Management of intrauterine adhesions using human amniotic mesenchymal stromal cells to promote endometrial regeneration and repair through Notch signalling

Abstract Intrauterine adhesions (IUAs) severely hamper women's reproductive functions. Human amniotic mesenchymal stromal cell (hAMSC) transplantation is effective in treating IUAs. Here, we examined the function of Notch signalling in IUA treatment with hAMSC transplantation. Forty‐five Sprague‐Dawley female rats were randomly divided into the sham operation, IUA, IUA + E2, IUA + hAMSCs and IUA + hAMSCs + E2 groups. After IUA induction in the rats, hAMSCs promoted endometrial regeneration and repair via differentiation into endometrial epithelial cells. In all groups, the expression of key proteins in Notch signalling was detected in the uterus by immunohistochemistry. The results indicated Notch signalling activation in the hAMSCs and hAMSCs + E2 groups. We could also induce hAMSC differentiation to generate endometrial epithelial cells in vitro. Furthermore, the inhibition of Notch signalling using the AdR‐dnNotch1 vector suppressed hAMSC differentiation (assessed by epithelial and mesenchymal marker levels), whereas its activation using the AdR‐Jagged1 vector increased differentiation. The above findings indicate Notch signalling mediates the differentiation of hAMSCs into endometrial epithelial cells, thus promoting endometrial regeneration and repair; Notch signalling could have an important function in IUA treatment.


| INTRODUC TI ON
Intrauterine adhesions (IUAs), resulting from damaged basal layer of the endometrium, partially or completely obliterate the uterus and/ or the cervix, 1,2 and cause infertility, oligomenorrhoea and recurrent abortions. [3][4][5][6] The aim of IUA treatment is to re-establish a healthy anatomy (by adhesion removal) and reinstate uterine function. 7,8 However, prognosis following surgical restoration of the uterine cavity is poor, with up to 62.5% cases recurring. 9 This can be attributed to a suboptimal regeneration of functional endometrum. 10 The transplantation of mesenchymal stem cells (MSCs) into the uterine cavity could efficiently induce endometrial regeneration and remove IUAs. 11,12 Human amniotic mesenchymal stromal cells (hAMSCs) represent an ethically nonproblematic, easily isolable, abundant and immune-privileged cell source. 13 In addition, similar to other mesenchymal stromal cells, hAMSCs can differentiate into cells of all three germ layers in animal and cell culture experiments, exhibiting immunomodulatory properties through paracrine effects. 14 Interestingly, hAMSCs can also be used to treat other female pathologies, including primary ovarian insufficiency and polycystic ovarian syndrome. 15 Consequently, hAMSCs are considered useful candidates for transplantation treatment approaches and might represent a superior option compared with classical stem cells for xenografts and allografts. 16,17 Previously, we demonstrated hAMSCs can differentiate into endometrial epithelial cells, with hAMSC transplantation potentially clearing uterine adhesions. 18 Notch receptor genes including NOTCH1, NOTCH2 and NOTCH3, the NOTCH ligand, JAG1 and downstream effectors HEY1, HEY2 and HEYL are reportedly upregulated in endometrial MSCs (eMSCs).
This strongly indicates that the Notch system is activated in eMSCs. 19,20 Notch1 and Jagged1 are also regulated in the endometria with UATs. A study has examined how Notch signalling regulates hAMSCs in the treatment of IUAs. 21 There are four Notch receptors (Notch1 to Notch4) and five Notch ligands (Jagged1 and Jagged2 and DLL1, DLL3 and DLL4) in vertebrates that encompass intracellular, extracellular and transmembrane domains, all of which control multiple physiological events, including cell division and differentiation. 22,23 After binding of the Notch ligand to its receptor in neighbouring cells, the receptor undergoes cleavage at two sites, thereby releasing the Notch intracellular domain (NICD). The NICD exerts its biological effects by entering the nucleus and binding to the DNA-binding protein CSL (RBPjk/CBF1) and the transcriptional coactivator MAM to regulate the target genes Hes1, Hes5 and Hey1, as shown in the figure below. 24,25 In the present study, we aimed to explore the function of Notch signalling and its molecular mechanisms in hAMSC differentiation into endometrial epithelial cells and to evaluate its significance in hAMSC transplantation for IUA treatment by overexpressing and competitively inhibiting Jagged1 (ligand) and Notch1 (receptor), respectively.

| Animals
Forty-five Sprague-Dawley (SD) rats (8 weeks old, 200-220 g) were provided by the Animal Experimental Centre of Chongqing Medical University. Five rats were housed in one cage; they were maintained in a controlled environment under a 12-h light and dark cycle at 22°C, with food and water provided ad libitum. The study protocol was approved by the Ethics Committee of Chongqing Medical University (20141230).

| Isolation, culture, and identification of hAMSCs
Amniotic membrane samples were provided by the University-town

Hospital of Chongqing Medical University and the First Affiliated
Hospital of Chongqing Medical University. Previously used methods for the isolation and culture of hAMSCs by our research group were applied in this study. 26 We isolated hAMSCs from amniotic membrane specimens obtained from healthy pregnant women during full-term caesarean section, after obtaining informed consent.
Fragments (3 mm 2 ) of the amniotic membranes were digested with 0.25% trypsin (BeyotimeBio, Shanghai, China) at 37°C for 30 min, followed by incubation with 0.1% type I collagenase (Meilun, Shanghai, China) at 37°C for 1 h. The obtained hAMSCs were cultured in DMEM/F12 medium (Gibco) with 10% foetal bovine serum (Gibco). Half of the medium was changed after 24 h, and the entire medium was changed after 72 h. hAMSCs were passaged when they reached >80% confluency and were selected for identification in the third passage. The expression of the markers CD29, CD44, CD73, CD105, CD19, CD34 and CD45 on the surface of hAMSCs was detected by immunofluorescence, and the differentiation characteristics of hAMSCs were examined by staining with Oil Red O and Alizarin Red (Solarbio, Shanghai, China).

| PKH26 labelling of hAMSCs
Third-passage hAMSCs were prepared as a single-cell suspension, which was washed with phosphate-buffered saline (PBS) (Solarbio, Shanghai, China) and centrifuged to obtain a cell pellet.
For PKH26 labelling (Sigma), the cells were resuspended in 1 ml of Diluent C as directed by the manufacturer. Next, the labelling solution was mixed with the single-cell suspension and incubated for 5 min. 10% foetal bovine serum (Gibco) was then added to terminate the reaction. The sample was washed thrice with 10 ml of complete medium. The labelled hAMSCs were then cultured until they became adherent. Thereafter, the nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI) (Boster Biological, Shanghai, China). An observation under a fluorescent microscope (Nikon, Japan) revealed the cells were well-labelled and adhered to the walls.
The obtained hAMSCs were finally stored in an incubator until transplantation into rat uteruses.

| Preparation of endometriumconditioned medium
The bilateral uteruses of 12-week-old female SD rats were removed under aseptic conditions after general anaesthesia with 10 ml/kg of 5% chloral hydrate (Biosharp, Shanghai, China). The uteruses were then immediately placed on ice and dissected longitudinally, cut transversely at approximately 0.3 cm in length and weighed. Next, 100 g of the prepared uterine segments was added to 1 L of icecold DMEM/F12 medium; the samples were sealed and mixed on a shaker at 4°C for 2 h. After overnight incubation at 4°C, the samples were centrifuged at 18,000 g for 30 min at 4°C. Finally, the solution was filtered with a sterile filter (0.22 μm filter; Millipore) and the filtrate was stored at −20°C. 27

| Generation of model IUA rats
To generate IUA models, the endometria of the rats were injured with 95% alcohol. Dioestrus rats were selected based on the vaginal smear analysis and were fully anaesthetized with 10 ml/kg of 5% chloral hydrate. The skin of the lower abdomen was then exposed and disinfected with iodophor, and a longitudinal incision of 2 cm was made at the site (3 cm above the vaginal orifice). Next, the abdomen was cut open layer-by-layer until the uterus was exposed. The entire exposed uterus was ligated by suturing the avascular area of the mesometrium using a small round needle (with a silk thread). Next, 95% ethyl alcohol was administered by injection into the uterus with a 1-ml empty syringe; an investigator performed clamping of the distal end of the right uterus using a pair of flat forceps. The injection was stopped after complete filling of the uterus. Uterus colour was subsequently assessed. When the uterine cavity turned white after approximately 3 min ( Figure 2C), ethanol was removed followed by saline washing. After these procedures, the uterus was unclamped, unstitched and placed back into the abdomen; thereafter, the muscle and skin were sutured. The abdomen was incised again after 2 weeks to observe the morphological changes on the modelled side ( Figure 2D).

| Sampling
All rats were euthanized after 4 weeks to harvest the bilateral uteruses. The uteruses of three rats randomly chosen from the IUA and Notch pathway effectors (Notch1-4, Jagged1 and Jagged2, Hes1

TA B L E 1 Primers utilized in quantitative real-time PCR
and NICD) was detected by immunohistochemistry.

| Immunofluorescence
The rat uteruses were harvested and used to prepare frozen sections 2 weeks after hAMSC transplantation. The nuclei were stained with DAPI and a fluorescent microscope was utilized to observe PKH26labelled hAMSCs in the uterus specimens. Frozen sections of the uteruses from the IUA models not injected with PKH26-labelled hAMSCs were used as negative controls.

| Histochemical staining and image analysis
The harvested uteruses were fixed with 4% formalin (

| Quantitative reverse transcription-PCR (qRT-PCR)
The total RNA was obtained from the induction and control group cells grown in six-well plates using the TRIzol method (Sangon

| Statistical analysis
Data were analysed using SPSS 21.0. Normally distributed measurement data are presented as mean (x) ± standard deviation (sd). The one-way ANOVA and t test were carried out to compare multiple groups and group pairs, respectively. Histograms were drawn with Image-Pro Plus 6.0. p < 0.05 indicated statistical significance.

| Identification of hAMSCs
The morphology of most adherent hAMSCs was similar to that of fibroblasts ( Figure 1B). Immunofluorescence revealed that hAMSCs expressed the stemness biomarkers CD29, CD44, CD73 and CD105, but did not express CD19, CD34 and CD45 ( Figure 1A). To confirm that hAMSCs have differentiation ability, we induced hAMSC differentiation to osteoblasts and adipocytes. hAMSCs in the osteogenic differentiation medium were stained with Alizarin Red after culturing for 21 days, and orange-red nodules were observed ( Figure 1C). hAMSCs in the adipocyte differentiation medium were stained with Oil Red O after being cultured for 14 days, and red fat drops were observed ( Figure 1D).

| Promotion of endometrial epithelial cell regeneration and repair by in vivo transplantation of hAMSCs
PKH26-labelled hAMSCs grew healthily with intact cell membranes and nuclei and were adherent. No suspended cells (Figure 2A) were observed. The uterus of normal SD rats is shaped like 'Y' (Figure 2B). The right uteruses of the rats turned white after injury with 95% alcohol for 3 min ( Figure 2C). The site became hard. Following the generation of the IUA model, the uterine cavity became harder and thinner after 2 weeks ( Figure 2D).
H&E staining demonstrated that the uterine cavity in sham- Immunohistochemistry revealed that E-cadherin expression was not significant in the IUA and IUA + E2 groups, but was markedly elevated in endometrial epithelial cells in the IUA + hAMSCs and IUA + hAMSCs + E2 groups (p < 0.05) ( Figure 2M,N), suggesting differentiation into endometrial epithelial cells.

| Notch pathway's function in the treatment of IUAs by hAMSC transplantation
To explore the function of Notch signalling in hAMSC transplantation, we detected the expression levels of NICD and its downstream target ( Figure 3E,F). Notch1 and Jagged1 expression was reduced in the IUA and IUA + E2 groups (p < 0.05) compared with that in the sham animals, but was elevated in the IUA + hAMSCs and IUA + hAMSCs + E2 groups compared with that in the IUA group (p < 0.05). In contrast, Notch3 expression in the uteruses was markedly elevated in the IUA and IUA + E2 groups compared with that in sham rats (p < 0.05) and decreased in the IUA + hAMSCs and IUA + hAMSCs + E2 groups compared with that in the IUA group (p < 0.05) ( Figure 3G).

| Differentiation of hAMSCs into endometrial epithelial cells in vitro
To simulate the uterine environment, endometrium-conditioned medium was used to promote hAMSC differentiation into target cells. Notch1, Notch3 and Jagged1 levels in hAMSCs were examined by immunofluorescence, and the results showed positive expression ( Figure 4F).

| DISCUSS ION
Common treatment options for IUA encompass hysteroscopic adhesiolysis, hormone therapy and physical barriers. Despite great progress in surgery, recurrent IUAs after treatment remain a challenge. 28 In our previous research, we found that the recurrence of treated IUAs could be effectively prevented by hAMSC transplantation. 18,29 Here, we studied the functions and In this study, we examined the Notch pathway's role in the treatment of IUAs by hAMSC transplantation, but whether hAMSCs differentiate into endometrial stem cells and whether their role in rats would change after interfering with Notch signalling remain to be explored.
In conclusion, the present study confirmed that the activation of Notch signalling promotes hAMSC differentiation into endometrial epithelial cells, which in turn promotes regenerative repair of the endometrium in uterine adhesions. In the future, the therapeutic benefits of Notch pathway intervention before hAMSC transplantation for endometrial injuries should be explored. in each group showed that the expression of E-cadherin increased in the induction group compared with that in the control group, but decreased in the inhibition group compared with that in the induction group and also elevated in the activation group compared with that in the inhibition group, with a significant difference (p < 0.05). (F) Analysis of the expression of vimentin in each group showed that the expression of vimentin was reduced in the induction group compared with that in the control group, with a significant difference (p < 0.05), and that it increased in the inhibition group compared with that in the induction group, but decreased in the activation group compared with that in the inhibition group, with a significant difference (p < 0.05)

CO N FLI C T S O F I NTE R E S T
The authors have no potential conflict of interest to declare.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data generated and/or analysed in the current study are included in this manuscript.