WNT5A Interacts With FZD5 and LRP5 to Regulate Proliferation and Self-Renewal of Endometrial Mesenchymal Stem-Like Cells

Endometrial mesenchymal stem-like cells (eMSC) reside in the basal layer of the endometrium and are responsible for cyclic regeneration during the reproductive lives of women. Myometrial cells act as a component of the niche and regulate the stem cell fate through the activation of WNT/β-catenin signaling via WNT5A. Since WNT5A-responsive mechanisms on eMSC are still uncertain, we hypothesize that the WNT ligand–WNT5A works to activate WNT/β-catenin signaling through binding to Frizzled receptors (FZDs) and co-receptor low-density lipoprotein receptor-related protein 5 (LRP5). Among the various receptors that have been reported to interact with WNT5A, we found FZD5 abundantly expressed by eMSC when compared to unfractionated stromal cells. Neutralizing the protein expression by using anti-FZD5 antibody suppressed the stimulatory effects on phenotypic expression and the clonogenicity of eMSC in a myometrial cell–eMSC co-culture system as well as in an L-Wnt5a conditioned medium. Gene silencing of FZD5 not only reduced the binding of WNT5A to eMSC but also decreased the TCF/LEF transcriptional activities and expression of active β-catenin. Inhibition of LRP coreceptors with recombinant Dickkopf-1 protein significantly reduced the binding affinity of eMSC to WNT5A as well as the proliferation and self-renewal activity. During postpartum remodeling in mouse endometrium, active β-catenin (ABC) was detected in label-retaining stromal cells (LRSCs), and these ABC+ LRSCs express FZD5 and LRP5, suggesting the activation of WNT/β-catenin signaling. In conclusion, our findings demonstrate the interaction of WNT5A, FZD5, and LRP5 in regulating the proliferation and self-renewal of eMSC through WNT/β-catenin signaling.


INTRODUCTION
Human endometrium undergoes cyclical changes between proliferation, differentiation, and shedding under the control of sex steroids (Jabbour et al., 2006;Gellersen and Brosens, 2014). Endometrial stem/progenitor cells are responsible for the cyclic regenerative capacity of the human endometrium (Chan et al., 2004;Gargett et al., 2016). A specific population of endometrial stromal cells termed endometrial mesenchymal stem cells (eMSC) localized in the perivascular regions of functionalis and basalis can be identified by the coexpression of CD140b and CD146 (Schuring et al., 2011). These cells are clonogenic, express typical BM-MSC phenotypic markers (CD29, CD44, CD73, CD90, and CD105), and are able to differentiate into the mesoderm lineages upon specific induction media (Schwab and Gargett, 2007;Spitzer et al., 2012). The percentage of eMSC was similar across the menstrual cycle, but interestingly, eMSC from menstrual underwent more rounds of self-renewal and enabled a greater total cell output than those from the secretory phase (Xu et al., 2017).
The WNT signaling pathway is a complicated and highly conserved signaling cascade that can be categorized into a βcatenin-dependent (WNT/β-catenin signaling) and β-cateninindependent signaling pathway (Wend et al., 2010). WNT/βcatenin signaling is a well-elucidated pathway that is featured in various self-renewing tissues and is involved in somatic stem cell maintenance (Nusse and Clevers, 2017;Steinhart and Angers, 2018). WNT/β-catenin signaling is activated by the interaction of WNT ligand with a functional receptor from the Frizzled family (FZD) and a coreceptor low-density lipoprotein receptor-related protein (LRP) (Nusse, 2005;Gordon and Nusse, 2006;Dijksterhuis et al., 2015). The WNT-FZD-LRP trimeric complex results in the inhibition of β-catenin phosphorylation and leads to the stabilization of β-catenin, which is then translocated to the nucleus and binds to transcription factors T-cell factor (TCF) and lymphoid enhancer-binding factor (LEF) for stimulating the transcription of WNT target genes (Cadigan and Waterman, 2012). Meanwhile, β-catenin-independent signaling can be further subcategorized into the planar cell polarity pathway or the WNT/Ca 2+ pathway according to its downstream effectors (Komiya and Habas, 2008).
Somatic stem cells reside in the niche. The neighboring niche cells provide soluble, adhesive, and physical signals to the stem cells, which are crucial for maintaining stem cell functions (Chacón-Martínez et al., 2018). Our group has demonstrated that WNT5A derived from myometrial cells can regulate eMSC by activating the WNT/β-catenin signaling pathway (Cao et al., 2019). To gain a better understanding of the WNT5A-responsive manner, we propose that specific FZD receptor(s) and LRP coreceptor(s) expressed on eMSC will bind with WNT5A to activate the WNT/β-catenin signaling pathway. Here, we report the expression of WNT5A-related FZD receptors and WNT coreceptors in different subpopulations of endometrial stromal cells and demonstrate the role of FZD5 and LRP5 in the WNT5Ainduced activation of WNT/β-catenin signaling in eMSC.

Human Tissues
Full-thickness human endometrial tissue and underlying myometrium were collected from 41 women aged between 38 and 53 years (mean age 46 years) who had regular menstrual cycles and underwent total abdominal hysterectomy for benign pathologies (Supplementary Table S1). The recruited subject took no exogenous hormonal therapy for at least 3 months prior to surgery. Written consents were obtained from all women before their participation of the research. Ethical approval was obtained from the Institutional Review Board of The University of Hong Kong/Hospital Authority Hong Kong West Cluster UW14-133 and UW19-851.
An experienced histopathologist defined the phase of the menstrual cycle as proliferative (n = 23) and secretory (n = 18) based on microscopic morphology in hematoxylin and eosinstained tissue sections.

Isolation of Endometrial Cells
Endometrial tissue was minced and digested with phosphatebuffered saline (PBS) containing 0.3 mg/ml collagenase III (Worthington Biochemical Corporation, Freehold, NJ, United States) and 40 μg/ml deoxyribonuclease type I (Worthington Biochemical Corporation) in a shaking water bath for 60 min at 37°C (Xiang et al., 2014). The endometrial tissue underwent two rounds of digestion, and the dispersed cells were filtered through 40-μm sieves (BD Bioscience, San Jose, CA, United States). The red blood cells, cell debris, and cell clumps were removed by Ficoll-Paque (GE Healthcare, Uppsala, Sweden) density-gradient centrifugation. Anti-CD45 antibody-coated Dynabeads (Invitrogen, Waltham, MA, United States) were used for the removal of leukocytes, and anti-CD326 antibodycoated microbeads (Miltenyi Biotec Inc., San Diego, CA, United States) were used to separate epithelial cells from stromal cell population. Freshly purified stromal cells were seeded onto 10-cm dishes (BD Biosciences, San Jose, CA, United States) coated with fibronectin (1 mg/ml, Invitrogen) containing growth medium (GM) with 1% penicillin (Invitrogen), 1% L-glutamine (Invitrogen), and 10% FBS (Invitrogen) in DMEM/F-12 (Sigma-Aldrich, St Louis, MA, United States). Stromal cells were cultured for 1-2 weeks in a humidified carbon dioxide incubator at 37°C. The medium was changed every 7 days.

Magnetic Selection of Endometrial Mesenchymal Stem-Like Cells
To obtain eMSC (CD140b + CD146 + cells), two separate positive magnetic bead selections were conducted (Xu et al., 2017). After in vitro expansion, stromal cells were trypsinized and incubated with PE-conjugated anti-CD140b antibody (R&D Systems, Minneapolis, MN, United States) for 45 min at 4°C and then with antimouse IgG1 coated magnetic microbeads (Miltenyi Biotec Inc.) for 15 min at 4°C. The CD140b + population was obtained by applying the cell suspensions to MS columns (Miltenyi Biotec Inc.) in a magnetic field. The CD140b + stromal cells were cultured in fibronectin-coated plates containing GM at 37°C in 5% CO 2 for 7-10 days to allow microbead degradation during cell expansion. When the plates were 80% confluent, the cells were trypsinized and incubated with anti-CD146 antibody-coated microbeads (Miltenyi Biotec Inc.) for 15 min at 4°C. CD140b + CD146 + eMSC were obtained from the column separation. Phenotypic study of eMSC shows their positive expression for CD140b and CD146 (Cao et al., 2021). Stromal cells at passages 1-3 were used in this study.

Isolation of Myometrial Cells
Myometrial tissue was minced into 1-mm 3 pieces and dissociated in PBS containing collagenase type III (300 μg/ml) and deoxyribonuclease type I (40 μg/ml) in a shaking water bath for 4 h at 37°C as described (Cao et al., 2019). Single cells were filtered through 100-µm sieves (BD Bioscience) and seeded into 100-mm dishes containing GM. Upon~80% confluence, the myometrial cells were trypsinized and passaged. Myometrial cells at passages 1-3 were used in this study.
Coculture eMSC were seeded at clonal density (300 cells per well) into a 6well plate, and myometrial cells pretreated with mitomycin C (2.7 × 10 4 ) were seeded into transwell insert (EMD Millipore, Billerica, MA, United States) and cultured separately. On the following day, the inserts were placed into the 6-well plates with eMSC for indirect coculture. All treatment groups were set up in duplicates, and eMSC cultured alone (Monoculture) served as the control. The culture medium was changed every 7 days.

In Vitro Colony-Forming Assay
After 15 days of culture, eMSC were fixed in 10% formalin, washed, and stained with Toluidine Blue (1 mg/ml, Sigma-Aldrich). Clonogenic efficiency was calculated as the number of colonies formed divided by the number of cells seeded multiplied by 100 (Cao et al., 2019).

Preparation of Myometrial Conditioned Medium
For myometrial conditioned medium, 1 × 10 6 cells were cultured in T-75 flasks for 3 days in GM, washed with PBS, and then cultured in 5 ml of serum-free medium (DMEM-F12 base medium). The myometrial conditioned medium (CM) was collected after 24 h and was filtered through a 0.45-μm filter.
The CM is concentrated using an Amicon ® Ultra-15 centrifugal filter (EMD Millipore) with a molecular weight cutoff of 10 kDa and centrifugation at 4,000 g for 45 min at 4°C. The myometrial concentrated conditioned medium (Myometrial CCM) was stored at −80°C. The amount of proteins from 5 ml of myometrial CM was considered as one unit of myometrial CCM proteins. Myometrial CCM at 1/6 unit was used for various bioassays (Cao et al., 2019). CCM collected from cells free DMEM-F12 medium was used as control.

Activation of WNT Signaling
L-Wnt5a CM was obtained from spent cultured medium of mouse fibroblasts L-M (TK-) cells transfected with a Wnt5a vector (CRL-2814, ATCC, Atlanta, United States). Mouse L-Wnt5A cells (~2 × 10 6 cells) were cultured in a T-75 flask containing GM until they reached 80% confluence. The following day, serum-free medium was replaced, cultured for 24 h and collected after sterile filtration as L-Wnt5A CM. Control CM was obtained from normal mouse fibroblast L-M (TK-) cells. All CMs were stored at −80°C. The ratio of 1:10 diluted WNT5A CM was used. Recombinant human WNT5A (0.01 μg/ml, R&D Systems) was supplemented to the GM of eMSC.
Gene Silencing and Luciferase Assay eMSC were seeded in 48-well plates at a density of 1 × 10 4 cells/ well in OptiMEM (Invitrogen). The next day, the cells were transfected with 10 pmol of siRNA directed against FZD5 (s15416 and s15417, Ambion) or siRNA control with scrambled sequence (Ambion) using 10 pmol of a Lipofectamine RNAiMax transfection reagent (Invitrogen). On the following day, the cells were cotransfected with 4 μg of a TOPflash/FOP-flash vector and 1 μg of pRL-TK (Renilla-TKluciferase vector, Promega, Madison, WI, United States) using Lipofectamine 3,000 (Invitrogen). Eighteen hours after transfection, the medium was changed to growth medium, and some wells were treated with L-cell or L-Wnt5A CM or myometrial CCM or human recombinant WNT5A (0.01 μg/ ml, R&D Systems) or human recombinant dickkopf-1 (0.2 μg/ ml, DKK-1, R&D systems) for 48 h. The luciferase activity was measured using a GLOMAX ™ 96 microplate luminometer. The TOP/FOP ratio was calculated. The knockdown efficiency was assessed by flow cytometry (Supplementary Figure S1A).

Protein Labeling Assay
Human recombinant WNT5A protein (20 μg, 645-WN/CF, R&D systems) was labeled with a fluorescence dye using the Alexa Fluor 488 microscale protein labeling kit according to the manufacturer's instruction (Thermo Fisher Scientific). eMSC were transfected with 10 pmol of siRNA directed against FZD5 or scrambled siRNA as described above. The next day, cells were harvested and stained with Alexa Fluor 488 labeled human recombinant WNT5A protein for 2 h at 4°C in the dark. Another set of cells were treated with recombinant DKK-1 (0.2 μg/ml, R&D systems) for 2 days. Cells were washed and analyzed by a CytoFlex ™ flow cytometer (Beckman Coulter, Ca, United States). Data were analyzed with the FlowJo software (Tree Star Inc., OR, United States).

Label-Retaining Cells in Mouse Endometrium
Mouse endometrial paraffin sections were obtained from our previous label-retaining cell (LRC) experimental setup (Cao et al., 2015). In brief, day 19 prepubertal CD57BL/6J female mice were administered with bromodeoxyuridine (BrdU) intraperitoneally (50 μg/g of body weight; Sigma Aldrich) twice daily for four consecutive days and allowed to grow without further labeling. After sexual maturity age, some of the mice were sacrificed at 6 weeks chased (n = 4). Other animals were mated with > 6-week-old fertile C57BL/6J male mice, and the presence of the vaginal plug indicates that the female mice are pregnant. These mice were sacrificed~8 weeks chase (postpartum day 1). All animals were euthanized, uterine horns collected, fixed overnight with 4% paraformaldehyde, and processed into paraffin blocks by standard method.

Statistical Analysis
Data were analyzed using the GraphPad PRISM software (version 8.0; GraphPad Software Inc., San Diego, CA, United States) and tested for normal distribution using D'Agostino and Pearson test. Mann-Whitney test was performed to determine the statistical significance between two groups. Kruskal-Wallis test followed by Dunn's post hoc test was used for multiple group comparisons. Data are presented as mean ± standard deviation (SD). p < 0.05 was considered statistically significant.
To investigate the role of LRP5 in regulating the WNT/βcatenin signaling, the TCF/LEF luciferase activities in eMSC were evaluated after treatment with DKK-1 protein. Treatment with DKK-1 and WNT5A proteins (1.38 ± 0.16 fold) or WNT5A alone (1.54 ± 0.63 fold) significantly increased the relative luciferase activity when compared to the DKK-1 protein group (0.82 ± 0.26 fold, p < 0.01, n = 7, Figure 5D). Dual immunofluorescent staining confirmed that the majority of the eMSC coexpress FZD5 and LRP5 ( Figure 5E).

Label-Retaining Stromal Cells in Mouse Endometrium Express FZD5 and LRP5
Using the BrdU label, the LRC approach has identified stem/ progenitor cells in the mouse endometrium (Chan and Gargett, 2006). Label-retaining stromal cells (LRSCs) were present after 6 weeks chase, and the percentage of these cells remained steady until parturition (Cao et al., 2015). In the mouse endometrial tissue, LRSCs (BrdU + cells) were found to coexpress with FZD5 and LRP5 at 6 weeks chase ( Figure 6C). Activated LRSCs in the postpartum endometrium expressed Wnt/β-catenin signal pathway components during uterine remodeling (Cao et al., 2015). The active β-catenin expression on LRSCs was only

DISCUSSION
Stem cells remain quiescent and undifferentiated when situated within the niche. Most of the work in the field of endometrium stem/progenitor cells focuses on identifying, characterizing, and locating the origin of the cells (Gargett et al., 2016;Santamaria et al., 2018). The molecular mechanisms controlling stem cell renewal and differentiation remain unknown. To gain a better understanding of the interplay between endometrial stem/ progenitor cells and their niche, we have identified myometrial cells as a niche component of eMSC. Coculture of myometrial cells or myometrial CCM with eMSC enhanced proliferation and renewal activities when compared to monoculture or cultured with growth medium alone. In addition, the participation of WNT/β-catenin signaling as a molecular component in the eMSC niche was observed in mice and humans (Cao et al., 2015(Cao et al., , 2019Xu et al., 2020). The impact of WNT/β-catenin signaling in stem cell regulation is well documented (Grigoryan et al., 2008;Haegebarth and Clevers, 2009;Bukowska et al., 2015). In the human endometrium, we have demonstrated that myometrialsecreted WNT5A can activate the WNT/β-catenin pathway in eMSC (Cao et al., 2019). WNT5A has been proven to be a classic noncanonical WNT ligand (Yamanaka et al., 2002;Dejmek et al., 2006). However, WNT5A can activate the canonical WNT pathway depending on the receptor context (Mikels and Nusse, 2006). For example, frizzled receptors such as FZD4, FZD5, and FZD7 can interact with WNT5A and activate canonical and noncanonical pathways (Umbhauer et al., 2000;Mikels and Nusse, 2006;Hii et al., 2015). Overall, endometrial stromal cells express Frizzled genes, and the expression was similar among the stromal subpopulations. Interestingly, eMSC contain a higher expression of FZD5 than unfractionated stromal cells. Protein evaluations further confirmed that this receptor was abundantly expressed by eMSC. Knockdown of FZD5 expression and blocking LRP5 significantly reduced the stimulatory effect of WNT5A on eMSC, providing evidence that these receptors can respond to WNT ligands in the endometrial microenvironment. Although LRP5 protein is higher in unfractionated stromal cells than eMSC, FZD5 and LRP5 have coexpression in a significant proportion of the eMSC population, suggesting the possibility of FZD5-LRP5 interaction in eMSC maintenance.  FZD5 is a common noncanonical WNT receptor (Ishikawa et al., 2001;Arderiu et al., 2014;Kim et al., 2019), and the blocking of WNT5A-FZD5 signaling pathway can inhibit the PKC pathway (Weeraratna et al., 2002). A recent study has demonstrated FZD5 as a regulator in the maintenance of human mesenchymal stem/ stromal cells (hMSCs) (Harada et al., 2021). The β-cateninindependent WNT5A-FZD5 signaling preserved the properties of hMSCs by preventing cell senescence. Nevertheless, FZD5 can also activate the WNT/β-catenin signaling pathway in hMSCs (Kolben et al., 2012). It has been reported that WNT5A can act via FZD5 to initiate an intracellular pathway leading to the accumulation of β-catenin (He et al., 1997). The FZD-LRP complex can regulate the activation of LRP5/β-catenin signaling (Hua et al., 2018;Chen et al., 2019). This FZD-LRP interaction also blocks the activation of noncanonical WNT pathways by inhibiting FZD-mediated signaling (Ren et al., 2015). Therefore, we postulate that the FZD5 and LRP5 expressing eMSC can activate the WNT/ β-catenin pathway by interacting with WNT ligands produced by niche cells, in particular the response to WNT5A.
Myometrial coculture enhanced the clonogenic activity and phenotypic expression of stem cell markers (CD140b and CD146) in eMSC. Similar stimulatory effect was observed when L-Wnt5A CM was used. The addition of anti-FZD5 antibody reduced the stimulatory action of myometrial coculture on colony formation and self-renewal of eMSC. When L-Wnt5a CM was blocked with anti-FZD5, a decreasing trend in the percentage of eMSC was also observed, but the result did not reach significance due to high variation between individuals. The use of antibodies may not efficiently neutralize binding of the ligand to the receptor. Thus, knockdown of the receptor with gene silencing will be more specific. Indeed, the stimulatory effect of myometrial CCM or recombinant WNT5A significantly reduced the TCF/LEF transcriptional activities of eMSC upon the knockdown with FZD5-siRNA. From the other aspect, blocking the LRP5 coreceptor also significantly reduced the clonogenicity and proportion of eMSC due to the inhibition of the canonical WNT signaling pathway. Although DDK1 can inhibit the effect of LRP5 and LRP6, we only evaluated LRP5 in this study because it was reported that WNT5A may activate the Wnt/β-catenin pathway in the presence of the LRP5 coreceptor and FZD5 (He et al., 1997). The role of LRP6 in eMSC regulation remains to be explored.
Coexpression of FZD5 and LRP5 was also observed in mouse endometrial stem/progenitor cells. According to our previous findings, BrdU + stromal cells after 6 weeks chase were termed LRSCs (Cao et al., 2015). The LRC approach can trace quiescent cells in the mouse endometrium, and hence we can study the changes of these functional LRSCs during pregnancy and postpartum. In addition, these FZD5 + LRSCs expressed active β-catenin at postpartum days 1 and 3, a period of dynamic physiological regeneration, indicating that WNT/β-catenin signaling is involved in stem cell activation during endometrial remodeling. WNT5A is an important Wnt ligand in female reproduction and is expressed in both the epithelial and stromal cells of uterine tissue (Hou et al., 2004;Latifi et al., 2018). These activities seen in WNT5A-activated cells fit well with our observation that endometrial LRSCs express FZD5, a prospective receptor for WNT5A. Whether WNT5A directly interacts with FZD5 expressing LRSC for uterine regeneration is yet to be determined. We postulate that FZD5 may also be a key molecule in regulating the stem cell fate in the mouse endometrium.
Overall, our study demonstrated that the interaction of WNT5A, FZD5, and LRP5 can regulate the proliferation and self-renewal of eMSC through the activation of the WNT/βcatenin signaling pathway. Further studies on the regulatory molecules can be done to develop an efficient protocol for the culture expansion of eMSC for cell-based therapy.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding authors.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by this study protocol, and informed consent forms were approved by the Institutional Review Board of the University of Hong Kong/Hospital Authority Hong Kong West Cluster (UW 14-133 and UW 19-851). The patients/ participants provided their written informed consent to participate in this study. The animal study was reviewed and approved by the Committee on Use of Live Animals in Teaching and Research, The University of Hong Kong, Hong Kong.

AUTHOR CONTRIBUTIONS
TL participated in most of the experimental work, analysis of data, and writing of the manuscript. RC contributed to the study concept, writing, and editing of the manuscript. C-LL and PC contributed to the conception and design of the study. RL and EN contributed to the recruitment of patients and sample collection for this study. WY contributed to the study design, critical discussion, and proofreading of the manuscript.

ACKNOWLEDGMENTS
We are grateful to all the women who agreed to donate their tissue samples for this study. We sincerely acknowledge Joyce Yuen, our Frontiers in Cell and Developmental Biology | www.frontiersin.org February 2022 | Volume 10 | Article 837827