Calle et al. Stem Cell Research & Therapy (2019) 10:23 https://doi.org/10.1186/s13287-018-1129-1 RESEARCH Open Access Bovine endometrial MSC: mesenchymal to epithelial transition during luteolysis and tropism to implantation niche for immunomodulation Alexandra Calle1, Soraya López-Martín2, Marta Monguió-Tortajada3, Francesc Enric Borràs3,4, María Yáñez-Mó2,5 and Miguel Ángel Ramírez1* Abstract Background: The uterus is a histologically dynamic organ, and the mechanisms coordinating its regeneration during the oestrous cycle and implantation are poorly understood. The aim of this study was to isolate, immortalize and characterize bovine endometrial mesenchymal stem cell (eMSC) lines from different oestrous cycle stages (embryo in the oviduct, embryo in the uterus or absence of embryo) and examine their migratory and immunomodulatory properties in an inflammatory or implantation- like environment, as well as possible changes in cell transdifferentiation. Methods: eMSCs were isolated and analysed in terms of morphological features, expression of cell surface and intracellular markers of pluripotency, inmunocytochemical analyses, alkaline phosphatase activity, proliferation and osteogenic or chondrogenic differentiation capacities, as well as their ability to migrate in response to inflammatory (TNF-α or IL-1β) or implantation (IFN-τ) cytokines and their immunomodulatory effect in the proliferation of T cells. Results: All eMSCs showed MSC properties such as adherence to plastic, high proliferative capacity, expression of CD44 and vimentin, undetectable expression of CD34 or MHCII, positivity for Pou5F1 and alkaline phosphatase activity. In the absence of an embryo, eMSC showed an apparent mesenchymal to epithelial transition state. eMSC during the entire oestrous cycle differentiated to osteogenic or chondrogenic lineages, showed the ability to suppress T cell proliferation and showed migratory capacity towards pro-inflammatory signal, while responded with a block in their migration to the embryo-derived pregnancy signal. Conclusion: This study describes for the first time the isolation, immortalization and characterization of bovine mesenchymal stem cell lines from different oestrous cycle stages, with a clear mesenchymal pattern and immunomodulatory properties. Our study also reports that the migratory capacity of the eMSC was increased towards an inflammatory niche but was reduced in response to the expression of implantation cytokine by the embryo. The combination of both signals (pro-inflammatory and implantation) would ensure the retention of eMSC in case of pregnancy, to ensure the immunomodulation necessary in the mother for embryo survival. In addition, in the absence of an embryo, eMSC showed an apparent mesenchymal to epithelial transition state. Keywords: Endometrial mesenchymal stem cells, Embryo implantation, Inflammation, Cell migration * Correspondence: ramirez@inia.es 1 Departamento de Reproducción Animal, Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria (INIA), Avenida Puerta de Hierro 12, local 10, 28040 Madrid, Spain Full list of author information is available at the end of the article © The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Calle et al. Stem Cell Research & Therapy (2019) 10:23 Page 2 of 12 Background oestrous cycle and implantation are poorly understood. In ruminants, embryo mortality is a key factor affecting fer- It has been proposed that bone marrow-derived cells tility [1]. Approximately 50% of pregnancy loss occurs be- contribute to uterine regeneration [10]. An alternative or tween blastocyst hatching and embryo attachment to additive mechanism of endometrial regeneration could endometrium in cattle [2]. Efficient communication be- be in situ cellular transdifferentiation, an example of tween cells and tissues is paramount in many physiological which is mesenchymal to epithelial transition (MET). processes, including embryo development. Preimplanta- During MET, mesenchymal cells are re-programmed, tional embryos of mammals develop within the female thereby gradually losing mesenchymal cell characteristics genital tract (oviduct and uterus) and their efficient com- while gaining epithelial cell traits [11]. MET and epithe- munication conditions their development and survival. In lial to mesenchymal transition (EMT) are fundamental ruminants, the bicornuate uterus is coated with the endo- processes that occur during the bovine embryo develop- metrium, the ultimate and unique biological layer facing ment and are also implicated in tumour metastasis [12]. the implanting embryo in normal early pregnancy, which Uchida et al. reported that the EMT might play an im- drives the development of the embryonic disc and portant role in human embryo implantation helping the extra-embryonic tissues [3]. The endometrium displays two embryo to penetrate through the endometrial epithelial specific areas, namely the caruncles and the intercaruncular cells into the endometrial stromal cell layer [13]. In areas. The caruncles represent aglandular structures of lim- addition, Zhang et al. suggested that MET might exist ited size that are distributed over the endometrial surface. during decidualization, providing a stable developmental The intercaruncular areas are large and contain the endo- environment and an anchor point for embryos to invade metrial glands, which produce histotroph, a collection of into the uterus [14]. Recently, Patterson et al. reported numerous and diverse factors including cytokines and that MET contributes to endometrial regeneration fol- growth factors [4]. lowing natural and artificial decidualization by sesame During the window of implantation, after blastocyst at- oil injection into the uterine lumen in mouse [15], but tach onto the endometrium, both in human and mice, the possibility that this cellular transdifferentiation also characterized by haemochorial implantation, the stromal serves as a mechanism to regenerate adult endometrium cells at the implantation site begin to undergo deciduali- in bovine has not been evaluated. zation, a process by which the fibroblast-like endomet- Although the presence of stem cells in the endomet- rial stromal cells differentiate into polygonal epithelial- rium was first speculated in 1978 by Prianishnikov [16], like cells [5]. The differentiation of decidual cells is one it was not until 2004 that stem cells were identified and of the earliest uterine adaptations to pregnancy in these characterized for the first time in the endometrial tissue species [6] and involves cell growth, change of shape, [17, 18]. The adult stem cells, mesenchymal stem cells multinucleation and establishment of inter-cellular junc- (MSCs), unlike pluripotent embryonic stem cells, are tions [7]. Many genes and signal pathways are involved multipotent and can produce a limited range of cell line- in the process of decidualization that is crucial for ages depending on their location. MSCs are isolated blastocyst implantation and maintenance of pregnancy. from virtually every tissue, so that their stem cell proper- Ruminants and pigs have noninvasive implantation; ties are difficult to predict in an in vivo setting. The however, they differ in the type of placentation. Pig con- presence in the bovine endometrium of mesenchymal ceptuses undergo epitheliochorial placentation in which stem cells (eMSCs) has been already described [19–21]. luminal epithelium (LE) remains morphologically intact Bovine eMSCs possess several MSC properties such as throughout pregnancy and the conceptus trophectoderm clonogenicity, ability to adhere to plastic, fibroblast simply attaches to the apical LE surface without contact- morphology and in vitro differentiation capacity to adi- ing uterine stromal cells [8]. In contrast, synepithelio- pocyte, chondrocyte and osteocyte [22]. Moreover, bo- chorial placentation in ruminants involves extensive vine eMSCs present high levels of CD29, CD44 and erosion of the LE due to syncytium formation with vimentin markers, low expression levels for CD34 and trophectoderm binucleate cells. After day 19 of preg- no MHC-II [21]. It is important to highlight that in nancy, the conceptus tissue is apposed to, but does not some cases MSCs are able to divide but to a limited penetrate, uterine stroma. extent in vitro before entering replicative senescence. Be- Johnson el al. reported that the uterine stroma of sheep tween passages 7 and 12, MSCs increase their cell size and undergoes a programme of differentiation similar to reduce the expression of certain pluripotency markers, decidualization in invasive implanting species, whereas leading to proliferative arrest [23]. porcine stroma exhibits differentiation that is more lim- MSCs have the ability to migrate to damaged and in- ited than that in sheep, rodents or primates [9]. flamed tissues where they proliferate and utilize their The uterus is a histologically dynamic organ, and the immunomodulatory properties as part of the regener- mechanisms coordinating its regeneration during the ation process of the tissue itself [24–26]. The immune Calle et al. Stem Cell Research & Therapy (2019) 10:23 Page 3 of 12 reaction has classically been functionally divided into The uterine horns ipsilateral to the luteum corpus were Th-1 and Th-2 responses. Although, in human, numer- removed and washed with fresh tap water, and the uter- ous studies attempted to support pregnancy as a Th-2 or ine lumen was washed three times using a 20-ml syringe anti-inflammatory state [27], a similar number of studies with Hank’s balanced salt solution (HBSS) supplemented supported its condition as a Th-1 or pro-inflammatory with 500 U/ml penicillin, 500 mg/ml streptomycin and state [28]. A two-phase model was thereafter proposed 0.1% bovine serum albumin (BSA) (Merck KGaA, Darm- by Mor et al., being the first and third trimesters charac- stadt, Germany) by flushing from the utero-tubal junc- terized by a pro-inflammatory ambient, while the second tion. To obtain bovine endometrial cells, a longitudinally trimester would represent an anti-inflammatory phase cut of the uterine horn was performed with sterile scis- [29]. MSC can suppress or reduce the proliferation of T sors. The intercaruncular endometrium was carefully cells and also have a role in the regulation of Th-1/Th-2 dissected from underlying tissues, cut in strips, minced balance shifting a Th-1 phenotype to a Th-2 [30]. MSC using sterile scissors and incubated in 25-ml type II col- interactions with immune cells have been studied widely lagenase solution: HBSS supplemented 0.05% type II col- in both humans and rodents even in pigs [24], although lagenase (Gibco by Life Technologies, Grand Island, NY, this research has not yet been extended to MSC from USA), 0.1% BSA (Merck KGaA, Darmstadt, Germany) the main veterinary species. and 30 mM CaCl2. The solution was homogenized and In ruminants, the major embryo-derived signal of incubated during 1 h at 37 °C shaking gently every 10 pregnancy recognition has been suggested to be type I min. The solution resulting from the digestion was fil- interferon [31]. The production of this interferon (inter- tered through a 70-μm cell strainer (SPL life science, feron-tau; IFN-τ) is unique as it is secreted by trophecto- Korea), and a volume of culture medium RPMI (Hyclone derm cells during the elongation phase and secretion Laboratories, Utah, USA), supplemented with 10% foetal stops when trophectoderm cells appose to the endomet- calf serum (PAA laboratories, Austria), 2% non-essential rial luminal epithelium. IFN-τ is crucial for maternal amino acids and antibiotics (100 U/ml penicillin, 100 recognition through the inhibition of endometrial mg/ml streptomycin), was added to block the action of prostaglandin-F2 alpha secretion, thus preventing luteo- collagenase and the obtained suspension centrifuged at lysis, an indispensable step for maintaining P4 luteal se- 300 × g for 5 min. The resulting pellets were resuspended cretion [32]. in culture medium and plated in 100-mm2 tissue culture Taking all this information into account, we have ad- dish (JetBiofil, Guangzhou, China) and incubated in an dressed the isolation, immortalization and characterization atmosphere of humidified air and 5% CO2 at 37 °C. Cul- of eMSC from different phases of the bovine oestrous cycle ture medium was changed every 48–72 h. in terms of their ability to migrate in vitro by agarose spot assay mediated by inflammatory cytokines (TNF-α or eMSC immortalization IL-1β) or the implantation cytokine (IFN-τ). We have also PA-317 LXSN-16E6E7 cells were acquired from the evaluated their possible immunomodulating function in a American Type Culture Collection (Manassas, VA). xenogeneic condition. Virus was harvested from 16-h conditioned media of PA-317 cells that and filtered through a 0.45-mm low Methods protein binding filter (Millipore, Bedford, MA) to re- Isolation and culture of bovine endometrial mesenchymal move cells and debris. Ten microlitres of viral super- stem cells (eMSCs) natant was obtained from a confluent P100 dish. Viral Heifer uterus and ovaries were collected from routine supernatant was diluted 1:3 with serum-free DMEM/ slaughtered at the local abattoir. Oestrous cycle stages RPMI medium containing polybrene (4 mg/ml; Sigma). were determined by a macroscopic examination of the Primary cultures of bovine endometrial cells obtained luteum corpus and ovarian morphology of the cycling from different cyclic stages which have been grown to heifers according to literature criteria [33] (see Table 1). 60% confluence in P100 culture dishes were overlaid Table 1 eMSC isolation and immortalization efficiency Oestrous cycle stage Embryo location Primary Immortalization Immortalized Oestrous cultures efficiency eMSC stage D2-D4 early luteal Embryo in the oviduct 4 1 (25%) 1A 1 D4-D8 medium luteal Embryo in the uterus with zona pellucid (ZP) 4 0 (0%) – 2 D9-D16 late luteal Embryo in the uterus w/o ZP 8 3 (37.5%) 3A, 3D, 3E 3 D17-D21 follicular phase Regression of corpus luteum if not embryo 6 4 (66.7%) 4B, 4C, 4D, 4H 4 Calle et al. Stem Cell Research & Therapy (2019) 10:23 Page 4 of 12 with 3 ml of this mixture and incubated in an atmos- solution (1 ml of Solution A (Fast Red 1 mg/ml), 1.6 μl of phere of humidified air and 5% CO2 at 37 °C for 3 h. Napthol AS-mx phosphate and 40 μl TRIS HCl 1 M, pH After 3–4 h, viral supernatant was removed; cells were 8.6) during 10–15 min at room temperature in the dark. washed twice with HBSS, and complete culture medium Solution B was finally removed, and the cells were washed was added to dishes. After 48 h, cells having integrated twice with PBS and covered with PBS to prevent drying. the vector were selected by resistance to the neomycin The colonies were studied for appearance of pink/red col- analogue G418 300 μg/ml (Gibco of Life Technologies, oration indicating alkaline phosphatase activity. The Grand Island, NY, USA) in complete culture medium. stained colonies were imaged using a Motic SMZ-171 Selection continued for 7 days, and surviving cells were stereomicroscope coupled to a Moticam BTU8 digital subsequently cultured with complete culture medium camera using Motic Image Plus software version 2.0 containing 100 μg/ml G418 for general cell maintenance. (Motic China Group Co., LTD). The G418-resistant cell populations were subcultured for more than 30 passages, and immortalized cells were Cell proliferation measurement cryopreserved each passage. Different mesenchymal cell lines of passages 10–15 were seeded at 2 × 105 cells per 60-mm tissue culture plates eMSC characterization by flow cytometry (JetBiofil, Guangzhou, China). The culture medium was Surface, cytoplasmic and nuclear cell antigens were ana- changed every 2 days. At each time point, a duplicate of lysed by flow cytometry using a Cell Lab Quanta SC sys- plates was detached by tripsinization and counted using tem of Beckman Coulter. Cell cultures at 80–90% a Bürker counting chamber (Paul Marienfeld GmbH & confluence were detached using 0.05% trypsin–EDTA Co. Lauda-Königshofen, Germany). Twenty microlitres (T/E) solution, collected and fixed with 4% paraformal- of cell suspension was placed on both sides of the cham- dehyde for 10 min and subsequently washed twice with ber and viewed using × 100 magnification under an PBS. For the analysis of the expression of vimentin inverted Nikon Diaphot phase contrast microscope (clone LN-6, Sigma-Aldrich) and cytokeratin (Clone coupled to a Jenoptik ProgRes CT1 digital camera. Im- C-11, Sigma-Aldrich) (cytoplasmic proteins) and ages were taken using ProgRes capture pro software ver- POU5F1 (rabbit polyclonal, Biorbit) (a nuclear protein), sion 2.7 (Jenoptik Laser, Optic Systeme GmbH). A total cell permeabilization was performed by incubation with of 5 × 1 mm2 squares per sample were analysed, and the 0.3–0.5% Triton X-100 for 10 min and washing with number of cells per millilitre was determined according PBS. Nonspecific binding of antibodies was blocked with to number of cells in 1 ml = N/Z × dilution × 104, where TNB-blocking solution during 30 min at 37 °C. Appro- N = the whole number of cells counted and Z = the num- priated dilutions, provided by manufacturers, of primary ber of counted squares. Cell population doubling time antibodies against the markers commonly used to define (PDT) was calculated using Roth V. 2006 Doubling Time mesenchymal stromal cells (MSCs): vimentin (clone Computing, available from http://www.doubling-time. LN-6, Sigma-Aldrich) and CD44 (clone IM7, Bio-rad), as com/compute.php. positive markers; cytokeratin, CD34 (Rabbit polyclonal, Biorbyt), and MHCII (clone CVS20, Bio-Rad) as negative In vitro differentiation potential assay markers; and POU5F1 as pluripotency marker, were eMSC lines at passages 10–15 were grown until 90% added to the cells and incubated overnight at 4 °C. Cells confluence on 24-well multidish (JetBiofil, Guangzhou, were then stained with the appropriated Alexa Fluor China). For osteogenic differentiation, The StemPro® 488-conjugated secondary antibodies (Jackson InmunoR- Osteogenesis Differentiation Kit (Thermo Fisher Scien- esearch Laboratories, Baltimore Pike, West Grove, PA). tific, Meridian Road Rockford, IL, USA) was used accord- Negative control samples were produced by the omis- ing to the manufacturer’s instructions. Differentiating sion of the primary antibody. Analysis of the samples media was changed every 3–4 days for 21 days. Simultan- was performed with Cell Lab Quanta SC system of Beck- eously, control cells were cultured in standard conditions. man Coulter using Flow-Jo X SOFTWARE version Cells were then fixed in 4% paraformaldehyde solution for 10.0.7r2. 30 min. After fixation, cells were incubated for 2–3 min in 2% Alizarin Red S solution (pH 4.2) to visualize the cal- Alkaline phosphatase (AP) activity cium deposits. eMSC lines at passages 10–15 were grown on 35-mm For chondrogenic differentiation, The StemPro® Chon- dishes (JetBiofil, Guangzhou, China) for 2 weeks. Cells drogenesis Differentiation Kit (Thermo Fisher Scientific, were washed twice with PBS and fixed with a solution of Meridian Road Rockford, IL, USA) was used according 4% paraformaldehyde during 10 min at room temperature. to the manufacturer’s instructions. Differentiating media Paraformaldehyde was aspirated, and the plates were was changed every 2–3 days for 14 days. Simultaneously, washed twice with distilled water and covered with B control cells were cultured in standard conditions. Cells Calle et al. Stem Cell Research & Therapy (2019) 10:23 Page 5 of 12 were then fixed in 4% paraformaldehyde solution for 30 at 3·105 cells/well in flat-bottomed well plates in which min. After fixation, cells were incubated for 30 min with eMSC-1A or eMSC-4H had been previously seeded 1% Alcian Blue solution prepared in 0.1 N HCl. Blue stain- (50.000 cells/well) with RPMI medium. T cells had a via- ing was corresponding with proteoglycans synthetized by bility of over 94% in all experiments performed and were chondrocytes. Cells under differentiation conditions were stimulated with anti-CD2/CD3/CD28-coated microbeads photographed, using a Motic SMZ-171 stereomicroscope (Pan T Cell Activation Kit; Miltenyi Biotech) or un- coupled to a Moticam BTU8 digital camera. coated microbeads as a negative control in a 1:10 bead to T cell ratio. Xenoproliferation was determined after Cell migration measurement: agarose spot assay 4.5 days by measuring the CellTrace™ Violet loss by flow The cell migration measurement by agarose spot assay cytometry in a LSR Fortessa Analyzer (BD Biosciences). was carried out following the procedures of Calle et al. [34]. Briefly, PBS-0.5% agarose solution was heated on a water bath until boiling to facilitate complete dissol- Statistical analysis ution. When the temperature cooled down to 40 °C, Statistical analysis was performed using GraphPad Prism 6 90 μL of agarose solution was pipetted into a 1.5-ml (GraphPad Software, La Jolla, CA, USA). One-way ANOVA Eppendorf tube containing 10 μL of PBS or PBS supple- for multiple comparisons by Fisher’s LSD tests were used mented with TNF-α, IL-1β or IFN-τ for a final concen- for cell proliferation and doubling time. Two-way ANOVA tration of 6 nM [33]. Five-microlitre spots of agarose- for multiple comparisons by Fisher’s LSD tests were used containing PBS, TNF-α or IL-1β were pipetted onto for cell migration. Values are presented in mean ± standard six-well plates (JetBiofil, Guangzhou, China), 16 drops error of the mean (SEM). Differences were regarded as sig- per well, 12 drops per MSC line, and allowed to cool for nificant when p < 0.05. Two-tailed Student’s t tests were 15 min at 4 °C. At this point, cells that had been treated used when two groups were compared for immunomodula- with 1 μg/ml C-Mitomycin overnight (Merck KGaA, tory assays. Values are expressed as mean ± standard error Darmstadt, Germany) to avoid cellular duplication were of the mean (SEM). Differences were considered to be sig- plated onto spot-containing dishes in the presence of nificant when p < 0.05. culture media and onto two dishes for mitomycin con- trol, using Bürker counting chamber. Imaging was per- formed at 24 and 48 h using a Motic SMZ-171 Results stereomicroscope coupled to a Moticam BTU8 digital MSC isolation and immortalization efficiency camera using Motic Image Plus software version 2.0 eMSCs were isolated from heifer uterus that were as- (Motic China Group Co., LTD). Motile cells penetrated cribed to one of the four bovine oestrus stage categories the agarose spot. The longest straight distance from the (1–4) (Table 1) proposed by [33] based on the morph- border of the spot was analysed for each cell using ImageJ. ology of the active ipsilateral ovary to the uterine horn from which the cells were isolated. To avoid the risk of In vitro xenostimulatory assays senescence by the maintenance of eMSC in vitro, iso- To measure immunomodulatory properties of eMSCs, a lated cells were immortalized using the retroviral vector co-culture of eMSCs and human T cells was performed. LXSN-16E6E7. From a total of 22 primary cultures of Human peripheral blood mononucleated cells (PBMCs) endometrial stromal cells, eight cell lines were success- from peripheral blood of healthy donors were obtained fully immortalized (eMSC-1A, eMSC-3A, eMSC-3D, by Ficoll Hypaque Plus (GE Healthcare) density centrifu- eMSC-3E, eMSC-4B, eMSC-4C, eMSC-4D, eMSC-4H) gation. T cells were then isolated from PBMCs using the (Table 1). Easy Sep Human T cell Enrichment Kit (Stem Cell Tech- nologies) following the manufacturer’s instructions. Briefly, cells were diluted in a column buffer at a con- Morphological features centration of 5·107 cells/ml, labelled with the tetrameric Pre-immortalized mesenchymal stem cell cultures at antibody complexes (TAC) recognizing human CD14, passage 0 adhered to the plastic surface of culture dishes CD16, CD19, CD20, CD36, CD56, CD66b, CD123, gly- exhibiting a mixture of round, spindle or elongated cophorin A and dextran-coated magnetic particles and shape morphology (Fig. 1—upper panels). However, after separated using the EasySep magnet (StemCell Tech- the first cell passage, the cells formed a more homoge- nologies). Enriched T cells were then washed and resus- neous population of fibroblast-like adherent cells, with pended in PBS with an equal volume of CellTrace™ the exception of eMSC-4D and eMSC-4H that showed Violet (Thermo Fisher Scientific) for 10 min. After an an epithelial-like morphology remained constant before overnight resting period, labelled T cells were washed and after the immortalization even after more than 20 twice with RPMI supplemented with 10% FBS and plated passages (Fig. 1—lower panels). Calle et al. Stem Cell Research & Therapy (2019) 10:23 Page 6 of 12 Fig. 1 Morphology of MSCs. Pre-immortalized mesenchymal stem cell cultures at passage 0 (upper panels) and immortalized mesenchymal stem cell lines at passages 10–15 (lower panels). Phase-contrast images were acquired with × 100 magnification Expression of cell surface, intracellular and pluripotent- follicular phase and with epithelial morphology, were posi- specific markers tive for the nuclear marker POU5F1 (Fig. 2c) and showed Some characteristic MSC surface and intracellular alkaline phosphatase activity (Fig. 2d). markers were assessed by flow cytometry (Fig. 2a–c). All cell lines were positive for cell surface CD44 and cytoplas- mic vimentin, both of them are characteristic markers of Proliferation capacity MSCs. Interestingly, cytokeratin, a typical cytoplasmic To analyse the cell proliferation capacity of eMSCs, the marker expressed by epithelium of ectoderm and endo- number of cells/dish was counted for each cell line at derm, and commonly used as a negative marker of mesen- days 3, 5, 7 and 10, starting in all cases from an initial chymal stem cells, was present in all stage 4 eMSC lines: seeding of 2 × 105 cells. As showed in Fig. 3a, the num- strongly detected in eMSC-4H, clearly positive in ber of cells increased for all cell lines along the entire eMSC-4C and slightly positive in eMSC-4B and eMSC-4D assay, demonstrating high proliferative capacity of these (Fig. 2b), correlating with the epithelial morphology of cells even at passages 10–15. eMSC-4H was the most two of these cell lines. No expression of haematopoietic proliferative line and at day 10 reached 19.74 × 105 ± markers, such as MHCII or CD34, was found in any of 2.13 × 105 cells. Figure 3b shows the proliferation rate the eMSC lines. Regarding pluripotency features, all eMSC of MSC lines in terms of the population doubling time lines, including those cell lines immortalized from the (PDT). Among eMSC lines, eMSC-4H and eMSC-4D Fig. 2 Expression of cell surface, intracellular and pluripotent specific markers and alkaline phosphatase activity. a–c Analysis by flow cytometry of the expression levels of cell surface markers CD34, CD44 and MHCII and intracellular markers cytokeratin, vimentin and POU5F1 in eMSC. Data correspond to the mean fluorescence intensity (folds of negative control) for each sample. d Analysis of AP activity: bright field images were obtained at × 50 magnifications, showing some red-stained cell groups after the action of alkaline phosphatase on Fast Red in the presence of Napthol AS-mx phosphate. Bars = 150 μm Calle et al. Stem Cell Research & Therapy (2019) 10:23 Page 7 of 12 A B Fig. 3 In vitro proliferation of mesenchymal stem cells. a Absolute number of cells/dish (mean ± SD) at indicated points. b Doubling time in days of each MSC line (mean ± SD). Different letters indicate significant differences (p < 0.05) showed on day 10 a significantly shorter PDT of 5.77 ± Migratory response of MSC to inflammatory and 0.47 and 3.50 ± 0.23 days, respectively (p < 0.05). implantational stimuli Assessment of the invasive capacity of all eMSC lines In vitro differentiation of MSCs was performed using the agarose spot assay [34], which As showed in Fig. 4, all eMSC lines cultured under allows the quantification of cell invasion by measuring osteogenic conditions presented a comparable amount their crawling underneath an agarose gel on a planar of calcium deposits, indicating a high osteogenic differ- surface. All eMSC lines showed migration capacity into entiation potential of these lines. Cells cultured under agarose spots without cytokine stimuli from day 2 chondrogenic conditions showed the presence of acidic (Fig. 5). In terms of migration without cytokine stimuli, proteoglycan that was demonstrated at monolayer cells eMSC-4C showed a significantly higher migration rate by Alcian blue staining, presenting in most cases stained than the rest of eMSC lines with a net migrated distance nodules typical from cartilaginous tissue. of 1008.0 ± 38.26 μm on day 7 (Fig. 5B). All this characterization demonstrates that all immor- The inflammatory cytokine TNF-α did not cause an talized lines complied with MSC features in terms of increase in the migration capacity of any of the eMSC marker expression and differentiation capacities, even lines studied compared to the control of PBS, but com- those obtained from the follicular phase that showed pletely abrogated that of eMSC-3D and eMSC-4B. In cytokeratin expression and epithelial morphology. contrast, the inflammatory cytokine IL-1β stimulated the Fig. 4 In vitro differentiation of eMSC to different lineages. Images show Alizarin Red S staining of calcium deposits in cells cultured in basal medium (control) or in osteogenic differentiation medium (top panel); Alcian blue staining of acidic proteoglycan in cells cultured in basal medium (control) or in chondrogenic differentiation medium (bottom panel). Bright field images were acquired with a × 3 magnification (bars = 150 μm) Calle et al. Stem Cell Research & Therapy (2019) 10:23 Page 8 of 12 Fig. 5 Migration analysis in an agarose spot assay. A Representative images of eMSC-4C migration assay into a PBS-, TNF-α-, IL-1β- or IFN-τ-containing agarose spot at the indicated times. Images were obtained in a light stereomicroscope at × 20 magnification. b and c Migratory response of MSC to inflammatory and implantational stimuli. b The distance migrated from the border of the PBS-agarose spot was measured in two independent experiments for all eMSC lines at the indicated times (mean ± SD). Different letters indicate significant differences: p < 0.005 for eMSC migration mediated by PBS at day 7 (a, b, c, d, e); C p < 0.05 for eMSC migration mediated by PBS (a, b, c), p < 0.005 for eMSC migration mediated by TNF-α (e, f, g) and p < 0.05 for eMSC migration mediated by IL-1β (i, j, k, l) and IFN-τ (p, q). *p < 0.05; ***p < 0.0005; ****p < 0.0001 migration capacity of cell lines eMSC-1A, eMSC-3A, influence on polyclonal xenogeneic proliferation. The eMSC-3E, eMSC-4C and eMSC-4H while caused a re- percentage of proliferating cells was determined by flow duction in the cell migration rate of only the eMSC- cytometry, quantifying FSChighVioletlow T cell blasts 4B line. (Fig. 6a, b). An almost complete abrogation of activated Regarding the preimplantation protein IFN-τ, it re- human T cell proliferation was observed when these duced cell motility of most cell lines (Fig. 5C). cells were co-cultured with either eMSC-1A (p < 0.0001) or eMSC-4H (p < 0.0001) (Fig. 6c). Immunomodulatory properties Although all eMSC immortalized lines showed MSC dif- Discussion ferentiation capacities, those obtained from the follicular We have successfully isolated, immortalized and charac- phase had unexpected features such as cytokeratin ex- terized eMSC from different phases of the bovine pression and epithelial morphology. Thus, we analysed oestrous cycle. The results of the present study clearly the immunomodulatory potential of two representative demonstrated that immortalized eMSC showed MSC lines: eMSC-1A with a pure mesenchymal pattern and properties such as adherence to plastic, high proliferative eMSC-4H with a phenotype consistent with MET. Co- capacity until passages 10–15 and surface expression of culture experiments with human T cells activated with CD44 and intracellular expression of vimentin, both anti-CD2/3/28 beads were established to analyse eMSC’s MSC markers. Moreover, all eMSC lines were negative Calle et al. Stem Cell Research & Therapy (2019) 10:23 Page 9 of 12 A C B Fig. 6 Measurement of T cell proliferation suppression capacity by eMSC. a, b Representative plots of non-stimulated and stimulated (anti-CD2/ CD3/CD28-coated microbeads) human T cell control conditions (in the absence of MSCs). c Percentage of proliferating stimulated T cells in the presence of eMSC-1A or eMSC-4H (X:X T:eMSC ratio ± 0.54 and ± 0.63 respectively); proliferative T cell were analysed by Violet CellTrace fluorescence loss after a 4.5-day culture. Results are reported as percentage of proliferating T cells relative to the proliferation of activated T cells in the absence of eMSCs (positive control). ****p < 0.0001 for haematopoietic specific markers CD34 and MHCII [11]. In both mouse and human, uterine regeneration and positive for pluripotency-related marker Pou5F1, a was shown to occur both by bone marrow-derived cells nuclear protein that is well known for pluripotency [10] and also by MET of resident cells [14, 15]. Cousins maintenance and is expressed in bovine MSCs [20]. Nu- et al. demonstrated that MET contributes to endometrial merous studies demonstrate that the ability to express epithelium after progesterone withdrawal-induced tissue alkaline phosphatase activity is also a pluripotency breakdown in a mouse model of menstruation [36]. This marker of stem cells [35]. In this report, immortalized finding is consistent with Yu et al. observations showing eMSC lines from different phases of the bovine oestrous that luteal hormone concentration and/or duration cycle showed positivity in AP activity. thresholds confer plasticity and reversibility of endomet- All eMSC lines from early and late luteal phases rial stromal phenotype via MET [37]. In bovine, proges- showed a fibroblast-like morphology characteristic of terone produced by the corpus luteum is essential for MSC and expressed vimentin. Interestingly, eMSC lines the establishment and maintenance of pregnancy. Rapid obtained from the follicular phase (absence of an em- regression of the cow’s corpus luteum is a key event in bryo) expressed cytokeratin and two of these lines the bovine oestrous cycle as it is responsible for the pre- showed a clear epithelial-like morphology. This observa- cipitous decrease in the blood concentration of proges- tion could reflect a mesenchymal to epithelial transition terone, the end of the cycle (follicular phase) [38]. Our (MET) of eMSC, a process in which mesenchymal cells eMSC lines established from the different phases of the are re-programmed, thereby gradually losing mesenchy- oestrous cycle show a clear mesenchymal pattern in the mal cell characteristics while gaining epithelial cell traits early luteal phase (eMSC-1A, corresponding to the Calle et al. Stem Cell Research & Therapy (2019) 10:23 Page 10 of 12 presence of the embryo in the oviduct) and late luteal (mesenchymal or epithelial). Although the eMSC lines ob- phase (eMSC-3A, eMSC-3D and eMSC-3E, correspond- tained from the follicular phase (absence of an embryo) ing to the presence of the embryo in the uterus), while expressed cytokeratin and two of these lines showed clear an apparent mesenchymal to epithelial transition state is epithelial type morphology, they also showed a high migra- observed in eMSC from the follicular phase (eMSC-4B, tion capacity and showed intracellular expression of eMSC-4C and eMSC-4D, and eMSC-4H corresponding vimentin. Vimentin is an important marker of mesenchy- to the regression of the corpus luteum if there was no mal to epithelial transition and a requisite regulator of embryo) since they express cytokeratin and in some mesenchymal cell migration. During MET, vimentin pro- cases show an epithelial morphology. Considering that tein and mRNA expression are downregulated as cell mo- both the epithelium and the bovine stroma present tility decreases and cells adopt epithelial characteristics oestrogen and progesterone receptors during the [40, 41]. Taking into account the above, MSCs show a cel- oestrous cycle [39], it would be interesting to analyse in lular plasticity characteristic of MET that allows them to future experiments the paracrine function performed by show characteristics of mesenchymal and epithelial cells. the eMSCs, thanks to the possible expression of hor- In all scenarios, either when the embryo is still in the ovi- mone receptors during the oestrous cycle. duct (phase 1), or when the embryo without zona pellucida The ability to differentiate in vitro to mesodermal is in the uterus (phase 3), or in case of regression of the lineages has been demonstrated for eMSC isolated corpus luteum as a consequence of the absence of an em- from the endometrium at the medium luteal phase bryo in the uterus (phase 4), most of the eMSCs present [20] or medium and late luteal phase [21], but it was an active migration towards a pro-inflammatory niche not known if the MSCs present in the uterus con- (Th-1) characterized by the presence of IL-1β and de- served such multipotency capacity during the entire scribed in the initial and final stages of human gesta- oestrous cycle. Our data on osteogenic and chondro- tion [29]. In contrast, most of the eMSCs from the genic induction with eMSC lines confirmed their mul- three scenarios described respond with a block in their tipotency, both in the early and late luteal phases as migration to IFN-τ, the major embryo-derived preg- well as in the follicular phase. nancy signal in bovine [31]. This is in clear contrast An important characteristic of MSCs is their ability to with the effect of this cytokine on bovine MSC derived migrate to damaged tissues or to inflammatory processes. from bone marrow, which are attracted by IFN-τ All eMSC lines showed migration capacity without cyto- (Calle et al., under revision). The combination of both kine stimulation. There was not a clear difference in terms IL-1 and IFN-τ signals would ensure the retention of of non-stimulated migration between eMSC lines from dif- eMSC in case of pregnancy, as well as further recruit- ferent oestrus phases or with different morphologies ment of MSC from the bloodstream, to ensure the Fig. 7 Model for bovine endometrial MSC: mesenchymal to epithelial transition during luteolysis and tropism to inflammation niche for immunomodulation Calle et al. Stem Cell Research & Therapy (2019) 10:23 Page 11 of 12 immunomodulation necessary in the mother for em- Consent for publication bryo survival. Not applicable. eMSCs could act there as primarily suppressive effec- Competing interests tors on Th1 cells by suppressing their proliferation, dif- The authors declare that they have no competing interests. ferentiation and activation [42, 43]. In accordance, both analysed lines; eMSC-1A, with a pure mesenchymal pat- Publisher’s Note tern and eMSC-4H with a phenotype consistent with Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. MET, showed the ability to suppress T cell proliferation in vitro. Author details 1 Departamento de Reproducción Animal, Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria (INIA), Avenida Puerta de Hierro 12, local Conclusion 10, 28040 Madrid, Spain. 2Departamento de Biología Molecular, UAM, Madrid, In sum, our results show that eMSC isolated and immor- Spain. 3REMAR Group and Nephrology Service, Germans Trias i Pujol Health Science Institute & University Hospital, UAB, Badalona, Spain. 4Department of talized from the uterus shared common MSC features, al- Cell Biology, Physiology and Immunology, Universitat Autònoma de though eMSCs isolated from the endometrium of the final Barcelona, Bellaterra, Spain. 5CBM-SO, Instituto de Investigación Sanitaria phase of the oestrous cycle show a phenotype compatible Princesa (IIS-IP), Madrid, Spain. with the mesenchymal-epithelium transition. Finally, the Received: 16 August 2018 Revised: 10 December 2018 migratory capacity of eMSCs towards an inflammatory Accepted: 27 December 2018 niche and their retention by interferon-τ secreted by the References embryo would ensure the immunomodulatory ambient 1. Wiltbank MC, Baez GM, Garcia-Guerra A, Toledo MZ, Monteiro PL, Melo LF, necessary for productive pregnancy (Fig. 7). et al. Pivotal periods for pregnancy loss during the first trimester of gestation in lactating dairy cows. Theriogenology. 2016;86:239–53. Abbreviations 2. Wolf E, Arnold GJ, Bauersachs S, Beier HM, Blum H, Einspanier R, et al. BSA: Bovine serum albumin; DMEM-LG: Dulbecco’s modified Eagle’s medium Embryo-maternal communication in bovine - strategies for deciphering a low glucose; eMSCs: Endometrial mesenchymal stem cells; EMT: Epithelial to complex cross-talk. Reprod Domest Anim. 2003;38:276–89. mesenchymal transition; HBSS: Hank’s balanced salt solution; IFN-τ: Interferon-τ; 3. Hue I, Degrelle SA, Campion E, Renard JP. Gene expression in elongating IL-1β: Interleukin-1β; LE: Luminal epithelium; MET: Mesenchymal to epithelial and gastrulating embryos from ruminants. Soc Reprod Fertil Suppl. 2007;64: transition; MHCII: Major histocompatibility complex II; PDT: Cell population 365–77. doubling time; POU5F1: POU class 5 homeobox 1; T/E: Trypsin–EDTA; 4. Spencer TE, Bazer FW. Uterine and placental factors regulating conceptus TAC: Tetrameric antibody complexes; TNF-α: Tumour necrosis factor-α growth in domestic animals. J Anim Sci. 2004;82(E-Suppl):E4–13. 5. Dunn CL, Kelly RW, Critchley HO. Decidualization of the human endometrial Acknowledgements stromal cell: an enigmatic transformation. Reprod BioMed Online. 2003;7: The authors thank Dr. Rizos from the Animal Reproduction Department, INIA 151–61. in Madrid, Spain, for his excellent assistance to determine oestrous cycle 6. Alam SM, Konno T, Dai G, Lu L, Wang D, Dunmore JH, et al. A uterine stages of the heifers. decidual cell cytokine ensures pregnancy-dependent adaptations to a physiological stressor. Development. 2007;134:407–15. 7. Li Q, Kannan A, Wang W, Demayo FJ, Taylor RN, Bagchi MK, et al. Bone Funding morphogenetic protein 2 functions via a conserved signaling pathway This work was supported by grants from the Spanish Ministerio de Economía involving Wnt4 to regulate uterine decidualization in the mouse and the Industria y Competitividad to Ramírez M.A. (AGL2015-70140-R) and European human. J Biol Chem. 2007;282:31725–32. Union’s Horizon 2020 Research and Innovation Programme under grant 8. Björkman N. Fine structure of the fetal-maternal area of exchange in the agreement N°731014 (Ramírez M.A.). epitheliochorial and endotheliochorial types of placentation. Acta Anat Suppl (Basel). 1973;61:1–22. Availability of data and materials 9. Johnson GA, Burghardt RC, Joyce MM, Spencer TE, Bazer FW, Pfarrer C, et al. The materials used and/or analysed during the current study are available Osteopontin expression in uterine stroma indicates a decidualization-like from the corresponding author on reasonable request. differentiation during ovine pregnancy. Biol Reprod. 2003;68:1951–8. 10. Du H, Taylor HS. Contribution of bone marrow-derived stem cells to Authors’ contributions endometrium and endometriosis. Stem Cells. 2007;25:2082–6. MAR contributed to the conception and design and financial support. 11. Hugo H, Ackland ML, Blick T, Lawrence MG, Clements JA, Williams ED, et al. AC and MAR contributed to the provision of study material, collection Epithelial--mesenchymal and mesenchymal--epithelial transitions in and assembly of data, data analysis and interpretation, manuscript writing and carcinoma progression. J Cell Physiol. 2007;213:374–83. final approval of manuscript; SLM contributed to the provision of study 12. Imakawa K, Bai R, Kusama K. Integration of molecules to construct the material; MYM contributed to the data analysis and interpretation, manuscript processes of conceptus implantation to the maternal endometrium. J Anim writing and final approval of manuscript; MMT contributed to the provision of Sci. 2018;96:3009–21. study material and critical review of manuscript; BFE contributed to the data 13. Uchida H, Maruyama T, Nishikawa-Uchida S, Oda H, Miyazaki K, Yamasaki A, interpretation and critical review of manuscript. All authors read and approved et al. Studies using an in vitro model show evidence of involvement of the final manuscript. epithelial-mesenchymal transition of human endometrial epithelial cells in human embryo implantation. J Biol Chem. 2012;287:4441–50. Ethics approval and consent to participate 14. Zhang XH, Liang X, Liang XH, Wang TS, Qi QR, Deng WB, et al. The Whole blood, used as source of T lymphocytes for in vitro allogenenic mesenchymal-epithelial transition during in vitro decidualization. Reprod proliferation experiments, was obtained from healthy donors after Sci. 2013;20:354–60. informed consent approved by the Germans Trias i Pujol Universitary 15. Patterson AL, Zhang L, Arango NA, Teixeira J, Pru JK. Mesenchymal-to- Hospital Ethics Committee. This article does not include any individual epithelial transition contributes to endometrial regeneration following patient’s data. natural and artificial decidualization. Stem Cells Dev. 2013;22:964–74. Heifer uterus and ovaries were collected from routine slaughtered at the 16. Prianishnikov VA. On the concept of stem cell and a model of functional- local abattoir. morphological structure of the endometrium. Contraception. 1978;18:213–23. Calle et al. Stem Cell Research & Therapy (2019) 10:23 Page 12 of 12 17. Chan RW, Schwab KE, Gargett CE. Clonogenicity of human endometrial 42. Ma C, Guo Y, Liu H, Wang K, Yang J, Li X, et al. Isolation and biological epithelial and stromal cells. Biol Reprod. 2004;70:1738–50. characterization of a novel type of pulmonary mesenchymal stem cells 18. Taylor HS. Endometrial cells derived from donor stem cells in bone marrow derived from Wuzhishan miniature pig embryo. Cell Biol Int. 2016;40:1041–9. transplant recipients. JAMA. 2004;292:81–5. 43. Cagliani J, Grande D, Molmenti EP, Miller EJ, Rilo HLR. Immunomodulation 19. Lara E, Rivera N, Rojas D, Rodríguez-Alvarez LL, Castro FO. Characterization by mesenchymal stromal cells and their clinical applications. J Stem Cell of mesenchymal stem cells in bovine endometrium during follicular phase Regen Biol. 2017;3:1–26. of oestrous cycle. Reprod Domest Anim. 2017;52:707–14. 20. Łupicka M, Bodek G, Shpigel N, Elnekave E, Korzekwa AJ. Identification of pluripotent cells in bovine uterus: in situ and in vitro studies. Reproduction. 2015;149:317–27. 21. de Moraes CN, Maia L, Dias MC, Dell’Aqua CP, da Mota LS, Chapwanya A, et al. Bovine endometrial cells: a source of mesenchymal stem/progenitor cells. Cell Biol Int. 2016;40:1332–9. 22. Schwab KE, Chan RW, Gargett CE. Putative stem cell activity of human endometrial epithelial and stromal cells during the menstrual cycle. Fertil Steril. 2005;84(Suppl 2):1124–30. 23. Wagner W, Horn P, Castoldi M, Diehlmann A, Bork S, Saffrich R, et al. Replicative senescence of mesenchymal stem cells: a continuous and organized process. PLoS One. 2008;3:e2213. 24. Ock SA, Baregundi Subbarao R, Lee YM, Lee JH, Jeon RH, Lee SL, et al. Comparison of immunomodulation properties of porcine mesenchymal stromal/stem cells derived from the bone marrow, adipose tissue, and dermal skin tissue. Stem Cells Int. 2016;2016:9581350. 25. Chow L, Johnson V, Coy J, Regan D, Dow S. Mechanisms of immune suppression utilized by canine adipose and bone marrow-derived mesenchymal stem cells. Stem Cells Dev. 2017;26:374–89. 26. Colbath AC, Dow SW, Phillips JN, McIlwraith CW, Goodrich LR. Autologous and allogeneic equine mesenchymal stem cells exhibit equivalent immunomodulatory properties in vitro. Stem Cells Dev. 2017;26:503–11. 27. Szekeres-Bartho J, Wegmann TG. A progesterone-dependent immunomodulatory protein alters the Th1/Th2 balance. J Reprod Immunol. 1996;31:81–95. 28. Saito S, Nakashima A, Shima T, Ito M. Th1/Th2/Th17 and regulatory T-cell paradigm in pregnancy. Am J Reprod Immunol. 2010;63:601–10. 29. Mor G, Cardenas I, Abrahams V, Guller S. Inflammation and pregnancy: the role of the immune system at the implantation site. Ann N Y Acad Sci. 2011;1221:80–7. 30. Perez-Sepulveda A, Torres MJ, Khoury M, Illanes SE. Innate immune system and preeclampsia. Front Immunol. 2014;5:244. 31. Spencer TE, Sandra O, Wolf E. Genes involved in conceptus-endometrial interactions in ruminants: insights from reductionism and thoughts on holistic approaches. Reproduction. 2008;135:165–79. 32. Roberts RM, Chen Y, Ezashi T, Walker AM. Interferons and the maternal- conceptus dialog in mammals. Semin Cell Dev Biol. 2008;19:170–7. 33. Miyamoto Y, Skarzynski DJ, Okuda K. Is tumor necrosis factor alpha a trigger for the initiation of endometrial prostaglandin F(2alpha) release at luteolysis in cattle. Biol Reprod. 2000;62:1109–15. 34. Calle A, Barrajón-Masa C, Gómez-Fidalgo E, Martín-Lluch M, Cruz-Vigo P, Sánchez-Sánchez R, et al. Iberian pig mesenchymal stem/stromal cells from dermal skin, abdominal and subcutaneous adipose tissues, and peripheral blood: in vitro characterization and migratory properties in inflammation. Stem Cell Res Ther. 2018;9:178. 35. Carlin R, Davis D, Weiss M, Schultz B, Troyer D. Expression of early transcription factors Oct-4, Sox-2 and Nanog by porcine umbilical cord (PUC) matrix cells. Reprod Biol Endocrinol. 2006;4:8. 36. Cousins FL, Murray A, Esnal A, Gibson DA, Critchley HO, Saunders PT. Evidence from a mouse model that epithelial cell migration and mesenchymal-epithelial transition contribute to rapid restoration of uterine tissue integrity during menstruation. PLoS One. 2014;9:e86378. 37. Yu J, Berga SL, Johnston-MacAnanny EB, Sidell N, Bagchi IC, Bagchi MK, et al. Endometrial stromal decidualization responds reversibly to hormone stimulation and withdrawal. Endocrinology. 2016;157:2432–46. 38. Gordon I, editor. Laboratory production of cattle embryos. 2nd ed. Wallingford: Cabi; 2003. 39. Kimmins S, MacLaren LA. Oestrous cycle and pregnancy effects on the distribution of oestrogen and progesterone receptors in bovine endometrium. Placenta. 2001;22:742–8. 40. Chaffer CL, Brennan JP, Slavin JL, Blick T, Thompson EW, Williams ED. Mesenchymal-to-epithelial transition facilitates bladder cancer metastasis: role of fibroblast growth factor receptor-2. Cancer Res. 2006;66:11271–8. 41. Mendez MG, Kojima S, Goldman RD. Vimentin induces changes in cell shape, motility, and adhesion during the epithelial to mesenchymal transition. FASEB J. 2010;24:1838–51.