Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Feb 21.
Published in final edited form as: Dev Biol. 2018 Aug 15;442(2):301–314. doi: 10.1016/j.ydbio.2018.08.006

The ERM family member Merlin is required for endometrial gland morphogenesis

Erin Williams Lopez a,b, Zer Vue a,c, Russell R Broaddus b,d, Richard R Behringer a,b,c, Andrew B Gladden a,b,*
PMCID: PMC12921682  NIHMSID: NIHMS2140489  PMID: 30118662

Abstract

Disruption of endometrial gland formation or function can cause female infertility. Formation of endometrial glands via tubulogenesis of luminal epithelial cells requires the establishment and maintenance of cell polarity and cell adhesion. The FERM domain-containing protein Merlin coordinates epithelial cell polarity and cell adhesion and is critical for epithelial tissue function in the skin and kidney. We now demonstrate a requirement for Merlin in endometrial gland development. Conditional deletion of Merlin in the endometrium results in female infertility caused by the absence of gland formation. Interestingly, we observed glandular epithelial markers within discrete groups of cells in the Merlin-deficient luminal epithelium. Wnt signaling, a pathway necessary for endometrial gland development is maintained in Merlin-deficient endometrium, suggesting the glandular fate program is active. Instead, we observe increased levels of apical actin and markers indicative of high membrane tension on the basal surface of the Merlin-deficient luminal epithelium. These findings suggest that the structural integrity of the luminal epithelium during gland formation is required for appropriate endometrial tubulogenesis and tissue function. Moreover, our work implicates Merlin-dependent regulation of mechanical tension in the proper formation of endometrial gland architecture and function.

Keywords: Merlin, Endometrium, Gland morphogenesis, NF2, Tension

1. Introduction

Female infertility is a complex problem that is still not well understood. Infertility is medically relevant as 17% of women in United States need infertility assistance (Chandra et al., 2014, 2013). In a multitude of animal models, the endometrium has been found to be a common cause of impaired fertility, specifically when endometrial glands are absent (Franco et al., 2010; Jeong et al., 2010; Su et al., 2016). The endometrium is composed of stromal cells that are adjacent to the luminal and glandular epithelium. The luminal epithelium, the epithelium surrounding the uterine cavity, gives rise to glandular epithelium that invaginates into the stroma after birth (Gray et al., 2001; Vue et al., 2018). The glandular epithelium secretes mucins and nutrients to protect the uterus from infection and facilitate proper uterine and blastocyst response during pregnancy (Braga and Gendler, 1993; Filant and Spencer, 2014).

Endometrial gland function has been found to be required for protein secretion correlated with successful embryo implantation and as a source of nutrients as the placenta forms (Boomsma et al., 2009; Burton et al., 2002; Filant and Spencer, 2013). Data examining fertile and infertile women demonstrated that increased levels of secreted factors from the endometrial glands were predictive of a successful implant in women undergoing in vitro fertilization (Karagouni et al., 1998; Vialard et al., 2013). Examining how the endometrial glands form may lead to a better understanding of how female infertility arises and can be potentially treated.

Endometrial glands are formed through a process called tubulogenesis that occurs postnatally in both mice and humans (Gray et al., 2001; Vue et al., 2018). In mice at postnatal day 0 (P0), the uterine lumen is lined by epithelial cells but no endometrial glands are present (Hu et al., 2004; Vue et al., 2018). Between P0 and P21 endometrial glands elongate from the luminal epithelium and begin branching. After puberty endometrial glands consist of branched structures that are connected to the luminal epithelium (Arora et al., 2016). Development of the endometrial gland involves Wnt signaling, however, what role this pathway has in regulating glandular elongation and/or branching is still not known. There are many genes that have been associated with gland formation including Foxa2, β-catenin, and Wnt7a (Dunlap et al., 2011; Goad et al., 2017; Jeong et al., 2010, 2009). While the molecular mechanisms underlying endometrial gland development have begun to be examined, there are still many unanswered questions that could help us understand both endometrial gland formation and infertility.

Endometrial gland formation is a collaboration of cell adhesion and apicobasal polarity to properly form glandular architecture. Deletion of E-cadherin (Cdh1) in the endometrium impairs cell:cell adhesion resulting in the epithelium not sustaining adhesion or the endometrial gland structure (Reardon et al., 2012). In other cell types cell polarity has been shown to be necessary for polarized migration and cell adhesion. Merlin, the protein product of the Nf2 tumor suppressor gene, couples apical junction maturation and polarity establishment in vivo in various types of epithelia (Gladden et al., 2010). What role Merlin has in endometrial gland development is still not known.

We recently found that proteins that regulate polarity are disrupted in early endometrial cancer leading to disruption of membrane receptor signaling (Williams et al., 2017). The polarity regulating protein Merlin is present in human endometrial glands (Uhlen et al., 2010; Uhlén et al., 2015). To understand how the regulation of cell adhesion and polarity affects endometrial development, we generated two mouse models with a conditional deletion of the Nf2 gene either in the endometrial epithelium (Wnt7a-Cre) or in the entire endometrium (PR-Cre). The endometrium of both mouse models lack endometrial gland formation, while markers for glandular cell fate are still observed. Additionally, we found that markers indicative of increased membrane tension were present in both models endometrial epithelium. Finally, female infertility was observed suggesting the cells that express gland markers are not functional. This data suggest that Merlin is required for endometrial formation and function.

2. Materials and methods

2.1. Mouse strains

We generated both the Nf2flox/flox; Wnt7a-Cre and the Nf2flox/flox; PR-Cre mice by crossing the previously generated Nf2flox/flox mice with either Wnt7a-Cre mice or PR-Cre mice and the Cre was contributed paternally. (Giovannini et al., 2000; Huang et al., 2012; Soyal et al., 2005). The heterozygous deletion of Nf2 (Nf2flox/wt; Wnt7a-Cre or Nf2flox/wt; PR-Cre) resulted in no phenotype in the uterus compared to wild-type mice and for control mice we included Nf2flox/wt, Nf2flox/flox, Nf2flox/wt; PR-Cre, Nf2flox/wt; Wnt7a-Cre littermates. The MD Anderson Cancer Center Institutional Animal Care and Use Committee (IACUC) approved all animal procedures. Mice were maintained on a mixed FVB and C57BL/6J background.

2.2. Fecundity testing

Six control (Nf2flox/flox, Nf2flox/wt; PR-Cre, Nf2flox/wt; Wnt7a-Cre) and 6 Nf2flox/flox; PR-Cre female mice were bred for 3 months with Nf2flox/flox males. On the day plugs were detected, weight was taken (E0.5) and at E7.5 to examine increased weights corresponding to early pregnancy. Females were left with males during the entire 3 months and litters were counted for number of pups.

2.3. Immunohistochemistry

Mouse tissue was fixed in either 3.7% formalin or 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS). Tissues were then either processed for paraffin-embedding (McCormick Scientific) or OCT-embedding (Fischer Healthcare). Hematoxylin & eosin staining: Paraffin-embedded samples were heated to 55 °C for 20 min. Samples were incubated in histoclear (3×) for 10 min each. Samples were progressively hydrated in (100% (×2), 95% (×2), and 70% EtOH) ethanol and then deionized water for 22 min. Samples were stained with Harris Hematoxylin modified (StatLab) for 7 min followed by washing in EtOH Acid (1% HCl in 70% EtOH) and bluing in 0.1% Sodium Bicarbonate. Samples were stained with 0.5% eosin Y (Sigma Aldrich) for 5 s before being washed in 4× 100% EtOH. Samples were incubated in HIstoclear (3×, 10 min) before being mounted with Permount (Fischer). BrdU and Cleaved Caspase 3 staining: All mice were pulsed with BrdU (100 μg per gram of body weight) for 2 h before being sacrificed. OCT-embedded samples were thawed and dried for 10 min. Samples were then washed in PBS 2× for 30 min each and permeabilized in 0.5% Triton/PBS for 20 min at room temperature. Samples were then washed in PBS, 3× before incubated in 1.5 N HCl in PBS for 10 min. Slides were briefly washed 3× in PBS. Samples were blocked for 1 h in PBST/10% FBS and then left overnight at room temperature with the following primary antibodies diluted in blocking buffer (BrdU [1:200, mouse, abcam, ab6326] and/or Cleaved Caspase 3 [1:400, rabbit, Cell Signaling, Asp175]). Samples were washed in PBST 3× for 20 min and then incubated in blocking buffer with secondary antibodies for 1 h at room temperature. Other immunofluorescent staining: OCT-embedded samples were thawed and dried for 10 min. Samples were then hydrated in PBS 2× for 30 min each and permeabilized in 0.5% Triton/PBS for 20 min at room temperature. Samples were then washed in PBS. Samples were blocked for 1 h in PBST/10% FBS/1% BSA and then left overnight at room temperature with primary antibodies (Sox9 [1:250, rabbit, Santa Cruz, sc-20095], E-cadherin [1:1000, mouse, BD, 610182], Par3 [1:500, rabbit, Millipore, 07–330], ZO-1 [1:500, rabbit, Invitrogen, 61–7300], Muc1 [1:100, rabbit, Dan Carson’s Laboratory (Pemberton et al., 1995)], Foxa2 [1:2000, rat, Seven Hills BioReagents, WRAB-1200], Vinculin [1:300, mouse, Millipore, MAB3574], Myosin IIB [1:500, rabbit, Biolegend, 909901], Rho [1:100, mouse, Santa Cruz, sc-418], pMLC [1:500, rabbit, abcam, ab2480], β-catenin [1:1000, mouse, BD, 610153], α-catenin [1:500, mouse, BD, 610193], Merlin [1:100, rabbit, Santa Cruz, sc-331], P-cadherin [1:200, rat, Takara, M109], Yap [1:300, mouse, Cell Signaling, 4912]) in blocking buffer. The mouse tissue samples were washed in PBST 3 times for 20 min and then incubated in blocking buffer with secondary antibodies and either smooth muscle actin-Cy3 (1:200, mouse, Sigma, C6198, SMA) or Phalloidin Cy3 (1:200, Invitrogen, A12380) for 1 h at room temperature. All samples were mounted with Fluoromount-G (Southern Biotech, 0100–01). Light sheet three-dimensional imaging was done using the protocol described by Vue et al. (2018).

2.4. Quantitative RT-PCR

RNA was extracted by homogenizing snap-frozen tissue in Trizol, according to the manufacturer’s instructions. RNA pellets were dissolved in 30–50 μL of Diethyl pyrocarbonate (DEPC) water at 55 °C. RNA purity was confirmed by spectrophotometer to have an A260/A280 ratio greater than 1.8. cDNA was synthesized using a SuperScript First-Strand Synthesis kit (Invitrogene). The cDNA was analyzed using SYBR green quantification with the 7900HT Sequence Detection System (Applied Systems). Primers used for qRT-PCR are listed in the following table. Samples were assayed in triplicate. Data were normalized to Gapdh and β-actin. Samples below the limit of quantification were not included. The fold change of the ΔCT compared to Hprt was utilized for analysis (2−ΔCT).

Table of qRT-PCR primers for isolated endometrium.

Name F/R Sequence
β Actin F CTAAGGCCAACCGTGAAAAG
R ACCAGAGGCATACAGGGACA
GAPDH F CCCTTCATTGACCTCAACTACA
R ATGACAAGCTTCCCGTTCTC
Ccnd1–1 F GCCCTCCGTATCTTACTTCAAG
R GCGGTCCAGGTAGTTCATG
Ccnd1–2 F CATCTACACTGACAACTCTATCCG
R TCTGGCATTTTGGAGAGGAAG
Myc-1 F GCTGTTTGAAGGCTGGATTTC
R GATGAAATAGGGCTGTACGGAG
Myc-2 F CGATTCCACGGCCTTCTC
R TCTTCCTCATCTTCTTGCTCTTC
Sox9–1 F CAAGACTCTGGGCAAGCTC
R GGGCTGGTACTTGTAATCGG
Sox9–2 F GCCGACTCCCCACATTC
R CGCTTCAGATCAACTTTGCC
Sox17–1 F CGATGAACGCCTTTATGGTG
R TTCTCTGCCAAGGTCAACG
Sox17–2 F AATATGGCCCACTCACACTG
R TTTCTCTGTCTTCCCTGTCTTG
Wnt7a-1 F ACGAGTGTCAGTTTCAGTTCC
R AATCGCATAGGTGAAGGCAG
Wnt7a-2 F TTACACAATAACGAGGCGGG
R TTGTCCTTGAGCACGTAGC
Hoxa11–1 F AGGAGAAGGAGCGACGG
R GGTATTTGGTATAAGGGCAGCG
Hoxa11–2 F CTAAACTAGCATCCCTACCCTG
R ATCAGTTCTTGCCTCTTCCG
Birc2 F GAAGAAAATGCTGACCCTACAGA
R GCTCATCATGACGACATCTTTC
Birc3 F AGAGAGGAGCAGATGGAGCA
R TTTGTTCTTCCGGATTAGTGC
CTGF F GGGCCTCTTCTGCGATTTC
R ATCCAGGCAAGTGCATTGGTA
Merlin-1 F CTCCTGCATACCTGCATATCTC
R CTAAGCCAGTCCACACTTCTAC
Merlin-2 F CAGGGAAGAGAAGGCTAGAAAG
R ATTGGGTTCATGGGTGGATAG
Open in a new tab

2.5. Image processing and analysis

All tissue samples and cells were visualized using a Nikon A1 laser scanning confocal microscope or a Nikon 80i upright fluorescent microscope. Images were processed using the Nikon-Elements software (Nikon). Quantitative analysis of the mouse tissue including myometrium, cell adhesion angles, BrdU, and Cleaved Caspase 3 (CC3) calculations was performed utilizing ImageJ (https://imagej.nih.gov/ij). Line intensity plots (Phalloidin, pMLC, [Vinculin, Myosin IIB], Sox9, [P-cadherin, E-cadherin], Muc1, and FoxA2) were also performed on ImageJ analysis.

3. Results

3.1. Deletion of Merlin in the endometrium blocks gland formation

We examined the role of the FERM-domain containing protein Merlin, the protein product of the tumor suppressor gene Nf2, in the endometrium by generating two conditional knockout mouse models targeting Merlin in the endometrial epithelium, Nf2flox/flox;Wnt7a-Cre, or in the endometrial epithelium, stroma and myometrium, Nf2flox/flox;PR-Cre. We collected uteri from Nf2flox/flox;Wnt7a-Cre and control littermates before puberty at P21. Control mice at P21 have an average of 8–10 endometrial gland cross sections visible in histological cross-sections, while Nf2flox/flox;Wnt7a-Cre mice have less than one (Fig. 1a, f). Histologically, the Nf2flox/flox;Wnt7a-Cre luminal epithelium was similar to the control luminal and glandular simple columnar epithelial cells (Fig. 1c–e). The absence of glands in the Nf2flox/flox;Wnt7a-Cre mice were not due to a delay in gland formation as no glands were observed during estrous at P60 (Supplemental Fig. 1c, d). Nf2flox/flox;Wnt7a-Cre mice were not able to be consistently examined following 60 days of age due to accelerated post-natal lethality likely due to loss of Merlin in tissues other then the endometrium. Similarly, analysis of Nf2flox/flox;PR-Cre uteri at P21 found a lack of endometrium glands compared to control littermates that had an average of 9–11 visible endometrial gland cross sections visible in histological cross-sections (Fig. 1g, Supplemental Fig. 1a, b). Consistent with the aged Nf2flox/flox;Wnt7a-Cre mice three month old Nf2flox/flox;PR-Cre endometrium did not display gland formation (Supplemental Fig. 1e, f). To determine if endometrial gland morphogenesis was initiated, whole mount uterine horns were collected at P7 and immunostained for cytokeratin 8 to examine endometrial epithelium. The control uterus had epithelial bud structures that will give rise to endometrial glands as has been previously described (Fig. 1h and h′, Supplemental Video 1) (Vue et al., 2018). Conversely, the Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre uterus had no evidence of endometrial buds, demonstrating initial glandular morphology was not present (Fig. 1i, i′, j, j′, Supplemental Videos 2–3). We verified that Nf2 was deleted in our Nf2flox/flox;Wnt7a-Cre endometrium (Fig. 1k). Uterine tissue was isolated from control or Nf2flox/flox;Wnt7a-Cre mice and analyzed by PCR to determine whether Cre-mediated recombination had occurred. Using three separate PCR primer sets, we determined whether the uterine tissue was heterozygous or homozygous for the Nf2 floxed allele (Fig. 1k top panel). In addition, we verified the presence of Cre and deletion of Exon 2 following Cre-mediated recombination. We found that Exon 2 had been deleted in tissue expressing Cre indicating deletion of Nf2 (Fig. 1k). Furthermore, we isolated RNA from control, Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre uteri and performed qRT-PCR on intact Nf2 transcripts and found Nf2 RNA was reduced further indicating we accurately deleted Nf2 in the endometrium (Fig. 1l, m).

Fig. 1. Loss of Merlin blocks endometrial gland development.

Fig. 1.

Open in a new tab

Control (a) and Nf2flox/flox;Wnt7a-Cre (b) H & E staining show the loss of glands on the Nf2flox/flox;Wnt7a-Cre uteri. Scale Bar: 50 μm. Magnification of the control (c) glandular epithelium, (d) luminal epithelium and the Nf2flox/flox;Wnt7a-Cre (e) luminal epithelium. Quantifications of the number of glands observed in a cross-section of control and Nf2flox/flox;Wnt7a-Cre mice (n = 13) (f) or control and Nf2flox/flox;PR-Cre mice (n = 3) (g). **P < 0.01, ***P < 0.001. Still shots from Supplemental videos 1–3 of three-dimensional images of Wild-type (h, h′) (Scale bar, 500 μM), Nf2flox/flox;Wnt7a-Cre (i, i′) (Scale bar, 200 μM), and Nf2flox/flox;PR-Cre (j, j′) (Scale bar, 300 μM). Merlin loss was examined by PCR for recombination of the targeted Nf2 allele (ΔExon2) in control (Nf2flox/wt and Nf2flox/flox) and three independent Nf2flox/flox;Wnt7a-Cre mice (k). qRT-PCR of uterine RNA examining the level of Nf2 transcript in control, Nf2flox/flox;Wnt7a-Cre (l) and Nf2flox/flox;PR-Cre (m).

Endometrial glands arise from endometrial luminal epithelium and are molecularly distinguished from luminal epithelium by expressing markers of glandular cell fate. To determine if Nf2flox/flox;Wnt7a-Cre or Nf2flox/flox;PR-Cre endometrium have markers of gland fate in the absence of endometrial gland structures, we examined glandular markers that are distinct from the luminal epithelium. Previous studies have shown that endometrial glands have nuclear Foxa2 (Jeong et al., 2010) and we confirmed this in control glandular but not luminal epithelium (Fig. 2a, a′, b, b′, e). Examination of Foxa2 immunostaining in both Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre luminal epithelium showed that the majority of cells did not have nuclear Foxa2 (Fig. 2c, c′, d, d′). The Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre luminal epithelium did have cells with nuclear FOXA2 at discrete locations, specifically near curves in the luminal epithelium (Fig. 2c, c′, d, d′, arrows) that had a similar intensity to the glandular epithelium of the control tissue (Fig. 2f). Nuclear Sox9 is another endometrial gland marker (Gonzalez et al., 2016). We observed Sox9 in the nucleus of control glandular epithelium (Fig. 2g, g′, j). We also examined Sox9 immunostaining in Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre endometrial epithelium and found that the majority of luminal epithelia cells had cytoplasmic Sox9 that had a similar intensity as the control luminal epithelium (Fig. 2h, h′, i, i′, j). Like the observations with Foxa2 immunostaining, we also saw a subset of cells with nuclear Sox9 in both Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre luminal epithelium that had the same intensity as nuclear Sox9 in control glandular epithelium (Fig. 2g–i, k). This data indicates that markers of glandular cell fate are present in the luminal epithelium of the Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre endometrium lacking endometrial glands.

Fig. 2. Disruption of endometrial gland markers in P21 endometrium lacking Merlin.

Fig. 2.

Open in a new tab

Foxa2 staining in control endometrial glands (a, a′), control (b, b′), Nf2flox/flox;Wnt7a-Cre (c, c′), and Nf2flox/flox;PR-Cre (d, d′) luminal epithelium. Line plots of Foxa2 intensity in control glandular epithelium compared to control, Nf2flox/flox;Wnt7a-Cre, and Nf2flox/flox;PR-Cre luminal epithelial cell (e), left region of line pot is the basal boundary of the cell. Line intensity of Foxa2 in control endometrial glands compared to Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre luminal epithelium with nuclear Foxa2 (f). White and red lines indicate where line intensity plot was quantified. Sox9 staining in control (g, g′), Nf2flox/flox;Wnt7a-Cre (h, h′), and Nf2flox/flox;PR-Cre (i, i′) mice. The line intensity of Sox9 in control glandular epithelium, as well as, control, Nf2flox/flox;Wnt7a-Cre, and Nf2flox/flox;PR-Cre luminal epithelium, the left area of the line represents the basal boundary of the cell (j). Line intensity of Sox9 in control endometrial glands compared to Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre luminal epithelium with nuclear Sox9 (k). White and red lines indicate where line intensity plot was quantified. Scale bar, 50 μM.

3.2. Apicobasal polarity and junction condensation is absent in mutant endometrial epithelium

With the robust change in endometrial epithelial morphology in the Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre mice, we wanted to examine whether cell polarity and cell:cell adhesion were altered. Because Merlin has previously been shown to disrupt apicobasal polarity, specifically the PAR complex, we examined the localization of Par3 in Nf2flox/flox;Wnt7a-Cre endometrial epithelium (Gladden et al., 2010). While Par3 localizes to the apical surface of the luminal and glandular epithelium in control mice (Fig. 3a, a′, b, b′), in the Nf2flox/flox;Wnt7a-Cre luminal epithelium Par3 was absent at the apical junctional region (Fig. 3c, c′). This indicates apicobasal polarity is disrupted in Nf2flox/flox;Wnt7a-Cre endometrial epithelium. Additionally, because Merlin is necessary in junctional maturation, we were interested in whether apical junction proteins like ZO-1, E-cadherin and P-cadherin were properly localized. The tight junction protein ZO-1 was examined by immunofluorescence and found to localize to the apical junctions in both control, Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre endometrial epithelium (Fig. 3d, d′, e, e′, Supplemental Fig. 2a, a′, b, b′). This was surprising because in other Nf2-deficient epithelial tissues, ZO-1 is mislocalized to the cytoplasm (Gladden et al., 2010). Proper localization of ZO-1 can indicate that tight junctions are able to form, however, it does not address whether the junctions are fully functional. The adherens junction protein E-cadherin was examined in the endometrial epithelium of control, Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre mice. E-cadherin appeared relatively similar between control, Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre mice at both P7 and P21 (Fig. 3f′–i′, Supplemental Fig. 2c′–f′). There was some intense punctate E-cadherin staining noted in a subset of the Nf2flox/flox;Wnt7a-Cre mice, not observed in control littermates but overall, proteins in the apical adherens and tight junctions, appear to localize properly.

Fig. 3. Merlin loss disrupts apicobasal polarity without altering apical junctions.

Fig. 3.

Open in a new tab

Par3 staining in control luminal epithelium (a, a′) and glandular epithelium (b, b′) show apical localization in control, however Nf2flox/flox;Wnt7a-Cre luminal epithelium (c, c′) has diffuse staining, * denotes either the luminal or glandular lumen. Scale bar, 50 μM. ZO-1 staining in control (d, d′) and Nf2flox/flox;Wnt7a-Cre (e, e′) show minimal changes to the apical localization. Scale bar, 100 μM. Staining of E-cadherin and P-cadherin in control (f, h) and Nf2flox/flox;Wnt7a-Cre (g, i) mice at P7 (f–g) and P21 (h–i) show an increased luminal staining of P-cadherin and minimal changes to E-cadherin at P21. (Scale bar, 10 μM. Line intensity graphs of P-cadherin for control and Nf2flox/flox;Wnt7a-Cre (j–k) show an increase of P-cadherin at the apical lumen at both P7 (j) and P21 (k) in Nf2flox/flox;Wnt7a-Cre. White and red lines indicate where line intensity plot was quantified.

P-cadherin is another cadherin that is known to be expressed within the endometrium (Kadokawa et al., 1989; Nose and Takeichi, 1986; van der Linden et al., 1995). P-cadherin is involved in mammary tubulogenesis where it is expressed on the cap cells in terminal end buds (Radice et al., 1997). Bazelliéres et al. (2015) showed that P-cadherin and E-cadherin can play different roles in mechanical stress and cellular tension. E-cadherin has been shown to strengthen the cell adhesions, while P-cadherin regulates the intercellular tension (Bazellières et al., 2015). Thus, we examined P-cadherin in the precycling uterus. In control uteri, P-cadherin is dispersed throughout the membrane of the epithelium, while in the Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre mice, P-cadherin localizes to the apical lumen (Fig. 3f″–i″, Supplemental Fig. 2c″–f″). This P-cadherin apical localization occurs early during gland development at P7 (Fig. 3f″, g″, j) and increases over time observable at P21 (Fig. 3h″, i″, k).

3.3. Increase in intercellular tension in mutant endometrial epithelium

To determine why endometrial gland architecture did not form, endometrial tissue was examined at an early stage of gland formation (P7). Gland formation has been examined in numerous organs, however the mechanisms by which the endometrial gland form are not well understood. In order for the buds to form and then go through tubulogenesis, proliferation and apoptosis may be involved in the cellular changes necessary for gland formation. Additionally, local changes in cell:cell and cell:matrix tension could play a role in the elongation of the gland structure. Proliferation was examined following a 2-h pulse of BrdU. In Nf2flox/flox;Wnt7a-Cre mice, there was a decrease in BrdU incorporation in both the epithelium and the stroma (Fig. 4a–b, e–f). The epithelium and the stroma showed about a 5% decrease in BrdU incorporation (Fig. 4e–f). This suggests there is a small decrease in cellular proliferation in the endometrium due to Merlin loss in the epithelium. In addition to proliferation, we examined apoptosis using the apoptotic marker, cleaved caspase 3 (CC3). CC3 levels were low in both control and Nf2flox/flox;Wnt7a-Cre mice and no significant difference was observed (Fig. 4c, d, g).

Fig. 4. Increased basal membrane tension is independent of changes in proliferation and apoptosis within the Merlin deficient endometrium.

Fig. 4.

Open in a new tab

BrdU staining labeling proliferating cells in control (a, a′) and Nf2flox/flox;Wnt7a-Cre (b, b′) endometrial epithelium and stroma. Cleaved Caspase 3 (CC3) staining marking apoptotic cells in the control (c, c′) and Nf2flox/flox;Wnt7a-Cre (d, d′) endometrial epithelium. Quantification of the amount of BrdU positive cells within the luminal epithelium (e) and endometrial stroma (f). Percent of cells that have CC3 in the luminal epithelium of control versus Nf2flox/flox;Wnt7a-Cre mice (g) show no significant change. Scale bar, 50 μM. Control (h) and Nf2flox/flox;Wnt7a-Cre (i) endometrium show increased co-localization of Myosin IIB (h′–i′) and Vinculin (h”-i”) in the Nf2flox/flox;Wnt7a-Cre endometrium suggesting an increase in basal membrane tension. Scale bar, 20 μM. Phosphorylated Myosin light-chain (pMLC) staining reveal increased levels at the basal membrane of Nf2flox/flox;Wnt7a-Cre (k, k′) versus control (j, j′) luminal epithelium. Scale bar, 50 μM. Comparison of the intensity of Myosin IIB and Vinculin in control (black and blue lines) and Nf2flox/flox;Wnt7a-Cre uteri (green and red lines (l). Quantification of pMLC intensity in control (black) versus Nf2flox/flox;Wnt7a-Cre uteri (green) (m). White and red lines indicate where line intensity plot was quantified.

Given the modest decreases in uterine epithelial proliferation in the mutant mice, we were interested in how tension was affected. Tension may be necessary to properly form the glandular structure. In addition, since P-cadherin is known to react to changes in tension in other tissues, the alterations in P-cadherin localization (Fig. 3f″–i″) may suggest changes in tension. Co-localization of Vinculin and Myosin IIB indicates an increase in cellular tension (Cohen et al., 2006; Dumbauld et al., 2010; Yonemura et al., 2010). When an epithelium has low amounts of tension at cell:cell or cell:matrix contacts there is a decreased association of actin, Myosin IIB, and Vinculin to the cell:cell or cell:matrix junction (Cohen et al., 2006; Dumbauld et al., 2010; Yonemura et al., 2010). When cell membranes are under tension then Myosin IIB and Vinculin associate at the cell membrane (Cohen et al., 2006; Dumbauld et al., 2010; Yonemura et al., 2010). Thus, co-localization of Vinculin and Myosin IIB can indicate cell:cell and cell:ECM contacts are under tension. Examination of control endometrium shows little to no co-localization of Vinculin and Myosin IIB except in small sections at the basal membrane of the luminal epithelium (Fig. 4h, h′, h″). The Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre mice however, had co-localization of Vinculin and Myosin IIB across a majority of the basal membrane of the luminal epithelium (Fig. 4i, i′, i″, Supplemental Fig. 3a–b). The co-localization was more intense in mutant endometrium compared to any co-localization observed in the control endometrium (Fig. 4h, i, l). In addition, pMLC (phospho-Myosin light-chain) is known to be associated with areas of high cellular tension. Rock (Rho-associated protein kinase) phosphorylates Myosin light-chain which causes Myosin ATPase to activate actin (Amano et al., 1996). We observed an increase in the intensity of pMLC at the basal membrane in Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre mice compared to control mice (Fig. 4j, j′, k, k′, Supplemental Fig. 3c–d). A comparison of line plots showed about a three times increase in the pMLC intensity on the basal membrane of mutant endometrium compared to control endometrium (Fig. 4m).

Cell culture-based modeling systems have found that utilizing the angle of cell:cell and cell:ECM interactions can calculate where high versus low tension regions are located (Cerchiari et al., 2015). Using this strategy, we measured the cell:cell and cell:ECM angles, in control samples (n = 3 mice, 30 angles/mouse) and found the average control cell:cell:angles was 98°, while the average Nf2flox/flox;Wnt7a-Cre (n = 3 mice, 30 angles/mouse) cell:cell angle was acute at 81° suggesting that the mutant cells are compressing the apical surface of the luminal epithelium (Fig. 5a–e).

Fig. 5. Increased apical constriction in Nf2flox/flox;Wnt7a-Cre endometrium.

Fig. 5.

Open in a new tab

Quantification of the angles observed between the basal and lateral membranes (Cell:ECM) or apical and lateral membranes (Cell:Cell) of the luminal epithelium in control (a–b, Cell:ECM median = 88°, Cell:Cell median = 98°) and Nf2flox/flox;Wnt7a-Cre (b–c, Cell:ECM median = 96°, Cell:Cell median = 81°) endometrium showing increased apical constriction in Merlin-deficient epithelial cells. Statistical summary of the angles measured at the cell:cell membrane (d) and the cell:ECM membrane (e). Phalloidin staining of F-actin in P7 (f, g) and P21 (h, i) control (f, f′, h, h′) and Nf2flox/flox;Wnt7a-Cre (g, g′, i, i′) endometrium. Scale bar (f, g) 50 μM, (h, i) 20 μM. Comparison of the line intensity of Phalloidin at the apical surface of control and Nf2flox/flox;Wnt7a-Cre mice from P7 (j) or P21 (k) images. White and red lines indicate where line intensity plot was quantified.

In addition, the cell:ECM interactions showed obtuse angles in the Nf2flox/flox;Wnt7a-Cre mice compared to acute angles in the control mice (Fig. 5a–e). In order for this switch of cell interaction angles to occur, we hypothesized that mutant endometrial luminal epithelium is under increased tension. Additionally, because of the direction of the angle changes, we postulated that this could be from increased apical constriction producing a contractile actin ring on the apical surface of the mutant endometrial epithelium. To examine whether an actin ring was present, we stained uterine tissue with Phalloidin and found that Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre tissue has more intense F-actin staining around the apical lumen then the control endometrium (Fig. 5f–k, Supplemental Fig. 5c–e). This indicates the potential changes in tension observed at the basal membrane may be from apical constriction thus pulling the cells in a manner where they are unable to form a glandular structure.

3.4. Wnt signaling is intact in mutant endometrium

To understand what signaling pathways may be involved in the lack of gland formation, we examined pathways regulated by Merlin including the Hippo and β-catenin signaling pathways. Merlin is well known for the tissue-specific role it can play as a positive regulator of Hippo signaling in Drosophila and in mammalian brain and liver (Lavado et al., 2013; Zhang et al., 2010). Hippo signaling is inactive when YAP (Yes-associated protein) or TAZ (transcriptional co-activator with PDZ-binding motif) translocates into the nucleus and interacts with transcription factors like TEAD (TEA domain-containing transcription factor family) proteins to increase target gene expression. Control and mutant mice at either P7 or P21 showed no nuclear YAP staining in the endometrial epithelium, suggesting that Hippo signaling is not increased in the mutant uterus (Supplemental Fig. 4b–e). We confirmed the staining with embryonic osteoblasts in cartilage primordium where nuclear YAP staining has been previously shown (Supplemental Fig. 4a, a′) (Dupont et al., 2011). To confirm these results, we examined YAP downstream targets Birc2 (Baculoviral IAP repeat containing 2), Birc3, and Ctgf (Connective tissue growth factor) by qRT-PCR in uteri isolated from control and mutant mice at P7. While Ctgf did show a significant decrease between control and mutant mice, Birc3 showed an increase in expression and Birc2 was not significantly different suggesting some targets of YAP may be affected by Merlin loss (Supplemental Fig. 4f).

Numerous observations indicate that Wnt signaling is involved in endometrial gland development so we also examined β-catenin localization, Wnt signaling regulators and downstream targets. β-catenin staining was examined for localization to the nucleus indicative of active β-catenin but no nuclear β-catenin was observed in the wild type or the Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre littermates on the endometrial luminal or glandular epithelium (Fig. 6a–c, a′–c′, Supplemental Fig. 5a–b). However, there were variable changes to the Wnt downstream targets and effectors. Myc and Sox9, Wnt downstream target genes were decreased in the Nf2flox/flox;Wnt7a-Cre endometrium (Fig. 6d–e). Sox17, is known in gastric cancer mouse models to be involved in a feedback loop with Wnt signaling. When Wnt signaling increases, it results in an increase in Sox17 that suppresses additional Wnt signals (Du et al., 2009). Sox17 expression was shown to be increased in Nf2flox/flox;Wnt7a-Cre uteri (Fig. 6f) (Chew et al., 2011; Guimarães-Young et al., 2016). Additionally, in Nf2flox/flox;Wnt7a-Cre mice we observed an increase in Cyclin D1 (Ccnd1) and Axin2 expression both Wnt downstream targets (Fig. 6h, i). Furthermore, the expression of the Wnt ligand gene, Wnt7a, was increased in the Nf2flox/flox;Wnt7a-Cre uterus at P7 (Fig. 6g). This indicates that Wnt signaling is not specifically lost within the Nf2flox/flox;Wnt7a-Cre endometrial epithelium. Taking into account previous studies showing Wnt/β-catenin signaling is necessary for gland formation and our data that gland fate markers are present in the luminal epithelium it would suggest that the glandular fate program is present in the mutant uteri but the glands are unable to morphologically form.

Fig. 6. β-catenin signaling is intact in the absence of glandular architecture in Nf2flox/flox;Wnt7a-Cre endometrium.

Fig. 6.

Open in a new tab

β-catenin staining in control glands (a, a′), control luminal epithelium (b, b′), and Nf2flox/flox;Wnt7a-Cre luminal epithelium (c, c′) show no change in β-catenin localization. Scale bar, 10 μM. Yellow dotted line outlines the epithelial tissue. qRT-PCR of downstream targets of β-catenin signaling, Myc (d), Sox9 (e), Sox17 (f), Wnt7a (g), Cyclin D1 (h), and Axin2 (i) in control and Nf2flox/flox;Wnt7a-Cre mice.

3.5. The mutant luminal epithelial cells expressing glandular markers do not support fertility

To understand whether the Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre luminal epithelium that expressed glandular markers function like control glandular epithelium, we examined two different gland functions. First, endometrial glands secrete more mucins then the endometrial luminal epithelium. Thus, we were interested if Muc1 was present in the endometrium of our Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre mice. We observed that the control mice exhibited a higher intensity of Muc1 immunostaining in the glandular epithelium than in the luminal epithelium (Supplemental Fig. 6a, a′, d). Compared to control luminal epithelial cells, Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre luminal epithelium showed increased Muc1 immunostaining on portions of the epithelium (Supplemental Fig. 6b, b′, c, c′). Interestingly, the intensity of Muc1 immunostaining in Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre luminal cells were at a similar level as the glandular staining of the control mice (Supplemental Fig. 6d). The increase in glandular markers provides evidence that the mutant luminal epithelium is capable of producing mucins associated with differentiated glandular epithelium but is unable to assemble the glandular architecture.

Endometrial glands are required for female fertility, therefore we determined if our mutant females were fertile. Nf2flox/flox;Wnt7a-Cre mice were not able to be consistently examined following 60 days of age due to accelerated post-natal lethality likely due to loss of Merlin in tissues other then the endometrium. Therefore, we performed a fecundity study over 3 months using 6 Nf2flox/flox;PR-Cre and 6 control female mice. The females were mated to control male mice. During this time period, the 6 control female mice had an average of 3 litters with 7 pups per litter, while the Nf2flox/flox;PR-Cre mice did not have any litters (Supplemental Fig. 6e–g). Vaginal plugs were observed in the Nf2flox/flox;PR-Cre mice similar to control littermates suggesting mating and estrous is intact in the mutant females. Weight was examined after plugs were observed and control mice showed an increase in weight around E7.5, Nf2flox/flox;PR-Cre mice never gained weight during the normal gestation period of 19 days. In addition, the gross morphology and histology of the ovaries were examined and were similar to control females. To determine if the Nf2flox/flox;PR-Cre endometrial stroma responds to pregnancy signals we examined BRDU incorporation in Nf2flox/flox;PR-Cre and control mice 3.5 days following the observation of a copulation plug. Control endometrial stroma showed high levels of proliferation indicating the stroma is preparing for implantation while the endometrial epithelium labeled with pan-cadherin shows almost no proliferation (Fig. 7a, a′, a”). The Nf2flox/flox;PR-Cre stroma has little to no BRDU incorporation and there is a high level of proliferation in the luminal epithelium (Fig. 7b, b′, b”). To determine if implantation occurs in the Nf2flox/flox;PR-Cre uterus we examined control and Nf2flox/flox;PR-Cre mice 8.5 days post copulation and observed multiple embryos implanted in the control (Fig. 7c) compared to no implanted embryos in the Nf2flox/flox;PR-Cre uterus (Fig. 7d). Further examination of the uterus at 8.5 days post-copulation we observed multinucleated cells in the control stroma (Fig. 7e, g) while the Nf2flox/flox;PR-Cre stroma has increased in density and the luminal epithelium appear to be stratifying (Fig. 7f, H). This indicates that Nf2flox/flox;PR-Cre female mice are infertile and do not respond to copulation signals.

Fig. 7. Loss of a stromal response in Nf2flox/flox;PR-Cre endometrium inhibits embryo implantation.

Fig. 7.

Open in a new tab

Staining for BRDU, SMA or Pan-cadherin in control (a) and Nf2flox/flox;PR-Cre endometrium 3.5 days (E3.5) following observation of a copulation plug showing a strong stromal proliferation response in the control stroma and increased proliferation in the epithelium of the mutant. Scale bar, 20 μM. Pictures of control (c) or Nf2flox/flox;PR-Cre (d) uteri 8.5 days (E8.5) following observation of a copulation plug showing no implanted embryos in the mutant uterus. H & E staining of E8.5 control (e, g) and Nf2flox/flox;PR-Cre (f, h) uterus showing increased number of polyploidy cells in the control endometrial stroma and a condensation of the Nf2flox/flox;PR-Cre endometrial stroma. Scale bar (e, f) 500 μM and (g, h) 100 μM. Summary model for gland formation in control (i) endometrium where the gland specific transcripts (green cells) turning on and the gland undergoing involution in to the stroma to form the gland morphology compared to mutant endometrium (j) where the glandular program is initiated but the gland structure can not form do to the increased basal tension.

4. Discussion

To understand the role of Merlin and the establishment of apicobasal polarity in endometrial development, we generated two mouse models in which Merlin was deleted within the endometrial epithelium, Nf2flox/flox;Wnt7a-Cre, or the entire endometrium, Nf2flox/flox;PR-Cre. Both mutants fail to form endometrial glands during postnatal development. Both β-catenin (Ctnnb1) and E-cadherin (Cdh1) deficient mice have been shown to lose endometrial glands as well but within the mutant tissue β-catenin and E-cadherin was properly localized (Jeong et al., 2009; Reardon et al., 2012). This suggests that adhesion was not affected similar to what has previously been observed in other types of Merlin-deficient epithelial tissues and indicates the loss of glandular structure was due to other defects (Gladden et al., 2010).

In the majority of other mouse mutants that affect endometrial gland development, endometrial specific transcription factors, Foxa2 and/or Sox9, are not expressed within the nuclei of the luminal epithelium (Dunlap et al., 2011; Filant and Spencer, 2013; Franco et al., 2010). While a majority of the luminal epithelium in Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre mice showed no nuclear Sox9 and Foxa2 similar to the control luminal epithelium, there were distinct regions in the Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre luminal epithelium that showed increased nuclear localization. Interestingly, the areas of increased nuclear Sox9 and Foxa2 had similar intensities as control glandular epithelium. In addition, there was a subset of the luminal epithelium of the Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre that had similar levels of Muc1 as the control glandular epithelium suggesting regions of the mutant luminal epithelium could secrete gland specific proteins. However, the loss of fertility observed in Nf2flox/flox;PR-Cre females suggest that the regions of the luminal epithelium with glandular markers are not able to fully function. This suggests that Nf2flox/flox;Wnt7a-Cre and Nf2flox/flox;PR-Cre luminal epithelium can initiate differentiation into glandular epithelium correctly but cannot form the proper architecture or function (Fig. 7h, i).

Since a majority of the aglandular mouse mutants do not show nuclear expression of Foxa2 or Sox9, this may mean that Merlin-deficient tissue is unable to from endometrial glands by a different mechanism (Dunlap et al., 2011; Filant and Spencer, 2013; Franco et al., 2010; Jeong et al., 2010). A Notch over activation conditional mutant also causes aglandular phenotype, but the entire luminal epithelium has nuclear Foxa2 expression and Notch signaling has been shown to be necessary for differentiation of gastric glandular epithelium (Su et al., 2016; Matsuda et al., 2005). However, the Merlin-deficient tissue only has subsets of cells that differentiate into glandular tissue, suggesting that Notch overexpression may not cause the Merlin-deficient phenotype. Additionally, we have recently shown that loss of apicobasal polarity, similar to what we observe in the Merlin-deficient endometrium, results in mislocalization of Notch receptors, decreased Notch signaling and altered differentiation of endometrial tumor cells further suggesting high levels of Notch signaling are not driving the aglandular phenotype observed in the Merlin-deficient (Williams et al., 2017). In addition, the cells that express Foxa2 or Sox9 in the nucleus are found at curves within the luminal epithelium, suggesting that potentially the structure of the tissue is necessary for proper differentiation into functional glandular epithelium. The glandular tissue function of the mammary gland have been shown to be affected by the proper architecture and interaction of surrounding cells (Kratochwil, 1969; Weaver et al., 1997). Potentially, while the Nf2flox/flox;Wnt7a-Cre endometrial glandular epithelium is able to initiate glandular differentiation, the inability to initiate the proper glandular architecture makes it unable to maintain and advance the glandular differentiation process

While there was a small decrease in proliferation in Nf2flox/flox;Wnt7a-Cre mice compared to control mice, the increase in cellular tension markers was robust. Increased markers of cellular tension were observed at the basal membrane of the luminal epithelium along with changes in the cell shape in Nf2flox/flox;Wnt7a-Cre mice compared to control mice. Co-localization of Myosin IIB and Vinculin at different junctions has been shown to increase when the junctions are under high tension (Cohen et al., 2006; Dumbauld et al., 2010; Yonemura et al., 2010). Since our data shows an increased co-localization at the basal membrane, we hypothesize this is at the focal adhesions. At focal adhesions, Vinculin is stabilized when actin-myosin related tension increases indicating increased pressure (Cohen et al., 2006; Dumbauld et al., 2010). In addition, cell shape changes for different cellular functions have been associated with accumulation of proteins like Myosin IIB (Schiffhauer and Robinson, 2017). Drosophila models have shown that alterations in force can cause cell shape change which is thought to lead to modifications to cell function (Butler et al., 2009). The changes in cell shape including an increase in acute angles at the apical surface of the cell, therefore we postulated that apical constriction could cause this change to apical angles. F-actin was found to increase at the apical surface of Nf2flox/flox;Wnt7a-Cre luminal epithelium compared to control mice indicative of an increase in apical constriction. The apical constriction would result in the lateral cell edges pulling on the basal membrane causing the increase in tension at the cell:ECM interface. In the epithelium of the chicken lung, apical constriction is necessary for the budding of epithelial glands (Kim et al., 2013). This suggests that rather than having localized apical constriction where budding is suppose to occur, the Merlin-deficient luminal epithelium has apical constriction across the entire epithelium. The global apical constriction may make it infeasible for the luminal epithelium to properly bud into the glandular epithelium causing the aglandular phenotype observed.

Cellular tension has been shown to affect both Hippo signaling and canonical Wnt signaling (Benham-Pyle et al., 2015). In addition, Merlin is known to regulate both Hippo signaling and Wnt signaling in a tissue-specific manner (Kim and Jho, 2016; Morrow et al., 2016; Lavado et al., 2013; Zhang et al., 2010). However, there was no change to YAP localization and Hippo signaling downstream targets showed varying changes depending on the target indicating no major changes in Hippo signaling. Besides being regulated by cellular tension and Merlin, Wnt signaling is also necessary for proper endometrial gland formation (Dunlap et al., 2011; Jeong et al., 2009). We found that β-catenin localization did not change. Additionally, while some canonical Wnt signaling downstream targets were decreased, others increased by qRT-PCR. The increase observed in Sox17 was surprising since previously a Sox17 conditional knockout mouse model caused a Wnt signaling-independent aglandular phenotype (Guimarães-Young et al., 2016). However, alterations in tension and cadherins can modulate the levels of Sox17. This indicates that the observed increase in apical P-cadherin or basal tension markers may be causing the increase in Sox17 expression. This data suggests that canonical Wnt signaling and Hippo signaling is not altered indicating the loss of endometrial glands in the Merlin-deficient uterus occurs in a Wnt- and Hippo-independent manner. Overall our studies shows that Merlin is necessary for the proper formation of glandular architecture without causing loss of signaling pathways known to be involved in endometrial gland development.

Supplementary Material

Supplemental Figure 1
Supplemental Figure 3
Supplemental Figure 2
Supplemental Figure 4
Supplemental Figure 6
Supplemental Figure 5
Supplemental Movie 1
Download video file (9.6MB, mp4)
Supplemental Movie 2
Download video file (7.7MB, mp4)
Supplemental Movie 3
Download video file (10.4MB, mp4)

Supplementary material related to this article can be found online at doi:10.1016/j.ydbio.2018.08.006.

Acknowledgements

We thank the members of the Gladden and Behringer lab for numerous discussions, comments, and suggestions on the manuscript. We would especially like to thank Alejandra Ontiveros for her technical assistance. We would like to thank Dan Carson’s lab for the Muc1 antibody gift. We thank Adriana Paulucci for microscopy assistance in the Genetics Microscopy Core. Supported by Department of Defense (DOD) Grant W81XWH-14–1-0053 to A.B.G. R.R.B. was supported by National Institutes of Health (NIH) Grant HD030284 and the Ben F. Love Endowment. Z.V. was supported by NIH NIGMS R25GM56929, NIH T32-CA009299–27, and NIH HD30284. Veterinary resources were supported by NIH Grant CA16672.

Appendix A. Supplementary material

Supplementary data associated with this article can be found in the online version at doi:10.1016/j.ydbio.2018.08.006.

Footnotes

Declarations of interest

None.

References

  1. Amano M, Ito M, Fukata Y, Chihara K, Nakano T, Matsuura Y, Kaibuchi K, 1996. Phosphorylation and activation of myosin by Rho-associated kinase (Rho-kinase). J. Biol. Chem. 271, 20246–20249. 10.1074/jbc.271.34.20246. [DOI] [PubMed] [Google Scholar]
  2. Arora R, Fries A, Oelerich K, Marchuk K, Sabeur K, Giudice LC, Laird DJ, 2016. Insights from imaging the implanting embryo and the uterine environment in three dimensions. Development 143, 4749–4754. 10.1242/dev.144386. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bazellières E, Conte V, Elosegui-artola A, Serra-picamal X, Bintanel-morcillo M, Roca-cusachs P, Muñoz JJ, Sales-pardo M, Guimerà R, Trepat X, 2015. Control of cell–cell forces and collective cell dynamics by the intercellular adhesome. Nat. Cell Biol. 17. 10.1038/ncb3135. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Benham-Pyle BW, Pruitt BL, Nelson WJ, 2015. Mechanical strain induces E-cadherin-dependent Yap1 and beta-catenin activation to drive cell cycle entry. Science 348 (80), 1024–1027. 10.1126/science.aaa4559. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Boomsma CM, Kavelaars A, Eijkemans MJC, Lentjes EG, Fauser BCJM, Heijnen CJ, MacKlon NS, 2009. Endometrial secretion analysis identifies a cytokine profile predictive of pregnancy in IVF. Hum. Reprod. 24, 1427–1435. 10.1093/humrep/dep011. [DOI] [PubMed] [Google Scholar]
  6. Braga VM, Gendler SJ, 1993. Modulation of Muc-1 mucin expression in the mouse uterus during the estrus cycle, early pregnancy and placentation. J. Cell Sci. 105, 397–405. [DOI] [PubMed] [Google Scholar]
  7. Burton GJ, Watson AL, Hempstock J, Skepper JN, Jauniaux E, 2002. Uterine glands provide histiotrophic nutrition for the human fetus during the first trimester of pregnancy. J. Clin. Endocrinol. Metab. 87, 2954–2959. 10.1210/jc.87.6.2954. [DOI] [PubMed] [Google Scholar]
  8. Butler LC, Blanchard GB, Kabla AJ, Lawrence NJ, Welchman DP, Mahadevan L, Adams RJ, Sanson B, 2009. Cell shape changes indicate a role for extrinsic tensile forces in Drosophila germ-band extension. Nat. Cell Biol. 11, 859–864. 10.1038/ncb1894. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Cerchiari AE, Garbe JC, Jee NY, Todhunter ME, Broaders KE, Peehl DM, Desai TA, LaBarge MA, Thomson M, Gartner ZJ, 2015. A strategy for tissue self-organization that is robust to cellular heterogeneity and plasticity. Proc. Natl. Acad. Sci. USA 112, 2287–2292. 10.1073/pnas.1410776112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Chandra A, Copen CE, Stephen EH, 2014. Infertility Service Use in the United States: Data From the National Survey of Family Growth, 1982–2010. Natl. Health Stat. Report, 1–21. [PubMed] [Google Scholar]
  11. Chandra A, Copen CE, Stephen EH, 2013. Infertility and Impaired Fecundity in the United States, 1982–2010: Data From the National Survey of Family Growth. Natl. Health Stat. Report 19. 10.1093/humrep/des207, (1–18, 1 p following 19). [DOI] [PubMed] [Google Scholar]
  12. Chew L-J, Shen W, Ming X, Senatorov VV, Chen H-L, Cheng Y, Hong E, Knoblach S, Gallo V, 2011. Sox17 regulates the Wnt/beta-catenin signaling pathway in oligodendrocyte progenitor cells. J. Neurosci. 31, 13921–13935. 10.1523/JNEUROSCI.3343-11.2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Cohen DM, Kutscher B, Chen H, Murphy DB, Craig SW, 2006. A conformational switch in vinculin drives formation and dynamics of a talin-vinculin complex at focal adhesions. J. Biol. Chem. 281, 16006–16015. 10.1074/jbc.M600738200. [DOI] [PubMed] [Google Scholar]
  14. Du YC, Oshima H, Oguma K, Kitamura T, Itadani H, Fujimura T, Piao YS, Yoshimoto T, Minamoto T, Kotani H, Taketo MM, Oshima M, 2009. Induction and down-regulation of Sox17 and its possible roles during the course of gastrointestinal tumorigenesis. Gastroenterology 137, 1346–1357. 10.1053/j.gastro.2009.06.041. [DOI] [PubMed] [Google Scholar]
  15. Dumbauld DW, Shin H, Gallant ND, Michael KE, Radhakrishna H, Garcia AJ, 2010. Contractility modulates cell adhesion strengthening through focal adhesion kinase and assembly of vinculin-containing focal adhesions. J. Cell. Physiol. 223, 746–756. 10.1002/jcp.22084.Contractility. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Dunlap KA, Filant J, Hayashi K, Rucker EB 3rd, Song G, Deng JM, Behringer RR, DeMayo FJ, Lydon J, Jeong JW, Spencer TE, 2011. Postnatal deletion of Wnt7a inhibits uterine gland morphogenesis and compromises adult fertility in mice. Biol. Reprod. 85, 386–396. 10.1095/biolreprod.111.091769. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Dupont S, Morsut L, Aragona M, Enzo E, Giulitti S, Cordenonsi M, Zanconato F, Le Digabel J, Forcato M, Bicciato S, Elvassore N, Piccolo S, 2011. Role of YAP/TAZ in mechanotransduction. Nature 474, 179–183. 10.1038/nature10137. [DOI] [PubMed] [Google Scholar]
  18. Filant J, Spencer TE, 2014. Uterine glands: biological roles in conceptus implantation, uterine receptivity and decidualization. Int. J. Dev. Biol. 58, 107–116. 10.1387/ijdb.130344ts. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Filant J, Spencer TE, 2013. Endometrial glands are essential for blastocyst implantation and decidualization in the mouse uterus. (93–93)Biol. Reprod. 88. 10.1095/biolreprod.113.107631. [DOI] [PubMed] [Google Scholar]
  20. Franco HL, Lee KY, Rubel C.a., Creighton CJ, White LD, Broaddus RR, Lewis MT, Lydon JP, Jeong J-W, DeMayo FJ, 2010. Constitutive activation of smoothened leads to female infertility and altered uterine differentiation in the mouse. Biol. Reprod. 82, 991–999. 10.1095/biolreprod.109.081513. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Giovannini M, Robanus-Maandag E, van der Valk M, Niwa-Kawakita M, Abramowski V, Goutebroze L, Woodruff JM, Berns A, Thomas G, 2000. Conditional biallelic Nf2 mutation in the mouse promotes manifestations of human neurofibromatosis type 2. Genes Dev. 14, 1617–1630. 10.1101/gad.14.13.1617. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Gladden AB, Hebert AM, Schneeberger EE, McClatchey AI, 2010. The NF2 tumor suppressor, Merlin, regulates epidermal development through the establishment of a junctional polarity complex. Dev. Cell 19, 727–739. 10.1016/j.devcel.2010.10.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Goad J, Ko YA, Kumar M, Syed SM, Tanwar PS, 2017. Differential Wnt signaling activity limits epithelial gland development to the anti-mesometrial side of the mouse uterus. Dev. Biol. 423, 138–151. 10.1016/j.ydbio.2017.01.015. [DOI] [PubMed] [Google Scholar]
  24. Gonzalez G, Mehra S, Wang Y, Akiyama H, Behringer RR, 2016. Sox9 overexpression in uterine epithelia induces endometrial gland hyperplasia. Differentiation 92, 204–215. 10.1016/j.diff.2016.05.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Gray CA, Bartol FF, Tarleton BJ, Wiley AA, Johnson GA, Bazer FW, Spencer TE, Al GET, 2001. Developmental biology of uterine glands. Biol. Reprod. 65, 1311–1323. [DOI] [PubMed] [Google Scholar]
  26. Guimarães-Young A, Neff T, Dupuy AJ, Goodheart MJ, 2016. Conditional deletion of Sox17 reveals complex effects on uterine adenogenesis and function. Dev. Biol. 414, 219–227. 10.1016/j.ydbio.2016.04.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Hu J, Gray CA, Spencer TE, 2004. Gene expression profiling of neonatal mouse uterine development1. Biol. Reprod. 70, 1870–1876. 10.1095/biolreprod.103.026336. [DOI] [PubMed] [Google Scholar]
  28. Huang C-C, Orvis GD, Wang Y, Behringer RR, 2012. Stromal-to-epithelial transition during postpartum endometrial regeneration. PLoS One 7, e44285. 10.1371/journal.pone.0044285. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Jeong J-W, Kwak I, Lee KY, Kim TH, Large MJ, Stewart CL, Kaestner KH, Lydon JP, DeMayo FJ, 2010. Foxa2 is essential for mouse endometrial gland development and fertility. Biol. Reprod. 83, 396–403. 10.1095/biolreprod.109.083154. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Jeong J-W, Lee HS, Franco HL, Broaddus RR, Taketo MM, Tsai SY, Lydon JP, DeMayo FJ, 2009. beta-catenin mediates glandular formation and dysregulation of beta-catenin induces hyperplasia formation in the murine uterus. Oncogene 28, 31–40. 10.1038/onc.2008.363. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Kadokawa Y, Fuketa I, Nose A, Takeichi M, Nakatsuji N, 1989. Expression pattern of E-cadherin and P-cadherin in mouse embryos and uteri during the periimplantation period. Dev. Growth Differ. 31, 23–30. [DOI] [PubMed] [Google Scholar]
  32. Karagouni EE, Chryssikopoulos A, Mantzavinos T, Kanakas N, Dotsika EN, 1998. Interleukin-1β and interleukin-1α may affect the implantation rate of patients undergoing in vitro fertilization-embryo transfer. Fertil. Steril. 70, 553–559. 10.1016/S0015-0282(98)00243-X. [DOI] [PubMed] [Google Scholar]
  33. Kim HY, Varner VD, Nelson CM, 2013. Apical constriction initiates new bud formation during monopodial branching of the embryonic chicken lung. Development 140, 3146–3155. 10.1242/dev.093682. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Kim S, Jho E, 2016. Merlin, a regulator of Hippo signaling, regulates Wnt/beta-catenin signaling. BMB Rep. 23, 1638–1647. 10.1038/cdd.2016.54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Kratochwil K, 1969. Organ specificity in mesenchymal in the embryonic gland induction development demonstrated of the mammary of the mouse. Dev. Biol. 71, 46–71. [DOI] [PubMed] [Google Scholar]
  36. Lavado A, He Y, Pare J, Neale G, Olson EN, Giovannini M, Cao X, 2013. Tumor suppressor Nf2 limits expansion of the neural progenitor pool by inhibiting Yap/Taz transcriptional coactivators. Development 140, 3323–3334. 10.1242/dev.096537. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Matsuda Y, Wakamatsu Y, Kohyama J, Okano H, Fukuda K, Yasugi S, 2005. Notch signaling functions as a binary switch for the determination of glandular and luminal fates of endodermal epithelium during chicken stomach development. Development 132, 2783–2793. 10.1242/dev.01853. [DOI] [PubMed] [Google Scholar]
  38. Morrow KA, Das S, Meng E, Menezes ME, Sarah K, Metge BJ, Buchsbaum DJ, Samant RS, Lalita A, 2016. Loss of tumor suppressor Merlin results in aberrant activation of Wnt/β-catenin signaling in cancer. Oncotarget 7, 17991–18005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Nose A, Takeichi M, 1986. A novel cadherin cell adhesion molecule: its expression patterns associated with implantation and organogenesis of mouse embryos. J. Cell Biol. 103, 2649–2658. 10.1083/jcb.103.6.2649. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Pemberton L, F CO, Julian J, Pimental RA, Wegner C, Dey SK, Carson D, Das SK, 1995. Expression and steroid hormonal control of Muc-1 in the mouse uterus. Society 136, 3639–3647. [DOI] [PubMed] [Google Scholar]
  41. Radice GL, Ferreira-Cornwell MC, Robinson SD, Rayburn H, Chodosh LA, Takeichi M, Hynes RO, 1997. Precocious mammary gland development in P-cadherin-deficient mice. J. Cell Biol. 139, 1025–1032. 10.1083/jcb.139.4.1025. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Reardon SN, King ML, MacLean J.a., Mann JL, DeMayo FJ, Lydon JP, Hayashi K, 2012. Cdh1 is essential for endometrial differentiation, gland development, and adult function in the mouse uterus. (141–141)Biol. Reprod. 86. 10.1095/biolreprod.112.098871. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Schiffhauer ES, Robinson DN, 2017. Mechanochemical signaling directs cell-shape change. Biophys. J. 112, 207–214. 10.1016/j.bpj.2016.12.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Soyal SM, Mukherjee A, Lee KY-S, Li J, Li H, DeMayo FJ, Lydon JP, 2005. Cre-mediated recombination in cell lineages that express the progesterone receptor. Genesis 41, 58–66. 10.1002/gene.20098. [DOI] [PubMed] [Google Scholar]
  45. Su R-W, Strug MR, Jeong J-W, Miele L, Fazleabas AT, 2016. Aberrant activation of canonical Notch1 signaling in the mouse uterus decreases progesterone receptor by hypermethylation and leads to infertility. Proc. Natl. Acad. Sci. USA 113, 2300–2305. 10.1073/pnas.1520441113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Uhlén M, Fagerberg L, Hallström BM, Lindskog C, Oksvold P, Mardinoglu A, Sivertsson Å, Kampf C, Sjöstedt E, Asplund A, Olsson I, Edlund K, Lundberg E, Navani S, Szigyarto CA, Odeberg J, Djureinovic D, Takanen JO, Hober S, Alm T, Edqvist P, Berling H, Tegel H, Mulder J, Rockberg J, Nilsson P, Schwenk JM, Hamsten M, Feilitzen K, Von Forsberg M., Persson L, Johansson F, Zwahlen M, Heijne G, Von Nielsen J., Pontén F, 2015. Tissue-based map of the human proteome. Science 347 (80), 1–9. 10.1126/science.1260419. [DOI] [PubMed] [Google Scholar]
  47. Uhlen M, Oksvold P, Fagerberg L, Lundberg E, Jonasson K, Forsberg M, Zwahlen M, Kampf C, Wester K, Hober S, Wernerus H, Björling L, Ponten F, 2010. Towards a knowledge-based human protein atlas. Nat. Biotechnol. 28, 1248–1250. 10.1038/nbt1210-1248. [DOI] [PubMed] [Google Scholar]
  48. van der Linden PJ, de Goeij AF, Dunselman GA, Erkens HW, Evers JL, 1995. Expression of cadherins and integrins in human endometrium throughout the menstrual cycle. Fertil. Steril. 63, 1210–1216, (〈https://doi.org/S0015-0282(16)57599-2[pii]〉). [DOI] [PubMed] [Google Scholar]
  49. Vialard F, El Sirkasi M, Tronchon V, Boudjenah R, Molina-Gomes D, Bergere M, Mauduit C, Wainer R, Selva J, Benahmed M, 2013. Tumor necrosis factor-308 polymorphism increases the embryo implantation rate in women undergoing in vitro fertilization. Hum. Reprod. 28, 2774–2783. 10.1093/humrep/det264. [DOI] [PubMed] [Google Scholar]
  50. Vue Z, Gonzalez G, Stewart CA, Mehra S, Behringer RR, 2018. Volumetric imaging of the developing prepubertal mouse uterine epithelium using light sheet microscopy. Mol. Reprod. Dev. 85, 397–405. 10.1002/mrd.22973. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Weaver VM, Petersen OW, Wang F, Larabell CA, Briand P, Damsky C, Bissell MJ, 1997. Reversion of the malignant phenotype of human breast cells in three-dimensional culture and in vivo by integrin blocking antibodies. J. Cell Biol. 137, 231–245. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Williams E, Villar-Prados A, Bowser J, Broaddus R, Gladden AB, 2017. Loss of polarity alters proliferation and differentiation in low-grade endometrial cancers by disrupting Notch signaling. PLoS One 12, 1–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Yonemura S, Wada Y, Watanabe T, Nagafuchi A, Shibata M, 2010. alpha-Catenin as a tension transducer that induces adherens junction development. Nat. Cell Biol. 12, 533–542. 10.1038/ncb2055. [DOI] [PubMed] [Google Scholar]
  54. Zhang N, Bai H, David KK, Dong J, Zheng Y, Cai J, Giovannini M, Liu P, Anders RA, Pan D, 2010. The Merlin/NF2 tumor suppressor functions through the YAP oncoprotein to regulate tissue homeostasis in mammals. (〈https://doi.org/S1534-5807(10)00305-9[pii]〉)Dev. Cell 19, 27–38. 10.1016/j.devcel.2010.06.015. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Figure 1
NIHMS2140489-supplement-Supplemental_Figure_1.pdf (3.1MB, pdf)
Supplemental Figure 3
NIHMS2140489-supplement-Supplemental_Figure_3.pdf (2.8MB, pdf)
Supplemental Figure 2
NIHMS2140489-supplement-Supplemental_Figure_2.pdf (4.4MB, pdf)
Supplemental Figure 4
NIHMS2140489-supplement-Supplemental_Figure_4.pdf (3.2MB, pdf)
Supplemental Figure 6
NIHMS2140489-supplement-Supplemental_Figure_6.pdf (1.2MB, pdf)
Supplemental Figure 5
NIHMS2140489-supplement-Supplemental_Figure_5.pdf (4.4MB, pdf)
Supplemental Movie 1
Download video file (9.6MB, mp4)
Supplemental Movie 2
Download video file (7.7MB, mp4)
Supplemental Movie 3
Download video file (10.4MB, mp4)

RESOURCES