BIOLOGY OF REPRODUCTION (2012) 86(6):184, 1–8
Published online before print 7 March 2012.
DOI 10.1095/biolreprod.112.099309
Identification of Label-Retaining Perivascular Cells in a Mouse Model of Endometrial
Decidualization, Breakdown, and Repair1
Tu’uhevaha J. Kaitu’u-Lino,3,5,6 Louie Ye,5,6 Lois A. Salamonsen,7 Jane E. Girling,4,6 and Caroline E.
Gargett2,5,6
5
The Ritchie Centre, Monash Institute of Medical Research, Monash University, Clayton, Victoria, Australia
Centre for Women’s Health Research, Monash Institute of Medical Research, Monash University, Clayton, Victoria,
Australia
7
Prince Henry’s Institute, Monash Medical Centre, Clayton, Victoria, Australia
6
stromal expansion during the extensive remodeling associated
with this process.
ABSTRACT
The human endometrium is incredibly dynamic, undergoing
monthly cycles of growth and regression during a woman’s
reproductive life. Endometrial repair at the cessation of
menstruation is critical for reestablishment of a functional
endometrium receptive for embryo implantation; however, little
is understood about the mechanisms behind this rapid and highly
efficient process. This study utilized a functional mouse model of
endometrial breakdown and repair to assess changes in
endometrial vasculature that accompany these dynamic processes. Given that adult endometrial stem/progenitor cells
identified in human and mouse endometrium are likely
contributors to the remarkable regenerative capacity of endometrium, we also assessed label-retaining cells (LRC) as
candidate stromal stem/progenitor cells and examined their
relationship with endometrial vasculature. Newborn mouse
pups were pulse-labeled with bromodeoxyuridine (BrdU) and
chased for 5 wk before decidualization, endometrial breakdown,
and repair were induced by hormonal manipulation. Mean
vessel density did not change significantly throughout breakdown and repair; however, significantly elevated endothelial cell
proliferation was observed in decidual tissue. Stromal LRC were
identified throughout breakdown and repair, with significantly
fewer observed during endometrial repair than before decidualization. A significantly higher percentage of LRC were
associated with vasculature during repair than before decidualization, and a proportion were undergoing proliferation,
indicative of their functional capacity. This study is the first to
examine the endometrial vasculature and candidate stromal
stem/progenitor cells in a functional mouse model of endometrial breakdown and repair and provides functional evidence
suggesting that perivascular LRC may contribute to endometrial
blood vessels, endometrium, label-retaining cells, menstruation,
progenitor cells, stromal cells
INTRODUCTION
1
Supported by the National Health and Medical Research Council of
Australia (ID 490995 to T.J.K., ID 465121 to C.E.G., ID 545992 to
C.E.G.), Royal Australian and New Zealand College of Obstetricians
and Gynaecologists (to T.J.K.), Monash University, and the Victorian
Government’s Operational Infrastructure Support Program.
2
Correspondence: Caroline E. Gargett, The Ritchie Centre, Monash
Institute of Medical Research, P.O. Box 5418, Clayton, VIC 3168,
Australia. E-mail: [email protected]
3
Current address: University of Melbourne Department of Obstetrics
and Gynaecology, Mercy Hospital for Women, 163 Studley Road,
Heidelberg, VIC 3084, Australia.
4
Current address: Royal Women’s Hospital, Parkville 3052, Australia.
Received: 20 January 2012.
First decision: 24 February 2012.
Accepted: 1 March 2012.
Ó 2012 by the Society for the Study of Reproduction, Inc.
eISSN: 1529-7268 http://www.biolreprod.org
ISSN: 0006-3363
1
Article 184
Downloaded from www.biolreprod.org.
The human endometrium undergoes cyclical changes each
month in preparation for implantation. This includes cell
proliferation, differentiation, and eventual breakdown and
shedding during menstruation. Menstruation is a process that
occurs in humans and a limited number of other species. The
process of menstruation is rapid and efficient, ensuring
complete endometrial repair month after month during a
woman’s reproductive life (for review see [1]). Although
largely governed by female hormones estrogen and progesterone, the dynamic endometrium is also significantly influenced
by the production of local regulators, including cytokines and
growth factors [1, 2].
It has long been hypothesized that endometrial stem/
progenitor cells may play a key role in mediating endometrial
repair and subsequent tissue regeneration following menstruation [3–6]. The first evidence for these concepts, however,
only occurred recently with the initial identification of small
populations of clonogenic endometrial epithelial and stromal
cells, indicating the presence of at least two types of adult stem/
progenitor cells within the endometrium [7]. Since this finding
in 2004, a number of studies have sought to identify markers
that may assist in identification of these rare cells [8]; however,
functional evidence for their role in endometrial repair
following menstruation is still limited.
Epithelial and stromal label-retaining cells (LRC) have
recently been identified as candidate stem/progenitor cells in
mouse endometrium [9–12]. The LRC approach identifies
stem/progenitor cells by their quiescent nature [13–15]. This
technique relies on pulse labeling the majority of cell nuclei
with a DNA synthesis marker such as bromodeoxyuridine
(BrdU) at a time when stem cells are proliferating. Labeling is
followed by a chase period where slower cycling stem cells
will retain the BrdU label for longer periods of time, while
more mature, rapidly cycling transit amplifying cells dilute the
label to undetectable levels. BrdU immunohistochemistry
localizes the LRC and identifies their surrounding niche [10,
15].
Using the LRC technique, we have previously identified
candidate epithelial stem/progenitor cells in a functional model
of endometrial breakdown and repair [11]. We showed that
glandular epithelial cells retain the BrdU label for longer
periods than the luminal epithelial population, that they are
distinct in their capacity to proliferate during endometrial
KAITU’U-LINO ET AL.
rehydrated through descending grades of ethanol. Sections for PECAM1 were
treated with 0.1% pepsin in 3% glacial acetic acid for 10 min at 378C for
antigen retrieval. Sections were immersed in Dako blocking buffer (Dako,
Glostrup, Denmark) for 10 min and incubated for 1 h at 378C with primary
antibody, rat monoclonal PECAM1 (IgG1; PharMingen, San Diego, CA) at 0.5
lg/ml, or mouse monoclonal ACTA2 (IgG2a; Dako) at 70 lg/ml, diluted in 1%
bovine serum albumin in PBS (BSA/PBS). A negative control was included for
each tissue section by substitution of primary antibody with a matching
concentration of rat IgG1 or mouse IgG2a. For PECAM1, biotinylated anti-rat
IgG (Chemicon, Temecula, CA) diluted 1:200 in 1% BSA/Tris-buffered saline
(TBS) was applied for 1 h at RT, and the slides were washed as described
above. One drop of Dako LSAB þ AP streptavidin (Dako) was then applied for
15 min at RT before sections were washed, and Vector Blue (Vector,
Burlingame, CA) was applied for 10 min at RT. Sections were washed with
distilled H2O, and slides were mounted with Ultramount (Dako). For ACTA2,
the mouse Envision (horseradish peroxidase) system (Dako) was applied for 30
min at RT, and slides were washed as above. Diaminobenzidine (DAB)
solution (Dako) revealed the ACTA2 staining. Sections were then lightly
counterstained with Harris hematoxylin (Accustain; Sigma Diagnostics, Castle
Hill, NSW, Australia), dehydrated, and mounted using DPX mounting medium
(BDH Laboratory Supplies, Poole, U.K.).
repair, and that during repair they strongly express estrogen
receptor-a (ESR1). In contrast, the luminal epithelium
underwent rapid proliferation and dilution of LRC. We further
characterized candidate epithelial stem/progenitor cells in a
separate model of estrogen-induced endometrial regeneration
[15]. Here our data indicated that candidate epithelial stem/
progenitor cells played an important role in glandular
development, and we confirmed that mature epithelial cells
were likely to undergo self-replication in cycling endometrium
[15].
In this study we have used a functional model of
endometrial breakdown and repair to identify stromal LRC as
putative stromal stem/progenitor cells to investigate their
potential role in mediating the tissue remodeling and
hormone-independent endometrial repair observed in this
model [16]. Given that human endometrial stromal/mesenchymal stem cells are perivascular [17, 18], we have also utilized
the mouse model to examine endometrial vasculature throughout endometrial breakdown and repair, with particular focus on
PECAM1 (also known as CD31)-positive endothelial cells. As
mice have a bicornuate uterus, this model has the unique
capacity to provide untreated uterine horns within the same
mouse during treatment. As a consequence, the control horns
are exposed to all the same hormonal conditions as the treated
horns. Thus, a secondary aim was to assess changes in
endometrial vasculature during endometrial breakdown and
repair in relation to estrogen and progesterone in the untreated
horns.
PECAM1/Proliferating Cell Nuclear Antigen
Immunohistochemistry—Analysis of Proliferating
Endothelial Cells
MATERIALS AND METHODS
Animals
Female C57BL/6 mice of 6–10 wk of age (Monash Animal Services) were
housed in standard conditions with food and water provided ad libitum and a
constant light cycle of 12L:12D (lights on from 0800 to 2000 h). Ethics
approval for this project was granted by the Monash Medical Centre Animal
Ethics Committee A.
Counting of Endothelial Cells
Endothelial staining as determined by PECAM1 immunohistochemistry
was quantified to distinguish differences in microvessel density during
endometrial breakdown (Fig. 1, time C) and repair (Fig. 1, time D).
Longitudinal tissue sections were only included in the analysis of their
respective groups if their morphology matched our previously published
morphological scoring system [20]. Vessel density was determined by
assessment of the number of vessel profiles per unit area by counting 8–15
different areas for each longitudinal section for each animal. To determine the
percentage of proliferating vessels, vessel profiles containing one or more
proliferating endothelial cells were counted and expressed as a percentage of
total vessel profiles [21]. Finally, the ratio of the number of proliferating
endothelial cells per vessel was determined by counting the total number of
proliferating endothelial cells and dividing it by the total number of vessel
profiles counted [22]. Counts were conducted on areas of decidua, endometrial
breakdown, and within the subepithelial stroma. Control horns (that did not
undergo decidualization, breakdown, or repair) were collected at the same times
as the treated horns were collected. Control horns were collected to assess
differences resulting from the hormonal profile rather than events of
decidualization and breakdown. For each group, n ¼ 3–6 animals were
assessed. Data was expressed as mean 6 SEM.
Mouse Model of Endometrial Breakdown and Repair
At 3 days of age pups were pulse labeled with BrdU as previously described
[10]. At 5 wk of age, the standard protocol to induce endometrial breakdown
and repair was started, as previously described [19]. Briefly, under xylazine/
ketamine-induced anesthesia, mice were ovariectomized 7 days prior to the first
of three daily s.c. injections of 100 ng 17b-estradiol (Sigma Chemical Co., St
Louis, MO) in arachis oil at approximately 0900 h. After resting the mice for 4
days, progesterone implants were inserted s.c. into the back of each mouse.
Simultaneously, the mice received a single s.c. injection of progesterone (50
ng) in arachis oil, and an injection of 5 ng 17b-estradiol in arachis oil was given
at approximately 0900 h on that and the subsequent 2 days. At approximately
1100 h on the day of the final 17b-estradiol injection, 20 ll of sesame oil was
injected into the lumen of the right uterine horn of each mouse to induce
decidualization. The left horn remained untreated as a control. The
progesterone implants were removed 49 h after oil injection. Groups of mice
(n ¼ 6) were killed 1 day after the final priming estrogen injections (Fig. 1, time
A), at the time of endometrial decidualization (Fig. 1, time B), following
endometrial breakdown (Fig. 1, time C), or at the time of endometrial repair
(Fig. 1, time D). Uteri were cleaned of fat and weighed. For studies examining
endometrial vasculature, the control horn (no decidual stimulant applied) was
also collected. Tissue was cut into pieces and fixed in 4% paraformaldehyde
(PFA) for 4 h at room temperature (RT) and processed to wax in an orientation
that would provide longitudinal sections [11].
BrdU Immunohistochemistry—Analysis of LRC
BrdU immunohistochemistry was conducted on longitudinal sections of
uteri from animals at four stages during the mouse model of endometrial
breakdown and repair, as previously described [11, 15], to identify endothelial
or stromal LRC. The primary antibody used was sheep anti-BrdU (Biodesign,
Saco, ME) diluted in 1% fetal calf serum (FCS)/PBS to 8 lg/ml. A negative
control was included for each tissue section by substitution of primary antibody
with a matching concentration of sheep IgG. Biotinylated donkey anti-sheep
IgG (Jackson ImmunoResearch, West Grove, PA) diluted at 1:200 (final
concentration of 4 lg/ml) in 1% FCS/TBS was applied for 30 min at RT, and
the slides were washed as described above. The StrepABC horseradish
peroxidase kit (Dako) was then applied according to the manufacturer’s
instructions for 30 min at RT, and slides were washed as described above. DAB
PECAM1 and Alpha Smooth Muscle Actin (ACTA2)
Immunohistochemistry—Analysis of Endometrial
Vasculature
PECAM1 and ACTA2 immunohistochemistry was conducted on longitudinal sections of uteri from animals at four different stages during the mouse
model of endometrial breakdown and repair to identify endometrial vasculature.
Paraffin sections (3 lm) of PFA-fixed tissues were dewaxed in Xylene and
2
Article 184
Downloaded from www.biolreprod.org.
Double PECAM1/proliferating cell nuclear antigen (PCNA) immunohistochemistry was conducted on longitudinal sections of uteri from animals at the
four stages of endometrial breakdown and repair. PECAM1 staining was
revealed using Vector Blue as described above. Sections were then microwaved
for 20 min followed by cooling to RT for 30 mins. They were washed for 10
min in PBS, and Dako protein block was applied for 10 min at RT. Excess
block was drained from sections before mouse monoclonal PCNA (Novocastra,
Newscastle, U.K.) was applied at 0.9 lg/ml in 1% BSA/PBS for 1 h at 378C.
For negative controls, primary antibody was substituted with mouse IgG2a
diluted to the same concentration. Sections were then washed as described
previously before mouse Envision system was applied for 30 min at RT.
Sections were then washed as described above and mounted with Ultramount.
PERIVASCULAR LRC AND ENDOMETRIAL REPAIR
FIG. 1. Timeline of mouse model of endometrial breakdown and repair. Mice were pulse labeled with BrdU at 3 days of age (3do), twice daily for 3 days,
and the label was then chased for 5 wk (5w), after which mice were ovariectomized (OVX). One week later (6w þ 0), estrogen (E2) injections were
administered for 3 consecutive days. Mice were then rested for 4 days before a progesterone (P) injection was administered and a progesterone implant
inserted. On the same day and for the following 2 days, 5 ng 17b-estradiol was injected s.c. to prime the uterus for decidualization. On Day 8, a decidual
stimulus was applied, and 49 h later (0h) the progesterone implant was withdrawn. Tissue was collected 24 h following the final estrogen injection (A),
when tissue had decidualized (B), at the time of endometrial breakdown (C), and during endometrial repair (D). At these times both the treated and control
horn were collected for ongoing analysis.
BrdU Counting System—Quantification of LRC
Statistical Analysis
Longitudinal tissue sections were only included in the analysis of their
respective groups if their morphology matched our previously published
scoring system [20] (n ¼ 3–6 animals per group). BrdU staining dilutes from
the nucleus during cell division and is observed as partially stained nuclei [11].
As we were interested only in identifying long-lasting LRC that had either not
undergone proliferation or had only undergone one cell cycle, we only counted
cells that had greater than 75% of their nucleus as BrdU positive. Data was
expressed as mean 6 SEM.
After testing for normal distribution using a Kolmogorov-Smirnov test,
D’Agostino and Pearson omnibus normality test, and Shapiro-Wilk normality
test, statistical analysis was performed using the Kruskal-Wallis test, followed
by a Dunn multiple comparisons post hoc test. For comparisons between two
groups, the Mann-Whitney test was used. All statistical analysis was carried out
using PRISM version 4.03 for Windows (GraphPad, San Diego, CA). P 0.05
was taken as significant.
BrdU/PCNA Immunofluorescence—Analysis of
Proliferating Stromal or Endothelial LRC
RESULTS
Microvessel Density During Decidualization, Breakdown,
and Repair
Double BrdU/PCNA immunofluorescence was conducted on longitudinal
sections of uteri from animals in which uteri were collected following estrogen
injections and during endometrial repair. Paraffin sections (3 lm) of PFA-fixed
tissues were dewaxed in Xylene and rehydrated through descending grades of
ethanol. Sections were microwaved for 20 min in citrate buffer (pH 6.0) before
cooling to RT for 30 mins. The protocol for BrdU immunohistochemistry was
followed as detailed earlier. Alexa Fluor 568-conjugated donkey anti-sheep IgG
(Invitrogen, Grand Island, NY) was diluted 1:800 in 1% FCS/PBS and applied
for 45 min at RT in a darkened chamber. Sections were washed in PBS for 10
min before 20% normal goat serum in PBS was applied for 20 min at RT.
Primary antibody (PCNA) was then diluted in 10% normal goat serum (NGS)/
PBS/0.1% Tween 20 and applied for 1 h 15 min at RT before sections were
washed with PBS. Mouse IgG was applied at a matching concentration for
negative controls. Biotinylated goat anti-mouse IgG (Jackson ImmunoResearch) was diluted 1:200 in 10% NGS/PBS/0.1% Tween 20 and applied for 45
min at RT. Sections were washed with PBS and Alexa Fluor 488-streptavidin
(Invitrogen) diluted 1:800 in 10% NGS/PBS/0.1% Tween 20 and applied for 45
min at RT. Sections were then washed in PBS and 4 0 ,6-diamidino-2phenylindole (DAPI; Invitrogen) diluted in deionized water to a concentration
of 25 lg/ml was applied for 5 min at RT. Sections were then rinsed with
deionized water before wet mounting in Fluormount and stored in the dark.
Imaging of sections was carried out using a Leica DMR upright fluorescence
microscope with images acquired using a 403 objective and a color camera.
Fluorescence filters for DAPI (ex360/20, em425LP), Alexa 488 (ex480/40,
em527/30), and Alexa 568 (ex560/40, em645/75) were used. Images were
acquired using a 403 1.2NA objective.
PECAM1 immunostaining was used to assess microvessel
density during decidualization, breakdown, and repair. Strong
PECAM1 staining was observed throughout the decidual zone,
with large, elaborate vascular networks apparent (Fig. 2A).
Following progesterone withdrawal, many of these vascular
networks were destroyed, and fewer PECAM1-positive vessels
were observed (Fig. 2E) within the breakdown region, although
this difference was not significant (Fig. 2M), possibly due to
the presence of vessels at the endometrial/myometrial junction
as observed previously [10]. During endometrial repair,
PECAM1-positive vessels were observed in the ‘‘basalis’’
stromal zone underlying the repaired luminal epithelium (Fig.
2I), but the microvessel density did not differ significantly from
breakdown and decidualization zones (Fig. 2M). In the
untouched control horns, no differences in PECAM1-positive
vessel profiles were apparent (Fig. 2, B, F, and J), despite the
hormonal profile of the mice altering from high progesterone
during decidualization to absent hormones during breakdown
and subsequent repair. This observation was reflected in the
microvessel densities, where no significant differences between
control groups were observed (Fig. 2M).
PCNA/BrdU Counting System—Quantification of
Proliferating Stromal Cells and Stromal LRC
ACTA2 Coverage of Vessels During Endometrial
Breakdown and Repair
For each animal examined, a minimum of 500 stromal cells were counted,
and the number of proliferating cells (PCNA positive) as a percentage of total
stromal cells was determined. The number of stromal LRC (BrdUþ) and
Pericytes and smooth muscle cells stabilize blood vessels
and can be detected by ACTA2 immunoreactivity [22]. During
3
Article 184
Downloaded from www.biolreprod.org.
proliferating LRC (PCNAþBrdUþ) cells were also counted, and the percentage
of total stromal cells was similarly calculated as previously described [15]. Data
was expressed as mean 6 SEM.
solution revealed the BrdU staining. Sections were lightly counterstained with
Harris hematoxylin, dehydrated, and mounted using DPX mounting medium.
KAITU’U-LINO ET AL.
Downloaded from www.biolreprod.org.
FIG. 2. Immunohistochemical analysis of endothelial cells (PECAM1 positive) and perivascular cells (ACTA2 positive) in the mouse model of
endometrial decidualization, breakdown, and repair. PECAM1 (A, B, E, F, I, J) and ACTA2 (C, D, G, H, K, L) immunhistochemistry in uterine tissues during
decidualization (A, C), breakdown (E, G), and early endometrial repair (I, K). Untouched control horns were also immunostained from the same time
points (B, D, F, H, J, L). b, basalis; bd, breakdown zone; d, decidua; l, lumen; m, myometrium. Dotted lines are included to delineate different areas of
tissue. Bar ¼ 50 lM (B, D, E–L) and 100 lm (A, C). Microvessel density was assessed from PECAM1-immunostained tissue throughout the MMBR (M).
Data are expressed as mean 6 SEM of n ¼ 3–5 mice per group.
decidualization, ACTA2 staining was not observed within the
decidual zone (Fig. 2C) but was only detected within the
‘‘basalis-like’’ stromal area that had not undergone decidualization (adjacent to the myometrium) and between decidual
regions [11]. During breakdown and repair, ACTA2 staining
was observed within the stroma adjacent to the myometrium
(Fig. 2K). ACTA2 immunostaining was evenly distributed and
consistently observed on blood vessels throughout the
endometrial stroma in control tissue at all times (Fig. 2, D,
H, and L).
Identification of Proliferating Endothelial Cells
A large number of PECAM1-positive vessel profiles were
present in decidual tissue (Fig. 2M). The mouse model is one
of extensive tissue remodeling during decidualization, endometrial breakdown, and repair. In order to ascertain whether
4
Article 184
PERIVASCULAR LRC AND ENDOMETRIAL REPAIR
Downloaded from www.biolreprod.org.
FIG. 3. Assessment of proliferating endothelial cells in the MMBR. PECAM1/PCNA double immunohistochemistry on uterine tissues during
decidualization (A), endometrial breakdown (B), and endometrial repair (C). Untouched control horns were also examined from the same time points
during the model (D, E, F). PCNA reactivity is shown in brown and PECAM1 in blue. The percentage of proliferating vessels (G) and proliferating
endothelial cells per vessel profile (H) were determined throughout the MMBR. Data are expressed as mean 6 SEM of n ¼ 3–5 mice per group. **P 0.01 between decidua in treated and control horns. Arrows indicate proliferating endothelial cells. Dotted lines are included to delineate different areas of
tissue. b, basalis; bd, breakdown zone; d, decidua; e, epithelial cells; l, lumen. Bars = 100 lm.
angiogenesis was occurring during this remodeling, we
examined endothelial cell proliferation (Fig. 3). In decidual
tissue a significantly higher (P , 0.01) percentage of vessel
profiles contained one or more proliferating endothelial cells
compared to the control horn (Fig. 3, A, D, and G). Likewise,
the number of proliferating endothelial cells per vessel profile
was also significantly higher in decidual tissue compared to
control (Fig. 3H). No significant differences were observed for
proliferating endothelial cells or vessels between treated and
controls at any other time points (Fig. 3, G and H).
Identification of Stromal LRC During Endometrial
Breakdown and Repair
To assess whether stromal stem/progenitor cells are present
following endometrial breakdown, we set out to identify the
5
Article 184
KAITU’U-LINO ET AL.
Downloaded from www.biolreprod.org.
FIG. 4. Identification of stromal LRC and their relationship to endometrial vasculature. BrdU immunohistochemistry revealed the presence of
candidate stromal stem/progenitor (LRC) cells at the time of estrogen injections in the MMBR (A) and during endometrial repair (B). Black arrows indicate
examples of LRC (brown). Significantly fewer stromal LRC were identified during repair (C). Double BrdU (brown)/PECAM1 (blue) (D, E) and BrdU
(brown)/ACTA2 (blue) (G, H) immunochemistry identified a small population of LRC that were perivascularly localized. Round arrowheads indicate LRC
associated with vasculature; arrows indicate LRC not associated with vasculature. To quantify perivascularly located LRC, the percentage of
LRCassociated with PECAM1-positive vasculature was determined at the time of 17b-estradiol injection and during endometrial repair and expressed as
percentage of total stromal cells (F) or as percent of total LRC (I). To identify proliferating LRC, double immunofluoresence for BrdU (red; J, M) and PCNA
(green; K, N) was carried out at the time of estrogen injection (J–L) and during endometrial repair (M–O). Double-positive cells were identified as yellow
in the merged images (L, O) and are indicated by a white arrow. The percentage of proliferating stromal cells was determined at the time of estrogen
injection, during decidualization, and during endometrial repair (P) and the percentage of proliferating LRC was determined at the time of estrogen
injection and during repair (Q). Data are expressed as mean 6 SEM, n ¼ 4–6 mice per group except in Q, where the repair group is n ¼ 3 mice. *P 0.05 between decidua and estrogen-injected samples. Bars = 100 lm.
location of candidate stromal stem/progenitor cells during
endometrial breakdown and repair using the LRC method.
Following a chase period of approximately 6 wk, BrdU
immunohistochemistry was assessed in the stromal compartment (Fig. 4, A–C) prior to endometrial breakdown, during
decidualization, and during endometrial repair. Following
estrogen injections, 11.5% 6 6.1% (n ¼ 6) of total stromal
cells were BrdU positive (Fig. 4, A and C). The BrdU-positive
stromal cells were scattered throughout the endometrial horn,
with no obvious pattern of localization. At the time of
6
Article 184
PERIVASCULAR LRC AND ENDOMETRIAL REPAIR
decidualization, only cells with a decidual morphology were
examined, with only 0.14% 6 0.11% (n ¼ 4) of decidual cells
identified as BrdU positive (data not shown). During
endometrial repair, 0.73% 6 0.28% (n ¼ 3) of stromal cells
were identified as BrdU positive (Fig. 4, B and C), which was
significantly (P , 0.05) fewer than those observed following
estrogen injections.
Some Stromal LRC Are Associated with PECAM1-Positive
Endometrial Vasculature
Given previous reports showing that approximately one
third of endometrial LRC can be found in a perivascular
location [10], we carried out double immunohistochemistry for
BrdU/PECAM1 (Fig. 4, D and E) and BrdU/ACTA2 (Fig. 4, G
and H). In the estrogen-injected group, we observed that LRC
associated with PECAM1-positive vessels comprised 0.9% 6
0.3% (n ¼ 6) of total stromal cells, which was not significantly
different to those observed during repair (0.4% 6 0.13%,
n ¼ 3; Fig. 4F). When we assessed LRC associated with
PECAM1 vasculature as a percentage of total LRC, we
observed a significantly higher percentage in the repair group
compared to the estrogen-injected group (9.4% 6 3.87% vs.
2% 6 0.44%, n ¼ 3, P ¼ 0.05; Fig. 4I). We also observed
some LRC associated with ACTA2-positive vasculature during
endometrial breakdown and repair (Fig. 4, G and H),
particularly near the endometrial/myometrial junction as
reported previously [10, 11, 23].
Identification of Proliferating LRC
Given that approximately 12% of stromal LRC proliferate in
response to estrogen-induced endometrial regeneration [15],
we investigated whether stromal LRC in this breakdown-andrepair model initiated proliferation. To assess this functional
capacity of the identified stromal LRC, we carried out dual
immunofluorescence for PCNA/BrdU to identify LRC undergoing proliferation (Fig. 4, J–O). Initially, we quantified the
percentage of proliferating stromal cells (PCNA positive)
following estrogen injection, in decidual tissue, and during
endometrial repair (Fig. 4P). We observed the highest
percentage of proliferating stromal cells during decidualization,
when 25.6% 6 3.82% (n ¼ 5) of decidual cells were PCNA
positive, significantly higher (P , 0.05) than that observed in
the estrogen-injected group (9.35% 6 1.45%, n ¼ 5; Fig. 4P).
No difference for percentage of PCNA-positive cells was
identified between the estrogen-injected group and repair group
(Fig. 4P). Very low percentages of double-positive proliferating stromal LRC (BrdU þPCNA þ) were identified in the
estrogen-injected group (0.3% 6 0.1%, n ¼ 5) when assessed
as a percentage of total stromal cells (Fig. 4, J, K, L, and Q);
however, this equated to 22.8% 6 7.4% (n ¼ 5) of total BrdU þ
stromal cells. In the repair group, BrdU/PCNA-positive cells
were only identified in one of three animals assessed, with a
mean of 0.05% 6 0.04% (n ¼ 3); Fig. 4, M, N, O, and Q).
DISCUSSION
In this study, we have demonstrated the dynamic remodeling of endometrial vasculature associated with endometrial
breakdown and repair. We have shown that endothelial cells
undergo extensive proliferation to support the rapidly expanding decidua and that many of these vessels are destroyed during
endometrial breakdown. We have also demonstrated that
vascular proliferation during repair is limited. To assess the
contribution of stromal stem/progenitor cells, we utilized an
LRC approach and demonstrated that a small population of
7
Article 184
Downloaded from www.biolreprod.org.
candidate stromal stem/progenitor cells undergo proliferation
during the repair period and are closely associated with
endometrial vasculature.
Our examination of endometrial vasculature provided
evidence of significant angiogenesis during decidualization,
but not during endometrial repair. In this mouse model of
decidualization, breakdown, and repair (MMBR), we examined
very early epithelial repair, and thus it is possible that the
vessels we observed may have remained following breakdown.
If this was the case, vascular cells may only proliferate during
later estrogen-driven growth and development, which could be
equated with that observed during the human proliferative
phase. Thus, in this model, as in humans, we demonstrate a
rapid vascularization of the decidual tissue. Following
endometrial breakdown we showed that the surface epithelium
repairs before the proliferative expansion of the endometrium
occurs. Although beyond the scope of this study, it is expected
that extension of the model into a proliferative phase with
estrogen administration would stimulate angiogenesis within
the expanding stromal compartment.
To identify whether candidate stromal stem/progenitor cells
are involved in endometrial remodeling, we examined stromal
LRC at different stages of our MMBR model. We have shown
that following a long-term chase, a small proportion of stromal
LRC are identifiable within the endometrium, consistent with
previously published studies [10, 15]. Stromal growth occurs
during the 6-wk chase period (Postnatal Days 3–56), and hence
BrdU dilution is due to postnatal growth and development of
the prepubertal uterus rather than cyclical growth and
regression. Since endometrial stromal cells are well known
for their estrogen responsiveness (for review see [24, 25]), we
first examined the stromal compartment immediately following
a series of estrogen injections required to prime the
endometrium for decidualization. Our previous data [10, 11]
showed a high labeling efficiency for endometrial stromal cells
(65%) in mice pulse labeled with BrdU at 3 days of age [10].
After a 6-wk chase period and following the estrogen injections
used in our model, we observed that 11.5% of stromal cells
retained the BrdU label (LRC), indicating significant dilution
consistent with previous reports [10, 15]. At this time,
approximately 10% of stromal cells and 22% of total LRC
proliferated (PCNA positive), consistent with a different
estrogen-induced endometrial regeneration model [15], indicating that LRC preferentially proliferate in response to
estrogen stimulation. This important finding provides snapshot
evidence that LRC are functional and capable of proliferation
and are not cells that have simply incorporated BrDU in their
penultimate cell division. We suggest that this population of
functional LRC with proliferative capacity represent a pool of
candidate stem/progenitor stromal cells.
The highest stromal proliferation rates were observed during
decidualization, when 25% of decidual cells were identified as
strongly positive for PCNA. In line with this high level of
proliferation, very few true decidual cells were identified as
stromal LRC, reflecting rapid dilution of the label to
undetectable levels within this highly proliferative population.
Thus it is likely that stromal LRC themselves are not recruited
to decidualize, but remain in the basalis region in readiness for
endometrial regeneration following menstrual repair, a notion
supported by our previous observations that stromal LRC are
located at the endometrial/myometrial junction [10].
In human endometrium, the basalis is suggested as the
likely niche for stromal stem/progenitor cells [6, 26]. Previous
work from our laboratory has identified CD146 (official
symbol MCAM) and W5C5 as endometrial stromal stem cell
markers [17, 18, 27]. Furthermore, human endometrial
KAITU’U-LINO ET AL.
8. Gargett CE, Masuda H. Adult stem cells in the endometrium. Mol Hum
Reprod 2010; 16:818–834.
9. Cervello I, Martinez-Conejero JA, Horcajadas JA, Pellicer A, Simon C.
Identification, characterization and co-localization of label-retaining cell
population in mouse endometrium with typical undifferentiated markers.
Hum Reprod 2007; 22:45–51.
10. Chan RW, Gargett CE. Identification of label-retaining cells in mouse
endometrium. Stem Cells 2006; 24:1529–1538.
11. Kaitu’u-Lino TJ, Ye L, Gargett CE. Reepithelialization of the uterine
surface arises from endometrial glands: evidence from a functional mouse
model of breakdown and repair. Endocrinology 2010; 151:3386–3395.
12. Szotek PP, Chang HL, Zhang L, Preffer F, Dombkowski D, Donahoe PK,
Teixeira J. Adult mouse myometrial label-retaining cells divide in
response to gonadotropin stimulation. Stem Cells 2007; 25:1317–1325.
13. Braun KM, Watt FM. Epidermal label-retaining cells: background and
recent applications. J Investig Dermatol Symp Proc 2004; 9:196–201.
14. Taylor G, Lehrer MS, Jensen PJ, Sun TT, Lavker RM. Involvement of
follicular stem cells in forming not only the follicle but also the epidermis.
Cell 2000; 102:451–461.
15. Chan RW, Kaitu’u-Lino T, Gargett CE. Role of label retaining cells in
estrogen-induced endometrial regeneration. Reprod Sci 2012; 19:102–114.
16. Kaitu’u-Lino TJ, Morison NB, Salamonsen LA. Estrogen is not essential
for full endometrial restoration after breakdown: lessons from a mouse
model. Endocrinology 2007; 148:5105–5111.
17. Schwab KE, Gargett CE. Co-expression of two perivascular cell markers
isolates mesenchymal stem-like cells from human endometrium. Hum
Reprod 2007; 22:2903–2911.
18. Masuda H, Anwar S, Buhring H, Rao J, Gargett C. A novel marker of
human endometrial mesenchymal stem-like cells. Cell Transplantation (in
press). Published online ahead of print 27 March 2012; DOI 10.3727/
096368911X637362.
19. Brasted M, White CA, Kennedy TG, Salamonsen LA. Mimicking the
events of menstruation in the murine uterus. Biol Reprod 2003; 69:
1273–1280.
20. Kaitu’u-Lino TJ, Morison NB, Salamonsen LA. Neutrophil depletion
retards endometrial repair in a mouse model. Cell Tissue Res 2007; 328:
197–206.
21. Gargett CE, Lederman F, Heryanto B, Gambino LS, Rogers PA. Focal
vascular endothelial growth factor correlates with angiogenesis in human
endometrium. Role of intravascular neutrophils. Hum Reprod 2001; 16:
1065–1075.
22. Girling JE, Lederman FL, Walter LM, Rogers PA. Progesterone, but not
estrogen, stimulates vessel maturation in the mouse endometrium.
Endocrinology 2007; 148:5433–5441.
23. Ludwig H, Metzger H. The re-epithelization of endometrium after
menstrual desquamation. Arch Gynakol 1976; 221:51–60.
24. Gargett CE, Chan RW, Schwab KE. Hormone and growth factor signaling
in endometrial renewal: role of stem/progenitor cells. Mol Cell Endocrinol
2008; 288:22–29.
25. Maruyama T, Yoshimura Y. Molecular and cellular mechanisms for
differentiation and regeneration of the uterine endometrium. Endocr J
2008; 55:795–810.
26. Gargett CE, Chan RW, Schwab KE. Hormone and growth factor signaling
in endometrial renewal: role of stem/progenitor cells. Mol Cell Endocrinol
2008; 288:22–29.
27. Schwab KE, Hutchinson P, Gargett CE. Identification of surface markers
for prospective isolation of human endometrial stromal colony-forming
cells. Hum Reprod 2008; 23:934–943.
28. Crisan M, Yap S, Casteilla L, Chen CW, Corselli M, Park TS, Andriolo G,
Sun B, Zheng B, Zhang L, Norotte C, Teng PN, et al. A perivascular
origin for mesenchymal stem cells in multiple human organs. Cell Stem
Cell 2008; 3:301–313.
29. Du H, Taylor HS. Contribution of bone marrow-derived stem cells to
endometrium and endometriosis. Stem Cells 2007; 25:2082–2086.
REFERENCES
1. Salamonsen LA. Tissue injury and repair in the female human
reproductive tract. Reproduction 2003; 125:301–311.
2. Salamonsen LA, Kovacs GT, Findlay JK. Current concepts of the
mechanisms of menstruation. Baillieres Best Pract Res Clin Obstet
Gynaecol 1999; 13:161–179.
3. Leyendecker G, Herbertz M, Kunz G, Mall G. Endometriosis results from
the dislocation of basal endometrium. Hum Reprod 2002; 17:2725–2736.
4. Padykula HA. Regeneration in the primate uterus: the role of stem cells.
Ann N Y Acad Sci 1991; 622:47–56.
5. Padykula HA, Coles LG, Okulicz WC, Rapaport SI, McCracken JA, King
NW Jr, Longcope C, Kaiserman-Abramof IR. The basalis of the primate
endometrium: a bifunctional germinal compartment. Biol Reprod 1989;
40:681–690.
6. Gargett CE. Uterine stem cells: what is the evidence? Hum Reprod Update
2007; 13:87–101.
7. Chan RW, Schwab KE, Gargett CE. Clonogenicity of human endometrial
epithelial and stromal cells. Biol Reprod 2004; 70:1738–1750.
8
Article 184
Downloaded from www.biolreprod.org.
CD146 þPDGFR-b þ and W5C5þ cells are clonogenic and
multipotent [17, 18]. Using these markers, a perivascular
location was identified in both the functionalis and basalis,
suggesting a perivascular niche for the endometrial mesenchymal/stromal stem/progenitor cell population. More specifically,
these perivascular cells are directly adjacent to endothelial
cells, and their expression of additional markers indicates they
are pericytes [17, 28]. This, together with our previous
observations in murine endometrium [10], led us to examine
LRC in relation to PECAM1- and ACTA2-positive endometrial vasculature. Our observation of significantly more
perivascular LRC (as a percentage of total LRC) during
endometrial repair than in the estrogen-injected group suggests
that LRC may be recruited to the perivascular location possibly
from the bone marrow [29] during or following the insult of
endometrial breakdown. Alternatively, many of the BrdUpositive stromal cells remaining after the 5-wk chase and
estrogen injections may not be true stromal LRC, due to the
slow rate of stromal turnover during endometrial development
(compared to epithelial LRC) [10, 15], and hence stromal BrdU
dilution is incomplete at this time point in the model. This
would account for the smaller number of stromal LRC
associated with the perivascular niche at this time. We believe
that this population of nonvascular-associated BrdU-positive
cells are unlikely to be stromal stem/progenitor cells, but are
more likely to have either been labeled on their penultimate cell
division and/or are leukocytes derived from the bone marrow.
It is likely that the newly identified lineage marker W5C5 [18]
will be a tool that will more clearly elucidate the location of
stromal stem/progenitor cells and distinguish them from those
LRC not associated with endometrial vasculature and not
considered to be stromal stem/progenitor cells.
In summary, this study has demonstrated the vascular
changes that accompany endometrial breakdown and repair and
provided evidence for the existence of a population of stem/
progenitor cells involved in proliferation and expansion of the
endometrial stroma during the extensive remodeling observed
throughout this process.