|
Received: 24 August 2016 Accepted: 31 October 2016 DOI: 10.1111/aji.12609
ORIGINAL ARTICLE
Platelets are a possible regulator of human endometrial re-epithelialization during menstruation Koh Suginami1 | Yukiyasu Sato2 | Akihito Horie1 | Hisanori Matsumoto1,3 | Satoru Kyo4 | Yoshihiko Araki5 | Ikuo Konishi1,6 | Hiroshi Fujiwara1,7 1 Department of Gynecology and Obstetrics, Kyoto University Graduate School of Medicine, Kyoto, Japan 2
Department of Obstetrics and Gynecology, Otsu Red Cross Hospital, Otsu, Japan 3 Department of Gynecology and Obstetrics, National Hospital Organization Osaka National Hospital, Osaka, Japan 4
Department of Obstetrics and Gynecology, Faculty of Medicine, Shimane University, Izumo, Japan 5
Institute for Environmental and Genderspecific Medicine, Juntendo University Graduate School of Medicine, Urayasu, Japan 6
National Hospital Organization Kyoto Medical Center, Kyoto, Japan 7
Department of Obstetrics and Gynecology, Kanazawa University Graduate School of Medical Science, Kanazawa, Japan Correspondence Hiroshi Fujiwara, Department of Obstetrics and Gynecology, Kanazawa University Graduate School of Medical Science, Kanazawa, Ishikawa, Japan. Emails:
[email protected]; fuji@ kuhp.kyoto-u.ac.jp
Problem: The human endometrium periodically breaks down and regenerates. As platelets have been reported to contribute to the tissue remodeling process, we examined the possible involvement of platelets in endometrial regeneration. Method of study: The distribution of extravasating platelets throughout the menstrual cycle was immunohistochemically examined using human endometrial tissues. EM-E6/E7/hTERT cells, a human endometrial epithelial cell-derived immortalized cell line, were co-cultured with platelets, and the effects of platelets on the epithelialization response of EM-E6/E7/hTERT cells were investigated by attachment and permeability assays, immunohistochemical staining, and Western blot analysis. Results: Immunohistochemical study showed numerous extravasated platelets in the subluminar stroma during the menstrual phase. The platelets promoted the cell- to-matrigel attachment of EM-E6/E7/hTERT cells concomitantly with the phosphorylation of focal adhesion kinase. They also promoted cell-to-cell contact among EM-E6/E7/hTERT cells in parallel with E-cadherin expression. Conclusion: These results indicate the possible involvement of platelets in the endometrial epithelial re-epithelialization process. KEYWORDS
endometrial epithelium, menstruation, platelets, re-epithelialization, tissue remodeling
1 | INTRODUCTION
of basal glands.1,2 Based on the findings of an electron microscopic
The endometrium is an essential tissue for pregnancy because of
epithelial cells occurs prior to stromal expansion from the necks of the
its role in accepting and feeding the implanted embryo. To prepare
glands to meet migrating cells from neighboring glands.3 This begins
analysis performed by Ludwig and Spornitz, the spread of glandular
for this process, the human endometrium repeats monthly cycles of
on menstrual day 2, with full coverage of the luminal surface being
growth, differentiation, and breakdown (menstruation). During men-
achieved by days 5-6. However, in contrast to the above concept, pre-
struation, the withdrawal of ovarian progesterone and estrogen pro-
vious studies proposed stromal metaplasia as the source of cells for
duction provokes a spasm in the spiral artery, causing shedding of
the new surface epithelium.4,5 Recently, it was proposed that stem/
the upper two-thirds of the endometrium (functional layer), and the
progenitor cell populations such as endogenous epithelial progenitor
denuded surface of the endometrial stroma is then covered by a re-
cells, mesenchymal stem/stromal cells, the side population cells, and
generating epithelium to prevent infection and the leakage of tissue
label-retaining cells play a key role in endometrial regeneration during
fluid. After shedding of the functional layer, stumps of the glands
menstrual and estrus cycles.6
are retained in the lower one-third of the endometrium (basal layer).
Evidence from a rodent model suggested that estrogen was not
The most widely accepted theory is that the new surface epithelium
essential for endometrial restoration.7 In humans, circulating estro-
is derived from the proliferation and migration of the exposed ends
gen levels are very low and endometrial epithelial cells lack estrogen
Am J Reprod Immunol 2017; 77: e12609; wileyonlinelibrary.com/journal/aji DOI: 10.1111/aji.12609
© 2016 John Wiley & Sons A/S. | 1 of 8 Published by John Wiley & Sons Ltd
|
SUGINAMI
2 of 8
receptors until menstrual days 3-4. These findings suggest that coverage of the luminal surface by endometrial epithelial cells, the initial step in endometrial restoration, is independent of the actions of estrogen. In vitro studies revealed that hepatocyte growth factor,
et al
2 | MATERIALS AND METHODS 2.1 | Tissues and cell lines
epidermal growth factor, and transforming growth factor-α (TGF-α)
Endometrial tissue samples were obtained from normally cycling
promoted the proliferation and/or migration of endometrial epithe-
pre-menopausal women without endometrial lesions who had under-
lial cells.8,9 In addition, it was proposed that the abundance of leu-
gone hysterectomy for the treatment of benign uterine diseases or
kocytes in the endometrium during menstruation is involved in the
uterine cervical carcinoma in situ at Kyoto University Hospital (men-
remodeling process of menstruation by producing and activating
strual phase: n=5, proliferative phase: n=3, secretory phase: n=5).
10
suggesting the
None of the patients had been treated with hormonal medication
positive roles of cytokines such as growth factors and chemokines in
in the 3 months prior to surgery. The stage of the menstrual cycle
the re-epithelialization phase. Although this evidence suggests that
was histologically assessed according to the criteria of Noyes et al.24
peripherally circulating cells including immune cells may play import-
Conforming to the Declaration of Helsinki, informed consent for the
ant roles in the reconstructive process, the physiological contribution
use of the tissues was obtained from each donor. The use of these ma-
of platelets, one of the main populations of peripherally circulating
terials in this study was approved by the Ethics Committee of Kyoto
cells, to remodeling the process of the human endometrium remains
University Hospital.
matrix metalloproteinases within the endometrium,
EM-E6/E7/hTERT cells, an immortalized human endometrial ep-
unknown. Platelets are 1-to 4-μm-sized anucleate cells that break off in large numbers from megakaryocytes in the bone marrow. Circulating plate-
ithelial cell line, were produced by introducing HPV16 E6/E7 and TERT.25
lets accumulate at sites of endothelial disruption and, by the processes of adhesion, activation, and aggregation, form the primary thrombus, which is the first essential step in hemostasis. In recent years, it has
2.2 | Immunostaining
been increasingly acknowledged that platelets play multiple roles be-
The
yond hemostasis under various physiological and pathological condi-
Immunohistochemical staining of formalin-fixed paraffin-embedded
tions,11,12 including wound healing,13 atherosclerosis,14 closure of the
tissue sections against CD61 (a platelet marker) was carried
ductus arteriosus,
15
liver regeneration,
16
rheumatoid arthritis,
17
and
hypoxia-induced angiogenesis.18
antibodies
used
in
this
study
are
listed
in
Table 1.
out using the streptavidin-biotin-peroxidase method (Histofine, Nichirei Bioscience, Tokyo, Japan). Signals were visualized using
During human luteum formation, large numbers of immune cells
3,3′-diaminobenzidine (DAB, Merck, Darmstadt, Germany). All sub-
are distributed among luteinizing granulosa cells and these immune
luminal regions of each endometrial tissue slide were examined and
cells were proposed to contribute to the luteinizing process of granu-
divided into ten regions by the magnified area (×100) for quantifica-
losa cells.19,20 Along with immune cells, abundant extravasated plate-
tion. The CD61-positive platelet-extravasation sites in each area were
lets were observed to be deposited among the luteinizing granulosa
counted, the mean numbers per area were calculated as a number of
cells and in the fibrin net layer surrounding the central cavity. Platelet-
the platelet deposition site, and differences among the menstrual cy-
derived factors induced morphological changes in granulosa cells
cles were statistically evaluated by the Kruskal-Wallis test.
and migration of endothelial cells in vitro, suggesting the novel role of platelets in centripetal neovascularization and luteinization during human corpus luteum formation.21 In accordance with these findings,
2.3 | Preparation of isolated platelets
most of the non-hemostatic roles of platelets are fulfilled by the re-
Human platelets were isolated from healthy volunteers in the men-
lease of various bioactive molecules from activated platelets and/or
strual phase (females, 22-38 years old, n=25), as described previ-
a platelet-leukocyte interaction via P-selectin induced on activated
ously.26 Isolated platelets were suspended in PBS plus (PBS containing
platelets.
13
Based on this background, platelet-rich plasma has been re-
0.9 mmol/L Ca2+ and 0.33 mmol/L Mg2+) at a concentration of 4 × 108 platelets/mL and were immediately used in subsequent experiments.
ported to be clinically effective for the healing of uncontrolled chronic wounds.22,23 For example, during the cutaneous wound healing process, beneath the wound surface covered by blood clots, platelets
2.4 | Cell attachment assay
are activated and release a wide range of cytokines that can enhance
A cell attachment assay was performed as described previously.27
re-epithelialization and angiogenesis along with immune cells. The
EM-E6/E7/hTERT cells (1.0 × 105 cells) were used to inoculate
apparent similarity between wound healing and endometrial repair
24-well plates that had been pre-coated with diluted matrigel
suggests that platelets may also play some role in the latter process.
(0.5 μg/well) and were incubated for 1 hour in the presence of the
Therefore, in this study, to investigate the possible relevance of plate-
collagen type I-coated cell culture insert with a 0.4-μm pore mem-
lets in endometrial restoration, we examined the effects of platelets on
brane containing 0.8 × 108 platelets. After the wells had been gen-
the re-epithelialization process such as attachment ability and cell-cell
tly washed twice with PBS plus, the remaining cells were detached
contact.
with 0.05% Difco trypsine 250 (BD Biosciences) and counted as the
SUGINAMI
|
3 of 8
et al
T A B L E 1 List of antibodies used in this study
Alternatively, cells were permeabilized with 0.25% Triton-X-100 and incubated with rabbit anti-human pFAK pAb followed by FITC-
Antibody
Company
Mouse anti-human CD61 (glycoprotein IIIa) mAb (clone 2f2)
Novocastra Laboratories (Newcastle, UK)
FITC-conjugated rabbit anti-mouse immunoglobulin (Ig) pAb
Dako (Glostrup, Denmark)
FITC-conjugated swine anti-rabbit Ig pAb
Dako (Glostrup, Denmark)
FITC-conjugated rabbit anti-human Ig pAb
Dako (Glostrup, Denmark)
Horseradish peroxidase (HRP)- conjugated goat anti-mouse Ig pAb
Dako (Glostrup, Denmark)
Mouse anti-human integrin β (clone P4C10)
Millipore (Billerica, MA, USA)
Rabbit anti-human focal adhesion kinase (FAK) pAb
Millipore (Billerica, MA, USA)
Briefly, EM-E6/E7/hTERT cells were incubated in the upper insert for
HRP-conjugated goat anti-rabbit Ig pAb
Santa Cruz Biotechnology (Santa Cruz, CA, USA)
lower well coated with collagen type I was then replaced with PBS
Rabbit anti-human E-cadherin pAb
Santa Cruz Biotechnology (Santa Cruz, CA, USA)
Mouse anti-human E-cadherin mAb (clone HECD-1)
Takara Bio (Otsu, Japan)
Rabbit anti-human phospho-specific FAK (pFAK) [pY861] pAb
Life Technologies (Carlsbad, CA, USA)
Rabbit anti-human integrin β1 pAb
Cell Signaling Technology (Danvers, MA, USA)
Goat anti-mouse IgG (H+L) for negative control
Jackson ImmunoResearch Laboratories (West Grove, PA, USA)
Rhodamine-conjugated phalloidin (phallotoxin R415)
Molecular Probe (Eugene, OR, USA)
Rabbit anti-human β-actin pAb
Abcam (Cambridge, UK)
conjugated swine anti-rabbit Ig pAb, or reacted with rhodamine- conjugated phalloidin. Immunostained slides were examined under a confocal laser scanning microscope (Carl Zeiss Inc., Jena, Germany). These experiments were performed in triplicate, and the average of the ratio of intracellular cortical actin ring formation (ring-formed cells/whole cells) was calculated. Differences in the ratio between platelet-treated and non-treated groups were analyzed by the paired t-test.
2.6 | Permeability assay The permeability assay was performed as described previously.28 4 hours to form the compact cell monolayer. The medium filling the plus with or without 2.8 × 108 platelets. After another 3-hour incubation, the upper inserts were moved to new wells containing 700 μL of PBS plus and the medium in the upper inserts was replaced with 200 μL of DMEM high glucose containing 1% BSA and 10% trypan blue (Takara Bio). After a 2-hour incubation, the optical absorbance of the medium from each lower well was measured at 650 nm using an automated ELISA plate reader (Molecular Devices, Menlo Park, CA, USA). From the absorbance, the average of concentrations in triplicate of the trypan blue that had passed through the cell monolayer was calculated from standard curve. The experiment was repeated four times using platelets from different individuals. Differences were analyzed by the paired t-test.
2.7 | Western blotting EM-E6/E7/hTERT cells were lysed in RIPA buffer containing 0.2 mg/mL
attached cell number. These experiments were performed in tripli-
phenylmethylsulfonyl fluoride (Wako Pure Chemicals, Osaka, Japan)
cate, and the average of the ratio (attached cells/inoculated cells)
and 10 μg/mL leupeptin (Peptide Institute, Osaka, Japan). Each su-
was defined as the percent of attached cell number under each ex-
pernatant was electrophoresed in 7.5% SDS-polyacrylamide gel with
perimental condition. This experiment was repeated five times using
10% glycerol and transferred onto an Immobilon PVDF membrane
platelets from different individuals. Differences were analyzed by
(Millipore). The PVDF membrane was incubated with rabbit anti-
the paired t-test.
human integrin β1, E-cadherin, FAK, pFAK, or β-actin pAb followed
2.5 | Co-culture of EM-E6/E7/hTERT cells with platelets
using an ECL Western Blotting System (GE Healthcare, Piscataway,
by HRP-conjugated goat anti-rabbit Ig pAb. HRP was visualized NJ, USA). These experiments were repeated six times, and the intensities of the bands were quantified using Image J software and nor-
EM-E6/E7/hTERT cells (1.0 × 105 cells) pre-cultured in the matrigel-
malized to that of β-actin. Differences were analyzed by the paired
coated 24-well plates were incubated in the presence of the cell
t-test.
culture insert with a 0.4-μm pore membrane containing 0.8 × 108 platelets. Regarding immunocytochemistry, EM-E6/E7/hTERT cells were inoculated on coverslips coated with diluted matrigel in the well and were cultured in the presence of collagen type I-coated cell culture inserts with or without 2.0 × 108 platelets. Cells fixed with 0.5% paraformaldehyde on the coverslips were reacted with
3 | RESULTS 3.1 | Localization of platelets in the endometrium at various stages of the menstrual cycle
mouse anti-human E-cadherin mAb or mouse anti-human integrin
During the menstrual phase, a large number of CD61-positive plate-
β1 mAb followed by FITC-conjugated rabbit anti-mouse Ig pAb.
lets were observed in the extravascular regions of endometrial
|
SUGINAMI
4 of 8
(A)
(B)
(C)
(D)
(E)
(F)
et al
F I G U R E 1 Localization of platelets in the endometrium at various stages of the menstrual cycle. A-D, Endometrial tissues in the menstrual phase. B and D are magnified figures of the black squares in A and C, respectively. During the menstrual phase, numerous CD61-positive platelets (arrows) were observed in the extravascular regions in endometrial stromal tissues (A-D). E and F, Re-epithelialized sites at the end of the menstrual phase. F is a magnified figure of the black square in E. Aggregated platelets were detected on the endometrial stroma (arrows) just beneath the re-epithelized region of the luminal surface (arrowheads), while red cells (asterisks) were abundantly observed among the stromal tissues (F). Bars show 100 μm stromal tissues (Figure 1A-D). Immediately after re-epithelialization of
consequence of integrin activation, was promoted by the 30-minute
the luminal surface, aggregated platelets were detected beneath the
co-culture with platelet-inoculated culture inserts (Figure 4A).
luminal surface on the endometrial stroma, while red cells were abun-
An increase in the expression level of phosphorylated FAK by the
dantly observed among the stromal tissues (Figure 1E,F). The platelets
platelet co-culture was confirmed by the Western blot analysis
observed in the endometrial sections were unlikely to be an artifact
(Figure 4B-D).
of the operative procedures, because no platelet was detected in the uterine cervical sections taken from the same hysterectomy specimens (data not shown). In contrast to the breakdown of the endometrium during menstruation, no platelet was detected throughout the endometrium during the proliferative or secretory phase (data not shown). The number of
3.3 | Effects of platelets on morphological changes in EM-E6/E7/hTERT cells in association with intracellular cortical actin ring formation and cell surface recruitment of E-cadherin
platelet-extravasation sites was significantly higher in the menstrual
As the above results suggested the involvement of platelets in the
phase than in the secretory and proliferative phases (Figure 2).
process of endometrial re-epithelialization, we examined morphological changes in EM-E6/E7/hTERT cells in the presence or absence of
3.2 | Co-culture with platelets promoted attachment of EM-E6/E7/hTERT cells to matrigel concomitant with FAK phosphorylation
the platelets. Although EM-E6/E7/hTERT cells assumed a widespread shape in the control culture, the intracellular actin stress fibers were aligned in a textile pattern (Figure 5Aa). On the other hand, after the 3-hour co-culture with platelets, EM-E6/E7/hTERT cells became po-
We performed a cell attachment assay using E6/E7/hTERT cells in the
lygonal in shape and intracellular cortical actin ring formation was
presence of platelets. The co-culture with platelets significantly en-
significantly increased (Figure 5Ab,B). Furthermore, immunocyto-
hanced their attachment to matrigel, which has a similar composition
chemistry of paraformaldehyde-fixed non-permeabilized EM-E6/E7/
to that of the basement membrane (Figure 3A).
hTERT cells showed that the cell surface expression of E-cadherin,
The cell surface expression (Figure 3B) as well as total expres-
which is involved in the homophilic adhesion of epithelial cells, was
sion levels (Figure 3C,D) of integrin β1, one of the major integrins
increased by the platelet co-culture (Figure 6A). A Western blot
expressed on endometrial epithelial cells,
29
were not significantly
analysis demonstrated that the total expression level of E-cadherin
affected by the 3-hour co-culture with platelets. However, the phos-
in E6/E7/hTERT cells was also increased by the platelet co-culture
phorylation of focal adhesion kinase (FAK), which mainly occurs as a
(Figure 6B,C).
SUGINAMI
|
5 of 8
et al
3.4 | Co-culture with platelets promoted cell-to- cell contact In the permeability assay, the amount of pigment that passed through the E6/E7/hTERT cell monolayer formed on the matrigel-coated membrane was significantly reduced after the co-incubation with platelets (Figure 6D). As cell proliferation was not affected by the platelet co-culture (data not shown), the reduction observed in the permeating pigment suggested an enhancement in cell-to-cell contact in the E6/E7/hTERT cell monolayer.
4 | DISCUSSION In the present study, we immunohistochemically confirmed that platelets were mainly deposited just beneath the regenerating endometrial F I G U R E 2 The number of the deposition sites of the extravasated CD61-positive platelets in subluminal regions of the endometrium during the menstrual, proliferative, or secretory phase. All subluminal regions of each endometrial tissue slide were examined and divided into ten regions by the magnified area (×100) for quantification. The CD61-positive platelet-extravasation sites in each area were counted, the mean numbers per area were calculated as a number of the platelet deposition site, and differences among the menstrual cycles were statistically evaluated by the Kruskal-Wallis test. In contrast to the breakdown of the endometrium during menstruation, no platelet was detected throughout the endometrium during the proliferative or secretory phase. The number of platelet-extravasation sites was significantly higher in the menstrual phase than in the secretory and proliferative phases. **, P
