Equine cellular therapy—from stall to bench to bedside?



Pioneering clinical stem cell research is being performed in the horse, a recipient of cutting edge veterinary medicine as well as a unique animal model, paving the way for human medical applications. Although demonstrable progress has been made on the clinical front, in vitro characterization of equine stem cells is still in comparatively early stages. To translate the promising results of clinical stem cell therapy in the horse, advances must be made in the characterization of equine stem cells. Aiming to improve communication between veterinarians and other natural scientists, this review gives an overview of veterinary “bedside” achievements, focusing on stem cell therapies in equine orthopedics as well as the current state of in vitro characterization of equine multipotent mesenchymal stromal cells (MSCs) and equine embryonic stem cells (ESCs). © 2012 International Society for Advancement of Cytometry

Stem cells have gained increased attention in human as well as veterinary biomedical research due to their potential as therapeutic agents for a broad range of diseases (1). The use of stem cell therapies in veterinary medicine has led to pioneering clinical stem cell research activities during the last decade (2–5). In the field of equine tendinopathy therapy, stem cell–based therapeutic approaches such as the intratendinous application of MSCs (2), ESCs (3) or embryonic-like fetal stem cells are being investigated (6). Moreover, several companies offer stem cell isolation and expansion for commercial use in equine practice, taking advantage of the growing demand for “stem cell therapy” by horse owners and equine practitioners.

According to present studies, stem cells seem to improve tendon regeneration (7, 8). However, there is still a lack of knowledge concerning the mechanisms of stem cell–mediated repair as well as the basic cellular properties of equine MSCs (9, 10). While many groups are working on the characterization of human or small laboratory animal stem cells, only a small number of veterinary scientists work in the field of equine stem cells. Furthermore, there are difficulties such as the lack of suitable immunological tools that impair the success and comparability of studies in “nonlaboratory” animals (10). Because of these handicaps, plastic-adherence and trilineage differentiation potential are still the properties available to most reliably define equine MSCs.

Large animals are essential preclinical models for human medicine (11). Unlike other large animals such as sheep or pigs that require an induced injury (12, 13), horses frequently suffer from naturally occurring orthopedic conditions. Consequently, the horse offers the unique opportunity to explore treatment strategies for musculoskeletal disorders under conditions that much more resemble the pathophysiology in human patients (14). Therefore, the use of horses as large model animals for musculoskeletal injury in people has been approved by the US Food and Drug Administration (FDA) (15). Before being able to take full advantage of the horse as a model animal in stem cell research, it is important to overcome difficulties encountered in equine cell characterization, so that the features of equine MSCs and ESCs can be properly defined and compared to those of their human counterparts.

This review provides insight into early and recent clinical stem cell research in the horse and documents the current state of equine stem cell characterization in vitro. The review includes both MSCs and ESCs since these cell types are currently most relevant for in vivo application in the equine model. Further, a section describing the role of extracellular matrix and macrophages in the regeneration of musculoskeletal tissues in other species is included as a thought-provoking exercise for equine science.


Orthopedic conditions such as tendinopathy and osteoarthritis are, due to their frequent natural incidence, the focus of clinical studies of equine stem cells. Both tendon and cartilage are tissues with poor intrinsic regenerative capacities, producing fibrous repair tissue once an injury or an inflammatory process has occurred. Fibrous tissue impairs the proper function of native tissue, and thus leads to recurring injuries with compromised athletic performance.

Therefore, current biomedical research aims at investigating therapeutic tools such as stem cells, growth factors or bioinductive scaffold material, which are expected to improve tissue regeneration (16).

Working Mechanism of Stem Cells

Stem cell therapy is based upon the concept that a multi or pluripotent cell population at the site of tissue injury can proliferate, differentiate, and spatially organize in a manner which results in structurally and functionally normal tissue. The methods by which such cells accumulate at the intended site are limited to exogenous placement or endogenous recruitment.

Exogeneous stem cell placement

Stem cells applied to the injury site were hypothesized either to differentiate into the required cell type and produce extracellular matrix themselves or to supply growth factors that stimulate cells residing near the healing tissue (17). Furthermore, MSCs display immunomodulatory properties which are thought to influence the healing process in a positive way, by reducing the inflammatory processes associated with the disease (9, 18). Supporting the first mentioned hypothesis, it was shown that GFP labeled MSCs injected into artificial equine tendon lesions assumed a tenocyte-like morphology and were orientated along the direction of mechanical force. Yet in the same study it appeared that survival or traceability of labeled MSCs in tendon lesions was poor (<5% at 10 days and 0.2% at 90 days post injection, respectively) (3), raising even more questions concerning the mechanism of action of MSCs in tendon regeneration.

It should be considered that many, but not all, research groups used autologous blood products such as plasma, serum, platelet-rich plasma, or bone marrow supernatant to suspend the MSCs prior to injection. This implies that autologous growth factors other than those synthesized by the injected cells are also transferred to the injury site and might contribute to the beneficial effects of the therapy.

Standardization of methodologies for cell harvest, culture, and application would increase comparability between different studies. Several studies aimed at improving harvesting and isolation procedures for equine MSCs from various tissue sources (19–24), yet no standardized protocols have been agreed upon. Further, equine stem cell research would benefit from a more exact definition of the cell types used in the clinical studies.

Extracellular matrix and macrophages in musculoskeletal regeneration

Extracellular matrix as source for chemotactic stimuli

The recruitment of endogenous stem and progenitor cells is less well understood, but is logically dependent upon appropriate chemotactic stimuli. The source of such chemotactic stimuli can either be the secreted products of cells (i.e., paracrine effects), release of bound factors from tissues following appropriate signals, or chemotactic signals/molecules produced at times of need by local tissue structures. It has been shown that degradation products of the extracellular matrix (ECM) possess a variety of biologic properties; among which is the ability to attract stem and progenitor cells (25, 26). Biologic scaffold materials composed of ECM have been shown to generate chemotactic molecules through scaffold degradative process, and the in vivo placement of such scaffolds, or derivatives of such scaffolds, have been associated with the accumulation of pluripotential cells at sites of injury (27–29).

The extracellular matrix (ECM) represents the secreted products of resident cells within each tissue and therefore ECM has unique tissue specific compositional and ultrastructural characteristics. These characteristics include cell friendly ligands and signaling molecules which direct tissue development (30, 31), and signals that can modulate the host innate immune response. Biologic scaffold materials composed of ECM, if prepared in a manner that retains native structure/function relationships, can provide a favorable substrate for cells with the potential to modify the default processes of wound healing. Rather than proinflammatory pathways leading to scar tissue formation, biologic scaffolds can promote a constructive remodeling process leading to functional tissue replacement (32, 33). The mechanisms by which the ECM alters the host healing response includes immunomodulation (34, 35), the release of sequestered growth factors (36–38), the provision of unique surface signal/ligands (39), and generation of degradation products that include chemotactic properties. The beneficial effects of these chemotactic properties have been most emphatically shown in the musculoskeletal systems where ECM enhances the recruitment of different populations of stem cell types and lengthens the period of time that stem cells are actively recruited to the site of remodeling.

Muscle possesses a remarkable capacity for regeneration, a trait likely due to the close proximity of stem cell populations such as muscle satellite cells and myoblasts (40, 41). In the event of muscle damage, these stem cell types are rapidly activated and migrate to the site of injury. Once at the site of injury, these cells proliferate, align with existing muscle, fuse together into multinucleate cells, and then form functional myotubes capable of contraction. However, several studies have shown that stem cells other than muscle satellite and myoblasts are capable of participating in both muscle regeneration and tendon repair when ECM scaffolds are implanted at the site of injury. Marrow derived cells populate the site of ECM scaffold remodeling and have been shown to differentiate into site appropriate tenocytes in a mouse model of Achilles tendon reconstruction (42). Although the number of marrow derived cells compared to other cells in the eventual remodeled tendon were low (less than 10%), this study showed that the concept of circulating cells participating in functional tendon reconstruction is feasible. It has also been shown that perivascular stem cells (25) are strongly attracted to degradation products of extracellular matrix (26, 43) and that physiologic mechanisms of scaffold degradation result in oligopeptides that are chemoattractant for this multipotential cell population (28, 44). In the absence of ECM scaffold placement, similar accumulations of these marrow derived cells and perivascular cells do not occur in sufficient numbers to constructively impact the functional remodeling outcome.

The role of macrophages in immunoskeletal regeneration

In addition to the ability of biologic scaffolds to positively influence the participation of endogenous stem and progenitor cells in musculotendinous tissue reconstruction, such scaffold materials have been shown to markedly alter the local innate immune response which contributes to tissue repair and scaffold remodeling. Macrophages consist of a phenotypically heterogeneous population of cells (45–48). The various macrophage phenotypes have distinctly different effects upon wound healing ranging from the long recognized proinflammatory events with associated secretion of chemokines and cytokines to a regulatory, noninflammatory constructive tissue remodeling response. Although there are many variants of these phenotypes, the simplest concept includes the M1 (proinflammatory) macrophages and the M2 (regulatory, tissue remodeling) phenotypes (46, 49, 50). Of relevance to the use of biologic scaffold materials in tissue reconstruction, it has been shown that ECM has a clear M2 modulatory effect upon macrophages (32, 33, 51). That is, when placed in vivo, the responding host macrophages show a dominant M2 phenotype at the site of scaffold remodeling. These M2 macrophages are associated with the formation of new innervated and vascularized functional skeletal muscle (49–53). In a recent review, the importance of a temporospatial transformation of early M1 macrophage responders to a later M2 macrophage phenotype for skeletal muscle regeneration was elegantly described (52). In fact, in the absence of macrophages, biologic scaffold remodeling is arrested (33) and constructive outcomes are eliminated. Such events should prompt a rethinking of long accepted “anti-inflammatory” therapies for musculotendinous injuries. The appropriate use of biologic scaffold materials has the potential to optimize the innate regenerative capability of musculoskeletal tissues.

Stem Cells in Equine Tendon Therapy

Adult stem cells in equine tendon therapy

Equine tendon or ligament injuries most commonly occur in the superficial digital flexor tendon (SDFT), a large, energy-storing tendon similar to the human Achilles tendon, or in the suspensory ligament (SL), a ligamentous muscle containing only small amounts of myofibers (54). Typical equine tendon lesions are located in the center of the SDFT and ultrasonographically appear as hypoechogenic areas, surrounded with more or less intact tendon tissue with similarities to human Achilles tendinopathy. The existence of these intratendinous lesions, filled with tissue that is less dense than the surrounding tendon tissue, facilitates a direct intralesional application of therapeuticals. Therefore, ultrasound-guided intralesional injection of a cell suspension is the common approach to introduce stem cells to the injury site, which has been used in all studies mentioned below (2–4, 6–8, 55–60).

In 2001, promising results were reported after application of autologous bone marrow into damaged SLs (61). In 2003, a more developed approach for the treatment of strain-induced SDFT injury was published which utilized autologous, culture expanded, bone marrow derived MSCs that were suspended in blood plasma for application (2). Re-examinations 10 days and 6 weeks after treatment showed no adverse effects such as lameness or changes in tendon substance, while the cross-sectional area of the tendon at maximum injury zone had decreased by 10% (2). Since then, this approach has been evaluated and improved upon in several clinical and experimental in vivo studies, but the basic therapeutic concept has remained unchanged.

Multiple MSC populations, ranging from culture-expanded bone marrow derived MSCs to culture-expanded adipose tissue or tendon derived MSCs (59, 60) have been used for cell therapy applications. So far, the clinical outcome after using these alternative MSC sources in tendon therapy seems to be comparable. Furthermore, with possible relevance for potential applications in human medicine, the use of the mononuclear cell (MNC) fractions without subsequent selection of MSCs by culture expansion is also being evaluated (7, 58). Moreover, the effects of allogeneic MSC application have been investigated, and no enhanced immunological response was found compared to autologous cell applications (4, 62, 63). However, due to species specific differences of the immune system, this approach may not be easily transferrable to human medicine.

Clinical results following treatment with autologous bone marrow derived MSCs are promising and re-injury rates are reduced compared to those observed after conventional therapy (8, 55, 56, 64). In the earliest of these studies, 9 out of 10 race horses that had received MSC treatment of SDFT core lesions could return to racing and did not reinjure within the observed period of 2 years. Ultrasonographic evaluation showed dense filling of the tendon lesions with parallel fibers (55) whereas horses in the control group (n = 15) showed ultrasonographic evidence of fibrosis within the tendon lesions, and all of those horses reinjured after a median time of 7 months (55). A clinical study including a larger number of cases was published in 2011 (8): 113 race horses had received an intratendinous injection of autologous MSCs suspended in bone marrow supernatant for treatment of SDFT lesions and could be followed up for a minimum of 3 years. One hundred eleven of these horses could return back to racing; however, horses that had been used otherwise (breeding, other sports, etc.) after the injury had already been excluded from the study. 27.4% of these 111 horses suffered a reinjury, which was significantly less than observed after conventional treatments (8). Analysis of our own clinical data revealed similar results as in the studies described above; only 15.5% of 58 horses with SDFT or SL lesions treated with MSCs had suffered a reinjury or had retired from sports within the follow-up period, while the other 84.5% had returned back to their previous level of performance (n = 43) or were in full training (n = 6) (56).

Besides the overall clinical outcome, these studies also showed that re-injuries tend to appear less frequently when a larger number of MSCs (at least ∼10 × 106 MSCs) is applied (8, 55), the patients are younger (8, 56) or when the time between the incidence of injury and cell application is shorter (8). Further, the athletic discipline for which the horse is used seems to have a major influence on treatment success (8, 56). However, it should be noted that MSC treatment does not replace or shorten the standard rehabilitation but is rather an additional treatment that improves the outcome.

Experimental in vivo studies lead to the assumption that this improvement might mainly be due to the fact that MSCs enhance the structural organization of the repaired tissue. Autologous adipose derived nucleated cell fractions, used in a collagenase-induced tendinopathy horse model, significantly improved tendon fiber architecture compared to the control group, as revealed by histology 6 weeks after cell injection (58). However, biochemical and molecular analyses of tendon compositions revealed no significant differences except for a higher cartilage oligomeric protein (COMP) expression in treated tendons (58). A comparably designed study conducted by the same group, using bone marrow derived MSCs and insulin-like growth factor-I gene-enhanced bone marrow derived MSCs, arrived at similar results. While histology scores were significantly better in both MSC treated groups compared to the control group, no changes in tendon composition could be found. Additionally, MSC treated tendons showed greater stiffness at mechanical testing than control group tendons, although this difference was not significant (57). In a study which directly compared the use of bone marrow derived MNCs or MSCs, it was found that both treatments equally enhanced tendon healing compared to the controls (7). Improvements were evident at ultrasonographic measurements as well as at the histological and immunocytochemical evaluations 21 weeks after cell application, revealing a better structural organization of the healing tissue, marked by dense filling of the defect and parallel alignment of collagen fibers (7). Interestingly, in this study, changes in tendon compositions could also be detected, with higher collagen I and lower collagen III contents in the treatment groups (7), which might be due to the fact that horses were euthanized after a longer period of time compared to the other studies described here (57, 58). Similarly, a recent study revealed that MSC treatment had no effect on collagen fibril size within the tendon lesions at 12 weeks posttreatment (65), although the light microscopical appearance of tendinopathy was shown to be improved as early as 6 weeks after cell application (58). Summarizing the results obtained in the horse so far, MSC application seems to improve tendon healing, as demonstrated by the decreased re-injury rates and the better structural organization of the healing tissue. However, it still remains unclear how the cells influence the healing process.

Embryonic and fetal stem cells in equine tendon therapy.

The current approach for embryonic stem cell therapy is to directly apply allogeneic, pluripotent embryonic or fetal derived cells (3, 6). It was shown that ESCs distribute more widely in the tissue and display significantly higher survival rates compared to MSCs after 90 days. Yet the histology and ultrasonographic assessment revealed no improvement of tendon architecture compared to the serum injected controls (3). In contrast, in a study using fetal derived embryonic-like cells, it was found that after 8 weeks, tendon architecture was significantly better than in controls, although again, no differences in tendon composition could be demonstrated (6). In neither of those studies, teratoma formation was observed. However, this might be unique to equine ESCs, as a previous study reported that these equine cells did not induce teratomas in SCID mice, either (66).

Stem Cells in Equine Osteoarthritis Therapy

Osteoarthritis, which can develop from several conditions such as osteochondrosis or acute trauma, is the most common joint disease in human as well as equine patients. The horse is considered a suitable large model animal for cartilage repair, as cartilage thickness of equine stifle joints is close to that in human beings (67). Still, in contrast to tendinopathy, few studies have investigated the effect of MSCs on equine cartilage regeneration.

One approach to introduce MSCs into focal chondral defects is to apply the cells using a scaffold or a comparable carrier material. It was shown that bone marrow MSCs in a self-polymerizing fibrin vehicle enhanced early chondrogenesis in arthroscopically created full thickness cartilage defects, with significantly better healing scores at second-look arthroscopy and a more cartilage-like histologic appearance of biopsies than the controls 30 days after implantation (5). However, post mortem assessment at 8 months revealed no significant beneficial effects of MSCs concerning tissue composition, showing similar DNA and proteoglycan content as well as collagen I and II content in both the MSC and control group. Still, histology scores and graft integration were slightly better in the MSC group (5).

Another approach is the intra-articular injection of a MSC suspension. Direct injection of MSCs seems more effective when larger areas within the joint are affected by osteoarthritis, given that the cells adhere to the lesion sites. However, few improvements could be demonstrated when using this approach in combination with microfracturing for repair of artificially created cartilage lesions, compared to injection of hyaluronan and microfracturing alone (68). Firmness and aggrecan content of the repair tissue in the MSC group were increased, but no other clinically or histologically evident effects of MSCs were observed after 6 or 12 months (68). Similarly, injection of bone marrow derived MSCs did not lead to major improvements of artificially created osteoarthritis over a period of 70 days. Yet, a significant decrease in prostaglandin E2 content in the synovial fluid was observed in this study (69), suggesting that immunomodulation is the most important mechanism of action of the MSCs in this early state of disease. Interestingly, this effect could not be observed when using the adipose derived stromal vascular fractions (69).

Although based on these preclinical studies, the evidence of beneficial effects of MSCs in the treatment of osteoarthritis is not yet convincing, this approach might be still useful for treatment of naturally occurring osteoarthritis. It was reported that 29 out of 40 horses suffering from joint disease, which had been treated with an injection of bone marrow MSCs and hyaluronic acid, could return back to their intended use (70). These first clinical data indicate that MSCs might positively influence the development of osteoarthritis, but certainly, more studies including a larger number of cases as well as control groups are needed.


Equine Mesenchymal Stromal Cells

Isolation of MSCs from equine bone marrow, their in vitro culture and chondrogenic differentiation, were first described in 1998 (71). Thereafter, few in vitro studies on equine MSCs have been published until the first clinical applications were described (2, 72). Since then, numbers of publications on equine MSCs have been increasing every year, exploring the use of several tissues as MSC sources and making efforts to characterize the properties of the respective MSCs (73–78).

In 2006, the international society for cellular therapy (ISCT) introduced the term “multipotent mesenchymal stromal cell” and proposed minimal criteria to define human MSCs: plastic-adherence when maintained under standard culture conditions, a panel of surface markers (CD105, CD73, and CD90) and lack of the expression of CD45, CD34, CD14, or CD11b, CD79alpha or CD19 and HDL-DR molecules, and trilineage differentiation potential into osteoblasts, adipocytes, and chondroblasts in vitro (79).

Immunophenotyping equine mesenchymal stromal cells

In the horse, plastic-adherence is the only common method used to separate the fibroblast-like MSCs from the other tissue cells. Analysis of surface marker expression of equine MSCs is still in an early stage, although it is a standard procedure when characterizing human or murine cells. This is mostly due to the lack of commercially available anti-horse antibodies (Abs), thus resulting in testing and identifying antibodies that crossreact with specific equine epitopes (75, 78, 80). The earlier publications report positive immunocytochemical staining of CD90, CD29, and CD44 in equine MSCs (74, 81), or the detection of CD90 and CD105 positive cells by flow cytometry (FCM) (82). However, it was not determined whether the utilized Abs had been evaluated regarding their specificity in equine material. More recent studies included the validation of Ab-specificity using appropriate positive and negative controls as well as more specific tools such as western blots and mRNA expression analysis (75–77, 78). The presence of CD29, CD44, and CD90, as well as the lack of CD11a/18 and CD45 could be demonstrated in cultured equine mesenchymal progenitor cells by FCM and on mRNA level, although the expression of these markers clearly varied in early cell cultures (75). The expression of CD73 and CD 105 was variable even in established cell cultures (78) and in two studies, CD73 could only be detected on mRNA level, but not by FCM (76, 77). In contrast, CD14, being among the negative markers defined by the ISCT, was detected in equine MSCs by FCM, but the associated mRNA was not expressed in appreciable quantities, suggesting a false positive FCM result (76). However, results of a recent study demonstrate that CD14 might not be an appropriate negative MSC marker, as selection of CD14 positive equine bone marrow derived mesenchymal progenitor cells led to higher yields of adherent cells than selection of their CD14 negative counterparts. Furthermore, in this study, high CD14 mean fluorescence intensity was detected in early cell cultures, which decreased over time but then remained constant at a lower level (83). The absence of other negative markers such as CD34 and CD45 seems to be confirmed (75–78). Table 1 summarizes the most relevant results of immunophenotyping cultured equine MSCs that have been published so far.

Table 1. Expression of potential stem cell markers by equine mesenchymal stromal cells (MSCs)
  • BM: bone marrow; AT: adipose tissue; UCM: umbilical cord matrix; UCB: umbilical cord blood; G: gingiva; PDL: periodontal ligament; ICC: immunocytochemistry; FCM: flow cytometry; RT-PCR: reverse transcription polymerase chain reaction.

  • a

    ISCT positive markers.

  • b

    ISCT negative markers.

CD105a − (77)+ (76,77,84)+ (76); − (77); (−) (85)+ (76,77,84)+ (82,87); − (23)+ (23,84) (+) (78,87) + (87)+ (84,88,89)(−) (85)+ (90)
CD73a − (77)+ (76,77)− (76,77,91)+ (76,77)− (82,87)  (+) (78); − (87) − (87)   
CD90a+ (74,81)+ (75,77,83)+ (75,76,77)+ (76,77,85,86,91,92)+ (76,77,86)+ (82,87) + (81)+ (78,87) + (87) + (85,93)+ (90)
CD29+ (81)+ (75,77)+ (75,77,84)+ (77)+ (77) + (23,84)+ (81)+ (78)  + (84,89)  
CD44+ (81)+ (75,83); − (77)+ (75,77,84)− (77); + (91,92)+ (77)+ (87)+ (23,84)+ (81)+ (78,87) + (87)+ (84,89)+ (93) 
CD13  + (77)− (92)+ (77)       + (93) 
CD106  + (77) + (77)         
CD146  + (77) + (77)+ (82)        
CD166    + (77) + (23)    + (89)  
MHC I/ HLA-ABC+ (81) + (84)  + (23,62)+ (23,84)+ (81)+ (62)  + (84)  
CD45b − (75,77)− (75,76,77)− (76,77,85)− (76,77)− (82,87)  − (78,87) − (87) − (93); (−) (85) 
CD34b − (77)− (76,77,84,94)− (76,77,85)− (76); + (77)− (23,82,87)− (23,84) − (87) − (87)(+) (89); − (84,88)− (93); (−) (85) 
CD14b− (81)+ (83)(+) (76)+ (76)(+) (76)− (87)− (23)− (81)− (87) − (87)   
CD79αb− (81)      − (81)− (78)     
CD 86     − (62)  − (62)     
MHC II/ HLA-DRb− (81) − (84)  − (62)− (84,23)− (81)− (62)  − (84)  
OCT4+ (84,94); − (81) + (94); (+) (95) + (86)+ (82,84)+ (23,82) + (96); − (81)+ (84) + (88,89)  
SOX2  + (94); (+) (95)        + (88)  
NANOG  + (94); (+) (95) + (86)      + (88)  
c-Kit      + (82)       
c-Myc      + (82,23)       
TRA-1-60(+) (84); −(81)    (+) (84)(+) (82); (−) (23) + (96); − (81)(+) (84) + (89)  
TRA-1-81− (81)       + (96); − (81)     
SSEA-1− (81)       + (96); − (81)     
SSEA-3− (81)     (+) (82); (−) (23) − (81)  + (89)  
SSEA-4+ (84); − (81)    + (84)+ (82); (+) (23) − (81)+ (84) + (89)  

When a variety of Abs were tested for reactivity and specificity with equine antigens, it was found that many of them, including even anti-horse CD13 Abs, did not label equine cells in this experimental setting (75). This finding seems to be alarming when considering that many studies working on equine cells might arrive at false negative results when performing FCM analysis of equine cells, if Abs are not explicitly validated for the respective experiment. Another study, in which 379 monoclonal Abs produced against human CD molecules were tested for crossreactivity with equine leukocytes, not only showed that false negative results are more than likely, but that false positive results also occasionally occur. After these 379 Abs were used for FCM in equine cells by 3 different laboratories, only 14 Abs remained that were regarded as truly crossreacting, as the staining patterns in human and equine cells were similar. While the majority of Abs were nonreactive, there were also several Abs that may have shown positive staining (80). However, these findings were not further validated on mRNA level. Abs for potential stem cell markers that have been used in the horse for FCM so far are listed in Table 2.

Table 2. Antibodies used in flow cytometry to characterize equine mesenchymal stromal cells (MSCs)
EpitopeAuthorsProviderCloneaSpecificityaReactivity with equine cells
  • BM: bone marrow; AT: adipose tissue; UCM: umbilical cord matrix; UCB: umbilical cord blood; A: amniotic membrane; AF: amnion fluid; PB: peripheral blood; G: gingiva; PDL: periodontal ligament.

  • a

    Clones and specificity are given only if stated in the respective literature.

CD105(23)Abcam  − (UCM-MSC)
(76,78)SerotecSN6Anti-human+ (endothelial cells, AT-MSC); (+) (UCB-MSC)
(77)R&D Systems Anti-human− (BM-MSC, AT-MSC)
(82)Becton Dickinson  + (UCM-MSC)
(85)Millipore Anti-human(−) (AT-MSC, PB-MSC)
(87)Beckman Coulter Anti-human+ (UCM-MSC, UCB-MSC, AF-MSC)
CD73(76)Becton DickinsonAD2Anti-human− (AT-MSC)
(77)Becton Dickinson Anti-human− (BM-MSC, AT-MSC)
(78)Abcam10f1Anti-human+ (lymphocytes); (+) (UCB-MSC)
(82)Becton Dickinson  − (UCM-MSC)
(87)Beckman Coulter Anti-human− (lymphocytes, UCM-MSC, UCB-MSC, AF-MSC)
(91)Abcam Anti-human+ (lymphocytes); − (AT-MSC)
CD90(75)VMRD Anti-canine+ (granulocytes, BM-MSC)
(76)R&D SystemsThy1-A1Anti-human+ (AT-MSC)
(77)Becton Dickinson Anti-human+ (BM-MSC, AT-MSC)
(78)VMRDDH24AAnti-canine+ (mononuclear cells, UCB-MSC)
(82)Chemicon  + (UCM-MSC)
(85)Invitrogen Anti-rat+ (AT-MSC, PB-MSC)
(86,93)Becton Dickinson5E10Anti-human+ (AT-MSC, PB-MSC)
(87)Beckman Coulter Anti-human+ (UCM-MSC, UCB-MSC, AF-MSC)
(91)VMRD Anti-equine+ (AT-MSC)
(92)Molecular probes Anti-rat+ (AT-MSC)
CD29(75)Becton Dickinson Anti-human+ (lymphocytes, monocytes, granulocytes, platelets, BM-MSC)
 (77)Caltag Laboratories Anti-human+ (BM-MSC, AT-MSC)
 (78)BiolegendTS2/16Anti-human+ (mononuclear cells, UCB-MSC)
CD44(75)SerotecCVS18Anti-equine+ (leukocytes, BM-MSC)
(77)Immunostep Research Anti-human− (BM-MSC, AT-MSC)
(78)Becton DickinsonIM7Anti-mouse+ (mononuclear cells, UCB-MSC)
(87)Beckman Coulter Anti-human+ (UCM-MSC, UCB-MSC, AF-MSC)
(91)Serotec Anti-equine+ (AT-MSC)
(92)Molecular probes Anti-equine+ (AT-MSC)
(93)Becton Dickinson515Anti-human+ (PB-MSC)
CD13(92)Molecular probes Anti-equine(−) (AT-MSC)
(93)Becton DickinsonWM15Anti-human+ (PB-MSC)
CD146(82)Chemicon  + (UCM-MSC)
CD45(75)VMRD Anti-bovine+ (granulocytes, lymphocytes); −(BM-MSC)
(76)R&D Systems2D1Anti-human− (AT-MSC)
(77)Becton Dickinson Anti-human− (BM-MSC, AT-MSC)
(78)SerotecF10-89-4Anti-human+ (mononuclear cells); −(UCB-MSC)
(82)Becton Dickinson  − (UCM-MSC)
(85)Millipore Anti-human(−) (AT-MSC, PB-MSC)
(87)Beckman Coulter Anti-human− (lymphocytes, UCM-MSC, UCB-MSC, AF-MSC)
(93)Becton Dickinson30-F11Anti-mouse− (PB-MSC)
CD34(23)Abcam  − (UCM-MSC)
(76)Santa Cruz Biotechnology43A1Anti-human− (AT-MSC)
(77)Becton Dickinson Anti-human− (BM-MSC, AT-MSC)
(82)Becton Dickinson  − (UCM-MSC)
(85)Millipore Anti-human(−) (AT-MSC, PB-MSC)
(87)Beckman Coulter Anti-human− (UCM-MSC, UCB-MSC, AF-MSC)
(93)SerotecMEC14.7Anti-mouse− (PB-MSC)
CD14(76)R&D Systems134620Anti-human+ (AT-MSC)
(83)B. Wagner, Cornell University105Anti-equine+ (neutrophils, monocytes, BM-MSC)
(87)Beckman Coulter Anti-human− (UCM-MSC, UCB-MSC, AF-MSC)
CD79alpha(78)SerotecHM57Anti-human+ (mononuclear cells); − (UCB-MSC)
CD 86(62)Becton DickinsonIT2.2 −(UCB-MSC, UCM-MSC)
MHC I(62)SerotecCVS22 + (UCB-MSC, UCM-MSC)
HLA-ABC(23)Chemicon  + (UCM-MSC)
MCH II(62)SerotecCVS20 − (UCB-MSC, UCM-MSC)
(82)Chemicon  − (UCM-MSC)
OCT4(23,82)Abcam  + (UCM-MSC)
c-Kit(82)Santa Cruz Biotechnology  + (UCM-MSC)
c-Myc(82)Santa Cruz Biotechnology  + (UCM-MSC)
(23)Abcam  + (UCM-MSC)
TRA-1-60(23)Abcam  (−) (UCM-MSC)
(82)Chemicon  (+) (UCM-MSC)
SSEA-3(23)Abcam  (−) (UCM-MSC)
(82)Chemicon  (+) (UCM-MSC)
SSEA-4(23)Abcam  (+) (UCM-MSC)
(82)Santa Cruz Biotechnology  + (UCM-MSC)

Because of the difficulties of finding cross-reactive Abs, it is not yet certain whether equine MSCs express or lack exactly the same surface markers as human or murine MSCs (10). Moreover, there may be differences in the expression profile between equine MSCs derived from different sources (77). As it is therefore still difficult to define a suitable surface marker set, plastic-adherence, cell morphology and trilineage differentiation potential are still the more reliable criteria to identify equine MSCs. Very recently, however, a potential marker set for equine umbilical cord blood MSCs was proposed, based on the characterization of these cells by FCM using Abs that had been evaluated for crossreactivity (78). This proposed set includes CD29, CD44 and CD90 as positive markers and MHC II, CD45, CD79α, and a monocyte marker as negative markers (78). Although questions remain concerning the suitability of CD73 and CD105 as positive markers for equine MSCs, the use of this marker set can be an important first step which will improve comparability between studies and facilitate translation into human medical applications.

Furthermore, the expression of markers associated with pluripotency by equine MSCs was reported, such as OCT4, TRA1-60, TRA1-81, and SSEA-1 in cord blood derived MSCs (96), OCT4 and SSEA4 in cord matrix derived MSCs (23, 82) and OCT4, SOX2, and NANOG in bone marrow derived MSCs (94). However, other authors could not detect expression of OCT4, SSEA-1, SSEA-3, SSEA-4, TRA1-60, or TRA1-81 in equine MSCs derived from bone marrow or cord blood (81). These discrepancies might be due to the different methods and reagents used for the assessment of the marker expression, which emphasizes once again the importance of a thorough validation and standardization of the protocols used. Furthermore, not only immunostaining, but also PCR results should be interpreted with caution, in particular because pseudogenes exist for NANOG and POU5F1 (OCT3/4 protein) (95). Low level gene expression of NANOG, SOX2, and POU5F1 was detected in equine fetal cells, adult bone marrow-derived mesenchymal progenitor cells and adult chondrocytes in contrast to induced pluripotent stem cells (iPS) (95), emphasizing the necessity of comparative quantification. Table 1 also gives an overview of the studies investigating the expression of pluripotency markers in equine MSCs.

Differentiation potential of equine mesenchymal stromal cells

Trilineage differentiation potential has been described for equine MSCs derived from bone marrow (74), adipose tissue (97, 98), peripheral blood (99, 100), umbilical cord blood (101), umbilical cord matrix (82), gingiva and periodontal ligament (90), amnion (88) and amnion fluid (84), although the extent of differentiation varied based on the cell source and the differentiation protocols used. Adipogenic differentiation of equine MSCs strongly depends on the components of the differentiation media (102). It was reported that adipogenic differentiation of cord blood or peripheral blood derived equine MSCs was poor when using standard media, but succeeded when rabbit serum was added (100, 101). Our own data showed that rabbit serum generally enhances adipogenic differentiation of equine MSCs, irrespective of the cell source (102). Osteogenic differentiation, however, seems to depend more on the origin of the MSCs. Bone marrow derived equine MSCs were reported to be most prone to osteogenic differentiation, whereas cord blood or matrix MSCs show a rather weak osteogenesis compared to MSCs from other sources (103, 104). However, hydroxyapatite scaffolds seeded with cord blood MSCs form a bone-like matrix (105). In contrast, chondrogenic differentiation potential of cord blood derived equine MSCs seems to be superior to MSCs from other sources (103, 106).

With regard to the clinical applications of equine MSCs, their tenogenic differentiation potential is of significant interest, yet tenogenic differentiation in vitro remains a major challenge. It was reported that tenogenic differentiation of equine MSCs was achieved by exposing the cells to BMP-12, and as the cells exhibited an elongated morphology, they expressed decorin and tenomodulin mRNA (94). However, it was shown that these markers are also expressed in other musculoskeletal tissues and therefore cannot be considered as specific tendon markers. Instead, the expression profile of collagen 1A2, scleraxis and tenascin-C was found useful to distinguish tenocytes from other musculoskeletal tissue cells (107). Our own data show that equine MSCs express not only collagen 1A2, but also scleraxis, even in an undifferentiated state, which makes quantitative analyses of these markers essential for assessment of tenogenic differentiation (103). Furthermore, we showed that there are differences in the expression levels of these genes between MSCs derived from different sources, potentially indicating that tendon, but also adipose or cord blood derived MSCs might display greater tenogenic differentiation potential than bone marrow derived MSCs (103).

Very few studies investigated the in vitro differentiation of equine MSCs into cell types of ecto- or endodermal lineages (82, 96). Neurogenic differentiation of equine umbilical cord matrix MSCs (82) and hepatogenic differentiation of umbilical cord blood derived MSCs (96) were reported, suggesting that MSCs derived from these birth-associated tissues might display more plasticity than other adult MSCs (82, 96).

Equine Embryonic Stem Cells

The first establishment of two equine embryonic stem-like cell lines by culturing the inner cell mass of day 6–7 blastocysts on a feeder layer was described in 2002 (108). The cells exhibited a morphology similar to murine or human ESCs, appeared to be immortal and expressed alkaline phosphatase activity, OCT4, STAT3, and, in contrast to human ESCs, SSEA1 but not SSEA3 or SSEA4. Further, these cells formed embryoid bodies in the absence of a feeder layer and were shown to differentiate into neuronal, hematopoietic and endothelial precursor cells in vitro (108).

Similar results were obtained in a study characterizing equine ESCs from day 7–8 blastocysts (66). This study also tested teratoma formation by injecting the ESCs into the testis of SCID/beige mice, but no evidence of teratoma formation could be observed after 8–10 weeks (66). A study characterizing equine iPS cell lines supported this finding: teratoma formation in NOD/SCID mice was so slow that it could not be observed until 6 months after iPS cell injection, unless the mice were treated with doxycycline for the first 4 weeks to prevent premature final differentiation of the injected cells (109). Although the reasons for these findings are not yet known, the slow teratoma formation of equine ESCs might represent an important difference to human or murine ESCs, and facilitate a safe clinical application of these promising cells in the horse. On the other hand if substantial differences in human or murine ESCs exist, it makes studies on ESCs in the horse incomparable to the potential clinical applications in humans.


Clinical stem cell research in the horse has led to many promising results. However, further in vivo and in vitro research is required to gain more in-depth understanding of equine stem cell biology and the mechanisms underlying the regenerative effects of the cells, before a translation of equine therapies into human medicine can safely be achieved. Here, we consider it important to aim for work that will benefit from the unique advantages the horse offers as a model animal. For this purpose, further investigation of basic equine stem cell characteristics, leading to uniform definitions of equine MSCs and ESCs, seems to be of major importance. A closer collaboration of veterinary and other natural scientists, as well as clinicians, should accelerate a rapid translation of equine therapies into humans.