Veterinary Research Communications

, Volume 35, Issue 2, pp 103–121

Comparison of equine bone marrow-, umbilical cord matrix and amniotic fluid-derived progenitor cells


    • Department of Veterinary Clinical Science, Equine Reproduction UnitUniversità degli Studi di Milano
  • Bruna Corradetti
    • Department of Biochemistry, Biology and GeneticsUniversità Politecnica delle Marche
  • Anna Lange Consiglio
    • Department of Veterinary Clinical Science, Equine Reproduction UnitUniversità degli Studi di Milano
  • Camilla Recordati
    • Mouse & Animal Pathology Laboratory (MAPLab)—FONDAZIONE FILARETE Viale Ortles 22/4
  • Elisa Bonacina
    • IRCCS Istituto Ortopedico GaleazziCell and Tissue Engineering Laboratory
  • Davide Bizzaro
    • Department of Biochemistry, Biology and GeneticsUniversità Politecnica delle Marche
  • Fausto Cremonesi
    • Department of Veterinary Clinical Science, Equine Reproduction UnitUniversità degli Studi di Milano
Original Article

DOI: 10.1007/s11259-010-9457-3

Cite this article as:
Lovati, A.B., Corradetti, B., Lange Consiglio, A. et al. Vet Res Commun (2011) 35: 103. doi:10.1007/s11259-010-9457-3


The aim of the study was to compare in vitro the stemness features of horse progenitor cells derived from bone marrow (BM-MSCs), amniotic fluid (AF-MSCs) and umbilical cord matrix (EUC-MSCs). It has been suggested that there may be a stem cell population within both umbilical cord matrix and amniotic fluid. However, little knowledge exists about the characteristics of these progenitor cells within these sources in the equine species. This study wanted to investigate an alternative and non-invasive stem cell source for the equine tissue engineering and to learn more about the properties of these cells for future cell banking. Bone marrow, umbilical cord and amniotic fluid samples were harvested from different horses. Cells were analyzed for proliferation, immunocytochemical, stem cell gene expression and multilineage plasticity. BM- and AF-MSCs took similar time to reach confluence and showed comparable plating efficiency. All cell lines expressed identical stem cell markers and capability to differentiate towards osteogenic lineage. Almost all cell lines differentiated into the adipogenic lineage as demonstrated by cytochemical staining, even if no adipose gene expression was detectable for AF-MSCs. AF- and EUC-MSCs showed a limited chondrogenic differentiation compared with BM-MSCs as demonstrated by histological and biochemical analyses. These findings suggest that AF-MSCs appeared to be a readily obtainable and highly proliferative cell line from an uninvasive source that may represent a good model system for stem cell biology. More studies are needed to investigate their multilineage potential. EUC-MSCs need to be further investigated regarding their particular behavior in vitro represented by spheroid formation.


Progenitor cellsEquineBone marrowUmbilical cord matrixAmniotic fluidCharacterization


The knowledge acquired in the last years from stem cell biology and regenerative medicine is leading to new ways of repairing injured tissues and organs. Mesenchymal stem cells, more recently described as mesenchymal stromal cells (MSCs), are fibroblast-like, highly proliferative, plastic-adherent cells with the ability to differentiate into various tissues as bone, cartilage and fat (Pittenger et al. 1999; Kolf et al. 2007).

MSCs have a great appeal for tissue engineering because of their high in vitro expansion potential, self-renewal capacity and multi-potentiality (Kolf et al. 2007; Sacchetti et al. 2007).

In veterinary medicine, cell therapy is used today in the treatment of equine orthopedic diseases, with particular attention to ligament and tendon injuries (Smith et al. 2003; Pacini et al. 2007; Wilke et al. 2007; Fortier and Smith 2008; Guest et al. 2008).

Currently, bone marrow derived mesenchymal stem cells (BM-MSCs) are favored in equine regenerative medicine. Considering the invasive procedure related to their availability in adequate quantities and at the right time, there is an increasing interest in investigating the presence of MSCs in adult and extra-embryonic sources, such as fetal membranes, amniotic fluid and umbilical cord matrix (Marcus and Woodbury 2008; Parolini et al. 2008; Secco et al. 2008). Nonetheless, the collection of extra-embryonic adnexa is safe, ethical and easy, whereas the collection of bone marrow, besides being invasive, has a number of cells and a differentiation capacity negatively correlated with the age of the donor (Mueller and Glowacki 2001; Fehrer and Lepperdinger 2005).

Since the initial identification of human umbilical cord mesenchymal stem cells, a limited number of studies have been conducted on their isolation and expansion and only two studies regarded equine umbilical cord matrix derived cells (Hoynowski et al. 2007; Passeri et al. 2009). The umbilical cord matrix or Wharton’s Jelly is a mucoid connective tissue that surrounds the two arteries and the vein present in this structure. It contains fibroblast-like cells, which show properties similar to the MSCs and may represent a rich source of primitive cells (Troyer and Weiss 2008; Petsa et al. 2009; Sarugaser et al. 2009). Besides sharing common surface markers with BM-MSCs, they also express low levels of embryonic stem cell markers like Oct4, SSEA-3, SSEA-4 and Nanog and according to these features; they may be classified between embryonic and adult stem cells (Carlin et al. 2006; Hoynowski et al. 2007). In vitro these cells are reported to differentiate toward the adipogenic, chondrogenic and osteogenic lineages. Furthermore, recent studies have shown that umbilical cord matrix MSCs are immune suppressive in mixed lymphocyte assays and inhibit T-cell proliferation (Jomura et al. 2007; Cho et al. 2008; Weiss et al. 2008). Allogenic transplantation with umbilical cord matrix MSCs is well tolerated in several experimental models and seems to avoid immunological rejection (Cho et al. 2008). Therefore, their clinical application has been considered promising for regenerative medicine in humans and animals.

Recently, MSCs have been described in the amniotic fluid of human being (In’t Anker et al. 2003; Delo et al. 2006; De Coppi et al. 2007). Based on their morphological and growth characteristics, amniotic fluid cells can be classified into three types: epithelioid, fibroblast-like and amniotic fluid-specific cells (Sessarego et al. 2008; Gucciardo et al. 2009). The amniotic fluid mesenchymal stem cells (AF-MSCs) are thought to originate from several fetal tissues, including skin, digestive, respiratory and urinary systems and are considered to be in an intermediate stage between embryonic stem cells and lineage-restricted adult stem cells (Prusa and Hengstschlager 2002; You et al. 2008; Da Sacco et al. 2010).

The presence of a sub-population of amniotic fluid cells with mesenchymal features, able to proliferate in vitro more rapidly than comparable fetal and adult cells, was described for the first time by Kaviani et al. (2001). These cells are described as broadly multipotent stem cells that can differentiate into a variety of cell types, such as adipogenic, osteogenic, myogenic, endothelial, neurogenic, and hepatogenic lineages, inclusive of all embryonic germ layers (De Coppi et al. 2007). The AF-MSCs express a number of surface markers characteristic of MSCs, such as CD29, CD44, CD90 and CD105 as well as stage-specific embryonic antigen SSEA-4, expressed also by embryonic stem cells and the transcription factor Oct-4, which is associated with the maintenance of the undifferentiated state and the pluripotency of embryonic stem cells (Pan et al. 2002). The AF-MSCs could be used for allogenic cell transplantation, but it remains possible that they would be rejected by the host immune system (Wang et al. 2006).

However, during pregnancy AF-MSCs are known to play a role in preventing rejection of the fetus and are thought to have low immunogenicity (Wang et al. 2006). The relative potency of these stem cell populations needs to be fully determined in the equine species (Park et al. 2010), thus further investigations are needed.

The aim of this study was to optimize the isolation and culture and to compare in vitro the stemness features of equine progenitor cells derived from the bone marrow, umbilical cord matrix and amniotic fluid.

Besides comparing the characteristics in terms of derivation, growth kinetics, phenotype, plasticity and multipotency, our study also highlights the possibilities of EUC- and AF-MSCs overcoming the limitations of BM-MSCs.

Materials and methods

Sample collection

Three bone marrow samples were obtained from three horses different for breed, gender and age (mean age: 7 ± 3 years) at the slaughterhouse intended for human food and unrelated to our experiments. About 30 mL of bone marrow were collected from the sternum by using a Jamshidi biopsy needle (10 cm; 11 G) (Byopsibell, Modena, Italy) into syringes containing 12 500UI/mL heparin (Schwarz Pharma S.p.A., Milan, Italy). The samples were stored at 4°C and processed within few hours.

Three amniotic fluid samples were collected into sterile syringes during the full-term delivery of three different mares. About 60 mL of amniotic fluid were aspirated directly from the amniotic sac protruding from the vulva before its spontaneous rupture.

The umbilical cords were obtained immediately after delivery. A cable tie was placed at the distal end of the cord (foal junction), a second tie was tightened above the placenta/cord junction, and then scissors were used to cut away the tie limited cord portion. The isolated umbilical cord was washed three times in sterile saline solution containing, 0.05% chlorhexidine (ICF s.r.l., Cremona, Italy) and 70% isopropanol (Sigma Aldrich, Milan, Italy), and then the cord was placed into an antibiotic solution with 2% penicillin/streptomycin (Sigma Aldrich, Milan, Italy), 2% gentamicin (Sigma Aldrich, Milan, Italy) and 1% amphotericin B (Sigma Aldrich, Milan, Italy). The amniotic fluid and umbilical cord samples were stored at 4°C and processed within 8 h.

Cell isolation and culture

BM-MSCs were isolated by centrifuging the bone marrow samples at 400Xg for 10 min, removing the supernatant, collecting the buffy layer of mononuclear cells and washing twice in Phosphate Buffered Saline (PBS).

Before seeding the primary culture (P0), the obtained mononuclear cells were counted using a Burker chamber with the Trypan Blue (Merck, Darmstadt, Germany) dye exclusion assay.

For expansion, 1000 cells/cm2 were seeded in Control Medium composed of high glucose Dulbecco’s Modified Eagle’s Medium (HG-DMEM) (Celbio-Corning, Milan, Italy) supplemented with 10% foetal bovine serum (FBS, Sigma Aldrich, Milan, Italy), 100U/mL penicillin (Sigma Aldrich, Milan, Italy), 100 μg/mL streptomycin (Sigma Aldrich, Milan, Italy), 0.25 μg/mL amphotericin B (Sigma Aldrich, Milan, Italy), 2 mM/L L-glutamine (Gibco, Milan, Italy) and 10 ng/mL of basic fibroblast growth factor (bFGF, Invitrogen, Milan, Italy).

AF-MSCs were isolated by centrifuging the samples of amniotic fluid at 400Xg for 10 min, removing the supernatant and collecting the pellet formed at the bottom of the tube. The pellet was washed three times in PBS and 1000 cells/cm2 were seeded in the same aforementioned Control Medium.

EUC-MSCs were isolated from the umbilical cord matrix after removing the blood vessels from the umbilical cord. The umbilical cord matrix was minced into 1–2 mm2 pieces and digested in HG-DMEM supplemented with 0.075% collagenase type I (Sigma Aldrich, Milan, Italy) at 37.5°C overnight (about 16–20 h). The digested solution was filtered through an 80 μm strainer (Millipore, Milan, Italy), centrifuged at 400Xg for 10 min and washed twice in PBS.

In this case, 1000 cells/cm2 were seeded in the aforementioned Control Medium but containing double dosages of penicillin, streptomycin and amphotericin B.

All cell cultures were maintained at 5% CO2, 90% humidity and 37.5°C for the experiments described below. The Control Medium, with standard dosage of antibiotics and antimycotic, was replaced after 48–72 h to remove non-adherent cells and successively replaced twice a week. Adherent cells were detached with 0.05% trypsin-EDTA (Celbio-Corning, Milan, Italy) just prior to reaching confluence and then reseeded for culture maintenance.

Colony forming unit assay

BM-, AF- and EUC-MSCs from each donor were seeded in Control Medium as primary culture (P0) at 10000 cells/cm2 and as third subculture (P3) at 650 cells/cm2 in 6 well-plates (Iwaki, Milan, Italy) for 8 days at 5% CO2, 90% humidity, and 37.5°C. Cells were then fixed with 10% formalin, stained with 1% methylene blue and counted under an Olympus BX51 microscope (Japan). Colonies, consisting of more than 16–20 nucleated cells, were counted and data are reported as plating efficiency (PE%), calculated as number of colonies/number of seeded cells × 100.

Cell population doubling

The proliferation capacity of BM-, AF- and EUC-MSCs was evaluated from the first passage (P1) to the sixth passage (P6) in all samples from the three tissue sources. The number of viable cells was counted by the Trypan blue dye exclusion method. The population doublings (PD) were obtained according to the formula CD = log (NH/Ni)/log2, PD = CD/CT, where Ni represents initial seeded cells, NH is the cell harvest number and CT is the culture time. The results are reported as PD per day (PD/day).

Mesenchymal stem cell characterization

Reverse transcription- polymerase chain reaction (RT-PCR)

RT-PCR analysis was carried out on undifferentiated cells isolated from the three tissues at P3 to evaluate some of the most common mesenchymal stem cell (CD44, CD29, and CD105) or hematopoietic (CD34) markers. The MHC class I (ELA-ABC) and class II (ELA-DR) were also investigated. Equine glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was employed as reference gene.

Equine specific primers were initially designed with Primer3 software (, based on NCBI Equus caballus available sequences or on mammal multi-aligned sequences, and manually improved. The designed primers are reported in Table 1.
Table 1

Oligonucleotide sequences used for RT-PCR analysis


Accession No.

Sequences (5′→3′)

Product size (bp)

Annealing temperature

Glyceraidehyde-3-phosphate dehydrogenase.(GAPDH)






Endoglin (CD105)






CD34 molecule (CD34)






Integrin β1 (CD29)






CD44 antigen (CD44)






Major histocompatibility complex I (ELA-ABC)






Major histocompatibility complex II (ELA-DR)






Osteocalcin (OCN)






Osteopontin (OPN)












Adiponectin (ADPN)






Total RNA from cultured cells was isolated using the TriZol Reagent kit (Invitrogen, Milan, Italy) according to the manufacturer’s instructions and treated with DNase (Sigma Aldrich, Milan, Italy). RNA concentration and purity were measured by a spectrophotometer (Nanodrop® ND1000, Thermo Scientific, USA) and cDNA was synthesized from 200 ng of total RNA using iScript retrotranscription kit (Sigma Aldrich, Milan, Italy). Qualitative PCR was performed in 25 μL final volume with DreamTaq™ DNA Polymerase (Fermentas Life Sciences, Burlington, Canada) at the following conditions: initial denaturation at 95°C for 2 min, 32 cycles at 95°C for 30 s (denaturation), 55–63°C for 15 s (annealing), 72°C for 30 s (elongation) and final elongation at 72°C for 5 min. Amplified PCR products were visualized with ethidium bromide on a 1.8% agarose gel.


Antibodies were chosen according to results obtained by Hoyonowsky (2007) and were used following the manufacturer’s instructions.

Undifferentiated cells used for the immunocytochemistry were grown in 24-wells (Iwaki, Milan, Italy) on coverslips and cultured for 7 days in Control Medium, then fixed in 3.7% paraformaldehyde (Sigma Aldrich, Milan, Italy) for 15 min at room temperature. Immunostaining of MSC surface and intracellular proteins was performed using the following primary antibodies (Abcam, Cambridge, UK): a mouse monoclonal antibody anti-SSEA-4 (15 μg/mL), a mouse monoclonal antibody anti-TRA-1-60 (15 μg/mL) and a rabbit polyclonal antibody anti-Oct-4 (dilution 1:100).

To detect the intracellular proteins, cells were permeabilized for 10 min at room temperature in 0.25% Triton-X100 (Sigma Aldrich, Milan, Italy) diluted in Tris saline buffer (TBS, 10 nM Tris–HCl, 150nM NaCl, pH 7.4; Sigma Aldrich, Milan, Italy). After washing 3 times, cells were blocked in PBS containing 2% bovine serum albumin (Sigma Aldrich, Milan, Italy) for 4 h at 4°C. Cells were incubated with primary antibodies overnight at 4°C. After washing 3 times, the cells were incubated with secondary antibodies conjugated to Alexa Fluor 488 (1:250 dilution, Invitrogen, Milan, Italy) for 2 h at 4°C. Nuclei were counterstained with Hoechst 33342 (1 mg/mL) diluted 1:100 in PBS for 15 min. Cells were washed several times in deionised H2O and examined by a fluorescence microscope (Olympus BX51, Japan). All experiments were performed on cells at P4.

Multilineage differentiation assay

BM-, AF- and EUC-MSCs underwent osteogenic, adipogenic and chondrogenic differentiation.

All experiments were performed in triplicate on three independent cell samples at P3. Every experiment consisted of a control and a treated group. In the treated group, cells were cultured in specific differentiation media. In the control group, cells were cultured in Control Medium.

Osteogenic differentiation

The cell monolayer, after reaching 65% of confluence, was cultured for 21 days in osteogenic medium composed of HG-DMEM supplemented with 10% FBS, 0.1 μM dexamethasone (Sigma Aldrich, Milan, Italy), 10 mM β-glycerophosphate (Sigma Aldrich, Milan, Italy), 250 μM ascorbic acid (Sigma Aldrich, Milan, Italy). Von Kossa staining (Sigma Aldrich, Milan, Italy) to determine the presence of mineralized matrix deposition, was used to evaluate osteogenic differentiation. Briefly, the medium was removed and cells were fixed using 10% formalin for 10 min at room temperature. Subsequently, cells were washed with distilled water and incubated with 5% silver nitrate for 60 min and then counter stained with safranin for 1 min.

RT-PCR, as previously described, was carried out to confirm the mRNA presence of osteocalcin (OCN), and osteopontin (OPN), on treated and undifferentiated control cells (Table 1). The efficiency of primers was validated on native equine bone.

Adipogenic differentiation

A monolayer of cells, after reaching 95% of confluence, was cultured for 25 days in adipogenic medium composed of HG-DMEM supplemented with 10% FBS, 1 μM dexamethasone, 10 μg/mL insulin (Sigma Aldrich, Milan, Italy), and 150 μM indomethacin (Sigma Aldrich, Milan, Italy). Oil Red O staining to determine the lipid droplet deposition (red coloured) evaluated the adipogenic differentiation. Briefly, the medium was removed and the cells were fixed using 10% formalin for 10 min at room temperature. Subsequently, the cells were washed with distilled water and incubated with a 0.4% Oil Red O solution (Sigma Aldrich, Milan, Italy) for 50 min.

RT-PCR was carried out to confirm the mRNA expression of leptin or adiponectin (ADPN) on treated and undifferentiated control cells (Table 1). The efficiency of primers was validated on native equine adipose tissue.

Chondrogenic differentiation

Cells were grown as high-density micromass cultures (pellets of 5 × 105 cells) for 15 days in serum-free medium composed of HG-DMEM, ITS+ (Sigma Aldrich, Milan, Italy), 0.1 mM sodium ascorbate 2-phosphate (Sigma Aldrich, Milan, Italy), 100 mM sodium pyruvate (Sigma Aldrich, Milan, Italy), 0.1 μM dexamethasone, 1 M HEPES (Sigma Aldrich, Milan, Italy), HSA 100X (Sigma Aldrich, Milan, Italy) and 10 ng/mL recombinant human transforming growth factor beta 1 (rhTGFβ-1, Peprotech, Milan, Italy). To assess the chondrogenic differentiation, pellets were evaluated through histology, viability assay, total DNA quantification and sulphated glycosaminoglycan (sGAG) quantification.

Histological analysis was performed by fixing pellets for 24 h in 10% buffered formalin, subsequently embedded in paraffin and sectioned at 4 μm. The sections were stained with haematoxylin and eosin (H&E) (Sigma Aldrich, Milan, Italy) to evaluate pellet morphology and with Alcian Blue (Sigma Aldrich, Milan, Italy) (pH 2.5) to evaluate GAG deposition.

Cell viability in the pellets was analyzed by MTT assay (Chemicon, Ca, USA), which estimates the activity of the enzyme dehydrogenase by converting the MTT compound [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyletrazolium bromide] into formazan into the mitochondria. MTT was measured with a spectrophotometer (Perking Elmer HTS 700 plus) at an absorbance lecture of 570 nm for each sample.

Total DNA and sGAG quantifications were carried out after pellet digestion with 100 μg/mL of papain (Worthington, Lakewood, NJ) in 0.4 mM dibasic sodium phosphate (pH 6.8), 10 mM NaEDTA, and 200 mM L-cysteine (Sigma Aldrich, Milan, Italy) at 60°C for 16 h.

Total DNA quantification within the pellets was carried out in triplicate by Quant-iT ™ PicoGreen assay (Invitrogen, Milan, Italy). Pellet total DNA was measured with a spectrophotometer (Perking Elmer HTS 700 plus) at 480 nm Ex and 520 nm Em. DNA content was derived from a standard curve of serial dilutions of calf thymus DNA.

Sulphated glycosaminoglycan synthesis within the pellets was measured in triplicate by Blyscan GAG assay (Biocolor Ltd., Magenta, Milan, Italy), using chondroitin-4-sulfate as a standard. Pellet total sGAG was measured with a spectrophotometer (Perking Elmer HTS 700 plus) at 650 nm Em and a fluorescent lecture was obtained for the samples. The sGAG content was expressed as μg sGAG normalized to μg DNA content (GAG/DNA).

Statistical analysis

All statistical analyses were performed using SAS 9.1.2 software package (SAS Institute, Inc.). Data are presented as mean ± SD. Three replicates for each experiment were performed and the results represent these replicates. One-way analysis of variance (ANOVA) for multiple comparisons or two tailed student t-test, whenever applicable, was used. A level of P < 0.05 was accepted as significant.


Plating efficiency

No statistical difference in plating efficiency was seen at P0 among the three lineages, whereas at P3 there was a significant difference between BM- and AF-MSCs versus EUC-MSCs, i.e. the CFU-F frequency of cells derived from bone marrow and amniotic fluid was higher than that of cells derived from the umbilical cord matrix (Table 2).
Table 2

Plating efficiency (PE%). Mean of PE% values of the three investigated cell lines at primary culture (P0) and at their third passage (P3)

Cell line

PE% at P0

PE% at P3


0.025 ± 0.004%

0.42 ± 0.15%


0.024 ± 0.0009%

0.22 ± 0.10%


0.030 ± 0.007%

0.085 ± 0.023%

Morphology and population doubling

Throughout the experimental period, BM-MSCs maintained a spindle-shaped fibroblast-like morphology. AF-MSCs observed during primary culture consisted of two types of cells: one was similar to fibroblasts and others were flat and circular, resembling epithelial cells. Starting from P2, the AF-MSCs became morphologically homogeneous exhibiting fibroblast-like morphology. EUC-MSCs in the early stage of the primary monolayer culture formed a morphologically homogeneous population of fibroblast-like cells and frequently formed spheroid clusters that appeared as three-dimensional structures (Fig. 1).
Fig. 1

Morphology of tissue derived progenitor cells at P0. Undifferentiated BM-MSCs (a), AF-MSCs (b) and EUC-MSCs (c) (Magnification: X 10; scale bar: 200 μm)

PD assay showed that both AF- and BM-MSCs were able to divide for an extensive period in vitro, whereas EUC-MSCs were affected by the trypsin detachment and they required much more time to re-adhere on plastic plate in the following passage. Comparing the PD/day values at each passage, no significant difference was found between BM-MSCs and AF-MSCs, whereas there was a significant difference between EUC-MSCs and the other two lines, with EUC-MSCs showing a longer doubling time at P1 and P2 (Fig. 2). The PD throughout all the analyzed passages further confirmed these results; a statistically significant difference was noted between the mean PD/day values of EUC-MSCs (0.555 ± 0.32) and both BM-MSCs (1.024 ± 0.24) and AF-MSCs (1.060 ± 0.34).
Fig. 2

Population doublings. PD/day comparison between BM-, AF- and EUC-MSCs from P1 to P6. Values express as means ± SD (n = 3); (*) P = 0.05, (**) P < 0.05

Mesenchymal stem cell characterization

RT-PCR characterization

RT-PCR analysis showed that undifferentiated MSCs from all three lines expressed CD105 mesenchymal marker, CD29 and CD44 adhesion molecules and were negative for haematopoietic marker CD34, which was expressed by PBMCs (data not shown). Moreover, they were positive for MHC I (ELA-ABC) and negative for MHC II (ELA-DR) (Fig. 3).
Fig. 3

RT-PCR analysis of undifferentiated MSCs. RT-PCR analysis of mesenchymal (CD44, CD105 and CD29), hematopoietic (CD34) specific gene expression on undifferentiated BM-, AF- and EUC-MSCs. Major Histocompatibility Complex (MHC) class I and II gene expression is also reported for BM-, AF- and EUC-MSCs. GAPDH was employed as reference gene

Immunocytochemical characterization

Based on immunocytochemical analysis, all cell lines expressed Oct-4, SSEA-4 and were weakly positive for TRA-1-60 (Fig. 4); a negative response appeared when cells were staining just with the secondary antibodies (Fig. 4 Supplement).
Fig. 4

Immunostaining characterization of BM-, AF- and EUC-MSCs. Cells were stained with antibodies directed against human Oct-4, SSEA-4 and TRA-1-60 (Magnification: 20X; scale bars: 20 μm)

Multilineage differentiation induction

Osteogenic differentiation

Exposure of all cell lines to osteogenic induction medium resulted in secretion of extracellular calcium crystals, identified by von Kossa staining after 21 days of osteogenic induction. BM-MSCs showed increased mineralization as compared to AF- and EUC-MSCs (Fig. 5).
Fig. 5

Cytochemical analysis of differentiation assay. BM-, AF- and EUC-MSC differentiation into osteoblasts and adipocytes. The panels display representative photomicrographs of undifferentiated cells as controls (cntr); osteogenic and adipogenic induction detected by von Kossa staining (vK) and by Oil Red O staining (ORO) respectively. Magnification X 10 (A, B, D, E, F, G) and X 20 (C, H, I); scale bars: 100 μm. In Fig. 5C, the box represents the enlargement of the lipid droplets

RT-PCR results showed, in all lines, the expression of the osteocalcin and osteopontin after 21 days of osteogenic induction with respect to the control groups (Fig. 6). Both markers were expressed in the equine native bone (data not shown).
Fig. 6

RT-PCR analysis of differentiated MSCs. a: RT-PCR analysis of osteogenic (osteocalcin – OCN and osteopontin – OPN) gene expression on BM-, AF- and EUC-MSCs. Increase in OCN and OPN gene expression respect to undifferentiated cells (controls) after 21 days of induction. b: RT-PCR analysis of adipogenic (adiponectin—ADPN and leptin) gene expression on BM-, AF- and EUC-MSCs. Leptin mRNA on BM-MSCs and adiponectin mRNA on EUC-MSCs was observed after 25 days of induction. Controls did not express any adipogenic marker. GAPDH is reported as reference gene

Adipogenic differentiation

All cell lines exposed to adipogenic inductive medium underwent morphological changes and appeared as round cells with cytosolic droplets, which were stained in red by the Oil Red O staining. After 25 days in adipogenic induction medium, the AF-MSCs had larger numbers of intracytoplasmic Oil Red O positive lipid droplets than BM-MSCs, but AF-MSCs had a higher detachment rate during the induction period when compared to BM-MSCs. During adipogenic differentiation, there was a reduction in size of the typical spheroid clusters formed by the EUC-MSCs and the surrounding monolayer of cells disappeared. These new compact clusters were positive by Oil Red O staining, as well as some isolated cells around the clusters, which contained a small number of lipid droplets (Fig. 5).

RT-PCR showed that the expression of leptin was induced after 25 days of adipogenic induction in BM-MSCs and was absent both in AF- and EUC-MSCs, while adiponectin mRNA was detectable only in EUC-MSCs after 10 days of induction. Expression of adipogenic markers was not observed in AF-MSCs or in the undifferentiated control groups (Fig. 6). Both markers were expressed in the equine native adipose tissue (data not shown).

Chondrogenic differentiation

Histological analyses showed that pellets from treated BM-MSCs underwent moderate and almost diffuse cartilaginous differentiation. They had a small, poorly differentiated hypercellular central area composed of sparse undifferentiated mesenchymal cells admixed with hematopoietic-like cells and necrotic debris, surrounded by a thick rim of spindle-shaped to rounded cells (occasionally within lacunae) embedded in abundant fibrillar to homogenous pale eosinophilic to pale basophilic extracellular matrix, with deposition of abundant GAGs. Control BM-MSC pellets were undifferentiated with no detectable extracellular matrix or deposition of GAGs.

Pellets from treated AF-MSCs underwent only occasional mild peripheral cartilaginous differentiation. They had a central area composed of hematopoietic-like cells and necrotic debris surrounded by a small rim of spindle-shaped to rounded cells embedded in moderate amount of fibrillar eosinophilic extracellular matrix, occasionally forming lacunae, with deposition of small amounts of GAGs. Control AF-MSC pellets were undifferentiated and had no detectable extracellular matrix or deposition of GAGs.

Pellets from treated EUC-MSCs remained undifferentiated. They had a central area composed of uncohesive rounded to spindle-shaped cells admixed with moderate amount of necrotic debris, peripherally surrounded by spindle-shaped cells embedded in moderate amount of fibrillar eosinophilic extracellular matrix with deposition of small amounts of GAGs. Control EUC-MSC pellets were usually undifferentiated and had no detectable extracellular matrix or deposition of GAGs, but occasionally they presented a moderate deposition of GAGs (spontaneous differentiation) (Fig. 7).
Fig. 7

BM-, F- and EUC-MSC pellet cultures. H&E and Alcian Blue staining of treated MSCs (a–b) and control groups (c–d). BM-MSC pellets with marked chondrogenic differentiation characterized by occasional lacunae containing chondrocytes embedded in abundant extracellular matrix with abundant GAGs. AF-MSC pellets with mild multifocal peripheral differentiation with deposition of small amounts of GAGs. EUC-MSC pellets remained undifferentiated with a poor peripheral deposition of GAGs. Control groups showed no chondrogenic differentiation, but deposition of GAGs was occasionally found in EUC-MSC pellets (Magnification: X 200; scale bars: 50 μm)

The MTT assay was useful to estimate the cell metabolic activity in chondrogenic treated pellets. There was a very significant difference in cell viability between treated BM-MSCs and both treated and control samples of AF- and EUC-MSCs (P < 0.001); moreover, treated AF-MSCs showed a significant difference versus their control group; no significant difference was noted between treated EUC-MSCs and their control group (Fig. 8).
Fig. 8

Cell viability values (MTT assay). The graph shows the comparison of cell viability between treated cells in chondrogenic medium and untreated cells in control medium (cntr). Bars bearing different letter(s) are significantly different. BM-MSCs were significantly different compared to their control and other cell lines for p < 0.001 (a). AF-MSCs were significantly different compared to their control for p < 0.05 (b). No significant difference was noted for EUC-MSCs versus their control, while a significant difference appeared between EUC-MSCs and BM-MSCs (p < 0.001)

The Quant-iT PicoGreen assay quantified the total amount of DNA in the chondrogenic pellets and was expressed as total nanograms. There was a very significant difference between treated BM-MSCs and both treated and control groups of AF- and EUC-MSCs (P < 0.001) and a mild difference between treated AF-MSCs and control EUC-MSCs (P < 0.05). No significant difference was noted for all the other groups, neither between treated groups nor between treated and control groups from AF-MSCs and EUC-MSCs (data not shown).

Sulphated glycosaminoglycan concentrations normalized to total DNA concentrations showed a significant difference between treated BM-MSCs and their control group and versus treated and control AF-MSCs (P < 0.001), no difference existed between treated and control EUC-MSCs, while a less difference is noted between treated BM- and EUC-MSCs (P < 0.05). No significant difference was noted among the other groups (Fig. 9).
Fig. 9

sGAG concentrations normalized to total DNA concentrations. Bars bearing different letter(s) are significantly different. BM-MSC treated group was considered statistically significant versus treated AF-MSCs and their controls for p < 0.001 (a) and versus both treated and control EUC-MSCs for p < 0.05 (b)


Mesenchymal stem cells are becoming a choice for cell-based therapies in regenerative medicine. The optimal cell dose for clinical applications is currently unknown, but it is likely that large amounts of MSCs would be needed for both immune-modulation and regenerative medicine. According to this, isolation and characterization of stem cells derived from different tissues represent an important issue for cell therapy (Hipp and Atala 2008).

The purpose of the study was to reflect upon the naturally occurring clinical situation of equine species where the choice would be either to store neonatal tissues and adult bone marrow for future autologous or allogenic clinical use.

According to this, we compared equine BM-MSCs and extra-embryonic derived cells as EUC- and AF-MSCs for their proliferation rate, characterization and differentiation ability to understand better the physiology of these progenitor cells in vitro. Extra-embryonic tissues are a viable, accessible and not risky source of MSCs. High initial sample volumes of amniotic fluid and umbilical cord could be easily collected at the delivery without any invasiveness for the mare or for the foal.

High numbers of viable nucleated cells from umbilical cord and amniotic fluid have been easily isolated, except for some umbilical cord failure samples occurring in fungal/bacterial contamination as previously reported by others (Ishige et al. 2009; Passeri et al. 2009). Our group overcame this problem by adding a double dose of antibiotics and antifungal agent in the primary culture, reset to the standard amounts during subsequent medium changes. Moreover, an interval < 12 h from sample collection to cell isolation was useful to decrease the risk of bacterial growth during MSC isolation.

The Wharton’s Jelly is the connective tissue of the umbilical cord and recently researchers have focused on it as a potential new source of MSCs both in humans and horses (Mueller and Glowacki 2001; Cremonesi et al. 2008; Barholomew et al. 2009; Passeri et al. 2009), but the efficient isolation of cells that truly express MSC characteristics has been somewhat controversial. To our knowledge, no studies have been so far performed on horse amniotic fluid derived progenitor cells, whereas it is considered a rich source of mesenchymal stem cells in human beings. Indeed, due to its contact with the developing fetus, amniotic fluid contains large numbers of suspended cells including stem cells (In’t Anker et al. 2003; Da Sacco et al. 2010).

In this study, we successfully isolated, expanded and compared the MSCs from sources, such as AF- and EUC-MSCs, with BM-MSCs.

During the primary culture, EUC-MSCs adhered to the plastic dish surface and possessed a spindle-shaped fibroblast-like morphology. After reaching confluence, some of these cells formed spheroid colonies that grew upward from the substratum surface and the growth of these cells occurred independently from contact inhibition. It is possible that these colonies represent more primitive progenitor cells, similar to the embryoid bodies found in cultures of embryonic stem cells (Reed and Johnson 2008) and amnion derived pluripotent stem cells (Miki and Strom 2006). This behavior has never been noted in AF- or in BM-MSCs that maintained a fibroblast-like morphology.

During the primary culture (P0), cells derived from all tissue sources exhibited a very similar CFU-F frequency; differently, at P3 both BM- and AF-MSCs showed a higher CFU-F frequency than EUC-MSCs. This could indicate that umbilical cord matrix is composed of a heterogeneous cell population and that the EUC-MSCs possess a lower potential as MSCs. A study reported that the immature cells that retain the ability to proliferate were located close to the amniotic surface, whereas highly differentiated, non-proliferating fibroblasts were located in closer proximity to the vessels (Nanaev et al. 1997). In our study, we isolated cells from the umbilical cord matrix without distinction between near or far from vessel’s wall portion, thus the cell population was very heterogeneous and this could explain the lower CFU-F frequency.

Regarding cell proliferation, it was observed that the BM- and AF-MSCs showed a similar fold expansion during all passages. EUC-MSCs showed a lower expansion rate at the early passages (P1 and P2), then a higher fold expansion at passage four, which reduced in the fifth and sixth passage. The EUC-MSCs, in fact, were affected by the trypsin detachment so they required a longer adaptation phase in monolayer culture.

Cells from all the three sources exhibited specific mesenchymal stem cell marker expression (CD29, CD44 and CD105) and MHC I (ELA-ABC), but lacked hematopoietic markers (CD34) and MHC II (ELA-DR) based on RT-PCR results.

To better characterize cell populations under experiment, we carried out immunophenotyping assays as by Hoynowski et al. (2007). Immunocytochemical analyses on all three cell types showed the presence of embryonic stem cell antigens such as Oct-4, SSEA-4, and TRA-1-60. These results suggest that AF- and EUC-MSCs represent an intermediate stage between pluripotent embryonic stem cells and lineage-restricted adult stem cells (De Coppi et al. 2007; Hipp and Atala 2008; You et al. 2008). Furthermore, also BM-MSCs possessed these factors as reported in other studies that detected Oct-4 in BM-MSCs (Pochampally et al. 2004; Zangrossi et al. 2007). This behavior could be probably explained by the culture medium components like basic FGF (Battula et al. 2007) or it could be due to the in vitro microenvironment that changes the typical niche behavior of these cells and makes them regress to a more embryonic unspecialized form. In fact, as in other cell populations, it occurs that BM-MSCs dedifferentiate in a progenitor cell subpopulation, which renders these cells sensitive to their microenvironment for subsequent differentiation (Barbero et al. 2003; De la Fuente et al. 2004). In our experiments, Oct-4 appeared in the cell cytoplasm and weakly in the nuclear compartment as demonstrated in literature (Miki and Strom 2006).

The obtained MSCs did not express MHC II antigens (ELA-DR), thus they could address several immunological issues in a clinical set up. However, further studies are needed to clarify these results for their potential use in applicative medicine.

According to the cytochemical analysis, MSCs from our three sources differentiated into osteogenic lineage, as confirmed by gene expression of osteocalcin and osteopontin respect to the control groups. Furthermore, a high intracellular lipid droplet accumulation upon exposure to the adipogenic medium was noted after Oil Red O staining in BM- and AF-MSCs. RT-PCR confirmed this result only for BM-MSCs that expressed leptin gene after 25 days of induction; whereas, adipogenic marker expression lacked in AF-MSCs. In contrast to these data, EUC-MSCs expressed adiponectin at 10 days of induction as shown by the RT-PCR analysis, but the detection of adipogenesis by Oil Red O staining was of doubtful evaluation. Chang et al. (2006) reported similar observations in human umbilical cord blood; whereas, robust lipid formation was evident by Oil Red O histology in equine umbilical cord blood cells cultured in a similar adipocyte induction medium (Koch et al. 2007). The discrepancy between these various reports may be attributed to the heterogeneity of the starting population and/or culture conditions.

For the tested chondrogenic conditions, our results indicate that cellular proliferation, viability and sGAG/DNA production by AF- and EUC-MSCs are lower compared to BM-MSCs in our culture system (Marcus and Woodbury 2008). This was confirmed by H&E and Alcian blue staining of pellets, in fact, BM-MSCs showed a stronger and more diffused chondrogenic differentiation compared to both AF- and EUC-MSCs, and this last one tended to a spontaneous differentiation in the control pellets. The EUC-MSC spontaneous chondrogenic transformation in control pellets could be due to the potential variation of proteins involved in genomic stability (Armesilla-Diaz et al. 2009) and embryonic stem cell antigens expressed in these cells. The sGAG concentrations in BM-MSCs samples were higher compared with both AF- and EUC-MSCs grown under the same conditions.

Compared to BM-MSCs, the AF- and EUC-MSCs produced less cartilaginous matrix after 15 days of pellet culture indicating either a slower or less robust chondrogenic differentiation response to TGFβ1. It is possible that an increased culture time or, according to the primitive nature of these cells, a different induction protocol may be needed to obtain robust chondrogenesis.

Based on scientific criteria, our data showed characteristics of mesenchymal stem cells for AF- and BM-MSCs, but further investigations will be needed to elucidate biological mechanisms involved in the active proliferation and cellular plasticity of the EUC-MSCs, also in dependence to their particular way to form aggregates similar to those formed in embryonic stem cell culture and their spontaneous chondrogenic differentiation when exposed to the Control Medium.


In conclusion, the ontological and anatomical origins of MSCs probably have a deep influence on their proliferation and differentiation capabilities, and hence affect their performance as cellular sources for tissue engineering.

This study provided a comparison between equine progenitor cells derived from extra-embryonic tissues, such as umbilical cord matrix and amniotic fluid, and bone marrow stromal cells.

The higher proliferative potential and plating efficiency of AF-MSCs, similar to the BM-MSCs, confers advantages for rapid expansion and consequent application. While with a more limited potential compared to BM-MSCs, AF-MSC capability to differentiate in multilineage classifies them as multipotent cells. Furthermore, the lack of intracellular ELA-DR should render them immunologically advantageous for future allogenic uses. These features suggest that AF-MSCs appeared to be a readily obtainable from a uninvasive source and may represent a good model system for stem cell biology and therapy.

Equine extra-embryonic mesenchymal stem cells may be useful in large animal injuries, and our study may be relevant in the development of regenerative cell based treatments.


The authors thank the Filarete Foundation and Galeazzi Orthopaedic Institute for financial support of this project and technical help with histology and chondrogenic differentiation analyses, respectively. B. Corradetti PhD fellowship is partially supported by a research grant from Università Politecnica delle Marche to Professor D. Bizzaro.

Conflict of interest

The authors declare that they have no conflict of interest.

Supplementary material

11259_2010_9457_MOESM1_ESM.tif (5.9 mb)
Fig. 4Immunostaining characterization of BM-, AF and EUC-MSCs. Cells were stained just with the secondary antibodies to evaluate their negative response (Magnification: 20X; scale bars: 20 μm) (TIFF 6008 kb)

Copyright information

© Springer Science+Business Media B.V. 2010