT-PCR (19) and
in situ hybridization (36). This difference in
Hp transcript levels may be due to hypoxia in endometrial tissues, as
low oxygen concentration stimulates Hp gene expression in a human
hepatoma cell line (37). This hypothesis was tested by
assessing the relative levels of Hp mRNA in nontreated endometrial
explant cultures at different incubation times. Unlike UE cells, Hp
mRNA was detectable in endometrial tissue explants, and its relative
levels increased with the incubation time (Table 4).
|
;
F2, 17 = 8.741; P = 0.002). PE
cells expressed significantly higher Hp relative transcript levels
(1.19 ± 0.62; n = 5) than P (0.10 ± 0.06; n = 7)
and UE cells (0.00 ± 0.00; n = 9; P <
0.05). In Fig. 3, UE-E and PE cells were matching specimens from the
same patient who has an Hp 2–1 phenotype. The patient who contributed
the P specimen has an Hp 1–1 phenotype.
|
Exogenous hrIL-6 in the presence of Dex stimulated Hp gene
expression by PE and P cells, but not by UE-E or UE-C cells (Fig. 3).
When cultured in the presence of 100 ng/mL hrIL-6 and Dex, the average
increase in Hp gene expression by PE and P cells was 3.1-fold over
nontreated levels (n = 9; P = 0.02). GAPDH
amplification products are depicted in Fig. 3B. The identity of the Hp
amplification product was further confirmed by Southern blot analysis
(Fig. 3C). In selected experiments (n = 3), a second set of
controls treated with Dex alone was included to determine the effect of
this glucocorticoid on Hp gene expression by P and PE cells. Similar to
the results presented above, IL-6 significantly increased Hp mRNA
levels over nontreated and Dex only-treated controls (F2,
2,4 = 20.147; P = 0.008) by 5-fold. No
statistical difference was detected between the the two sets of
controls.
Overall, nontreated P cells produced low levels of IL-6 (0.2 ± 0.1
ng/mL·24 h; n = 5), whereas PE cells secreted 1.8 ± 0.3 ng
IL-6/mL·24 h (n = 5). Exogenous hrIL-6 further stimulated Hp
production by P and PE cells in a dose-dependent manner (Fig. 4). When P and PE cells were incubated
with 10 ng/mL hrIL-6 in the presence of Dex for 24 h, Hp relative
transcript levels significantly increased over endogenous levels by
2.4-fold. Additional stimulation was achieved by the addition of 100
ng/mL hrIL-6 (F2, 14 = 11.294; P
= 0.001). Likewise, 10 ng/mL hrIL-6 significantly increased Hp protein
accumulation in these cell-conditioned media by 3-fold over
basal levels (Fig. 5), whereas 100 ng/mL
hrIL-6 stimulated Hp accumulation 3.8-fold over basal levels
(F2, 11 = 5.295; P = 0.025).
|
|
).
|
P cells produced only 0.2 ng IL-6/mL·24 h, which is
equivalent to 0.7 ng IL-6/1 x 106
cells·24 h. As the amount of IL-6 needed to elicit a significant
stimulation of Hp production over basal levels by P and PE cells was
much higher (10 ng IL-6/mL·24 h), we reasoned that the source of
endogenous IL-6 had to be other than P cells. Therefore, IL-6 relative
transcript levels were determined in nontreated P, UE-C, UE-E, and PE
cells (Fig. 7). The average relative IL-6
mRNA levels in PE and UE-E cells were 5.0- and 3.8-fold greater than
that in P cells, respectively, and 2.6- and 1.9-fold greater than that
in UE-C cells (F3, 19 = 9.237; P
< 0.001). Likewise, PE and UE-E cells secreted 10 and 6 times,
respectively, more IL-6 protein than UE-C and P cells (Fig. 8; F3, 17 = 9.537;
P < 0.001). No statistical differences were detected
between PE cells and UE-E cells or between P cells and UE-C cells for
both IL-6 mRNA and protein. Therefore, with the exception of cells from
established endometriotic lesions, only UE-E cells secreted enough IL-6
(4.7 ± 1.3 ng/1 x 106 cells/24 h) to
stimulate Hp production by P and PE cells (see below).
|
|
Our results have shown that only UE-E cells secreted sufficient
quantities of IL-6 to significantly stimulate Hp mRNA expression by P
cells over nontreated levels. These findings were confirmed when normal
P cells incubated with UE-E-conditioned medium (containing 3 ng
endogenous IL-6/mL) and Dex expressed Hp mRNA (Fig. 9, lane 1). No Hp PCR product was
visualized when these cells were incubated with serum-free medium
containing vehicle only (Fig. 9, lane 2).
|
This study has demonstrated for the first time that IL-6 stimulates Hp production by both human PE stromal/subserosal and normal P subserosal cells. Further, we have shown that PE stromal/subserosal and UE-E stromal cells produce significantly greater amounts of IL-6 than either normal P subserosal or UE-C stromal cells. As Hp interacts with a variety of immune cells to alter their response (38, 39, 40) and stimulates angiogenesis (40), localized production of Hp in the peritoneal cavity may contribute to the altered peritoneal immune status in women with endometriosis as well as participate in neovascularization and establishment of a blood supply to the developing endometriotic lesion. The fact that established peritoneal endometriotic lesions produce substantial amounts of Hp (19, 23) and IL-6 (this study) indicates that the lesions possess a mechanism to potentially escape peritoneal immune surveillance and maintain angiogenic support for their survival in this ectopic location.
Our observations suggest that eutopic endometrium from women with endometriosis, the purported precursor of the endometriotic lesion, has unique intrinsic characteristics that may help determine whether a woman will develop endometriosis. The most widely accepted theory to explain the etiopathology of peritoneal endometriosis (41) proposes that menstrual endometrial tissue reaches the peritoneal cavity through the fallopian tubes and implants on the surrounding pelvic surfaces. This theory is supported by the reported presence of menstrual tissue (42) and viable endometrial epithelial cells (43) in the peritoneal cavity during the early follicular phase in the majority of cycling women. As only a small number of menstruating women with patent tubes develop endometriosis (44), additional processes must be taken into account to explain this disparity. In this regard, we showed that UE-E stromal cells possess an IL-6 mechanism that is greatly diminished in UE-C stromal cells, to up-regulate Hp production in the peritoneal cavity.
Our results support the hypothesis that Hp production by peritoneum is the result of a local interaction between this tissue and the refluxed menstrual endometrial tissue. Compared with PE, normal P subserosal cells produce very low levels of IL-6. Further, peritoneal fluid of women with and without endometriosis has only picogram quantities of IL-6 per mL peritoneal fluid (16, 45, 46, 47). If peritoneal fluid and peritoneum are ruled out as sources of IL-6, then endometrial tissue may provide sufficient amounts of this inflammatory cytokine to stimulate peritoneal production of Hp. This may, in turn, contribute to the establishment and further maintenance of the endometriotic lesion. This idea is further supported by the fact that in the human endometrium, IL-6 reaches its maximal concentration during the late secretory phase and menses (48).
A very small amount of endometrial tissue is required to produce sufficient IL-6 to significantly increase Hp production by peritoneal cells. Using this cell culture protocol, endometrium routinely yields approximately 3 x 105 cells/mg (23). In our experimental conditions, endometrial stromal cells from women with endometriosis produced 4.7 ng IL-6/1 x 106 cells·24 h. Further, IL-6 from endometrial origin was biologically active, as UE-E cell-conditioned medium stimulated Hp gene expression by P cells. If we consider that 10 ng IL-6 significantly up-regulated peritoneal Hp production over basal levels, it will take as little as 8–10 mg endometrial tissue to produce sufficient IL-6 to stimulate Hp production by peritoneum. We were able to corroborate prior reports stating that purified endometriotic cells produce substantially more IL-6 than endometrial stromal cells from normal women (17). Similarly, in our P, PE, and UE cell cultures, endogenous IL-6 was mostly from stromal (17, 49) and subserosal origins, as the numbers of epithelial cells and immune cells were low. The high proportion of stromal or subserosal cells compared with epithelial or mesothelial cells in our cultures may be due to several factors, such as a higher original proportion of the former, a better plating efficiency, or a faster growth rate. PE and UE-E cells secreted 10 and 6 times more IL-6, respectively, but expressed only 2.6 and 1.9 times more IL-6 RNA than UE-C cells. This apparent discrepancy may be due to differences in the limit of detection between RT-PCR and ELISA. Our IL-6 ELISA will not detect any signal above background in samples, with an IL-6 relative transcript level of about 0.2 or less. Therefore, differences between culture media with high and low IL-6 concentrations will be more dramatic when assessed by ELISA.
Interestingly, Hp gene expression by eutopic endometrial cells was not detected in this cell culture model and cannot be stimulated under our experimental conditions, even though the IL-6 receptor is reportedly present on both the cell membrane and the cytoplasm of endometrial stromal cells (50). In vivo, eutopic endometrial tissues have low levels of Hp RNA that can be detected by RT-PCR (19) and in situ hybridization (36), and endometrial Hp can be also detected by immunocytochemistry (36). We showed that these differences may be due to lower O2 concentrations, as Hp gene expression by endometrial explant cultures increased with time in culture. Alternatively, our endometrial cell cultures may lack some Hp-stimulating factor that is present in the intact and explanted tissue.
In conclusion, peritoneal Hp production, triggered by an inflammatory cytokine of endometrial origin such as IL-6, may contribute to the establishment and persistence of endometrial tissue in an ectopic location. These findings suggest a mechanism by which locally synthesized Hp and IL-6 may participate in the etiology and pathophysiology of endometriosis.
Acknowledgments
We acknowledge the participation of the faculty and the resident physicians at the Department of Obstetrics and Gynecology, School of Medicine, University of Missouri Health Sciences Center, in the procurement of the human tissues for this study.
Footnotes
1 This work was supported by NIH Office for Research in Women’s
Health HD-29026-05A2 and TAP Holdings, Inc. (to K.S.T.). 
Received September 22, 2000.
Revised February 21, 2001.
Accepted March 21, 2001.
References
-chain
phenotypes. Proc of the 32nd Annual Meet of the Soc for the Study of
Reprod. 1999; 261.
2FSß
and 1Sß variants. FEBS Lett. 168:103–107.[CrossRef][Medline]
(IFN-) and interleukin-6 (IL-6) in
peritoneal fluid and macrophage-conditioned media of women with
endometriosis. Am J Reprod Immunol. 32:180–183.
This article has been cited by other articles:
 |
|
 |
|
  P. D'Amora, T. T. Maciel, R. Tambellini, M. A. Mori, J. B. Pesquero, H. Sato, M. J. B. C. Girao, I. D. C. Guerreiro da Silva, and E. Schor Disrupted Cell Cycle Control in Cultured Endometrial Cells from Patients with Endometriosis Harboring the Progesterone Receptor Polymorphism PROGINS Am. J. Pathol., July 1, 2009; 175(1): 215 - 224. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Defrere, J.C. Lousse, R. Gonzalez-Ramos, S. Colette, J. Donnez, and A. Van Langendonckt Potential involvement of iron in the pathogenesis of peritoneal endometriosis Mol. Hum. Reprod., July 1, 2008; 14(7): 377 - 385. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. W. Laschke and M. D. Menger In vitro and in vivo approaches to study angiogenesis in the pathophysiology and therapy of endometriosis Hum. Reprod. Update, July 1, 2007; 13(4): 331 - 342. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S Hombach-Klonisch, A Kehlen, P A Fowler, B Huppertz, J F Jugert, G Bischoff, E Schluter, J Buchmann, and T Klonisch Regulation of functional steroid receptors and ligand-induced responses in telomerase-immortalized human endometrial epithelial cells J. Mol. Endocrinol., April 1, 2005; 34(2): 517 - 534. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Ye, D. W. Cramer, S. J. Skates, S. P. Gygi, V. Pratomo, L. Fu, N. K. Horick, L. J. Licklider, J. O. Schorge, R. S. Berkowitz, et al. Haptoglobin-{alpha} Subunit As Potential Serum Biomarker in Ovarian Cancer: Identification and Characterization Using Proteomic Profiling and Mass Spectrometry Clin. Cancer Res., August 1, 2003; 9(8): 2904 - 2911. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Wieser, G. Fabjani, C. Tempfer, C. Schneeberger, M. Sator, J. Huber, and R. Wenzl Analysis of an Interleukin-6 Gene Promoter Polymorphism in Women With Endometriosis Polymorphism in Women With Endometriosis By Pyrosequencing Reproductive Sciences, January 1, 2003; 10(1): 32 - 36. [Abstract] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
)
for three independent experiments. Vertical bars
indicate the SEM. a vs. b,
P < 0.05. Inset, Ethidium
bromide-stained agarose gel electrophoresis, showing the RT-PCR
analysis of a representative experiment.
