Close this search box.

De novo reconstruction of a functional in vivo-like equine endometrium using collagen-based tissue engineering – Scientific Reports

The animal care and use protocol was approved and performed according to the CU-ACUP of Chulalongkorn University (protocol 2131028) which complied with the ARRIVE guidelines. (The full form of CU-ACUP was in the Supplementary Document.) The biosafety procedures followed the guidelines of the Institutional Biosafety Committee of the Faculty of Veterinary Science, Chulalongkorn University (No. IBC1631022). The schematic diagram of materials and methods in this study was illustrated in Fig. 1.

Figure 1
figure 1

A schematic diagram of the three different parts of this study.

Isolation and characterization of equine eECs and eMSCs

Animals and sample collection

Eight healthy cyclical mares (1.5–7 years old) with normal estrus cycles were selected. The estrus cycles were consecutively monitored during breeding season by transrectal palpation and ultrasonography of the ovaries and uterus in order to detect the ovulation day (Day 0)23. Subsequently, the uterine biopsy was performed on days 2–7 post ovulation. The mares underwent anesthesia using 0.01 mg/kg detomidine. Pre-operative analgesia was administered intravenously with butorphanol tartrate at a dosage of 0.1 mg/kg. Additionally, localized epidural anesthesia was provided using 2% lidocaine. The biopsy samples were collected using EQUIVET uterine biopsy forceps (Cat. No 141965, KRUUSE), and transported at 26–28 °C to the laboratory within 6 h in a 0.9% saline solution supplemented with antibiotics (100 IU/ml penicillin and 100 µg/ml streptomycin). Post-operative care included four days of intravenous Flunixin meglumine (1.1 mg/kg) for pain management. The endometrial tissue biopsies were evaluated and categorized according to Kenney24 as part of the screening process for reproductive health before isolating the eECs and eMSCs. Tissues classified as grade I were chosen for further analysis.

Isolation of the eECs and eMSCs

The endometrial tissue was first cut into 3 × 3 mm pieces and cultured as explants in a petri dish using specific culture media: eMSCs medium and either semi-defined or defined epithelial medium. The eMSCs medium included 88% low glucose Dulbecco’s Modified Eagle Medium (DMEM, 31600-034, Gibco®), 10% (v/v) fetal bovine serum (FBS, 10270-106, Gibco®), 1% (v/v) l-glutamine (25030-081, Gibco®), and 1% (v/v) antibiotic/antimycotic solution 100× (15240-062, Gibco®). The epithelial medium, known as FBS-EC medium, featured MSC medium supplemented with 0.01 µg/ml epidermal growth factor (EGF) and 2.436 mM hydrocortisone. Alternatively, a defined culture medium, KSR-EC medium, was used that was similar to FBS-EC except that the FBS was replaced by 10% (v/v) knockout serum replacement (KSR). The explanted tissues and cells were incubated at 37 °C in a humidified atmosphere of 5% CO2-in-air until the expanded cells reached 80–90% confluence. Subsequently, the cell outgrowths were isolated by time-dependent trypsinization. Stromal cells underwent digestion for 2 min, while epithelial cells were digested for 5 min by exposure to 0.25% trypsin EDTA (25200-072, Gibco) at 37 °C. Following trypsinization, the two cell types were cultured separately to purify the population in their respective specific culture media. Eventually, the cells were sub-cultured and then trypsinized and cryopreserved at − 80 °C in the cryopreserved medium containing 10% (v/v) DMSO and 90% (v/v) fetal bovine serum. The morphology of eECs and eMSCs was examined under a phase contrast microscope (CK X41 Olympus, Japan).

Characterization of eECs and eMSCs

The eECs and eMSCs were characterized using a modification of the approach described by Rink et al.25. Immunofluorescence was used to detect proteins specific for eECs (Pan-cytokeratin) or for eMSCs (Vimentin). Flow cytometry was then used to assess eEC purity after staining. The eECs were gated according to side scatter to assess cellular complexity or granularity, and forward scatter to indicate cell size. Subsequently, the same population from both scatter sides within each cell line was selected for analysis. The samples were analyzed utilizing the BDFACSCalibur flow cytometer (BD, USA) with the BD CellQuest™ Pro software. Conventional polymerase chain reaction (conventional PCR) was also used to detect the expression of genes specific to eECs (Muc1) and eMSCs (CD29, CD44, CD90). For eMSCs, adipogenic and osteogenic differentiation was induced following the criteria outlined for MSC properties described by the International Society for Cellular Therapy26. Tissue-specific gene expression markers (LPL for adipogenic and COL1A1 for osteogenic) were assessed using conventional PCR. Alizarin red staining was used to detect extracellular calcium deposits during osteogenic differentiation, and alkaline phosphatase (ALP) activity was measured using an alkaline phosphatase test kit. Oil-red O staining was used to visualize lipid droplets in cells differentiated towards an adipogenic phenotype. This characterization approach confirmed the distinct identities of the eECs and eMSCs. Additional details of this process are illustrated in the Supplementary Material.

Primer design, RNA extraction and conventional PCR

Primer sequences for target genes were designed using Primer3 Web-based software in the National Center for Biotechnology Information (NCBI; with reference to Rink et al.25, as illustrated in Table 1; expression of a housekeeping gene (GAPDH), an epithelial marker (Muc1), and MSC markers (CD29, CD44, CD90), osteogenic markers and adipogenic markers were all examined. The RNA extraction and protocol for conventional PCR were modified from Setthawong et al.27, and described in the Supplementary Material.

Table 1 Primer pairs for conventional and real-time PCR of genes expressed by equine endometrial epithelial cells (eECs) and mesenchymal stromal cells (eMSCs) (*see Rink et al.25.

Optimization of culture conditions for eECs using the ROCK inhibitor, Y-27632

The effect of the ROCK inhibitor, Y-27632, on eECs derived from the defined culture medium (KSR-EC medium; n = 3 cell lines) was analyzed. The cells were first examined by conventional PCR in order to validate the presence of the Rho/ROCK pathway. The eECs cells were then cultured in a medium supplemented with different concentrations of Y-27632 (0, 5, and 10 µM). The Effect of the ROCK inhibitor on eECs function was determined via the following experiments.

Detection of Rho/ROCK pathway in eECs


Expression of genes associated with the Rho/ROCK downstream pathway (RHOA, ROCK1, ROCK2, LIMK1, and LIMK2) was detected using conventional PCR. Primer sequences were derived from NCBI and listed in Table 1. Samples were amplified using 100 ng/ml of cDNA. conventional PCR was performed following a previously established protocol.

Proliferation assay

Doubling time

eECs (15,000 cells/well) were cultured in triplicate in a 24-well plate with varying concentrations of Y-27632 at 37 °C in a humidified atmosphere of 5% CO2-in-air. At 24, 48, and 72 h, cells were detached from the plastic by trypsinization with 0.25% trypsin–EDTA at 37 °C for 5 min, and cell number was determined by staining with trypan blue followed by counting with a hemocytometer. Doubling time was calculated using V. Roth MD’s Doubling Time Calculator (2006).

Edu assay staining

The effect of Y-27632 on de novo DNA synthesis was evaluated after 72 h of culture using Click-it Edu (Click-iT™ Edu Alexa Fluor™ 488 Imaging Kit, Lot 1939601 (Invitrogen by Thermo Fisher Scientific, Oregon USA) Cell proliferation assays following the manufacturer’s instructions. Briefly, the Edu labeling was added at 48 h (1 µl/1 ml of culture medium) and incubation was continued for a further 24 h. At the end of the 24 h incubation with label, all samples were collected and permeabilized. The Click-it reaction cocktail was added into the cell pellet (500 µl/cell line) which was incubated for 1 h at room temperature, with light protection. Flow cytometry was used to detect the Edu (labelled with AlexaFluor 488® azide) with the help of a green emission filter (wavelength 519 nm). The proliferation rate was calculated as the proportion of Edu-positive nuclei among the cells.

Viability and apoptosis validation

FITC annexin V-propidium iodide staining

The eECs were cultured in triplicate (50,000 cells/cm2) in a 24-well plate with varying Y-27632 concentrations (0, 5, 10 µM). Apoptosis was induced by exposing the eECs to 200 mM hydrogen peroxide (H2O2) for 24 h28. Subsequently, all samples were collected and stained with the FITC-Annexin V Apoptosis Detection Kit with PI (Lot B337692, BioLegend®) according to the manufacturer’s protocol, and flow cytometry was used to assess the percentage of viable cells.

Quantitative PCR (qPCR, real-time PCR)

The expression of genes related to apoptosis (Pro-apoptotic: BAX, and Anti-apoptotic: BCL2) was assessed in control and treatment eEC populations by real-time PCR. RNA extraction and cDNA conversion were performed as described previously. The primer sequences for BAX and BCL2 are listed in Table 1. GAPDH was selected as a reference gene. For optimization of the qPCR, the primer was initially tested by conventional PCR. Standards for the qPCR were prepared from purified PCR products using a QIAquick PCR purification kit (QIAGEN) that was quantified by spectrophotometry and diluted over at least 9 orders of magnitude. The standard curve was achievable with R2 > 0.995 and the efficiency of the reaction close to 1.0 which means that 100% cDNA has occurred after each cycle. Real-time PCR was performed following the protocol described by Swangchan-Uthai29. All experiments were repeated in duplicate with the KAPA SYBR® FAST qPCR kit, KAPA Biosystems. The relative expression was developed from Livak and Schmittgen30, then levels of the BAX and BCL2 genes were normalized to GAPDH. Data of relative gene expression and the ratio of BAX: BCL2 were expressed as means ± SEM and compared by one-way ANOVA.

Functional assessment of equine endometrial cells (eECs)

For functional assessment, the eECs were cultured under optimal Y-27632 conditions in the defined KSR-EC medium. As determined by the expression of the specific functional secretory gene Muc1, these cells could maintain functional status up to passage 6, with the chosen ROCK inhibitor concentration. Confirmation was based on conventional PCR to examine the expression of the Muc1 gene.

Effect of fetal bovine serum on eECs proliferative activity

Due to limited proliferation of eECs when cultured in KSR-EC medium, the effect of FBS on cell proliferation was examined. The eECs were cultured in 10% FBS-EC medium containing the optimal ROCK inhibitor concentration. The cells were then further sub-cultured (Passage 6). Conventional PCR was used to analyze Muc1 gene expression.

Reconstruction of equine in vitro 3D-ET

For this experiment, the 3D-ET was maintained under the optimal long-term culture conditions (FBS-EC medium with 10 µM Y-27632) established in the preceding experiment.

Collagen scaffold preparation

The sterilized collagen basement membrane (BM) (Nitta Gelatin Inc., Japan, Lot no. 190709) was neutralized using 200 mM HEPES buffer and adjusted to pH 7–7.5 with 10 mM sterile NaOH. The collagen was gently stirred using a sterilized magnetic stirrer for approximately 5 min and maintained at 4 °C until used.

Reconstruction of the 3D-ET

Before reaching confluency, the eECs and eMSCs at passages 3–4 were digested and mixed at a ratio of 1:1 and then centrifuged at 157×g at 4 °C for 5 min to generate a cell pellet. After removing the supernatant, the prepared endometrial cells were mixed gently with the collagen (total of 150,000 cells/well) volume (500 µl or 0.965 cm3). This mixture was carefully introduced into a 24-well plate, followed by the addition of 200 µl of PBS to cover the collagen within each well. The plate was then placed in a 37 °C incubator in a humidified atmosphere of 5% CO2-in-air for 1.5 h to allow the collagen to polymerize. After gel formation, the PBS was removed and 500 µl of 10% FBS-EC medium was added into each well. On day 3 of culture, the supernatant was removed, the collagen was washed with PBS, and a layer of Matrigel™ was overlain onto the collagen as an artificial basement membrane and incubated at 37 °C for 15 min. Subsequently, eECs were seeded onto the Matrigel™, at approximately 1.5 × 104 cells/cm2 This culture configuration facilitated cell growth over a 7–14 day period, with medium renewal every 2–3 days. Morphological changes in the reconstructed tissue were monitored daily using a phase contrast microscope (CK X41 Olympus, Japan).

3D-ET characterization

Specific gene and protein characterization

Conventional PCR Expression of the endometrial gland marker (Foxa2) and the secretory marker (Muc1) in in vitro 3D-ET were assessed via conventional PCR. The primer sequences are listed in Table 1. The conventional PCR protocol was conducted as described previously.

Immunofluorescent and confocal microscopy The whole section of in vitro 3D-ET was fixed by immersion in 4% paraformaldehyde (PF) solution for 30 min. Subsequently, the in vitro 3D-ET was cut with a surgical blade into 3 × 3 mm pieces and placed onto glass slides. To identify the eECs and eMSCs, antibodies against Pan Cytokeratin and Vimentin were used, as described previously. Wheat germ agglutinin was additionally used to identify the cell membrane boundary. The nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI). For the identification of endometrial gland-like structures, a primary antibody against FOXA2 was used (see Table 2 for details of primary and secondary antibodies). Additionally, the secretory function of the endometrial gland-like structures was assessed by staining for MUC1 (sc-7313, Santa Cruz Biotechnology), with a protocol adapted from Piña et al.31. The details are described in the Supplementary Information. The anti-fade reagent was applied to the in vitro tissue blocks before they were covered with a cover glass and mounted using a mounting solution. Visualization of the overall structure was achieved using an immunofluorescent microscope, whereas z-stacks created via a confocal microscope (AX/AXR with NSPARC, Nikon Japan) were used for 3D structure assessment.

Table 2 Antibodies used to characterize eECs, eMSCs, and to detect endometrial markers.
Functional testing of 3D-ET via a lipopolysaccharide (LPS) challenge

For functional testing, the 3D-ETs were divided into triplicates. The control group was cultured in normal 10% FBS EC-medium, whereas the treatment group was cultured in FBS-EC supplemented with Escherichia coli O55:B5 bacterial LPS (5 µg/ml, Sigma Chemical Inc.). All 3D-ETs were incubated at 37 °C in 5% CO2-in-air for 24 h. After the 24 h incubation, the culture medium and in vitro 3D-ET samples were collected separately and frozen at − 80 °C for subsequent investigations. Enzyme-linked immunosorbent assays (ELISA) were employed for the quantitative analysis of secretory products (IL6 and PGF2α) in the culture media. qPCR was used to determine and quantify gene expression for the inflammatory cytokine IL6 and for the prostaglandin synthase enzyme, PGFS, within the collected in vitro 3D-ET samples (for primers see Table 1).

ELISA IL6 concentrations in the medium from the in vitro 3D-ETs was measured in duplicate using the Human Th Cytokine Panel 13-plex assay (Biolegend, USA) and analyzed via flow cytometry (BD FACS Calibur, Becton Dickinson, USA). The full protocol is described in the supplementary materials.

The PGF2α concentration was determined in duplicate using an ELISA Kit (Abnova) in accordance with the manufacturer’s guidelines. The intra-assay coefficient of variation (CV) was 2.82%, and the detection limit was 0.326 pg/ml.

Statistical analysis

Statistical analysis was performed using SPSS version, and all graphs were produced using GraphPad Prism 9.5.1. Continuous data were tested for normality and equivalence of variance and reported as means ± SEM. ANOVA was used to determine differences between treatments in experiments II and III. Independent sample t-tests were conducted to compare relative gene expression for IL6 and PGFS, and secretion of PGF2α in LPS-treated versus non-treated in vitro 3D-ET. Chi-square test of Fisher’s exact test was used to examine the effect of LPS on IL6 secretion.