ArticlePDF Available

Transcriptional heterogeneity of stemness phenotypes in the ovarian epithelium

Authors:

Abstract and Figures

The ovarian surface epithelium (OSE) is a monolayer of epithelial cells surrounding the ovary that ruptures during each ovulation to allow release of the oocyte. This wound is quickly repaired, but mechanisms promoting repair are poorly understood. The contribution of tissue-resident stem cells in the homeostasis of several epithelial tissues is widely accepted, but their involvement in OSE is unclear. We show that traits associated with stem cells can be increased following exposure to the cytokine TGFB1, overexpression of the transcription factor Snai1, or deletion of Brca1. We find that stemness is often linked to mesenchymal-associated gene expression and higher activation of ERK signalling, but is not consistently dependent on their activation. Expression profiles of these populations are extremely context specific, suggesting that stemness may not be associated with a single, distinct population, but rather is a heterogeneous cell state that may emerge from diverse environmental cues. These findings support that the OSE may not require distinct stem cells for long-term maintenance, and may instead achieve this through transient dedifferentiation into a stem-like state.
TGFB1 promotes stemness in the OSE a Primary (left; n = 3) and secondary (right; n = 5) sphere-forming capacity of mOSE cells treated with TGFB1 (10 ng/mL). Data points represent the average number of spheres per 4 fields of view for each replicate after 14 days of spheroid culture. b Left: Phase contrast images of control and TGFB1-treated spheroids (left). Scale bar = 100 μm. Right: Relative sizes of spheroids in control and TGFB1 conditions (primary spheroids n = 3; secondary spheroids n = 4). Each point on the plot represents the average size of spheroids from 4 fields of view for 4 separate wells. c Primary sphere-forming capacity of human OSE treated with TGFB1 (n = 4). hOSE spheroids were cultured in methylcellulose to prevent aggregation. Spheroids were cultured for 28 days. Scale bar = 100 μm. d Fold change values for a panel of putative stem cell markers in mOSE treated with TGFB1 for 4 days. e Relative protein quantifications of CD44 throughout a time course of TGFB1 treatment in mOSE (n = 3). Quantifications represent Western blot pixel densitometry, normalized to B-actin and scaled to the mean intensity in untreated samples. A representative blot is included in Supplemental Fig. 2b. f Immunofluorescence co-stain of CD44 and Ki67 in control and TGFB1-treated spheroids. White arrows highlight CD44 + cells in control spheroids. Scale bar = 100 μm, inlet scale bar = 15 μm. Separate channels are shown in Supplementary Fig 3g. Primary sphere-forming capacity of CD44- and CD44 + mOSE cells (n = 3). All boxplots show median value (horizontal black line), estimated 25th and 75th percentiles, and whiskers represent 1.5 times the interquartile range. Linear regression models were used for all statistical tests. *p < 0.05, **p < 0.01.
… 
Snail overexpression promotes stemness with minimal gene expression changes a Primary (left) and secondary (middle) sphere forming capacity of mOSE cells overexpressing Snail (n = 3). Data points represent the average number of spheres per 4 fields of view for each replicate. Boxplots show median value (horizontal black line), estimated 25th and 75th percentiles, and whiskers represent 1.5 times the interquartile range. Right: Phase contrast images of control (inducible GFP) and Snail-overexpressing spheroids. Spheroids were cultured for 14 days. Scale bar = 100 μm. b Plot showing the distribution of differentially expressed genes following Snail overexpression. Each point corresponds to a single gene. Selected genes related to stemness and/or the EMT are highlighted on the plot. Dashed lines correspond to significance criteria (absolute log fold change >0.5, p < 0.05). c Inferred pathway activity in control and Snail-overexpressing mOSE cells. P-values were computed from the t statistic of a linear regression model and were adjusted using the Benjamini–Hochberg FDR method. No pathway is significantly different between conditions. d GSEA results for selected gene sets enriched in differentially expressed genes following Snail overexpression. All gene sets are significantly enriched (p < 0.05) and NES values are shown. e UpSet plot showing overlaps in differentially expressed genes between all conditions assessed ranked by the condition/overlap with the largest number of genes. The top chart shows the intersection size for the conditions highlighted in the middle grid. A single, unconnected point corresponds to genes unique to only that condition. The total number of differentially expressed genes in each condition is shown in the left chart.
… 
This content is subject to copyright. Terms and conditions apply.
ARTICLE
Transcriptional heterogeneity of stemness
phenotypes in the ovarian epithelium
Lauren E. Carter1,2,3, David P. Cook1,2,3, Curtis W. McCloskey1,2, Melanie A. Grondin1,2, David A. Landry1,2,
Tiffany Dang1,2, Olga Collins1,2, Lisa F. Gamwell1,2, Holly A. Dempster1,2 & Barbara C. Vanderhyden 1,2
The ovarian surface epithelium (OSE) is a monolayer of epithelial cells surrounding the ovary
that ruptures during each ovulation to allow release of the oocyte. This wound is quickly
repaired, but mechanisms promoting repair are poorly understood. The contribution of tissue-
resident stem cells in the homeostasis of several epithelial tissues is widely accepted, but their
involvement in OSE is unclear. We show that traits associated with stem cells can be
increased following exposure to the cytokine TGFB1, overexpression of the transcription factor
Snai1, or deletion of Brca1.Wend that stemness is often linked to mesenchymal-associated
gene expression and higher activation of ERK signalling, but is not consistently dependent on
their activation. Expression proles of these populations are extremely context specic, sug-
gesting that stemness may not be associated with a single, distinct population, but rather is a
heterogeneous cell state that may emerge from diverse environmental cues. These ndings
support that the OSE may not require distinct stem cells for long-term maintenance, and may
instead achieve this through transient dedifferentiation into a stem-like state.
https://doi.org/10.1038/s42003-021-02045-w OPEN
1Cancer Therapeutics Program, Ottawa Hospital Research Institute, Ottawa, ON, Canada. 2Department of Cellular and Molecular Medicine, University of
Ottawa, Ottawa, ON, Canada.
3
These authors contributed equally: Lauren E. Carter, David P. Cook. email: bvanderhyden@ohri.ca
COMMUNICATIONS BIOLOGY | (2021) 4:527 | https://doi.org/10.1038/s42003-021-02045-w |www.nature.com/commsbio 1
1234567890():,;
Content courtesy of Springer Nature, terms of use apply. Rights reserved
It is thought that stem cell populations are responsible for long-
term maintenance of many adult tissues. The characterization
of stem cells associated with epithelial tissue maintenance has
been an active eld of research for the past few decades. While
several distinct stem cell populations have been functionally
described, such as an LGR5 +population at the base of intestinal
crypts1, it is unclear if all epithelial tissues are maintained by such
dened populations. For example, it has been shown that fol-
lowing stem cell depletion, differentiated airway epithelial cells
can dedifferentiate and become functional multipotent stem
cells2. It is also unclear if stem cells are necessarily required to
maintain epithelial tissues comprising a single cell type, as some
baseline capacity for proliferation could maintain the entire tis-
sue. In the mesothelium, for example, there have been reports of
putative stem/progenitor cells for over two decades, but a well-
dened stem cell population has yet to be identied3.
The ovarian surface epithelium (OSE) is a promising tissue for
studying stemness dynamics in tissue maintenance. It is a
monolayer of cells surrounding the ovary and, during each ovu-
lation, this tissue is ruptured to facilitate release of an oocyte.
Afterwards, the OSE layer is rapidly repaired47. Post-ovulatory
wound repair is a poorly understood process despite ovulation
being the primary non-hereditary risk factor for ovarian cancer.
Several putative OSE stem cell populations have been descri-
bed, each dened by different cell surface markers, including
ALDH1A1, LGR5, LY6A (Sca-1), and more811. However, rela-
tionships between the populations described in these studies are
still unclear, with little known about co-expression or mutual
exclusivity of markers. Further, it is unclear if these populations
are static or can emerge from differentiated OSE in response to
ovarian dynamics. We have previously shown that, like the
mammary epithelium12, induction of an epithelial-to-
mesenchymal transition (EMT) can transiently promote fea-
tures of stem cells (stemness) in differentiated OSE10. This is
particularly relevant in the context of ovulation, as the EMT is
thought to be an important component of wound repair and the
EMT-promoting cytokine TGFB1 is present in follicular uid,
bathing adjacent OSE at ovulation10,13. It is also secreted by
macrophages at the ovulatory wound and granulosa cells during
follicular development14,15.
Here, we further demonstrate that features of stemness can be
altered in OSE cells. Proling gene expression of different
populations with enhanced stemness, we demonstrate that some
features are relatively common, including EMT-associated
expression patterns and enhanced activity of ERK and NFkB
signaling, but global expression proles are widely variable and
stemness is not exclusively dependent on these common features.
Together, this work supports that OSE tissue maintenance may
not require a distinct stem cell population, but can emerge in
response to their environment.
Results
CD44 is a marker of EMT-associated stemness in OSE cells.We
have previously demonstrated that mouse OSE (mOSE) cells
undergo an EMT through canonical SMAD signaling when
exposed to TGFB1 and acquire stem cell characteristics10,16.To
conrm these ndings, we rst assessed the ability of TGFB1-
treated mOSE cells to form self-renewing spheroids in suspension
culture. Treated cells formed over twice as many primary
spheroids and, when dissociated and cultured, were more efcient
at successfully generating secondary spheroids, conrming their
capacity for self-renewal (Fig. 1a). Morphologically, mOSE
spheres were large and compact, with no difference in their size
regardless of whether they had been treated with TGFB1 (Fig. 1b).
This suggests that the phenotype is not simply the result of
increased proliferation rates, but rather reects an increased
ability for clonogenic growth. Consistent with this, we have
previously shown that TGFB1 reduces the growth rate of
mOSE10.
To validate this enhanced stemness in human cells, we also
performed these experiments on primary cultures of human OSE
(hOSE) cells. Since hOSE cultures have a low proliferation rate
in vitro, we used methylcellulose-based suspension culture to
immobilize the cells and minimize the impact of aggregation in
our quantications. While these conditions, along with the slower
proliferation, resulted in smaller spheroids, TGFB1-treated hOSE
cells formed 3 times as many spheroids as untreated cells
(Fig. 1c).
To determine if this enhanced stemness is associated with the
expression of previously reported markers of OSE stem cells, we
measured their expression throughout 7 days of TGFB1 treatment
in mOSE cells. Aldh1a1, Lgr5, and Nanog did not increase with
TGFB1 treatment, and in some cases decreased over time
(Supplementary Fig. 1). While this does not preclude the
possibility of these genes being valid markers of stem cell
populations in vivo, these results suggest that their regulation is
independent from TGFB1-associated stemness.
We next assessed the expression of a larger panel of markers
from a commercial Stem Cell MarkerqPCR array in order to
identify putative markers that are associated with this stemness
(Fig. 1d). This identied several highly upregulated markers
following 7 days of TGFB1 treatment, including Ncam1 (14-fold),
Cd44 (13-fold), and Ascl2 (6-fold) (Fig. 1d). CD44 has long been
associated with stemness in mammary epithelial cells17 and more
recently in the oviductal epithelium18.Werst validated CD44
RNA and protein levels throughout TGFB1 treatment and found
that it increases after four days of treatment (Fig. 1e; Supple-
mentary Fig. 2). We then stained both control and TGFB1-treated
spheroids for CD44 and Ki67 to assess the distribution of
proliferation and this putative stemness phenotype within
spheroids (Fig. 1f; Supplementary Fig. 3). CD44 was detectable
in many control spheroids but was expressed only in rare cells
throughout the population. Cells comprising TGFB1-treated
spheroids, however, ubiquitously expressed CD44. Proliferative
cells were diffuse throughout the spheroids with no clear
association with spatial location within the spheroid or relative
to CD44 +cells. Both CD44 expression patterns are consistent
with the possibility that CD44 expression enriches for stem-like
cells capable of seeding clonogenic growth of spheroids. To test
this, we sorted CD44high cells from TGFB1-treated mOSE by
uorescence-activated cell sorting (FACS). When placed in
suspension culture, CD44high cells formed approximately 2.5
times as many spheroids compared to CD44low cells (Fig. 1g).
Transcriptional proling of mOSE stemness. We next sought to
dene a global prole of stemness, beyond a small number of
markers. Spheroids themselves have been demonstrated to be
enriched with stem/progenitor populations and challenge cells to
exhibit stemness traits, including clonogenic growth and self-
renewal1921. Since untreated mOSE cultures are capable of
sphere formation, albeit at a lower frequency than TGFB1-treated
mOSE cells, we reasoned that the transcriptional prole of these
spheroids may represent an intrinsic stemness program, inde-
pendent from exogenous factors. To compare this with TGFB1-
induced stemness, we performed RNA-seq on mOSE cells cul-
tured as a monolayer or as spheroids, each with and without
TGFB1 treatment.
Untreated mOSE cells cultured as spheroids exhibited striking
differences from those cultured in a monolayer, with 4950
differentially expressed genes between the conditions (p< 0.05,
ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-021-02045-w
2COMMUNICATIONS BIOLOGY | (2021) 4:527 | https://doi.org /10.1038/s42003-021-02045-w | www.nature.com/commsbio
Content courtesy of Springer Nature, terms of use apply. Rights reserved
absolute log fold change >0.5) (Fig. 2a; Supplemental Data 1).
Using an aggregate reference of GO terms, KEGG pathways,
Reactome pathways, and MSigDB Hallmark gene sets from the
Molecular Signatures Database (MSigDB)22,23, we used gene set
enrichment analysis (GSEA) to identify biological features
associated with these changes (Fig. 2c; Supplemental Data 2).
Spheroids were associated with decreased cell cycle, epithelial cell
adhesion, and, interestingly, DNA repair. Along with these
changes, spheroid culture activated expression of chemokine
signaling and wound repair programs. We also note that CD44,
which we had used as a selection marker for stemness in TGFB1-
treated mOSE, was also expressed over 4-fold higher in spheroids,
whereas Aldh1a1, Lgr5, and Ly6a (Sca-1) were unchanged
(Fig. 2b). We next used the PROGENy algorithm to infer
changes in signaling pathway activity across these samples that
may be contributing to these differences. Spheroids were
associated with increased activity of many signaling pathways,
with the largest increases in Hypoxia, NFkB, and MAPK signaling
(Fig. 2d). While GSEA results suggest several EMT-related
changes, TGFB1 and WNT signaling are interestingly reduced,
suggesting that this EMT program may be activated through
NFkB or ERK (Fig. 2d).
While TGFB1 signaling was decreased in untreated spheroids
when compared to monolayer cultures, exogenous TGFB1
treatment of monolayers enhanced stemness, increasing the
proportion of cells capable of forming self-renewing spheroids.
While these results are seemingly contradictory, week-long
exposure to exogenous TGFB1 may activate similar expression
programs through secondary effects or signaling crosstalk24.We
next assessed expression changes associated with TGFB1
0
50
100
Ctrl
TGFB1
Spheres per 4 fields
0
2
4
6
Ctrl
TGFB1
Spheres per 4 fields
0
5
10
15
20
Ctrl
TGFB1
Spheres per 4 fields
Ncam1
Cd44
Ascl2
Jag1
Col2a1
Myc
Bmp1
T
Ter t
Dhh
Pparg
Frat1
Hspa9
Gja1
Dll1
Fgf3
Dtx2
Gdf2
Ccne1
Abcg2
Foxa2
Ccna2
Cxcl12
Cdc2a
Bmp3
Krt15
Bmp2
-40 -30 -20 -10 0 10
Fold Change
a
d
bc
0
20
40
CD44
low
CD44
high
Spheres per 4 fields
Control
TGFB1
TGFB1
Control TGFB1
Control
Mouse OSE (mOSE) Human OSE (hOSE)
1
2
0 6 12 24 48 96
TGFB1 treatment (hours)
Relative CD44 protein
Primary Secondary
e
f
g
** * *
*
*
0
5
10
Ctrl
TGFB1
Average sphere size
(Pixel area/1000)
Primary spheres
Secondary spheres
DAPI CD44 KI67 DAPI CD44 KI67
Fig. 1 TGFB1 promotes stemness in the OSE. a Primary (left; n=3) and secondary (right; n=5) sphere-forming capacity of mOSE cells treated with
TGFB1 (10 ng /mL). Data points represent the average number of spheres per 4 elds of view for each replicate after 14 days of spheroid culture. bLeft:
Phase contrast images of control and TGFB1-treated spheroids (left). Scale bar =100 μm. Right: Relative sizes of spheroids in control and TGFB1 conditions
(primary spheroids n=3; secondary spheroids n=4). Each point on the plot represents the average size of spheroids from 4 elds of view for 4 separate
wells. cPrimary sphere-forming capacity of human OSE treated with TGFB1 (n=4). hOSE spheroids were cultured in methylcellulose to prevent
aggregation. Spheroids were cultured for 28 days. Scale bar =100 μm. dFold change values for a panel of putative stem cell markers in mOSE treated with
TGFB1 for 4 days. eRelative protein quantications of CD44 throughout a time course of TGFB1 treatment in mOSE (n=3). Quantications represent
Western blot pixel densitometry, normalized to B-actin and scaled to the mean intensity in untreated samples. A representative blot is included in
Supplemental Fig. 2b. fImmunouorescence co-stain of CD44 and Ki67 in control and TGFB1-treated spheroids. White arrows highlight CD44 +cells in
control spheroids. Scale bar =100 μm, inlet scale bar =15 μm. Separate channels are shown in Supplementary Fig 3g. Primary sphere-forming capacity of
CD44- and CD44 +mOSE cells (n=3). All boxplots show median value (horizontal black line), estimated 25th and 75th percentiles, and whiskers
represent 1.5 times the interquartile range. Linear regression models were used for all statistical tests. *p< 0.05, **p< 0.01.
COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-021-02045-w ARTICLE
COMMUNICATIONS BIOLOGY | (2021) 4:527 | https://doi.org /10.1038/s42003-021-02045-w | www.nature.com/commsbio 3
Content courtesy of Springer Nature, terms of use apply. Rights reserved
exposure. Treatment of monolayer cultures with TGFB1 for
7 days resulted in 1508 differentially expressed genes (p< 0.05,
absolute log fold change >0.5) (Fig. 3a; Supplemental Data 3).
This involved the activation of EMT-associated gene sets as
expected, as well as a reduction in oxidative phosphorylation
(Fig. 3b; Supplemental Data 4). While TGFB1 signaling was the
only pathway inferred to have signicantly altered activity, the
estimated activity of EGFR and MAPK was higher in TGFB1-
treated cells (p=0.1 and 0.06, respectively) (Fig. 3c). Consistent
with this, the GO term ERK1 and ERK2 signaling cascadewas
signicantly enriched in upregulated genes following TGFB1
treatment (Fig. 3b).
This suggests that TGFB1 treatment initiates sequential or
parallel signals similar to those present in spheroids. Consistent
with this, untreated spheroids and TGFB1-treated mOSE
monolayers have a signicant overlap in expression changes
relative to untreated mOSE cells cultured as a monolayer, sharing
270 upregulated genes and 293 downregulated genes (Fisher exact
p=3.0e-66 and 2.8e-117, respectively) (Fig. 3d). Conserved
upregulated genes were strongly enriched for EMT-associated
genes, NFkB signaling, and angiogenesis (Fig. 3e). Interestingly,
very few gene sets were enriched in the conserved downregulated
genes, with only interferon response and substrate adhesion genes
being enriched (Fig. 3e). We then explored the gene expression
patterns of TGFB1-treated spheres using a linear model with an
interaction term to identify expression changes that were not
simply the additive effects of TGFB1 and spheroid culture
(Supplementary Fig. 4a). We performed GSEA on genes ranked
by the interaction coefcient of this model and found that
TGFB1-treated spheres were associated with higher expression of
genes associated with EMT, MAPK/ERK, and inammatory
pathways (Supplementary Fig. 4b). They also had notable
differences in the expression of metabolic genes, with an increase
in oxidative phosphorylation and reduced cholesterol biosynth-
esis. Together, these ndings suggest that various responses may
contribute to OSE stemness, including increased MAPK/ERK
signaling, activation of inammatory response pathways such as
NFkB and STAT, and EMT-associated gene expression.
Snail activation promotes a unique stemness program in
mOSE cells. Signaling pathways are highly pleiotropic and it is
unclear if TGFB1-enhanced stemness is activated from core EMT
regulatory networks or alternative components regulated by
TGFB1. The EMT transcription factor Snai1 (Snail) was upre-
gulated in both TGFB1-treated mOSE cells and spheroids
(Figs. 2a and 3a), and so to determine if EMT activation without
exogenous cytokines could promote stemness, we derived mOSE
cell lines with doxycycline-inducible Snail expression. Following
Snail induction, cells had a higher sphere forming capacity than
cells without doxycycline exposure, generating up to twice as
many primary spheres and 3 times as many secondary spheres
when passaged (Fig. 4a). Despite a higher sphere-forming ef-
ciency, spheroid size was not different between control and Snail-
overexpressing cells and the baseline proliferation rate of the cells
was also unchanged (Supplementary Fig. 5a, b).
We next assessed the expression of the putative stem cell
markers Cd44 and Sca-1, which are both increased with TGFB1
treatment, and found that Snail induction had no effect on their
expression, suggesting that their validity as markers of stemness
may be context specic. To determine if Snail induction activates
similar expression patterns to TGFB1-treated mOSE and
GO: Cell division
KEGG: Oxidative phosphorylation
GO: Recombinational repair
GO: Epithelial cell cell adhesion
HALLMARK: Inflammatory response
GO: Regulation of chemotaxis
GO: Chemokine mediated signaling pathway
−2 −1 0 1 2
NES
PI3K
TGFb
WNT
JA K - STAT
NFkB
TNFa
Hypoxia
MAPK
EGFR
VEGF
Androgen
Trail
Estrogen
p53
a
cd
b
Krt19
Zeb1
Zeb2
Brca1 Snai1
Cd44
Aldh1a1
Greb1
0
100
200
300
−4 0 4 8
log2(Fold Change)
log10(p−value)
Aldh1a1 Cd44 Lgr5 Ly6a
Control
Spheroid
Control
Spheroid
Control
Spheroid
Control
Spheroid
1.0
1.2
1.4
1.6
1.8
0.05
0.10
0.15
2
3
4
5
8.8
9.0
9.2
9.4
log(TPM+1)
-1.5
1.5
0
Relative activity
(Z-score)
Monolayer Spheroid
Sig?
**
*
***
***
***
***
***
***
**
**
**
*
**
**
Spheroid culture
UpregulatedDownregulated
Fig. 2 Transcriptional prole of intrinsic mOSE stemness. a Plot showing the distribution of differentially expressed genes in mOSE cells cultured as
spheroids for 14 days relative to a monolayer (n=3). Each point corresponds to a single gene. Selected genes related to stemness and/or the EMT are
highlighted on the plot. Dashed lines correspond to signicance criteria (absolute log fold change >0.5, p< 0.05). bBoxplots showing the expression values of
putative mOSE stemness markers in mOSE cells cultured in a monolayer (Control) or as spheroids (n=3). Boxplots show median value (horizontal black line),
estimated 25th and 75th percentiles, and whiskers represent 1.5 times the interquartile range cGSEA results for selected gene sets enriched in differentially
expressed genes in mOSE spheroids. All gene sets are signicantly enriched (p< 0.05) and normalized enrichment scores (NES) are shown. dInferred pathway
activity in monolayer- and spheroid-cultured mOSE cells. Linear models were used for statistical testing for band d.*p< 0.05, **p< 0.01, ***p< 0.001.
ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-021-02045-w
4COMMUNICATIONS BIOLOGY | (2021) 4:527 | https://doi.org /10.1038/s42003-021-02045-w | www.nature.com/commsbio
Content courtesy of Springer Nature, terms of use apply. Rights reserved
spheroids, including higher ERK and NFkB activity, we
performed RNA-seq on these cells with and without doxycycline.
Snail-induced changes were more modest than with TGFB1
treatment or in spheroid culture, with only 85 upregulated and 44
downregulated genes (p< 0.05, absolute log fold change >0.5;
Fig. 4b; Supplemental Data 5). Interestingly, inferred pathway
activity scores associated with EGFR, NFkB, MAPK, and TGFB1
were all unchanged following Snail induction (Fig. 4c). Further,
no relevant gene sets associated with these pathways were
enriched in the differentially expressed genes (Fig. 4d). The only
gene sets associated with upregulated genes were largely related to
cell morphology and extracellular matrix (ECM) remodeling
(Fig. 4d; Supplemental Data 6). Consistent with TGFB1-treated
mOSE and spheroids, the MSigDB Hallmark Interferon Alpha
Responsewas the only gene set enriched in the downregulated
genes following Snail induction (Fig. 4d). We note that of the 85
upregulated genes following Snail induction, 13 are shared with
those commonly regulated in TGFB1 treatment and spheroids
(Fig. 4e). These genes largely represent components of the ECM,
including Col18a1 and the metalloproteinases Mmp9 and
Adamts4. As the conditions share no consistently activated
downstream signal that could be induced by ECM changes, these
ndings suggest that expression programs associated with
stemness phenotypes are heterogeneous. Given frequent enrich-
ment of gene sets associated with a mesenchymal phenotype,
stemness may consistently involve higher levels of these traits,
which can emerge from variable expression patterns25.
BRCA1 loss promotes EMT-independent stemness in mOSE.
Spheroids were associated with higher expression of many genes
that were not similarly induced by TGFB1 treatment. We noted
that among spheroid culture-induced genes were several changes
typically associated with ovarian cancer, including activation of
the transcription factor Pax8, which is present in approximately
80% of ovarian tumors but not typically expressed in murine
OSE26; activation of Greb1, which promotes ovarian cancer
growth27; and loss of Brca1, which, along with Brca2, is mutated
in approximately 22% of high-grade serous ovarian tumors.
Interestingly, loss of BRCA1 has been associated with promoting
dedifferentiation and activation of EMT expression patterns in
mammary epithelial cells28.
As the association between BRCA1 loss and stemness in the
OSE had not been assessed, we next derived a primary mOSE line
from Brca1tm1Brn mice harboring oxed Brca1 alleles. To
determine if BRCA1 loss enhanced stemness in these cells, we
HALLMARK: Interferon alpha response
HALLMARK: Interferon gamma response
GO: Cell substrate adhesion
0.0 0.5 1.0 1.5 2.0
−log10(p−value)
GO: Angiogenesis
HALLMARK: Epithelial−mesenchymal transition
HALLMARK: TNFa signaling via NFkB
024
−log10(p−value)
HALLMARK: Oxidative phosphorylation
HALLMARK: Interferon alpha response
GO: ERK1 and ERK2 cascade
GO: Mesenchymal cell differentiation
GO: Extracellular structure organization
HALLMARK: Epithelial-mesenchymal transition
-2 -1 0 1 2
NES
−2.5
0.0
2.5
5.0
−4 0 4
Spheroid logFC
TGFB1 logFC
293 genes
270 genes
93 genes
Pearson’s r=0.19
187 genes
Trail
JAK-STAT
p53
WNT
TNFa
NFkB
PI3K
Hypoxia
VEGF
EGFR
MAPK
Estrogen
Androgen
TGFb
a
c
d
e
b
Snai1
Cd44
Krt19
Zeb1
Brca1
Zeb2
Greb1
Aldh1a1
0
50
100
150
-4 -2 0 2 4 6
log2(Fold Change)
-log10(p-value)
-1.5
1.5
0
Relative activity
(Z-score)
Ctrl TGFB1
Sig?
Gene sets enriched in
common upregulated genes
Gene sets enriched in
common downregulated genes
***
UpregulatedDownregulated
TGFB1 treatment
Fig. 3 TGFB1 promotes a distinct stemness phenotype. a Plot showing the distribution of differentially expressed genes in monolayers of mOSE cells
treated with TGFB1 compared to untreated samples (n=3). Each point corresponds to a single gene. Selected genes related to stemness and/or the EMT
are highlighted on the plot. Dashed lines correspond to signicance criteria (absolute log fold change >0.5, p< 0.05). bGSEA results for selected gene sets
enriched in differentially expressed genes following TGFB1 treatment. All gene sets are signicantly enriched (p< 0.05) and NES values are shown. c
Inferred pathway activity in untreated and TGFB1-treated mOSE cells. P-values were computed from the tstatistic of a linear regression model and were
adjusted using the BenjaminiHochberg false discovery rate (FDR) method. dPlot comparing log fold-change values for spheroid-cultured and TGFB1-
treated mOSE. Dashed lines correspond to fold change cutoffs used to assess signicance. ePlots showing gene sets enriched in commonly up- or
downregulated genes following both TGFB1 treatment and spheroid culture. P-values were calculated using a Fisher exact test and were adjusted using the
BejaminiHochberg FDR method.
COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-021-02045-w ARTICLE
COMMUNICATIONS BIOLOGY | (2021) 4:527 | https://doi.org /10.1038/s42003-021-02045-w | www.nature.com/commsbio 5
Content courtesy of Springer Nature, terms of use apply. Rights reserved
infected the cells with adenovirus containing either Cre
recombinase (Ad-Cre) or GFP (Ad-GFP) as a control. Cre
delivery, while not perfectly efcient, resulted in an approxi-
mately 60% reduction in BRCA1 levels across the population
(Supplementary Fig. 6). When placed in suspension culture, cells
with reduced BRCA1 formed over 5 times as many primary
spheres and 3 times as many secondary spheres than control
mOSE cells, suggesting that BRCA1 loss also enhances stemness
in mOSE cells (Fig. 5a). There was no difference in the size of the
spheres and BRCA1 deletion modestly decreased the growth rate
of cells in monolayer culture (p=0.008, linear regression on cell
counts from days 15) (Supplemental Fig. 7a, b).
To assess if BRCA1 loss enhances stemness phenotypes in vivo,
we crossed the Brca1tm1Brn mice with B6.129×1-Gt(ROSA)
26Sortm1(EYFP)Cos/J mice to generate a Brca1/YFP mouse line,
allowing us to track Brca1-null cells following exposure to Ad-
Cre. These mice were injected intrabursally (IB) with Ad-Cre or
PBS, and injected intraperitoneally (IP) with bromodeoxyuridine
(BrdU). Ovaries were collected after a 30-day chase period and
assessed for retention of the BrdU label and activation of the YFP
reporter. Ad-Cre injection IB in Brca1/YFP mice showed
successful activation of the YFP reporter, compared to the PBS
injection (Fig. 5b). When combining the IB injections with an IP
BrdU injection, Ad-Cre treatment increased the number of
label-retaining OSE cells (Fig. 5c, d). Given the high frequency of
ovulations of the mouse ovary over the 30-day chase period,
this increased label retention following BRCA1 loss is consistent
with BRCA1 deletion leading to expansion of quiescent stem-
like cells in the OSE. Whether this is through expansion of an
existing stem-like population or dedifferentiation of mOSE
is unclear.
To determine if BRCA1 loss results in similar expression
patterns to other conditions associated with stemness, we
performed RNA-seq on mOSE cells isolated from these mice
infected with Ad-Cre or Ad-GFP in vitro. BRCA1 loss resulted in
a large shift in gene expression, with 1499 signicantly upregu-
lated genes and 1881 downregulated (p< 0.05, absolute log fold
change >0.5; Fig. 6a; Supplemental Data 7). In mammary
epithelium, the induction of EMT through Brca1 deletion was
presumed to be due to loss of BRCA1-mediated repression on the
promoter of the EMT transcription factor Twist1. In contrast, we
found that Twist1 was approximately 8-fold lower in Brca1-null
mOSE cells (Fig. 6a). There were also no EMT-associated gene
sets enriched in upregulated genes. Rather, upregulated genes
were largely enriched for gene sets associated with cell membrane
transporters and downregulated genes were associated with cell
cycle, oxidative phosphorylation, and DNA repair (Fig. 6b;
Supplemental Data 8). Brca1 deletion did not result in other
features of stemness we observed in previous conditions,
including activation of ERK and NFkB, and repression of
interferon alpha response genes (Fig. 6c). Brca1-null cells were
associated with reduced PI3K signaling and higher levels of
estrogen signaling (Fig. 6c). Estrogen has been linked to EMT and
stemness in other cell types29, but this is presumed to be through
JAK-STAT
Androgen
Hypoxia
PI3K
EGFR
MAPK
WNT
Tra i l
NFkB
TNFa
TGFb
VEGF
Estrogen
p53
HALLMARK: Interferon alpha response
GO: Sterol biosynthetic process
GO: Biological adhesion
GO: Organ morphogenesis
GO: Specification of symmetry
REACTOME: Extracellular matrix organization
-2 -1 0 1 2
NES
UpregulatedDownregulated
ac
e
b
-1.5
1.5
0
Relative activity
(Z-score)
Ctrl Snail
Ctrl
Snail
Snai1
Cd44 Brca1
Aldh1a1
Zeb1
Zeb2
Lgr5
Krt19
0
5
10
15
20
25
-3 -2 -1 0 1 2 3
log2(Fold Change)
-log10(p-value)
0
4
8
12
Ctrl
Snail
Spheres per 4 fields
0
20
40
Ctrl
Snail
Spheres per 4 fields
Snail overexpression
Primary Secondary
p=0.06 *
d
1959
408 289 256 223 175 88 18 16 16 13 11 10 99
0
500
1000
1500
2000
Sig Gene
Intersection Size
Snail - Down
TGFB1 - Down
Spheroid - Down
Snail - Up
TGFB1- Up
Spheroid - Up
0500200015001000
Sig Gene Count
Mmp9
Col18a1
Adamts4
Tnc
Sema4g
Fig. 4 Snail overexpression promotes stemness with minimal gene expression changes. a Primary (left) and secondary (middle) sphere forming capacity
of mOSE cells overexpressing Snail (n=3). Data points represent the average number of spheres per 4 elds of view for each replicate. Boxplots show
median value (horizontal black line), estimated 25th and 75th percentiles, and whiskers represent 1.5 times the interquartile range. Right: Phase contrast
images of control (inducible GFP) and Snail-overexpressing spheroids. Spheroids were cultured for 14 days. Scale bar =100 μm. bPlot showing the
distribution of differentially expressed genes following Snail overexpression. Each point corresponds to a single gene. Selected genes related to stemness
and/or the EMT are highlighted on the plot. Dashed lines correspond to signicance criteria (absolute log fold change >0.5, p< 0.05). cInferred pathway
activity in control and Snail-overexpressing mOSE cells. P-values were computed from the tstatistic of a linear regression model and were adjusted using
the BenjaminiHochberg FDR method. No pathway is signicantly different between conditions. dGSEA results for selected gene sets enriched in
differentially expressed genes following Snail overexpression. All gene sets are signicantly enriched (p< 0.05) and NES values are shown. eUpSet plot
showing overlaps in differentially expressed genes between all conditions assessed ranked by the condition/overlap with the largest number of genes. The
top chart shows the intersection size for the conditions highlighted in the middle grid. A single, unconnected point corresponds to genes unique to only that
condition. The total number of differentially expressed genes in each condition is shown in the left chart.
ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-021-02045-w
6COMMUNICATIONS BIOLOGY | (2021) 4:527 | https://doi.org /10.1038/s42003-021-02045-w | www.nature.com/commsbio
Content courtesy of Springer Nature, terms of use apply. Rights reserved
crosstalk, activating growth factor signaling pathways, which we
do not see in Brca1-null OSE cells.
Comparing the expression proles of each condition associated
with stemness phenotypes in this study, we nd minimal overlap
in the specic genes activated or repressed in each (Fig. 6d).
Ranking genes by the number of conditions they are activated or
repressed in, we found that Adamts4 and Pnmal2 are the only
genes upregulated in all four conditions (Fig. 6e). Adamts4 has
been linked to stemness in uveal melanoma through modulating
crosstalk between the cells and their adjacent ECM30. While this
may be relevant here, conserved downstream signals promoting
stemness remain elusive. We note high frequency of EMT-
associated changes, including activation of Snail, various
collagens, and repression of cytokeratins (Fig. 6e). Notably,
however, specic EMT transcription factors and putative OSE
stemness markers (Lgr5,Aldh1a1,Ly6a,Nanog, and Cd44) are
only activated in 1-2 conditions, and are even repressed in some
conditions (Fig. 6e).
Discussion
Several studies have reported putative stem cell populations in the
OSE, but the relationships between these populations are unclear.
The ability of differentiated epithelial cells to dedifferentiate and
full functional roles of stem cells has now been observed in
several tissues, suggesting that static stem cell populations may
not be required to maintain all tissues. In this study, we have
further explored the ability of OSE cells to acquire features of
stemness and have demonstrated it can be promoted by a variety
of conditions. It may be expected that a common gene expression
program would underlie the specic stemness phenotype we have
assessed with these experiments, but we demonstrate that
expression proles are context specic.
While transcriptional responses were variable, several patterns
were recurrent across multiple conditions. We observed that
induction of an EMT with TGFB1 treatment or Snail over-
expression could promote stemness in OSE, but expression pro-
les of spheroids naive to TGFB1 treatment also showed EMT
activation, which has also been observed with ovarian cancer cells
cultured as spheroids31. As spheroids have been shown to enrich
for cells with stem cell properties, these ndings suggest that
intrinsic stemnessindependent of exogenous treatmentsmay
be associated with a more mesenchymal phenotype. The rela-
tionship between the EMT and stemness is well documented12,32,
but there is growing evidence that stemness and EMT are not
inextricably linked. For example, the EMT-promoted transcrip-
tion factor PRRX1 suppresses stemness in breast cancer cells33.
Further, transcriptional dynamics of the EMT have been shown
to be highly context-specic, which explains why it does not
consistently promote stemness25. Just as the EMT can occur
without promoting stemness, we have shown that deletion of
Brca1 promotes stemness in OSE without activating any EMT-
associated expression, including Twist1 activation, which had
been shown to drive stemness following Brca1 deletion in
mammary epithelial cells28. Instead, Brca1 loss caused many
changes in cell membrane transport and metabolic genes. While
the mechanism of induced stemness following Brca1 loss is
unclear, this provides strong evidence that stemness is not
dependent on a mesenchymal expression prole and is perhaps as
context specic as the EMT response25.
Several alterations in signaling pathway activity were also
common across conditions. TGFB1 treatment and spheroids were
associated with higher levels of ERK activity, which have both
been linked to stemness in epithelial34 and carcinoma cells35.In
EGF-free media, paracrine/autocrine signaling is established,
maintaining ERK activity in stem cell populations of intestinal
organoids35. While these mechanisms may contribute to stemness
in OSE spheroids or those treated with TGFB1, increased ERK
activity was not enhanced following Snail overexpression or Brca1
loss. Similarly, a gene set comprising interferon alpha response
genes was downregulated following spheroid culture, TGFB1
treatment, and Snail overexpression. While it is unlikely that
interferon alpha itself was present, it is possible that various
signaling pathways may affect common target genes. Consistent
with this, disruption of type 1 interferon signaling promotes
stemness in breast cancer cells36. None of these patterns, how-
ever, are consistent across all conditions, further supporting that
mechanisms promoting stemness may vary considerably
acb
0
5
10
15
20
Ctrl
Brca1 KO
Spheres per 4 fields
0
20
40
60
80
Ctrl
Brca1 KO
Spheres per 4 fields
Brca1 deletion
Primary
PBS
Ad-Cre
PBSAd-Cre
Secondary
****
d
0
5
10
PBS Ad−Cre
Intrabursal Injection
BrdU+ OSE (%)
Wild type
BRCA1 KO
Fig. 5 Brca1 deletion in vivo promotes increased label retention. a Primary (left; n=3) and secondary (middle; n=5) sphere forming capacity of mOSE
cells following Brca1 deletion by infection with Ad-Cre (Brca1 KO). Cells infected with Ad-GFP were used as a control (Ctrl). Data points represent the
average number of spheres per 4 elds of view for each replicate. Right: Phase contrast images of control (OSE from PBS-injected mice) and BRCA KO
spheroids. Spheroids were cultured for 14 days. Scale bar =100 μm. bImmunohistochemical staining of ovaries following intrabursal injection of PBS or
Ad-Cre. Staining shows BRCA1 (red) and the YFP (green) reporter activated upon delivery of Cre recombinase. Nuclei are stained with DAPI (blue). Scale
bar =25 μm. cBrdU label retention (red) and YFP (green) signal in ovaries following intrabursal injection of either PBS or Ad-Cre. Scale bar =100 μm, inlet
scale bar =25 μm. dQuantication of BrdU+OSE cells in ovaries (n=4 ovaries from independent mice). All boxplots show median value (horizontal black
line), estimated 25th and 75th percentiles, and whiskers represent 1.5 times the interquartile range.
COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-021-02045-w ARTICLE
COMMUNICATIONS BIOLOGY | (2021) 4:527 | https://doi.org /10.1038/s42003-021-02045-w | www.nature.com/commsbio 7
Content courtesy of Springer Nature, terms of use apply. Rights reserved
depending on environmental conditions (e.g., ovulatory wound
repair, tissue expansion during folliculogenesis, natural cell
turnover).
While we have relied heavily on in vitro models here, this has
enabled us to explore the ability of OSE cells to acquire stemness
following various experimental perturbations. This suggests that
differentiated epithelial cells may be capable of self-regulating
tissue maintenance in response to environmental cues, such as
tissue damage. The expression proles of this emergent stemness
may be variable, depending on the specic properties of the cells
microenvironment. This model is particularly interesting because
it is a stark contrast to how stem cells and differentiation hier-
archies have been viewed for the last several decades. The OSE is
a promising tissue to explore this further as it undergoes regular
rupture and repair throughout reproductive cycles, and is a
simple tissue comprising a single cell type, which may be the most
likely to exhibit this behavior. Designing strategies to monitor
stemness dynamics in vivo will be critical to understand these
behaviors in a normal physiological context.
Methods
OSE cell isolation and culture. The isolation and culture of mOSE cells was done
as previously described10, in accordance with the guidelines of the Canadian
Council on Animal Care and under a protocol approved by the University of
Ottawa Animal Care Committee. Briey, ovaries from randomly cycling female
mice (FVB/N, 6 weeks old) were collected and incubated in 0.25% Trypsin/PBS
(Invitrogen) (37 °C, 5% CO
2
, 30 min) to facilitate OSE removal. mOSE cells
were isolated by centrifugation and plated onto tissue culture plates (Corning) in
mOSE media [a-Minimum Essential Medium (Corning) supplemented with 4%
FBS, 0.01 mg/mL insulin-transferrin-sodium-selenite solution (ITSS; Roche), and
2 µg/mL EGF (R&D Systems)]. hOSE cells were isolated and cultured as previously
described37, with patient consent and under a protocol approved by the Ottawa
Health Science Network Research Ethics Board (Protocol #1999540). Briey,
ovaries from 5 different women were collected during surgery for reasons other
than ovarian pathology. Using a scalpel, hOSE cells were scraped from the ovarian
surface and isolated by centrifugation in hOSE media (Wisent Bioproducts) sup-
plemented with 10% FBS. All mouse and human OSE cells were passaged 23 times
prior to experimental use and experiments were conducted with cells of a passage
number less than 25.
Quantitative reverse transcription polymerase chain reaction (RT-PCR). The
RNeasy Mini Kit (Qiagen) was used to extract RNA and the OneStep RT-PCR Kit
(Qiagen) was used to synthesize cDNA. Quantitative PCR was done using the ABI
7500 FAST qRT-PCR machine (Applied Biosystems) using the Taqman gene
expression (Life Technologies) and SsoFast gene expression (Bio-rad) assays uti-
lizing Tbp as an endogenous control. Primer sequences are listed in Supplemental
Table 1. RQ (relative quantity) was determined using the cycling threshold for the
gene of interest in control or untreated samples compared to the cycling threshold
in experimental samples, calculated using the Applied Biosystems 7500 FAST
v2.3 software.
Greb1
Twist1
Cd44
Zeb1
Aldh1a1
Snai1
Lgr5
Krt19
0
20
40
60
80
-4 0 4
log2(Fold Change)
-log10(p-value)
EGFR
PI3K
Estrogen
Trail
JAK−S TAT
WNT
p53
Androgen
Hypoxia
VEGF
MAPK
TNFa
NFkB
TGFb
-1.5
1.5
0
Relative activity
(Z-score)
Ctrl TGFB1
Sig?
*
*
a
c
d
e
b
GO: Oxidative phosphorylation
REACTOME: Cell cycle
HALLMARK: DNA repair
GO: Transmembrane transport
KEGG: ABC transporters
GO: Organic acid transport
-3 -2 -1 0 1 2
NES
1607
1009927
492
307 229 201 191 161 123 115 112 61 57 47
0
500
1000
1500
Sig Gene
Intersection Size
Brca1 KO - Down
Snail - Down
TGFB1 - Down
Spheroid - Down
Brca1 KO - Up
Snail - Up
TGFB1 - Up
Spheroid - Up
0500200015001000
Sig Gene Count
1
2
3
4
0 1000 2000
Gene Count
Differential Expression
Frequency (# of conditions)
1
2
3
4
0 1000 2000 3000 4000
Gene Count
Differential Expression
Frequency (# of conditions)
Upregulated genes
UpregulatedDownregulated
BRCA1 Deletion
44
844
1806
20
Adamts4, Pnmal2
Krt7, Krt18, Vcam1, Igfbp6
Krt19, Tead2, Mmp14, Fgf5
Krt14, Twist1, Snai1,
Cd44, Aldh1a1, Nanog
Wnt5a, Snai1, Col18a1, Mmp9
Cd44, Alcam, Fzd2, Fn1, Lhx4,
Tcf7, Itgav, Wdr54, Jag1, Ptgs2
Lgr5, Ly6a, Nanog,
Zeb1, Zeb2, Twist1
49
579
3485
Downregulated genes
Fig. 6 Stemness phenotypes are transcriptionally diverse. a Plot showing the distribution of differentially expressed genes following Brca1 deletion by
infection with Ad-Cre. Each point corresponds to a single gene. Selected genes related to stemness and/or the EMT are highlighted on the plot. Dashed
lines correspond to signicance criteria (absolute log fold change >0.5, p< 0.05). bGSEA results for selected gene sets enriched in differentially expressed
genes following Brca1 deletion. All gene sets are signicantly enriched (p< 0.05) and NES values are shown. cInferred pathway activity in control and Snail-
overexpressing mOSE cells. P-values were computed from the tstatistic of a linear regression model and were adjusted using the BenjaminiHochberg FDR
method. dUpSet plot showing overlaps in differentially expressed genes between all conditions assessed ranked by the condition/overlap with the largest
number of genes. ePlots showing the number of assessed conditions that genes are either activated or repressed in. Selected genes are listed and putative
stemness markers are underlined.
ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-021-02045-w
8COMMUNICATIONS BIOLOGY | (2021) 4:527 | https://doi.org /10.1038/s42003-021-02045-w | www.nature.com/commsbio
Content courtesy of Springer Nature, terms of use apply. Rights reserved
Western blot. M-PER mammalian protein extraction reagent (GE Healthcare) was
used to extract protein from mOSE cells and run on NuPAGE 412% Bis-Tris
gradient gels (Life Technologies). Polyvinylidene diuoride membranes were used
to transfer protein samples. Membranes were blocked in 5% non-fat milk prior to
antibody incubation. Antibody conditions are described in Supplemental Table 2.
Western blots were developed using ClarityTM Western ECL Substrate (Bio-Rad)
and the FluorChem FC2 imaging system (Alpha Innotech).
Stem cell PCR array. mOSE cells (1 × 106cells) were plated 24 h prior to treatment
with TGF B1 (10 ng/mL, R&D Systems). RNA was collected 7 days post TGFB1
treatment (RNAeasy Kit, Qiagen). cDNA synthesis was performed using RT2First
Strand Kit (Qiagen) and run on the RT2First Strand Kit (Stem cell PCR array)
(Qiagen). The array was run in triplicate (N=3) and analyzed using the DataA-
nalysis Excel platform provided with the array kits.
Snail-overexpressing mOSE cells. mOSE cells stably expressing reverse
tetracycline-controlled transactivator (rtTA) protein were transduced with a len-
tiviral construct (pWPI) expressing the murine Snai1 or eGFP under the control of
a doxycycline-inducible promoter and the hygromycin resistance gene under the
control of the PGK promoter. Transduced cells were selected for resistance to
Hygromycin B. 200 ng/mL of doxycyc line was added to cultu res for 4 days prior to
all experiments overexpressing Snail.
Primary sphere-forming assays. For free-oating spheres, mOSE cells were
cultured in stem cell media [Dulbeccos Modied Eagles Medium: Nutrient
Mixture F-12 (Sigma) supplemented with 1 X B27 supplement (Invitrogen),
0.02 µg/mL EGF (R&D Systems), 0.04 µg/mL broblast growth factor (FGF; R&D
Systems), 4 µg/mL heparin (Sigma) and 0.01 mg/mL ITSS (Roche), and 2 µg/mL
EGF (R&D Systems)] at 5 × 104cells/mL in non-adherent 24-well culture plates
(Corning) and incuba ted at 37 °C, 5% CO
2
for 14 days. For TGFB1-treated spheres,
cells were pre-treated with recombinant TGFB1 (10 ng/mL) for 4 days prior to
being placed in spheroid culture and was replenished when plating cells in spheroid
culture. Spheres were quantied using ImageJ using a pixel cutoff of >1000 pixels
and a circularity limit of 0.51.0. For spheres cultured in methylcellulose, mOSE
cells were placed in a 1:1 mixture of methylcellulose and stem cell media at 5 × 104
cells/mL in 24-well culture plates (Corning), and incubated at 37 °C, 5% CO
2
for
28 days. Methylcellulose-embedded spheres were quantied using ImageJ using a
pixel cutoff of >500 pixels and a circularity limit of 0.51.0. For each experiment, a
minimum of 3 replicates were performed, each replicate was performed in three
independent wells, spheres were counted in 4 elds per well, and the average count
was reported.
Secondary sphere-forming assay. Primary free-oating mOSE spheres were
collected and washed in PBS. Spheres were dissociated by rst incubating in
trypsin/PBS (Invitrogen) at 37 °C for 10 min, then by passing cells through a 25
gauge needle to obtain a single cell suspension. Single cell suspension was veried
using phase contrast microscopy. Cells were washed in PBS, counted using a
hemocytometer, and plated in stem cell media at 5 × 104cells/mL. Cells were
incubated in non-adherent 24-well culture plates (Corning) at 37 °C, 5% CO
2
for
14 days. Spheres were quantied using ImageJ using a pixel cutoff of >500 pixels
and a circularity limit of 0.51.0.
Spheroid immunouorescence. Control and TGFB1-treated spheroids were col-
lected after 14 days in free-oating spheroid culture. and xed in 4% paraf-
ormaldehyde for 30 min. Spheroids were then embedded in Epredia HistoGel
Embedding Media prior to being transferred into 70% ethanol overnight. HistoGel
blocks were then embedded in parafn. Parafn were sectioned at 5 μm and dried
overnight on slides at room temperature. Sections were deparafnized in Xylene
and rehydrated in a series of graded ethanol baths. Antigen retrieval and per-
meabilization was performed using heat-mediated antigen unmasking solution (H-
3300, Vector Laboratories, Burlingame, Ca, USA) according to manufacturers
instructions followed by incubations in 0.2% Triton X-100 diluted in PBS for
15 min at room temperature, respectively. Slides were rinsed three times PBS-T and
incubated in blocking solution [10% goat serum (G9023, Sigma-Aldrich, St. Louis,
MS, USA) with 1% BSA diluted in PBS] for 30 min at room temperature followed
by an overnight incubation at 4 °C in primary antibodies [rabbit anti-Ki-67
(ab9260, MilliporeSigma, Burlington, MA, USA) at 1:50 and rat anti-Cd44
(ab119348, Abcam, Cambridge, UK) at 1:100 diluted in blocking buffer]. Slides
were rinsed three times in PBS-T and incubated in 1:200 dilution of secondary
antibodies in blocking buffer [Goat anti-Rabbit IgG AlexaFluor 488 (A-11006,
Invitrogen, Carlsbad, CA, USA) and Goat anti-Rat IgG AlexaFluor 594 (A-11012,
Invitrogen, Carlsbad, CA, USA)] for 1 h at room temperature and wash three times
with PBS. The slides were mounted with ProLongDiamond Antifade Mountant
with DAPI (P36962, Thermosher, Waltham, MA, USA) and coverslip. We used
the Zeiss Axioskop 2 MOT (Oberkochen, Germany) uorescence imaging micro-
scope at 20x and 40x magnication. Intensity thresholds were set according to the
highest intensity image and a minimum of three images was analyzed for each
treatment (n=3).
Brca1 deletion in mOSE cells. mOSE cells were isolated from homozygous
Brca1tm1Brn mice as described above and then infected with Ad-Cre to achieve
Brca1 knockout. Ad-GFP was used as a control. Cells were cultured for 1 week after
infection prior to experimental use.
BrdU pulse-chase.Brca1tm1Brn mice were bred to B6.129×1-Gt(ROSA)26Sortm1
(EYFP)Cos/J mice to produce Brca1/YFP mice. Six week-old Brca1/YFP mice were
injected IB with Ad-Cre (8 × 107PFU) or PBS on day 1 and injected IP with BrdU
(0.25 mg daily) on days 710. Ovaries were collected on day 40 and frozen in
Optimal Cutting Temperature Compound.
BrdU immunouorescence. Frozen sections (5 µm) were xed using formalin-
vapor xation38 overnight at 20 °C. Samples were then hydrated in PBS and
antigen retrieval performed using an antigen unmasking solution (pH 6.0, Vector)
in a steam chamber (Hamilton Beach). Slides were then washed in PBS and blocked
with 5% goat serum for 1 h at room temperature. Primary antibodies against
BRCA1 (1:200, H-300, rabbit), GFP (1:1000, ab13970, chicken), and BrdU (1:200,
ab6326, rat) were added and incubated overnight at 4 °C. Following a PBS wash,
species-appropriate secondary antibodies (1:250, Alexauor 594 nm or 488 nm)
were incubated for 1 h at room temperature. Slides underwent a nal PBS wash and
were mounted using Prolong Gold with DAPI (ThermoFisher). Quantications of
BrdU+cells were performed by taking cross-sections of ovaries and manually
counting OSE nuclei (DAPI) around the perimeter of the tissue. Experimenters
were blinded to the experimental group of each tissue section and YFP channel was
not shown during quantication.
RNA-seq sample preparation. For TGFB1 treatment of monolayer cultures,
mOSE cells were plate d 24 h prior to the addition of TGFB1 (10 ng/mL, R&D
Systems) and cells were collected after 4 days of treatment. For inducible Snail
expression and Brca1 deletion, cells were plated for 24 h prio r to the addition of
doxycycline (200 ng/mL, Sigma), and RNA was collected 4 days later. For sphere-
forming conditions, mOSE cells (1 × 106) were rst plated as monolayer cultures
24 h prior to treatment with TGFB1 (10 ng/mL, R&D Systems). Four days after the
addition of TGFB1, mOSE cells were then plated in free-oating sphere-forming
conditions. Cells were maintained in sphere-forming cultures for 2 weeks prior to
RNA collection (RNAeasy Kit, Qiagen). TGFB1 was replenished when placing
mOSE cells in sphere-forming conditions.
Library preparation and sequencing. Total RNA was quantied using a Nano-
Drop Spectrophotometer ND-1000 (NanoDrop Technologies, Inc.) and its integ-
rity was assessed on a 2100 Bioanalyzer (Agilent Technologies). Libraries were
generated from 250 ng of total RNA as follows: mRNA enrichment was performed
using the NEBNext Poly(A) Magnetic Isolation Module (New England BioLabs).
cDNA synthesis was achieved with the NEBNext RNA First Strand Synthesis and
NEBNext Ultra Directional RNA Second Strand Synthesis Modules (New England
BioLabs). The remaining steps of library preparation were done using the NEBNext
Ultra II DNA Library Prep Kit for Illumina (New England BioLabs). Adapters and
PCR primers were purchased from New England BioLabs. Libraries were quantied
using the Quant-iTPicoGreen®dsDNA Assay Kit (Life Technologies) and the
Kapa Illumina GA with Revised Primers-SYBR Fast Universal kit (Kapa Biosys-
tems). Average size fragment was determined using a LabChip GX (PerkinElmer)
instrument.
The libraries were normalized, denatured in 0.05 N NaOH, and then diluted to
200 pM and neutralized using HT1 buffer. ExAMP was added to the mix and the
clustering was done on an Illumina cBot and the owcell was run on a HiSeq 4000
for 2 × 100 cycles (paired-end mode) following the manufacturers instructions. A
phiX library was used as a control and mixed with libraries at 1% level. The
Illumina control software was HCS HD 3.4.0.38 and the real-time analysis program
was RTA v. 2.7.7. The program bcl2fastq2 v2.18 was then used to demultiplex
samples and generate fastq reads.
RNA-seq processing and differential expression. Transcript quantication for
each sample was performed using Kallisto (v0.45.0)39 with the GRCm38 tran-
scriptome reference and the -b 50 bootstrap option. The R package Sleuth
(v0.30.0)40 was then used to construct general linear models for the log-
transformed expression of each gene across experimental conditions. Walds test
was used to test for signicant variables for each gene and the resultant p-values
were adjusted to q-values using the BenjaminiHochberg false discovery rate
method. Signicant genes were dened as genes with a q-value <0.05. An effect size
(beta coefcient of the regression model) cutoff of >0.5 or < 0.5 was also used for
each data set. To compare TGFB1-treated spheres with monolayer conditions and
the untreated spheroids, we used a regression model including an interaction term
between spheroid culture and TGFB1 treatment to identify expression patterns
associated with TGFB1-treated spheres that are not simply additive changes due to
the two variables.
Gene set enrichment analysis and pathway activity inference. GSEA was
performed with the R package fgsea (v1.13.5)41. GO terms, KEGG pathways,
COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-021-02045-w ARTICLE
COMMUNICATIONS BIOLOGY | (2021) 4:527 | https://doi.org /10.1038/s42003-021-02045-w | www.nature.com/commsbio 9
Content courtesy of Springer Nature, terms of use apply. Rights reserved
Reactome pathways, and Hallmark genesets were collected from the Molecular
Signatures Database (MSigDB)22,23 and used to query differential expression
results ranked by fold change. All gene sets discussed in the manuscript have a
signicant enrichment (BenjaminiHochberg adjust p-value <0.05). For pathway
activity inference, we used the R package PROGENy (v1.9.6)42. Pathway activity
was compared between experimental conditions using a simple linear model and p-
values were adjusted using the BenjaminiHochberg false detection rate method.
CD44 cell sorting. mOSE cells were treated with TGFB1 (10 ng/mL, 2 days) prior
to collecting cells for FACS. Cells (1 × 107) were trypsinized and a single-cell
suspension was made using a 40 µm cell strainer. Cells were labeled and sorted as
previously described18. Briey, cells were resuspended in a ow buffer (4% FBS in
PBS) and incubated with anti-CD44 conjugated to allophycocyanin (1:5000;
eBioscience, San Diego, CA) for 15 min at 4 °C. Unbound antibody was removed
with washing buffer and the fraction of cells with surface protein labeled with
CD44 antibody was determined using a MoFlo cell sorter (Dako Cytomation).
Statistics and reproducibility. Statistical analyses were conducted in R (v4.0.3).
For all comparisons of means, data were assessed with linear models and two-sided
Studentsttests. For experiments involving multiple comparisons, p-values were
controlled using the BenjaminiHochberg false discovery rate method. To ensure
reproducibility, all experiments were conducted with independent biological
replicates. The number of replicates are specied in gure legends. Individual
values for data plots are included in the public GitHub repository at https://github.
com/dpcook/ose_stemness.
Reporting summary. Further information on research design is available in the Nature
Research Reporting Summary linked to this article.
Data availability
Raw sequencing les have been deposited and are available along with processed
transcript quantications at GSE122875. Source data for all plots in main gures is
provided with this paper.
Code availability
All code used to process data and generate gures is available in a public GitHub
repository43. Code can be viewed at https://github.com/dpcook/ose_stemness.
Received: 15 June 2020; Accepted: 26 March 2021;
References
1. Barker, N., van de Wetering, M. & Clevers, H. The intestinal stem cell. Genes
Dev. 22, 18561864 (2008).
2. Tata, P. R. et al. Dedifferentiation of committed epithelial cells into stem cells
in vivo. Nature 503, 218223 (2013).
3. Herrick, S. E. & Mutsaers, S. E. The potential of mesothelial cells in tissue
engineering and regenerative medicine applications. Int. J. Artif. Organs 30,
527540 (2007).
4. Singavarapu, R., Buchinsky, N., Cheon, D. J. & Orsulic, S. Whole ovary
immunohistochemistry for monitoring cell proliferation and ovulatory wound
repair in the mouse. Reprod. Biol. Endocrinol. 8, 98 (2010).
5. Murdoch, W. J. Ovarian surface epithelium during ovulatory and anovulatory
ovine estrous cycles. Anat. Rec. 240, 322326 (1994).
6. Bjersing, L. & Cajander, S. Ovulation and the role of the ovarian surface
epithelium. Experientia 31, 605608 (1975).
7. Wright, J. W., Jurevic, L. & Stouffer, R. L. Dynamics of the primate ovarian
surface epithelium during the ovulatory menstrual cycle. Hum. Reprod. 26,
14081421 (2011).
8. Ng, A. & Barker, N. Ovary and mbrial stem cells: biology, niche and cancer
origins. Nat. Rev. Mol. Cell Biol. 16, 625638 (2015).
9. Auersperg, N. The stem-cell prole of ovarian surface epithelium is
reproduced in the oviductal mbriae, with increased stem-cell marker density
in distal parts of the mbriae. Int. J. Gynecol. Pathol. 32, 444453 (2013).
10. Gamwell, L. F., Collins, O. & Vanderhyden, B. C. The mouse ovarian surface
epithelium contains a population of LY6A (SCA-1) expressing progenitor cells
that are regulated by ovulation-associated factors. Biol. Reprod. 87, 80 (2012).
11. Flesken-Nikitin, A. et al. Ovarian surface epithelium at the junction area
contains a cancer-prone stem cell niche. Nature 495, 241245 (2013).
12. Mani, S. A. et al. The epithelial-mesenchymal transition generates cells with
properties of stem cells. Cell 133, 704715 (2008).
13. Haensel, D. & Dai, X. Epithelial-to-mesenchymal transition in cutaneous
wound healing: where we are and where we are heading. Dev. Dyn. 247,
473480 (2018).
14. Shaw, T. J. & Martin, P. Wound repair: a showcase for cell plasticity and
migration. Curr. Opin. Cell Biol. 42,2937 (2016).
15. Knight, P. G. & Glister, C. TGF-βsuperfamily members and ovarian follicle
development. Reproduction 132, 191206 (2006).
16. Carter, L. E. et al. COX2 is induced in the ovarian epithelium during
ovulatory wound repair and promotes cell survival.Biol. Reprod. 101,
961974 (2019).
17. Hebbard, L. et al. CD44 expression and regulation during mammary gland
development and function. J. Cell Sci. 113, 26192630 (2000). (Pt 14).
18. Alwosaibai, K. et al. PAX2 maintains the differentiation of mouse oviductal
epithelium and inhibits the transition to a stem cell-like state. Oncotarget 8,
7688176897 (2017).
19. Herheliuk, T., Perepelytsina, O., Ugnivenko, A., Ostapchenko, L. &
Sydorenko, M. Investigation of multicellular tumor spheroids enriched for a
cancer stem cell phenotype. Stem Cell Investig. 6, 21 (2019).
20. Robertson, F. M. et al. Imaging and analysis of 3D tumor spheroids
enriched for a cancer stem cell phenotype. J. Biomol. Screen. 15, 820829
(2010).
21. Liao, M.-J. et al. Enrichment of a population of mammary gland cells that
form mammospheres and have in vivo repopulating activity. Cancer Res. 67,
81318138 (2007).
22. Subramanian, A. et al. Gene set enrichment analysis: a knowledge-based
approach for interpreting genome-wide expression proles. Proc. Natl Acad.
Sci. U.S.A. 102, 1554515550 (2005).
23. Liberzon, A. et al. The Molecular Signatures Database (MSigDB) hallmark
gene set collection. Cell Syst. 1, 417425 (2015).
24. Luo, K. Signaling Cross Talk between TGF-β/Smad and Other Signaling
Pathways. Cold Spring Harb. Perspect. Biol. 9, a022137 (2017).
25. Cook, D. P. & Vanderhyden, B. C. Context specicity of the EMT
transcriptional response. Nat. Commun. 11,19 (2020).
26. Tacha, D., Zhou, D. & Cheng, L. Expression of PAX8 in normal and neoplastic
tissues: a comprehensive immunohistochemical study. Appl.
Immunohistochem. Mol. Morphol. 19, 293299 (2011).
27. Hodgkinson, K. et al. GREB1 is an estrogen receptor-regulated tumour
promoter that is frequently expressed in ovarian cancer. Oncogene 37,
58735886 (2018).
28. Bai, F. et al. BRCA1 suppresses epithelial-to-mesenchymal transition and stem
cell dedifferentiation during mammary and tumor development. Cancer Res.
74, 61616172 (2014).
29. Di Zazzo, E. et al. Estrogen receptors in epithelial-mesenchymal transition of
prostate cancer. Cancers 11, 1418 (2019).
30. Peris-Torres, C. et al. Extracellular protease ADAMTS1 is required at early
stages of human uveal melanoma development by inducing stemness and
endothelial-like features on tumor cells. Cancers 12, 801 (2020).
31. Rafehi, S. et al. TGFβsignaling regulates epithelial-mesenchymal plasticity in
ovarian cancer ascites-derived spheroids. Endocr. Relat. Cancer 23, 147159
(2016).
32. Wilson, M. M., Weinberg, R. A., Lees, J. A. & Guen, V. J. Emerging
mechanisms by which EMT programs control stemness. Trends Cancer Res.
https://doi.org/10.1016/j.trecan.2020.03.011 (2020).
33. Ocaña, O. H. et al. Metastatic colonization requires the repression of the
epithelial-mesenchymal transition inducer Prrx1. Cancer Cell 22, 709724
(2012).
34. Beumer, J. & Clevers, H. Regulation and plasticity of intestinal stem cells
during homeostasis and regeneration. Development 143, 36393649 (2016).
35. Brandt, R. et al. Cell type-dependent differential activation of ERK by
oncogenic KRAS in colon cancer and intestinal epithelium. Nat. Commun. 10,
2919 (2019).
36. Castiello, L. et al. Disruption of IFN-I signaling promotes HER2/Neu tumor
progression and breast cancer stem cells.Cancer Immunol. Res. 6, 658670
(2018).
37. Tonary, A. M., Macdonald, E. A., Faught, W., Senterman, M. K. &
Vanderhyden, B. C. Lack of expression of c-KIT in ovarian cancers is
associated with poor prognosis. Int. J. Cancer 89, 242250 (2000).
38. Jockusch, H., Voigt, S. & Eberhard, D. Localization of GFP in frozen sections
from unxed mouse tissues: immobilization of a highly soluble marker protein
by formaldehyde vapor. J. Histochem. Cytochem. 51, 401404 (2003).
39. Bray, N. L., Pimentel, H., Melsted, P. & Pachter, L. Near-optimal probabilistic
RNA-seq quantication. Nat. Biotechnol. 34, 525527 (2016).
40. Pimentel, H., Bray, N. L., Puente, S., Melsted, P. & Pachter, L. Differential
analysis of RNA-seq incorporating quantication uncertainty. Nat. Methods
14, 687690 (2017).
41. Korotkevich, G., Sukhov, V. & Sergushichev, A. Fast gene set enrichment
analysis. Preprint at bioRxiv https://doi.org/10.1101/060012.
42. Schubert, M. et al. Perturbation-response genes reveal signaling footprints in
cancer gene expression. Nat. Commun. 9, 20 (2018).
43. Cook, D. P. dpcook/ose_stemness: comms_bio_release, v1.0. Zenodo, https://
doi.org/10.5281/zenodo.4610521.
ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-021-02045-w
10 COMMUNICATIONS BIOLOGY | (2021) 4:527 | https://doi.org /10.1038/s42003-021-02045-w | www.nature.com/commsbio
Content courtesy of Springer Nature, terms of use apply. Rights reserved
Acknowledgements
We thank the tissue donors for making this research possible. We also thank Dr. Ken
Garson for generating mOSE cells with inducible Snail and inducible GFP expression,
and Dr. Diane Lagacé for providing B6.129×1-Gt(ROSA)26Sortm1(EYFP)Cos/J mice. We
wish to acknowledge the contribution of staff of the McGill University and Génome
Québec Innovation Centre (Montreal, QC) for performing library preparation and
sequencing associated with RNA-seq experiments. We also wish to acknowledge Stem-
Core Laboratories (Ottawa, ON) for performing the FACS of CD44-positive mOSE cells.
This work was supported by grants from the Canadian Institutes of Health Research
and the National Science and Engineering Research Council (BCV). L.E.C. and L.F.G.
were supported by Ontario Graduate Scholarships, D.P.C. by a Frederick Banting and
Charles Best Doctoral Award (CIHR), and C.W.M. by the Vanier Canada Graduate
Scholarship.
Author contributions
L.E.C. and B.C.V. conceived the study. L.E.C., D.P.C., and B.C.V. interpreted results and
wrote the manuscript. L.E.C., M.A.G., L.F.G., O.C., H.A.D., and T.D. performed cell
culture experiments, qPCR analysis, and Western blots. C.W.M and D.A.L.
performed immunouorescence and its analysis. O.C. derived mOSE and hOSE cultures.
C.W.M. performed mouse experiments. D.P.C. and L.E.C. performed all computational
analysis.
Competing interests
The authors declare no competing interests.
Additional information
Supplementary information The online version contains supplementary material
available at https://doi.org/10.1038/s42003-021-02045-w.
Correspondence and requests for materials should be addressed to B.C.V.
Reprints and permission information is available at http://www.nature.com/reprints
Publishers note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional afliations.
Open Access This article is licensed under a Creative Commons
Attribution 4.0 International License, which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative
Commons license, and indicate if changes were made. The images or other third party
material in this article are included in the articles Creative Commons license, unless
indicated otherwise in a credit line to the material. If material is not included in the
articles Creative Commons license and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this license, visit http://creativecommons.org/
licenses/by/4.0/.
© The Author(s) 2021
COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-021-02045-w ARTICLE
COMMUNICATIONS BIOLOGY | (2021) 4:527 | https://doi.org /10.1038/s42003-021-02045-w | www.nature.com/commsbio 11
Content courtesy of Springer Nature, terms of use apply. Rights reserved
1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com
... In recent years, several studies have demonstrated that Snai1 (Wang et al., 2021a), an epithelial mesenchymal transition (EMT) regulator, plays an important role in tumor cell invasion and metastasis (Choi et al., 2015;Greening et al., 2015). Snai1 functions as a transcription factor by binding to the E-box element (5'-CANNTG-3') in the promoter region of many tumor-associated genes (Carter et al., 2021;Nieto, 2002;Wang et al., 2021a). Snail promotes or inhibits the transcription of E-cadherin, IL-8, and CD147 to regulate EMT in various epithelial tumor cells (Yin et al., 2021;Zada et al., 2021). ...
Article
Full-text available
Malignant meningiomas often show invasive growth that makes complete tumor resection challenging, and they are more prone to recur after radical resection. Invasive meningioma associated transcript 1 (IMAT1) is a long noncoding RNA located on Homo sapiens chromosome 17 that was identified by our team based on absolute expression differences in invasive and non-invasive meningiomas. Our studies indicated that IMAT1 was highly expressed in invasive meningiomas compared with non-invasive meningiomas. In vitro studies showed that IMAT1 promoted meningioma cell invasion through the inactivation of the Krüppel-like factor 4 (KLF4)/hsa-miR22-3p/Snai1 pathway by acting as a sponge for hsa-miR22-3p, and IMAT1 knockdown effectively restored the tumor suppressive properties of KLF4 by preserving its tumor suppressor pathway. In vivo experiments confirmed that IMAT1 silencing could significantly inhibit the growth of subcutaneous tumors and prolong the survival period of tumor-bearing mice. Our findings demonstrated that the high expression of IMAT1 is the inherent reason for the loss of the tumor suppressive properties of KLF4 during meningioma progression. Therefore, we believe that IMAT1 may be a potential biological marker and treatment target for meningiomas.
Article
Full-text available
Epithelial–mesenchymal plasticity contributes to many biological processes, including tumor progression. Various epithelial–mesenchymal transition (EMT) responses have been reported and no common, EMT-defining gene expression program has been identified. Here, we have performed a comparative analysis of the EMT response, leveraging highly multiplexed single-cell RNA sequencing (scRNA-seq) to measure expression profiles of 103,999 cells from 960 samples, comprising 12 EMT time course experiments and independent kinase inhibitor screens for each. We demonstrate that the EMT is vastly context specific, with an average of only 22% of response genes being shared between any two conditions, and over half of all response genes were restricted to 1–2 time course experiments. Further, kinase inhibitor screens revealed signaling dependencies and modularity of these responses. These findings suggest that the EMT is not simply a single, linear process, but is highly variable and modular, warranting quantitative frameworks for understanding nuances of the transition. It is unclear if a common EMT expression program exists. Here, the authors perform multiplexed single-cell RNA sequencing across 12 EMT time courses and 16 kinase inhibitor screens, and find that EMT transcriptional responses are context specific and EMT is not a single, linear transition.
Article
Full-text available
Tissue regeneration relies on adult stem cells (SCs) that possess the ability to self-renew and produce differentiating progeny. In an analogous manner, the development of certain cancers depends on a subset of tumor cells, called cancer stem cells (CSCs), with SC-like properties. In addition to being responsible for tumorigenesis, CSCs exhibit elevated resistance to therapy and thus drive tumor relapse post-treatment. The epithelial–mesenchymal transition (EMT) programs promote SC and CSC stemness in many epithelial tissues. Here, we provide an overview of the mechanisms underlying the relationship between stemness and EMT programs, which may represent therapeutic vulnerabilities for the treatment of cancers.
Article
Full-text available
Extracellular matrix remodeling within the tumor microenvironment has been recognized as a relevant dynamic framework during tumor growth. However, research on proteases that trigger this remodeling keeps revealing a wide range of actions including both pro- and anti-tumorigenic. The extracellular protease ADAMTS1 exemplifies this dual role. In this work, we first confirmed a positive correlation of ADAMTS1 with endothelial-like phenotype of human melanoma cells together with the finding of associated signatures, including key genes such as endothelial CDH5. Using a CRISPR-Cas9 approach, we observed that the inhibition of ADAMTS1 in an aggressive uveal melanoma model compromised its endothelial-like properties, and more importantly, caused a robust blockade on the progression of tumor xenografts. Although vasculature emerged affected in ADAMTS1-deficient tumors, the most relevant action implied the downregulation of endothelial CDH5 in tumor cells, in association with stemness markers. Indeed, melanoma sphere assays also revealed a deficient commitment to form spheres in the absence of ADAMTS1, directly correlating with stemness markers and, remarkably, also with CDH5. Finally, taking advantage of advanced bioinformatics tools and available public data of uveal melanomas, we disclosed new prognosis factors, including endothelial elements and ADAMTS proteases. Our findings support the key role of ADAMTS proteases for uveal melanoma development since earlier stages, modulating the complex crosstalk between extracellular matrix and the induction of stemness and endothelial-like features. To our knowledge, this is the first report that supports the development of therapeutic targets on the extracellular matrix to overcome uveal melanoma.
Article
Full-text available
Prostate cancer (PC) remains a widespread malignancy in men. Since the androgen/androgen receptor (AR) axis is associated with the pathogenesis of prostate cancer, suppression of AR-dependent signaling by androgen deprivation therapy (ADT) still represents the primary intervention for this disease. Despite the initial response, prostate cancer frequently develops resistance to ADT and progresses. As such, the disease becomes metastatic and few therapeutic options are available at this stage. Although the majority of studies are focused on the role of AR signaling, compelling evidence has shown that estrogens and their receptors control prostate cancer initiation and progression through a still debated mechanism. Epithelial versus mesenchymal transition (EMT) is involved in metastatic spread as well as drug-resistance of human cancers, and many studies on the role of this process in prostate cancer progression have been reported. We discuss here the findings on the role of estrogen/estrogen receptor (ER) axis in epithelial versus mesenchymal transition of prostate cancer cells. The pending questions concerning this issue are presented, together with the impact of the available data in clinical management of prostate cancer patients.
Article
Full-text available
Background: Cancer stem cells (CSCs) provide self-renewal of the tumor after radiation and chemotherapy. These cells are important during tumor development. The in vitro model of avascular tumor that enriched of cells with stem like characteristics is critical to understanding of the role CSCs in the tumor. Methods: Cell viability was evaluated by MTT assay. The expression of cancer stem cells markers (CD133, CD44, CD24 and bmi-1) in 2D cell culture and multicellular tumor spheroids (MCTS) of MCF-7 cells was evaluated. The Stemi2000 software AxioVisionRed 4.7 was used for image processing. The volume of spheroids was calculated by Bjerkvig formula. Results: The highest expression of CD133, CD44, CD24 and bmi-1 receptors was detected in MCTS, enriched with cancer stem cells (eMCTS). Cell aggregates of eMCTS culture were returned from suspension to adhesive conditions. It was found that the cells of the MCTS surface layers were enriched with CD133, CD44, CD24, bmi-1, EpCAM, vim markers, but not adherent cells. eMCTS are less sensitive to anticancer drugs (cisplatin, methotrexate and doxorubicin), than adhesive cell culture and MCTS cultured under standard conditions in a complete nutrient medium (P<0.05). Conclusions: We observed that eMCTS population possesses aggressive phenotypic characteristics such as invasion, cancer stem cell markers and chemoresistance. eMCTS model could improve the screening efficiency of therapeutical agents against CSCs.
Article
Full-text available
The ovarian surface epithelium (OSE) is a monolayer of cells surrounding the ovary that is ruptured during ovulation. After ovulation, the wound is repaired, however this process is poorly understood. In epithelial tissues, wound repair is mediated by an epithelial-to-mesenchymal transition (EMT). Transforming Growth Factor Beta-1 (TGFβ1) is a cytokine commonly known to induce an EMT and is present throughout the ovarian microenvironment. We therefore hypothesized that TGFβ1 induces an EMT in OSE cells and activates signaling pathways important for wound repair. Treating primary cultures of mouse OSE cells with TGFβ1 induced an EMT mediated by TGFβRI signaling. The transcription factor Snail was the only EMT-associated transcription factor increased by TGFβ1 and, when overexpressed, was shown to increase OSE cell migration. A PCR array of TGFβ signaling targets determined Cox2 to be most highly induced by TGFβ1. Constitutive Cox2 expression modestly increased migration and robustly enhanced cell survival, under stress conditions similar to those observed during wound repair. The increase in Snail and Cox2 expression with TGFβ1was reproduced in human OSE cultures, suggesting these responses are conserved between mouse and human. Finally, the induction of Cox2 expression in OSE cells during ovulatory wound repair was shown in vivo, suggesting TGFβ1 increases Cox2 to promote wound repair by enhancing cell survival. These data support that TGFβ1 promotes ovulatory wound repair by induction of an EMT and activation of a COX2-mediated pro-survival pathway. Understanding ovulatory wound repair may give insight into why ovulation is the primary non-hereditary risk factor for ovarian cancer.
Article
Full-text available
Oncogenic mutations in KRAS or BRAF are frequent in colorectal cancer and activate the ERK kinase. Here, we find graded ERK phosphorylation correlating with cell differentiation in patient-derived colorectal cancer organoids with and without KRAS mutations. Using reporters, single cell transcriptomics and mass cytometry, we observe cell type-specific phosphorylation of ERK in response to transgenic KRASG12V in mouse intestinal organoids, while transgenic BRAFV600E activates ERK in all cells. Quantitative network modelling from perturbation data reveals that activation of ERK is shaped by cell type-specific MEK to ERK feed forward and negative feedback signalling. We identify dual-specificity phosphatases as candidate modulators of ERK in the intestine. Furthermore, we find that oncogenic KRAS, together with β-Catenin, favours expansion of crypt cells with high ERK activity. Our experiments highlight key differences between oncogenic BRAF and KRAS in colorectal cancer and find unexpected heterogeneity in a signalling pathway with fundamental relevance for cancer therapy.
Article
Full-text available
Estrogenic hormone replacement therapy increases the risk of developing ovarian cancer, and estrogen promotes tumour initiation and growth in mouse models of this disease. GREB1 (Growth regulation by estrogen in breast cancer 1) is an ESR1 (estrogen receptor 1)-upregulated protein which may mediate estrogen action. GREB1 knockdown prevents hormone-driven proliferation of several breast and prostate cancer cell lines and prolongs survival of mice engrafted with ovarian cancer cells, but its mechanism of action remains unclear. In this study, we explored GREB1 function in ovarian cancer. GREB1 overexpression in ovarian cancer cell lines increased cell proliferation and migration and promoted a mesenchymal morphology associated with increased Col1a2, which encodes a collagen I subunit. GREB1 knockdown inhibited proliferation and promoted an epithelial morphology associated with decreased Col1a2. In human tissues, GREB1 was expressed in all ESR1-expressing tissues throughout the normal female reproductive tract, in addition to several tissues that did not show ESR1 expression. In a TMA of ovarian cancer cases, GREB1 was expressed in 75–85% of serous, endometrioid, mucinous, and clear cell carcinomas. Serous, endometrioid, and mucinous ovarian cancers were almost always positive for either ESR1 or GREB1, suggesting a possible reliance on signalling through ESR1 and/or GREB1. Targeting GREB1 may inhibit tumour-promoting pathways both downstream and independent of ESR1 and is therefore a possible treatment strategy worthy of further investigation.
Article
Full-text available
Type I interferon (IFN-I) is a class of antiviral immunomodulatory cytokines involved in many stages of tumor initiation and progression. IFN-I acts directly on tumor cells to inhibit cell growth, and indirectly by activating immune cells to mount antitumor responses. To understand the role of endogenous IFN-I in spontaneous, oncogene-driven carcinogenesis, we characterized tumors arising in Her2/neu transgenic (neuT) mice carrying a nonfunctional mutation in the IFN-I receptor (IFNAR1). Such mice are unresponsive to this family of cytokines. Compared to parental neu+/- mice (neuT mice), IFNAR1-/- neu+/- mice (IFNAR-neuT mice) showed earlier onset and increased tumor multiplicity with marked vascularization. IFNAR-neuT tumors exhibited deregulation of genes having adverse prognostic value in breast cancer patients, including the breast cancer stem cell (BCSC) marker aldehyde dehydrogenase-1A1 (ALDH1A1). An increased number of BCSCs were observed in IFNAR-neuT tumors, as assessed by ALDH1A1 enzymatic activity, clonogenic assay, and tumorigenic capacity. In vitro exposure of neuT+ mammospheres and cell lines to antibodies to IFN-I resulted in increased frequency of ALDH+ cells, suggesting that IFN-I controls stemness in tumor cells. Altogether, these results reveal a role of IFN-I in NeuT-driven spontaneous carcinogenesis through intrinsic control of BCSCs.
Article
Full-text available
Aberrant cell signaling can cause cancer and other diseases and is a focal point of drug research. A common approach is to infer signaling activity of pathways from gene expression. However, mapping gene expression to pathway components disregards the effect of post-translational modifications, and downstream signatures represent very specific experimental conditions. Here we present PROGENy, a method that overcomes both limitations by leveraging a large compendium of publicly available perturbation experiments to yield a common core of Pathway RespOnsive GENes. Unlike pathway mapping methods, PROGENy can (i) recover the effect of known driver mutations, (ii) provide or improve strong markers for drug indications, and (iii) distinguish between oncogenic and tumor suppressor pathways for patient survival. Collectively, these results show that PROGENy accurately infers pathway activity from gene expression in a wide range of conditions.