Bone marrow mesenchymal stem cells (BM-MSCs) have been recently investigated as important source of undifferentiated cells in the field of regenerative medicine. As they can potentially differentiate into chondrocytes, adipocytes, osteocytes, myocardial and neural cells, they have become promising candidates as starting cells in tissue engineering studies aimed at improving wound healing and tendon tissue repair [35, 36]. The technique of equine MSC isolation from bone marrow, ex vivo culture and expansion, has been previously reported . Recently, Vidal described adipose-derive stem cell (ADSC) characteristics compared to BM-derived stem cells , mentioning their clinical relevance for tissue engineering applications in equine veterinary medicine, making adipose tissue a possible alternative source to BM.
Though ADSCs have been reported to display similar characteristics to BM-MSCs in humans in terms of gene expression profiles and phenotypes [39, 40], their potential for both chondrogenic differentiation in 3D culture and proteoglycan synthesis and thus their quality in tissue engineering is controversial [40, 41]. Winter and co-workers  found 90% reduced chondrogenic potential of ADSCs in micromass cultures compared to BM-MSCs and these observations were confirmed by Sakaguchi and co-workers , who analyzed various stem cell sources in chondrogenesis studies. Scientific literature has already shown that the methods to isolate BM-MSCs are faster than ADSCs and the yield in terms of purity is higher as in the latter case cells might be contaminated by mixed population of fibroblasts and adipocytes . Moreover, Richardson and co-workers  reported MSCs isolation from fat is riskier than from bone marrow as there is more donor-site morbidity because of the surgery.
In general, tendon lesions in horses have been treated with BM-MSCs . These cells have been largely used since they have been shown to be successful in repairing more specifically tendons and ligaments .
Adipose stem cell plasticity seems to be limited though compared to BM, as differentiation capabilities in tendon-like cells is still to be proven .
The focus of this study was to evaluate growth and differentiation of equine BM-MSCs. Adherent cells from bone marrow were found to actively proliferate in vitro and to maintain their morphological and growth characteristics for over 10 passages. Moreover, our results show that adherent equine BM-MSCs could proliferate rapidly in DMEM with FBS and EGF added with a doubling time of 1.8 day. This finding is comparable to previous studies that found an average cells doubling time of 1.4 days, with no significant difference in the doubling rate between foals and young horses BM-MSCs . In our study cells grown in standard media showed an elongated fibroblast-like morphology, large cell size and capabilities of continuously dividing as has been observed in bone marrow stroma and tissue-specific MSC cells from other species [45, 46]. RT-PCR analysis showed that Oct4, Nanog and Sox-2 genes were expressed by the stem cells. The following QRT-PCR showed these "stemness" markers were expressed at comparable level with the constitutive gene GAPDH. This data confirms that MSCs from bone marrow of adult animals also expresses embryonic stem cell markers.
Immunocytochemistry experiments confirmed that BM-MSCs expressed the embryo stem cell marker Oct4. These cells did not express the haematopoietic lineage marker CD34, thus confirming their characterisation as mesenchymal stem cells . Interestingly, so far only embryonic stem cells had been found to express Oct4 and Nanog . Recently, two papers reported the expression of Oct4 in equine umbilical cord cells [44, 48]. In our study we show that also bone marrow-derived mesenchymal stem cells do express both Oct4 and Nanog. Literature data have considered embryo and adult stem cell markers as part of two different groups of markers. Our discovery that horse bone marrow MSCs express Oct4 and Nanog, together with a documented expression of SSEA-1 in murine MSCs by Anjos-Afonso , might indicate that the distinction between embryo and adult markers is not so strict. Having assessed in the present study that embryo stem cell markers are also expressed by adult stem cells, further investigations in the field are needed to identify which still unknown markers are unique and peculiar for embryo stem cells and to characterize embryological patterns in terms of differentiation capabilities. It should be noted that in our study the expression of equine cell markers was assessed by a PCR based panel of specifically designed oligonucleotide primers, instead of using antigen/antibodies based flow cytometry. This panel was used not only to assess the stemness of bone marrow-derived cells, but also to investigate their differentiation in tenocytes. This DNA marker-based panel represents a powerful tool in equine stem cells research, as many positive stem-cell marker antibodies, so far described in other species, show little or no cross-reactivity and thus cannot be used in the horse [45, 46].
In this paper we show that horse BM-MSCs can be induced to differentiate in tenocytes. Following exposure to BMP-12 the BM-MSCs expressed two tendon-related markers, tenomodulin and decorin [31–33]. To rule out the possibility that BMP-12 had induced differentiation into other cell types, rather than tenocytes, P19 lipocalin expression was assayed. In the BM-MSCs no expression of this gene, which is known to be expressed in reproductive system tissues like endometrium and uterus  was assessed. The BM-MSCs also maintained their capability to differentiate into osteoblast lineage, which was confirmed by two different staining techniques to detect the presence of calcium deposits and positivity for alkaline phosphatase. In literature many reports have shown that MSCs have multilineage differentiation capabilities [7–10]. However, up to now there has been no report of tenocytes induction. Although the possibility of using transplanted mesenchymal stem cells for tissue repair has been suggested in rabbits , little it is known about capability of mesenchymal stem cells to differentiate into tissue-specific cell types in vivo. However, a combination of mechanical stimuli and proximity to tenocytes and tendon matrix are believed to be important as stimuli for differentiation into tendon cells, as shown by direct implantation of cells into the tendon. The transplantation of mesenchymal stem cells into injured tendons has been shown to promote tendon healing not only in laboratory animal models  but also in horses [51, 52].
PLGA (poly(lactic-co-glycolic acid) fibres have been used as scaffolds in therapeutic approaches to tendon repair, owing to their biodegradability and biocompatibility . However, 3D scaffolds have not shown any results in the case of tendon repair in horses. So far, many approaches have been investigated for improvements in tendon injury repair, but most are not completely understood and much further effort is necessary to develop the technology into a highly efficient treatment. The promise of functional tissue engineering to replace damaged organs or tissues has boosted research interest. At present, however, it is important to balance the understanding of our current limitations with a desire to progress the technology. The possibility to use MSCs that have been pre-differentiated into tendon cells for transplantation may represent a significant improvement over the use of undifferentiated cells.
For instance, there are some evidences of tumor induction by undifferentiated cells; more investigations in this matter, though, are needed .