The study “Therapeutic potential of human endometrial stem cells” was approved by the Ethics Committee of Biomedical Research of Vilnius Region, No. 158200-18/7-1049-550 (29/06/2018). All endometrial tissue samples were collected during routine procedure at Vilnius University Hospital Santaros Klinikos Obstetrics and Gynaecology Center Santaros Fertility Center. The female subjects of the present study signed consent forms to confirm their agreement to providing their endometrium specimens.

Endometrial tissue samples were taken from a total 20 females undergoing ART procedures due to couple infertility. The inclusion criteria were as follows: (a) age of female at the time of enrollment is 18–45 years; (b) the minimum duration of infertility—1 year; (c) female is undergoing ART procedures due to couple infertility; (d) female confirms the participation in the study and signed informed consent. The exclusion criteria were as follows: (a) confirmed oncological disease for female during the last 3 years; (b) female smokes or is addicted to alcohol or other substances; (c) pregnancy is contraindicated for female; (d) female is diagnosed with uncontrolled endocrine or other medical conditions, such as hyperprolactinemia or thyroid diseases; (e) female was diagnosed with mental disorders or illness in the past medical history.

A reproductive medicine professional collected the specimens using 3 mm wide pipelle, endometrial sampling catheter, while performing a scratching procedure on the inner wall of uterine lining. The procedure was carried out once on Day 20–22 of the menstrual cycle prior to in vitro fertilization (IVF)/intracytoplasmic sperm injection (ICSI) procedure.

The isolation of hEnMSCs was performed according to Tavakol et al. (2018). Shortly, endometrial tissue samples were washed with Hank’s media, cut into small pieces and incubated with collagenase I (1 mg/ml) for approximately 60 min. The growth medium was then added and the sample was passed through a 70 µm Falcon cell strainer. Passed cells were centrifuged and purified from red blood cells (RBCs) by using Ficoll gradient. Isolated cells were cultured in a growth medium DMEM/F12 (DMEM, Dulbecco’s Modified Eagle Medium/Nutrient Mixture F-12) supplemented with 10% Fetal Bovine Serum (FBS), 100 U/mL penicillin and 100 μg/mL streptomycin (Gibco, Thermo Fisher Scientific, Waltham, MA, United States) at 37°C in a 5% CO₂ atmosphere with humidity. Cells were plated into T75 cm² culture flasks at a density of 0.5–0.6 × 10⁶ cells. The medium was replaced every 2–3 days and cells were replated according to their confluence. Cell viability was evaluated during passaging using Neubauer’s chamber and Trypan Blue test. Preparation of hEnMSCs for the injection into female mice was done by washing cells twice with sterile phosphate-buffered saline (PBS) solution.

To determine hEnMSCs surface markers, cells were collected (about 0.1 × 10⁶ cells for one marker), centrifuged at 500 g speed for 5 min and washed twice with 1xPBS and 1% Bovine serum albumin (BSA) solution. Washed cells were then incubated for 30 min in the dark on ice with appropriate antibodies against cell surface markers (CD44, CD146, CD166, CD140b, CD34, CD45, HLA-DR). After that, the labelled cells were washed twice with PBS +1% BSA solution by centrifuging at 550 g for 5 min, 4°C. Since flow cytometry analysis was not done immediately, cells were fixated with 2% paraformaldehyde solution and later analyzed with Partec flow cytometer (Sysmex Corporation, Cobe, Hioko, Japan). Results were processed with Flowing Software 2 software.

Animal testing “Possibilities of stem cells application for restoration of endometrium” was approved by the State Food and Veterinary Service of the Republic of Lithuania. Veterinary approval for that study was No. G2-115 (20/05/2019).

Procedures with animal were performed by trained specialists using appropriate equipment.

A total of 58 NOD. CB-17-Prkdc scid/Rj female mice, 6 weeks of age, 18 g ± 0.5 g were purchased from JANVIER LABS (France) and allowed to adapt to the new environment for at least a week. The mice were maintained in specific-pathogen free conditions.

The anesthetized mice were euthanized by cervical dislocation procedure. Mice were removed from their cages, anesthetized by 1%–3% isoflurane inhalation (100 mg/g, Isoflurin, Vetpharma, Spain) and gently restrained while resting on the operating tables with aseptic pads. Cervical dislocation was performed mechanically and resulted in euthanasia within approximately 10 s.

The Control mice group (n = 11) did not receive any interventions. The group of female mice with mechanically damaged endometrium and the chemotherapy-induced female mice were designed by the following procedures.

Quick-DNA/RNA Miniprep Kit (Zymo Research, CA, United States) was used to facilitate the total RNA extraction from mice uterine horns. Frozen tissue was first submerged in liquid nitrogen and ground to a powder using a mortar and pestle. Then powder was mixed with about 600 µl of DNA/RNA Lysis Buffer, pipetted and transferred to a Zymo-Spin™ Column and centrifuged at 13,000 g for 1 min at room temperature. The following isolation steps were done in accordance with manufacturer’s recommendations to ensure high yield and purity of the RNA samples. cDNA was synthesized using Luna^® Universal One-Step the reverse transcription—quantitative polymerase chain reaction (RT-qPCR) Kit (New England Biolabs, Ipswich, MA, United States) in accordance with the manufacturer’s recommendations. cDNA was then amplified by RT-qPCR using Luna^® Universal qPCR Master Mix (New England Biolabs, Ipswich, MA, United States) in accordance with the manufacturer’s guidelines. RT-qPCR was performed using Rotor-Gene 6,000 Real-time Analyzer (Corbett Life Science, QIAGEN, Hilden, Germany) and cycling conditions were set as follows: initial denaturation 95°C 1 min (1 cycle), denaturation 95°C 15 s, primer Tm 60°C 30 s (40 cycles). The sequences of forward and reverse primers used in this study are detailed in Table 1. Relative gene expression of Col1a1 (Collagen, type I, alpha 1), Col3a1 (Collagen, type III, alpha 1), Acta2 (Actin alpha 2, smooth muscle) and CD44 was calculated using ΔΔCt method and mRNA levels were normalized according to 18S expression.

Samples of uterine horns collected from female mice were fixed in 10% neutral buffered formalin with subsequent paraffin embedding. 3 μm-thick sections were stained with hematoxylin and eosin (H&E) and Masson’s trichrome original in accordance with standard protocol. Digital whole slide images were recorded using a ScanScope XT Slide Scanner (Leica Aperio Technologies, Vista, CA, United States) under ×20 objective magnification (0.5-μm resolution) and subsequently subjected to digital image analysis by using HALOTM software (version 3.0311.174; Indica Labs, Corrales, New Mexico, United States). The positive pixel counts based HALO Area Quantification v2.1.3 algorithm was calibrated to recognize fibrosis areas and other surrounding tissues. Pixel intensity values were manually determined by the pathologist. Examples of image analysis output are presented in Figure 4 and Figure 6. The repeated manual measurements of the thickness of the total uterine horn wall and its separate layers (10 measurements per each tissue sample) were performed by using HALOTM software.

Statistical analysis was performed using GraphPad Prism version 8.0.1 for Windows, GraphPad Software, San Diego, California United States, www.graphpad.com. Data in graphs are represented as mean ± standard deviation (S.D.), while triangular data points indicate outliers determined by ROUT (Q = 5%). The statistical significance of difference was calculated using Mann-Whitney U test, Kruskal-Wallis test and Bonferroni test, significance was set at p ≤ 0.05 (*), p ≤ 0.01 (**), p ≤ 0.001 (***).

The morphology of hEnMSCs was observed during cultivation and prior to the injection into different female mice models (Figure 2). The images indicate that cells had spindle-shaped morphology, were adherent and formed a consistent monolayer. The cells were cultured and further characterized at early passages.

Cell surface analysis by flow cytometry was carried out to ensure that our isolated cells corresponded with basic MSCs properties. We have founded that hEnMSCs were identified as positive with CD44, CD73, CD90, CD105, CD146, CD166 and CD140b cell surface markers (Figure 3). CD44 and CD166 were determined to be highly positive, exceeding 95%. Positive expression of CD146 and CD140b markers are characteristic to endometrial stromal cells and positive expression of CD44, CD73, CD90, CD105 surface markers confirms mesenchymal origin of stem cells. Moreover, hEnMSCs did not express hematopoietic cell surface markers CD34, CD45 and HLA-DR, which were all measured at less than 1%. Histograms of analyzed surface markers are presented in Supplementary Figure S1. In order to further characterize hEnMSCs by their differentiational abilities, we induced hEnMSCs to specialize into adipogenic, osteogenic, myogenic, neurogenic, and chondrogenic lineages in our previous research (Žukauskaitė et al., 2023). According to our results, we showed that hEnMSCs were able to successfully differentiate into adipogenic, osteogenic, neurogenic, and chondrogenic directions.

The mechanically damaged endometrium model of NOD. CB-17-Prkdc scid/Rj female mice was established via endometrial scratching of the right uterine horn of female mice. Meanwhile, chemotherapy-induced female mice model was established using intraperitoneally chemotherapy. The Control mice model did not receive any interventions.

Uterine horns were collected and histologically evaluated 7 days after the endometrium injury (the MechI mice group) or intraperitoneally chemotherapy induction (the CheI mice group), and 7 days after hEnMSCs-treatment usage for endometrium regeneration in the MechI-hEnMSCs mice group and the CheI-hEnMSCs mice group, accordingly. Masson’s trichrome original staining and H&E staining was used to evaluate histological changes in layers of uterine horns: the total wall, the myometrium-endometrium, the myometrium and the endometrium.

A fertility assessment was performed by visually counting of embryo ISs after 14 days of housing female mice and male mace together in these groups: female mice with mechanically damaged endometrium (MechI) or chemotherapy induction (CheI) and female mice with mechanically damaged endometrium or chemotherapy induction which received hEnMSCs-treatment usage for endometrium regeneration—the MechI-hEnMSCs mice group and the CheI-hEnMSCs mice group, accordingly. In the mechanically injured endometrium model, we counted the number of embryo ISs only in the one uterine horn which was affected. Meanwhile, in the chemotherapy-induced female mice, the number of embryo ISs was counted in two uterine horns due to the systematic effect of the medication.

Expression analysis of genes involved in fibrosis process (Col1a1, Col3a1, Acta2 and CD44) was performed with study’s female mice groups using RNA extracted from their uterine horn tissue samples.

An evaluation of fibrosis in different layers of uterine horns (the total wall, the myometrium-endometrium, the myometrium and the endometrium) was performed in 3 different female mice groups (the Control, the MechI and the MechI-hEnMSCs) using Masson’s trichrome original staining (Figures 4D, H, L) and was subsequently statistically analyzed (Figure 5).

Statistical analysis showed significant differences in the fibrotic area between the Control mice group (Figure 4D), the MechI mice group (Figure 4H) and the MechI-hEnMSCs mice group (Figure 4L) in separate layers of mechanically injured uterine horns such as the total wall of uterine horn (Kruskal-Wallis; H = 8.38; df = 2; p = 0.02), the myometrium-endometrium (Kruskal-Wallis; H = 7.28; df = 2; p = 0.03) and the myometrium (Kruskal-Wallis; H = 7.53; df = 2; p = 0.02) (Figure 5). Conversely, the endometrial fibrosis was significant similar (Kruskal-Wallis; H = 5.58; df = 2; p = 0.06), although this related to the tendencies of the widest fibrotic area in the MechI mice group (Figure 4H) and the narrowest fibrotic area in the MechI-hEnMSCs mice group (Figure 4L) (Figure 5). Overall, the lowest spread of fibrosis was observed of all layers of uterine horns in MechI-hEnMSCs mice group (Figure 4L)—the total wall 6.3% (±1,3%), the myometrium-endometrium 4.1% (±2.9%), the myometrium 6.0% (±0.8%) and the endometrium 1.1% (±1.7%) compared to other study’s groups (Figure 5).

The Bonferroni post hoc test revealed that the percentage of fibrosis spread in the total wall and its separate layers—the myometrium and the myometrium-endometrium—was significantly higher in Mech I mice group (Figure 4H) compared to MechI-hEnMSCs mice group (Figure 4L) (Bonferroni test in all cases; p < 0.05), although the differences in mean values between the Control mice group (Figure 4D), the MechI mice group (Figure 4H) and the MechI-hEnMSCs mice group (Figure 4L) were not statistically significant (in all cases Bonferroni test; p > 0.05) (Figure 5).

Evaluation of fibrosis in different layers of uterine horns (the total wall of uterine horn, the myometrium-endometrium, the myometrium and the endometrium) was performed in 3 different female mice groups (the Control, the CheI and the CheI-hEnMSCs) using Masson’s trichrome original staining (Figures 6D, H, L) and then they were statistically analyzed (Figure 5).

Analysing the intensity of fibrotic area, we detected no significant differences in separate uterine horn layers within the different female mice groups: the total wall of uterine horn (Kruskal-Wallis; H = 1.62; df = 2; p = 0.45), the myometrium-endometrium (Kruskal-Wallis; H = 1.06; df = 2; p = 0,59), the myometrium (Kruskal-Wallis; H = 1.66; df = 2; p = 0.44) and the endometrium (Kruskal-Wallis; H = 0.35; df = 2; p = 0.84) (Figure 5). However, the highest level of fibrosis was observed in the CheI mice group (Figure 6H) in all layers of uterine horns: the total wall of uterine horn 16.9% (±8.4%), the myometrium-endometrium 14.8% (±11.3%), the myometrium 21.4% (±9.8%) and endometrium 8.4% (±8.3%) (Figure 5).

An evaluation of the thickness of different uterine horns layers (the total wall of uterine horn and the endometrium) was performed in the Control mice group (Figure 4B; Figure 6B), mechanically injured endometrium model (Figures 4F, 4J) and chemotherapy-induced model (Figures 6F, J) using Masson’s trichrome original staining and, after that, statistically analyzed (Figure 7).

During the study, it was revealed that the widest total wall thickness of uterine horns was observed in the Che-EnMSCs mice group (1645.3 ± 203.0 µm) (Figure 6J), a slightly lower thickness was detected in the CheI group (1539.7 ± 190.5 µm) (Figure 6F) and the Control mice group (1500.7 ± 189.7 µm) (Figure 6B; Figure 7). Overall, the total wall thickness was significantly smaller in the MechI mice group (1315.1 ± 222.0 µm) (Figure 4F) and the MechI-EnMSCs mice group (1220.8 ± 241.5 µm) (Figure 4J; Figure 7).

An evaluation of the endometrial area in all female mice groups revealed that the widest endometrial area, as well as the total wall thickness of uterine horns, was observed in the Che-EnMSCs mice group (485.1 ± 50.0 µm) (Figure 6J; Figure 7). In contrast, the result of the Control mice group (367.2 ± 50.6 µm) (Figure 4B), the MechI mice group (388.0 ± 72.6 µm) (Figure 4F) and the MechI-hEnMSCs mice group (320.5 ± 69.3 µm) (Figure 4J) showed a thickening of the narrower endometrial area (Figure 7). It was also seen that the endometrial diameter in all female mice groups was significantly reduced compared to the total wall thickness (in all cases Mann-Whitney U; p ≤ 0.05).

The nonparametric one-way analysis of variance revealed that the endometrial thickness of the Control mice group (Figure 6B) and the CheI mice group (Figure 6J) was statistically significant (Kruskal-Wallis; H = 6.56; df = 2; p = 0.04) and, moreover, the endometrial thickness in the Control group (Figure 6B) was significantly narrow in comparison with the CheI mice group (Figure 6F) and the Che-EnMSCs mice group (both Bonferroni test; p < 0.05) (Figure 7). Contrary to the endometrial area, the total wall thickness of uterine horns was not statistically different between the Control mice group (Figure 6B), the CheI mice group (Figure 6F) and the Che-EnMSCs mice group (Kruskal-Wallis; H = 1.62; df = 2; p = 0.45) (Figure 6J; Figure 7). There was also a similarity between the Control mice group (Figure 4B), the MechI mice group (Figure 4F) and the MechI-hEnMSCs mice group (Figure 4J) in relation to the thickness of endometrium (Kruskal-Wallis; H = 1.67; df = 2; p = 0.44) and the total wall thickness of uterine horns (Kruskal-Wallis; H = 3.72; df = 2; p = 0.16) (Figure 7).

Some tendencies could be seen according to the histology assessment by H&E, also. It was observed that in the MechI mice group, the smooth cavity of uterine horn was damaged with the detaching of the endometrium and glandular dilatation (Figures 8D–F). Meanwhile, these findings were less pronounced in the MechI-EnMSCs mice group (Figures 8G–I) and not expressed in the Control mice group (Figures 8A–C). Assessing the intensity of inflammatory cells infiltration, absent or scant number of lymphocytes and polymorphonuclear cells (PMNs) was observed in the Control mice group (Figures 8A–C) and the MechI-EnMSCs mice group (Figures 8G–I) in contrast with the MechI mice group (Figures 8D–F), where moderate amount of PMNs were seen. Evaluating results of apoptosis rate, the moderate number of apoptotic bodies in the MechI mice group (Figures 8D–F) were observed. None of these pathological changes were detected in the Control mice group (Figures 8A–C) and they showed improvement in the MechI-hEnMSCs mice group (Figures 8G–I). However, the distribution results of inflammatory cells and apoptotic bodies when comparing chemotherapy-induced mice groups (Figures 8J–O) and the Control mice group (Figures 8A–C) were quite scattered. Overall, the similar tendencies were also seen in the evaluation of mitotic bodies formation in different study’s groups (Figure 8).

A fertility assessment was performed to evaluate the direct effect of stem cell-based treatment on reproductive function in 5 different female mice groups (the Control, the MechI, the MechI-hEnMSCs, the CheI and the CheI-hEnMSCs) (Supplementary Table S1). In the mechanically injured endometrium model, we counted the number of embryo ISs only in the one uterine horn which was affected. Meanwhile, in the chemotherapy-induced female mice, the number of embryo ISs was counted in two uterine horns due to the specific effect of the medication. Chemotherapy drugs exert cytotoxic effects systemically and therefore can damage the ovaries, leading to infertility, premature ovarian failure, and, have indirect or direct deleterious effects on the uterus, that can be recognized clinically.

First of all, results from mechanically injured endometrium model showed a lower number of mice embryo ISs in the MechI mice group (2.6 ± 1.1 units) and the MechI-hEnMSCs mice group (3.8 ± 1,3 units) in comparison to the Control mice group (5.2 ± 0.8 units) (Figure 9). The Bonferroni post hoc test analysis revealed that the number of mice embryo ISs was significantly different between the Control mice group and the MechI mice group, also (Bonferroni test; p < 0.05) (Figure 9).

During the study of chemotherapy-induced female mice, the Control mice group was observed to have the highest number of mice embryo IS (10.8 ± 0.8 units), more than twice the number of embryo ISs was determined in the CheI-hEnMSCs mice group (4.6 ± 0.5 units) while the lowest number of embryo ISs was observed in the CheI mice group (0.2 ± 0.4 units) (Figure 10). Statistical analysis revealed that the number of embryo IS in the Control mice group, the CheI mice group and the CheI-hEnMSCs mice group was significantly different from each other (Kruskal-Wallis; H = 12.89; df = 2; p = 0.002; Bonferroni test; p < 0.001) (Figure 10).

Expression analysis of genes involved in fibrosis process was performed with 5 different female mice groups (the Control, the MechI, the MechI-hEnMSCs, the CheI and the CheI-hEnMSCs) using RNA extracted from their uterine horn tissue samples (Figure 11).

The discussed genes Col1a1, Col3a1, Acta2 have been shown to have an association with excess fibrosis and wound healing (Liu et al., 1997; Cutroneo et al., 2007; Szóstek-Mioduchowska et al., 2019), and CD44 gene is associated with the binding of extracellular matrix molecules, such as collagen, fibronectin and osteopontin (Foster et al., 1998). Gene expression analysis revealed that Col1a1 gene was upregulated in all of the female mice groups receiving interventions, compared to Control mice group, although, only two of which were considered to be statistically significant (Figure 11). Col1a1 gene was upregulated significantly in the MechI mice group and the CheI mice group (Figure 11). The expression of Col3a1 gene throughout different group was determined to be at a very similar level to the Control mice group, although a slight upregulation can be seen in the CheI-hEnMSCs mice group when compared to the CheI group. In all groups, CD44 was downregulated in comparison to the Control mice group while values of relative gene expression remained at a similar level between the different mice groups. Acta2 gene was upregulated in the MechI-hEnMSCs mice group in comparison to the MechI mice group, although the results were not significant.