Isolation of equine bone marrow-derived mesenchymal stem
cells: A comparison between three protocols
C. BOURZAC, L. C. SMITH†, P. VINCENT, G. BEAUCHAMP, J.-P. LAVOIE and S. LAVERTY*
Département de Sciences Cliniques,†Département de Biomédecine vétérinaire, Faculté de Médecine Vétérinaire, Université de Montréal,
Keywords: horse; stem cells; multipotent stromal cells; Percoll; Ficoll; density gradient separation
Reason for performing the study: There is a need to assess
and standardise equine bone marrow (BM) mesenchymal
stem cell (MSC) isolation protocols in order to permit valid
comparisons between therapeutic trials at different sites.
Objective: To compare 3 protocols of equine BM MSC isolation:
adherence to a plastic culture dish (Classic) and 2 gradient
density separation protocols (Percoll and Ficoll).
Materials and methods: BM aspirates were harvested from
the sternum of 6 mares and MSCs isolated by all 3 protocols.
The cell viability after isolation, MSC yield, number of
MSCs attained after 14 days of culture and the functional
differentiation capacity) were determined for all 3 protocols.
Results: The mean ? s.d. MSC yield from the Percoll protocol
was significantly higher (6.8 ? 3.8%) than the Classic
protocol (1.3 ? 0.7%). The numbers of MSCs recovered after
14 days culture per 10 ml BM sample were 24.0 ? 12.1,
14.6 ? 9.5 and 4.1 ? 2.5 ¥ 106for the Percoll, Ficoll and
Classic protocols, respectively, significantly higher for the
Percoll compared with the Classic protocol. Importantly,
no significant difference in cell viability or in osteogenic or
chondrogenic differentiation was identified between the
protocols. At Passage 0, cells retrieved with the Ficoll protocol
had lower self-renewal capacity when compared with the
Classic protocol but there was no significant difference
between protocols at Passage 1. There were no significant
differences between the 3 protocols for the global frequencies
of CFUs at Passage 0 or 1.
Conclusions and clinical relevance: These data suggest that the
Percoll gradient density separation protocol was the best in
terms of MSC yield and self-renewal potential of the MSCs
retrieved and that MSCs retrieved with the Ficoll protocol
had the lowest self-renewal but only at passage 0. Then, the 3
protocols were equivalent. However, the Percoll protocol
should be considered for equine MSC isolation to minimise
Mesenchymal stem cells (MSCs), also known as multipotent
stromal cells or mesenchymal progenitor cells (Dominici et al.
2006) are of increasing interest in the regenerative medicine field.
They represent a heterogeneous population of highly proliferative
cells, with a characteristic of self-renewal and in vitro multilineage
differentiation capacity (Pittenger et al. 1999; Fortier 2005). They
can differentiate into lineages of mesenchymal tissues, such as
bone, cartilage, adipose tissue (AT), muscles and tendon (Tuan
et al. 2003). Although it was originally believed that they
participate in tissue homeostasis by replacing damaged or
senescent cells, it is now postulated that they contribute to healing
by playing a trophic role, producing cytokines and growth factors
Equine MSCs have now been evaluated for their therapeutic
potential in a variety of experimental models of musculoskeletal
injury: articular cartilage (Wilke et al. 2007; Frisbie et al. 2009)
and tendon (Del Bue et al. 2008; Guest et al. 2008; Nixon et al.
2008; Schnabel et al. 2009; Violini et al. 2009). Furthermore, they
are currently employed in cell therapy to enhance repair in horses
with clinical tendinitis (Smith et al. 2003; Smith 2008a,b) or
suspensory ligament desmitis (Herthel 2001; Rosenbrock et al.
2004; Clegg 2008), as these conditions are difficult to treat because
of poor intrinsic healing quality in these tissues in mature horses.
Preliminary results are promising.
Several tissue sources of equine MSCs, such as bone marrow
(BM) (Fortier et al. 1998; Koerner et al. 2006; Vidal et al. 2006;
Arnhold et al. 2007; Violini et al. 2009),AT (Vidal et al. 2007; Del
Bue et al. 2008; Nixon et al. 2008), peripheral blood (Koerner
et al. 2006) and umbilical cord blood (Koch et al. 2007; Reed and
Johnson 2008), have been investigated. For practical reasons, the
most common harvest method employed clinically is BM or AT
harvest, although umbilical cord blood banks are rapidly
expanding. BM-MSCs are easy to harvest, isolate and expand but
pneumopericardium is a potential severe, but rarely reported,
complication (Durando et al. 2006). AT-MSCs are also easy to
harvest but require a more invasive standing surgical procedure.AT
yields a greater quantity of MSCs per volume without expansion
when compared to BM, but their osteogenic and chondrogenic
differentiation capacity has been reported to be lower and they are
less studied than BM-MSCs (Vidal et al. 2007, 2008; Kisiday et al.
2008; Colleoni et al. 2009).
In man, BM-MSCs are currently under investigation in clinical
trials to treat patients with acute myocardial infarction (Assmus
et al. 2002; Wollert et al. 2004; Janssens et al. 2006; Schachinger
et al. 2006) with variable success rates reported to date. A recent
*Corresponding author email: firstname.lastname@example.org
[Paper received for publication 10.08.09; Accepted 25.10.09]
EQUINE VETERINARY JOURNAL
Equine vet. J. (2010) 42 (6) 519-527
© 2010 EVJ Ltd
demonstrated that cell numbers and viability were not sufficient
parameters to assess the quality of cells retrieved, as even small
changes in protocols could affect the functional capacity of cells.
Thus, they postulated that the clinical outcome after MSC therapy
could also be affected by the cell isolation protocol (Seeger et al.
2007). These findings highlight the need to assess and standardise
equine MSC isolation protocols before embarking on therapeutic
trials in order to compare results at different sites. Relatively little
information has been reported on equine MSCs to date and
information has been extrapolated from what is known in other
Three equine BM-MSC isolation protocols have been
described in the literature to date. The first, referred to as Classic
protocol herein, is based on the capacity of MSCs to adhere to
plastic culture dishes (Fortier et al. 1998; Arnhold et al. 2007).
Mononuclear cells (MNCs, including MSCs) contained in BM are
plated on plastic culture dishes and allowed to attach for 5 days.
When the culturemedium
haematopoietic cells are removed. After 14 days expansion,
adherent MSCs are detached (Fortier et al. 1998; Arnhold et al.
2007). Density gradient solutions separate the MNC fraction from
the red blood cells, granulocytes, platelets and myeloid immature
precursors (Fortier 2005; Lennon and Caplan 2006; Pountos et al.
2007). The BM sample is placed on a density column and,
following centrifugation, MNCs are harvested at the interface
between the plasma and gradient density solution bands (Fig 1). A
Ficoll solution (polysaccharide solution, density of 1.077) has been
frequently used for BM-MSC isolation (Smith et al. 2003; Vidal
et al. 2006; Pacini et al. 2007; Smith 2008b). However, a Percoll
solution (colloidal solution of silica particles, density of 1.084–1.
088) has also been recently employed in studies to isolate equine
MSCs (Hegewald et al. 2004; Wilke et al. 2007; Colleoni et al.
The objective of the present study was to compare 3 protocols
for equine BM-MSC isolation, the Classic protocol and 2 gradient
density protocols (Percoll and Ficoll), by evaluating MSC yield and
the functional characteristics (self-renewal and differentiation
capacity) of the retrieved population. It was hypothesised that the
gradient density protocols would allow recovery of a higher
number of MSCs than the Classic protocol and that MSC functional
characteristics would be affected by the isolation protocol used.
is changed, the nonadherent
Materials and methods
The experimental protocol was approved by the institutional animal
care and use committee.
Bone marrow aspirates were obtained from healthy Standardbred
mares (n = 6) from a research herd with a mean ? s.d. age of 6.4 ?
4.7 years and weighing 400–550 kg. They were kept in pasture and
had not been used for other research projects for at least 5 months
prior to the experiment.
Bone marrow harvest
Bone marrow aspirates were obtained as previously described
(Orsini and Divers 1998). Horses were sedated and the sternum was
aseptically prepared. Under local anesthesia, four 10 ml samples of
BM were collected, using a Jamshidi biopsy needle1in syringes
preloaded with 1000 iu heparin2. Horses were administered TMS3
(5 mg/kg bwt, per os) and phenylbutazone4(2.2 mg/kg bwt, per os)
prophylactically 1 h before the procedure.
Mesenchymal stem cell isolation
A 20 ml sample was allocated to the Classic protocol and a 10 ml
sample each to the Percoll and Ficoll separation techniques. The
experimental protocol is outlined in Figure 2.
Classic protocol: The BM samples were centrifuged (AllegraTM
X-12 R)5at 1000 g for 15 min. The supernatant was aspirated and
the pellet was washed twice with PBS6(modified from Fortier et al.
1998). Following centrifugation and resuspension in 10 ml of
culture medium (DMEM low glucose with pyruvate6, 10% FBS6
and 1% penicillin, streptomycin, amphotericin B and fungizone6),
20 ml aliquots of MNC (containing MSCs) were counted using a
haemacytometer (magnification ¥10). Viability of the MNCs was
determined by trypan blue dye staining7(Strober 2001). MNCs
were then plated in 100 mm plastic culture dishes8at a cell density
of 1.2 ¥ 106/cm2with 10 ml of culture medium (0.13 ml/cm2).
centrifugation at 1000 g for 15 min, the supernatant was aspirated
gradientseparation (Percollor Ficoll):
2-Mononuclear cell band
3-Gradient density solution
(red blood cell...)
Fig 1: Separation of mononuclear cells by density centrifugation in Percoll and Ficoll. 5 ml (Percoll) or 15 ml (Ficoll) of bone marrow samples were
layered over a 7 ml preformed continuous Percoll density gradient (in a polycarbonate tube) or a 7 ml Ficoll density gradient (in a polypropylene tube),
© 2010 EVJ Ltd
520 Assessment of equine MSC isolation methods
and the pellet was resuspended in culture medium to attain either a
2 ¥ 5 ml (Percoll, in polycarbonate tubes)9or 15 ml (Ficoll, in a
polypropylene tube)8stock cell solution and were layered over two
7 ml preformed continuous Percoll10or a 7 ml Ficoll10density
gradients, respectively. They were centrifuged at 400 g for 25 min
at 20°C. The MSC-enriched MNC population was then aspirated
with a Pasteur pipette (approximately 1.5 and 1 ml for the Percoll
and the Ficoll, respectively) and washed in culture medium by
further centrifugation at 350 g for 10 min at 20°C (modified from
Smith et al. 2003; Hegewald et al. 2004). The pellets were
resuspended in 10 ml of culture medium. Then, 20 ml aliquots were
aspirated for MNC counts. Viability of the MNCs was determined
as described above. MNCs were then plated in 100 mm plastic
culture dishes at a cell density of 200,000/cm2(0.13 ml/cm2).
In each protocol, cells were incubated at 37°C in a humidified
atmosphere containing 5% CO2, allowed to attach for 5 days, after
which the medium was changed every 2 days. After 80–90%
confluence was reached (14 days for each group), adherent cells
were washed twice in PBS, trypsinised6(Passage 1) and
centrifuged at 350 g for 10 min at 20°C. Cells were then
resuspended in 10 ml of culture medium. A 20 ml aliquot was
aspirated and MSCs were counted.
Protocol MSC yield
With all 3 protocols, as all cells were not plated for culture, the
adjusted number of MSCs recovered was calculated by multiplying
the cell yield of the protocol by the total number of MNCs in a
10 ml BM sample. As density gradient protocols involved 2 steps,
the MSC yield was calculated as the multiplication of cell yield of
Step 1 and Step 2 (Fig 3).
Colony-forming unit (CFU) assays
A CFU limit dilution assay assesses the potential for cell self-
renewal – a stem cell’s capacity to divide into 2 daughter cells,
one of which remains in an undifferentiated stem-like state, while
the other differentiates into a more specific cell (Dennis and Caplan
2004). Each colony derives from the division of a single cell
(Friedenstein et al. 1974). CFU assays were performed at Passage
0 (P0 = cells retrieved from the BM and cultured before
trypsinisation) and Passage 1 (P1 = cells retrieved from the first
trypsinisation), in 100 mm plastic culture dishes in triplicates.As a
lower frequency of MSCs was anticipated at P0, MNCs were plated
at a higher density than at P1. At P0, MNCs were seeded at
densities of 5 ¥ 104and 104cells per plastic culture dish, and
incubated in culture medium as described above. After 14 days,
adherent cells were washed twice in PBS, fixed and stained with
1% crystal violet7in 10% ethanol for 10 min. CFUs per plate were
counted macroscopically and the mean of triplicate dishes recorded
for each of the densities.
The frequency of CFU at P0 was estimated by dividing the
number of CFUs with the number of MNCs plated and expressed as
a percentage. The global frequency of CFUs at P0 was determined
for each protocol as the mean of the frequency of CFUs for the
different densities (Arnhold et al. 2007).
At P1, the seeding densities were 100, 50 and 10 cells per
plastic culture dish. They were incubated and then washed, fixed,
stained and CFUs, frequency of CFUs and global frequency of
CFUs at P1 determined as described at P0.
Tri-lineage cell differentiation
Tri-lineage cell differentiation (osteocyte, chondrocyte and
adipocyte) is a standard method for identification of MSCs and
stemness in vitro (Dominici et al. 2006). For each differentiation
assay, P1-MSCs in a control group (unstimulated cells) and
differentiation group (stimulated cells) were seeded in a 6-well
plate, in duplicates. The control groups were cultured with culture
medium described above.
Osteogenesis assay: An osteogenesis assay was performed as
described previously with minor adjustments (Tondreau et al.
2005; Wagner et al. 2005). P1-MSCs were plated at a density of
3 ¥ 103cells/cm2. At 24 h, osteogenic differentiation was induced
dexamethasone7, 10 mmol/l b-glycerophosphate7and 0.1 mmol/l
ascorbic acid7(0.40 ml/cm2) for 7 days. Cells were then fixed in
70% ethanol for 10 min at -20°C and stained with 5-bromo-4-
chloro-3-indolyl phosphate-nitro blue tetrazolim11for 5 min for
the detection of alkaline phosphatase (ALP) activity (Arnhold
et al. 2007).
with 0.1 mmol/l
Chondrogenesis assay: A chondrogenesis assay was performed as
described previously (Ahrens et al. 2004). P1-MSCs were plated
at the density of 5 ¥ 103cells/cm2. When they reached 80%
DAY0 1421 2839
Bone marrow harvest
MNC isolation and viability:
Classic, Percoll, Ficoll
Classic: 1.2×106 MNC/dish
Percoll, Ficoll: 2×105 MNC/
CFU plating at P0:
Classic, Percoll, Ficoll
50,000, 10,000 MNC/dish
Classic, Percoll, Ficoll
CFU at P0 count
Classic, Percoll, Ficoll
CFU plating at P1
Classic, Percoll, Ficoll
100, 50, 10 MNC/dish
End of osteogenic
End of chondrogenic
End of adipogenic
Fig 2: Experimental protocol timelines.
© 2010 EVJ Ltd
C. Bourzac et al. 521
confluence, chondrogenic differentiation was induced in culture
medium supplemented with 10 ng/ml TGF-b17and 100 nmol/l
dexamethasone (0.40 ml/cm2) for 7 days. Cells were then fixed and
stained with 0.8% toluidine blue7for 4 min for the detection of
highly sulphated proteoglycans (Arnhold et al. 2007).
Adipogenesis assay: An adipogenesis assay was performed as
previously described (Wagner et al. 2005). P1-MSCs were plated at
the density of 3 ¥ 103cells/cm2. When they reached 100%
confluence, cells were cultured in induction medium (culture
medium supplemented with 1 mmol/l dexamethasone, 60 mmol/l
indomethacin7, 0.5 mmol/l 3-isobutyl-1-methylxanthin7
10 mg/ml recombinant human insulin7, 0.40 ml/cm2) for 2 days
followed by maintenance medium (culture medium, with 10 mg/ml
insulin (0.40 ml/cm2) for 4 days. After 3 cycles of induction/
maintenance, cells were fixed with 10% neutral buffered formalin
for 10 min, then stained with 0.5% oil red O7in isopropyl
alcohol : distilled water (60 : 40) for 30 min for the detection of
neutral lipid droplets (Arnhold et al. 2007).
After staining, for each differentiation assay, in all the 3
protocols of MSC isolation, images of 10 random fields in each of
2 plates of the differentiation group (magnification x20) were
acquired. Positive stained areas were measured in each field
using NIH Image J software (http://rsb.info.nih.gov/ij/index.html)
(Rangan and Tesch 2007). The 10-field areas were added in each
plate and the mean of the 2 plates was recorded. After reviewing
images of osteogenic differentiation, cells were seen to aggregate in
nodules or rows. Positive stained rows and nodules were counted
in the 10 random fields in each plate and overall percentages of
positive stained rows and nodules were determined.
For comparisons between the 3 methods, a repeated-measures
linear model with the protocol (Classic, Percoll, Ficoll) as within-
subject factor was used. Post hoc Tukey’s test was employed
when a significant difference was determined for the model.
For comparisons between P0 and P1, a double repeated-measures
linear model with the protocol and time (P0 and P1) as
within-subject factors was used. Statistical significance was set
at a value of P?0.05. All statistical analysis was performed
with SAS v.9.112.
Density gradient protocols
Total MNCs recovered
in a 10 ml bone marrow sample
Step 1 = centrifugation
Yield 1 = MNCs recovered after centrifugation/Total MNCs
MNC recovered after
Step 2 = expansion
Yield 2 = MSCs recovered after culture/MNCs plated
MSCs recovered after
14 days of culture
Yield = MSCs recovered after culture/MNCs plated
MSCs recovered after
14 days of culture
Fig 3: Mesenchymal stem cell yield calculation for Classic and density gradient protocols (Percoll and Ficoll). In the classic protocol, the MSC yield was
calculated as the ratio of the number of MSCs recovered after 14 days of culture to the number of Mononuclear cells (MNCs) plated (14 day MSC
count/number of MNC plated). As density gradient protocols involved 2 steps (centrifugation of sample with a density gradient solution and expansion of
the recovered cells after centrifugation), the MSC yield was calculated as the multiplication of cell yield of step 1 and step 2. Step 1: MNC separated/MNC
in a 10 ml BM sample. Step 2: 14 day MSC count/ number of MNC plated).
© 2010 EVJ Ltd
522 Assessment of equine MSC isolation methods
Mesenchymal stem cell isolation
The total number (mean ? s.d.) of MNCs in an untreated 10 ml BM
sample was 360 ? 114 ¥ 106(Table 1).
The total number of MNCs recovered from a 10 ml BM
sample, immediately after density gradient separation protocols,
was significantly lower compared to naive BM samples
(P<0.0001). The total MNC numbers recorded were 119 ? 68 for
the Percoll protocol and 103 ? 62 ¥ 106for the Ficoll protocol but
this difference was not significant (Table 1).
Protocol MSC yield
The MSC yield was significantly higher with the Percoll protocol
(6.8 ? 3.8%), but not the Ficoll protocol (4.2 ? 3.1%), compared
to the Classic protocol (1.3 ? 0.7%) (P = 0.005 and P = 0.1,
respectively). No significant difference was detected between the
density gradients methods (Table 1).
culture were 4.1 ? 2.5, 24.0 ? 12.1 and 14.6 ? 9.5 ¥ 106
for the Classic, Percoll and Ficoll protocols, respectively. The
difference was statistically significant for Percoll vs. Classic
(P = 0.004) but not for Percoll vs. Ficoll or Ficoll vs. Classic
Interestingly, the expansion yield (step 2 yield; Fig 3) was
significantly higher with the Percoll protocol compared to the
Ficoll protocol (25.3 ? 15.5% and 16.9 ? 11.4%, respectively,
P = 0.04).
Cell viability was similar with all 3 isolation protocols
Colony-forming unit assays
Colony-forming unit assays were successful in 5 and 6 horses at P0
and P1, respectively. No result was obtained in one horse at P0
because of a probable contamination in plastic culture dishes
leading to cell death.
The self-renewal capacity of MSCs isolated with the Ficoll
protocol was significantly lower than that of MSCs isolated with
the Classic protocol at P0 (density of 50,000 cells/plastic culture
dish, P = 0.02 for both number and frequency), but was similar
between the 3 protocols at P1 (Fig 4).
There were no significant differences between the 3 protocols
for the global frequencies of CFU at P0 or P1 (Table 2). The global
frequency of CFUs at P1 was higher than at P0 for all 3 protocols,
as the MSC population was enriched by passage. However this
difference was only significant within the Classic (P = 0.0005) and
the Percoll protocols (P = 0.0004) (Table 2).
Osteogenic differentiation capacity: Osteogenic differentiation was
observed in the 5 horses (Fig 5a). Quantitative determination of
ALP positive-stained cell areas, the percentage of positive-stained
nodules or the percentage of positive stained rows revealed no
significant differences between protocols (Fig 5d).
differentiation was observed in the 5 horses (Fig 5b). Quantitative
determination of toluidine blue-stained cell areas revealed no
significant differences between protocols (Fig 5d). Stained cell
area was, however, greater in the osteogenic compared to the
chondrogenic assays as cells aggregated into dense nodules in
chondrogenic assays, as opposed to both nodules and larger rows in
the osteogenic assays (Fig 5a, 5b, right panel).
Adipogenic differentiation capacity: The initial protocol employed
did not yield satisfactory adipogenesis. We subsequently repeated
the protocol described earlier, but employing 15% rabbit serum
(Janderova et al. 2003) using BM aspirates from 2 additional
horses. Satisfactory adipogenesis was attained (Fig 5c) but as the
numbers of horses were inadequate, a statistical comparison was
The major finding of this study is that the number of MSCs
recovered following in vitro expansion of equine bone marrow
aspirates is significantly affected by the isolation protocol
employed. The number of colony forming units (CFUs) is also
significantly affected by protocol choice but, importantly, from a
clinical perspective, functional, or at least differentiation capacity,
does not appear to be affected by the isolation method selected.
Specifically, the number of MSCs recovered 14 days post
culture, combined with CFUs were isolation-protocol dependant.
Although the number of MSCs recovered with gradient density
protocols was higher than with the Classic protocol, only the
Percoll protocol resulted in a significantly higher number of MSCs
compared to the Classic protocol (an approximate 6-fold increase
in yield). These findings could be explained by the difference in
cell seeding methods: first, the MNCs in the Classic protocol were
seeded at a higher density than in the density gradient protocols
which could alter behaviour in culture; second, ‘contaminant’cells
such as red blood cells, granulocytes, platelets and myeloid
TABLE 1: Summary of equine mononuclear cell (MNC) and mesenchymal stem cell (MSC) yield and kinetics obtained from bone marrow samples with
different isolation protocols
Classic protocolPercoll protocol Ficoll protocolP value
Recovery of MNC (¥106)
MSC yield of the protocol (%)
Centrifugation yield of the protocol (%)
Expansion yield of the protocol (%)
Adjusted MSC number (from 10 ml BM sample) 14 days post culture (¥106)
*360 ? 114
1.3 ? 0.7
119 ? 68¶
6.8 ? 3.8¶
33.7 ? 16.8
25.3 ? 16.0§
24.0 ? 12.1¶
103 ? 62¶
4.2 ? 3.1
28.7 ? 15.9
16.9 ? 11.0
14.6 ? 9.5
99 ? 1
P = 0.005
P = 0.04
P = 0.004 4.1 ? 2.5
Comparison of the MSC isolation protocols from a 10 ml bone marrow aspirate (n = 6 horses). Data are expressed as mean ? s.d. *The total number of MNC
recovered in a 10 ml bone marrow sample was 360 ? 114 ¥ 106.¶Indicates statistically significant differences for Percoll or Ficoll vs. Classic protocols.§indicates
statistically significant differences for Percoll vs. Ficoll protocol.
© 2010 EVJ Ltd
C. Bourzac et al. 523
precursors are present with the former technique (Fortier 2005;
Lennon and Caplan 2006; Pountos et al. 2007), which could
potentially impair MSC multiplication. Harvest and expansion
methods previously described in the literature were followed in
the present study because the goal was to compare the currently
employed methods for future practical applications.
Although the difference was not significant, the number of
MNCs retrieved with the Percoll protocol was higher than with the
Ficoll protocol. The Percoll and Ficoll solutions investigated differ
in density: 1.088 and 1.077, respectively. Consequently, we
postulate that some MNCs may have separated in the Ficoll
gradient solution at sites other than the harvested site. Thus, it
would be of interest to quantify and characterise MNCs recovered
in each of the bands of the gradients to determine if they are mainly
MSCs and confirm this hypothesis.
The CFUs in the Ficoll protocol were less numerous than in the
expansion yield of the Ficoll protocol compared to the Percoll
protocol at P0. Our findings corroborate a previous investigation
of protocols to separate MSCs from human BM, comparing Percoll
and Ficoll as it was also similarly reported that the number of
CFUs were higher with the Percoll protocol compared to the Ficoll
protocol (Georgiou et al. 1983). In the latter study, however, a
discontinuous Percoll gradient was used and the Percoll solution
density was lower than that of the Ficoll. Similar numbers of CFUs
with sodium metrizoate (Vannier et al. 1980).As Percoll and Ficoll
solutions differ in composition, we postulate that the Ficoll solution
could diminish or retard their capacity to multiply when compared
to Percoll but this needs to be investigated further.
The number of colonies in the CFU assay at P1 was higher than
the number of cells that were seeded, particularly with Classic and
Percoll protocols.An explanation may be mechanical disruption of
colonies induced by culture medium changes. Some MSCs may
unstick and subsequently adhere, and create new satellite colonies
(clonicity is an inherent property of MSCs), though this remains to
be proven. This could also explain the significant difference in
global frequencies of CFUs identified between P0 and P1 with the
Classic and Percoll protocols compared to the Ficoll protocol. At
P1, although the global frequency of CFU in the Percoll protocol
was more than twice the Ficoll protocol, the difference was not
significant. As our statistical power was good, this means that the
effect of the isolation protocol on MSC multiplication capacity is
lost at P1.
In the present study, MSC characterisation was based on
adherence to plastic culture dishes, CFU numbers and trilineage
2.9 (± 1.5)
0.006 (± 0.003) %
2.3 (± 0.7)
0.004 (± 0.001) %
0.7 (± 0.8)
0.007 (± 0.008) %
0.5 (± 0.3)
0.005 (± 0.003) %
0.9 (± 0.6)¶
0.002 (± 0.001) %¶
65.6 (± 45.6)
66 (± 46) %
65.1 (± 19.6)
65 (± 20) %
31.8 (± 24.9)
32 (± 25) %
19.5 (± 15.9)
39 (± 31) %
43.0 (± 26.0)
86 (± 46) %
34.6 (± 9.1)
69 (± 18) %
7.4 (± 2.9)
74 (± 29) %
6.8 (± 5.8)
68 (± 58) %
3.1 (± 2.0)
31 (± 20) %
Fig 4: Colony-forming unit (CFU) assays for the 3 protocols of MSC isolation from a bone marrow sample. These images are representative of the mean
? s.d. number and frequency of colonies obtained in CFU assays (a) at P0 (stained at Day 14) and (b) at P1 (stained at Day 28) (n = 6 horses). Colonies
were counted macroscopically and the mean of the triplicates recorded for each of the densities in each horse.¶Indicates statistically significant differences
for Percoll or Ficoll vs. Classic protocols.
TABLE 2: Mean ? s.d. global frequencies of CFUs (%) at P0 and P1 for the
Classic, Percoll and Ficoll protocols
0.006 ? 0.005¥
70 ? 20
0.005 ? 0.002¥
73 ? 37
0.0009 ? 0.0006
34 ? 24
This table summarises the global frequencies of colony-forming units (CFUs)
at Passage 0 (P0, n = 5) and Passage 1 (P1, n = 6). The global frequency of
CFUs (%) was determined at each passage and for each protocol as the mean
of the frequencies of CFUs for the different cell plating densities of the
passage.¥indicates a statistically significant difference between the passages
within a protocol.
© 2010 EVJ Ltd
524 Assessment of equine MSC isolation methods
differentiation. Satisfactory adipogenesis was attained with the use
of 15% rabbit serum in the differentiation medium. Although
adipogenic differentiation of BM-MSCs has been achieved (Fortier
et al. 1998; Arnhold et al. 2007) without the addition of rabbit
serum in the horse, numerous other studies have reported that it is
required to achieve reliable adipogenic differentiation of BM,
umbilical cord blood and AT MSCs (Vidal et al. 2006, 2007; Koch
et al. 2007; Reed and Johnson 2008). Rabbit serum has also been
found to enhance adipogenesis in BM-MSC in vitro in man
(Diascro et al. 1998; Houghton et al. 1998; Janderova et al. 2003),
and rodents (Diascro et al. 1998).
Importantly, cell viability and differentiation to osteocytes and
chondrocytes were similar between the 3 protocols. From a
functional perspective, where cell therapy for musculoskeletal
conditions is the clinical goal, it is intuitively important that they
do not lose, or have reduced capacity to differentiate, as it
would seem a desirable property for tissue regeneration goals.
However, as it has also been hypothesised that stem cells’
principal role in healing is to provide trophic factors (Caplan
2009) to enhance healing, the ability to differentiate may not be
an essential character of these cells in respect to enhancing
Additional studies assessing gene expression (Menicanin et al.
2009) and proteomic profiles (Mareddy et al. 2009) of these cells
would be required to determine if cell function is altered by
the isolation protocols. Cell migration capacity in response to
cytokines is also an important step in the healing process and it has
been recently reported that MSC migration capacity is impaired in
vitro by the protocol used to isolate cells (Seeger et al. 2007). This
illustrates and underpins that isolation protocols have the potential
to alter important functional aspects of the cell therapy healing
process and should be taken into consideration when choosing
Although MSCs hold great promise for future stem cell-based
therapeutic strategies and indeed are currently employed clinically
in equine athletes, further research is required to understand their
mechanisms of action in order to effectively dose and enhance
these effects. There remains a huge knowledge gap in this field, as
there is still a lack of rigorous definitions, characterisation and
standardisation to permit meaningful comparisons to be made
between sites. Important essential knowledge is consequently
lacking on how to tailor therapy.
In conclusion, this study of 3 different protocols to isolate
equine MSCs has shown that the Percoll protocol had the highest
MSC yield and a better self-renewal capacity when compared to the
Ficoll protocol. MSCs retrieved with the Ficoll protocol had the
lowest self-renewal potential at Passage 0. Importantly, no
difference in differentiation capacity was shown between the 3
protocols. For these reasons, a Percoll isolation protocol should be
considered for equine MSC isolation.
The authors would like to thank Kathleen Théroux, Pascal
Fontaine, Nadine Bouchard and Hélène Richard for technical
assistance. Sheila Laverty is funded by the Canadian Arthritis
Network (CAN) National Sciences and Engineering Research
Council (NSERC) and the Canadian Institutes of Health Research
1Cardinal Heath Canada, Vaughan, Ontario, Canada.
2McKesson Canada, Saint Laurent, Quebec, Canada.
3Novopharm Animal Health, Scarborough, Ontario, Canada.
4Professional Veterinary Laboratories, Winnipeg, Manitoba, Canada.
5Beckman Coulter, Brea, California, USA.
6Invitrogen Canada Inc., Burlington, Ontario, Canada.
7Sigma-Aldrich Canada Ltd, Oakville, Ontario, Canada.
8Sarstedt Inc., Quebec, Canada.
9Nalge Nunc International, Rochester, New York, USA.
Positive stained area (mm2)
Fig 5: Multilineage differentiation of equine MSCs. These images are
representative of multilineage differentiation of equine MSCs for each of the
3 protocols. a) Osteogenesis and alkaline phosphatase activity staining.
Cells transformed to a polygonal morphology and aggregated into nodules
or rows that stained positive for ALP activity compared to the control
group. b) Chondrogenesis and toluidine blue staining. Cells aggregated
into nodules that gradually increased in size over the 2 week culture period
and stained positive for toluidine blue dye compared to the control group.
c) Adipogenesis and oil red O staining. Lipid droplets (stained in red)
accumulated within the cytoplasm over the 3 week culture period.
Adipogenesis was present in the control group (left panel) although it was
inferior to the differentiation group. Bars represented a scale of 400 mm.
d) Graph of positive stained areas measured with Image J software. There
was no difference in multilineage differentiation capacities between the
3 MSC isolation protocols.
© 2010 EVJ Ltd
C. Bourzac et al. 525
10VWR International, Ville Mont Royal, Quebec, Canada.
11Fisher Scientific Company, Ottawa, Ontario, Canada.
12SAS, Cary, North Carolina, USA.
Ahrens, N., Tormin, A., Paulus, M., Roosterman, D., Salama, A., Krenn, V., Neumann,
U. and Scheding, S. (2004) Mesenchymal stem cell content of human vertebral
bone marrow. Transplantation 78, 925-929.
Arnhold, S.J., Goletz, I., Klein, H., Stumpf, G., Beluche, L.A., Rohde, C., Addicks, K.
and Litzke, L.F. (2007) Isolation and characterization of bone marrow-derived
equine mesenchymal stem cells. Am. J. vet. Res. 68, 1095-1105.
Assmus, B., Schachinger, V., Teupe, C., Britten, M., Lehmann, R., Dobert, N.,
Grunwald, F., Aicher, A., Urbich, C., Martin, H., Hoelzer, D., Dimmeler, S. and
Zeiher, A. (2002) Transplantation of progenitor cells and regeneration
enhancement in acute myocardial infarction (TOPCARE-AMI). Circulation 106,
Caplan, A.I. (2009) Why are MSCs therapeutic? New data: new insight. J. Pathol. 217,
Clegg, P.D. (2008) Biological therapies for treatment of tendon and ligament injuries.
In: Proceedings of the 47th British Equine Veterinary Association Congress,
Equine Veterinary Journal Ltd, Newmarket. pp 263-264.
Colleoni, S., Bottani, E., Tessaro, I., Mari, G., Merlo, B., Romagnoli, N., Spadari, A.,
Galli, C. and Lazzari, G. (2009) Isolation, growth and differentiation of equine
mesenchymal stem cells: effect of donor, source, amount of tissue and
supplementation with basic fibroblast growth factor. Vet. Res. Commun. 33, 811-
Del Bue, M., Ricco, S., Ramoni, R., Conti, V., Gnudi, G. and Grolli, S. (2008) Equine
adipose-tissue derived mesenchymal stem cells and platelet concentrates: their
association in vitro and in vivo. Vet. Res. Commun. 32 (Suppl 1), S51-S55.
Dennis, J.E. and Caplan, A.I. (2004) Bone marrow mesenchymal stem cells. In: Stem
Cells Handbook, Ed: S. Sell, Humana Press Inc., Totowa. pp 107-117.
Diascro, D.D., Jr, Vogel, R.L., Johnson, T.E., Witherup, K.M., Pitzenberger, S.M.,
Rutledge, S.J., Prescott, D.J., Rodan, G.A. and Schmidt, A. (1998) High fatty acid
content in rabbit serum is responsible for the differentiation of osteoblasts into
adipocyte-like cells. J. Bone Miner. Res. 13, 96-106.
Dominici, M., Le Blanc, K., Mueller, I., Slaper-Cortenbach, I., Marini, F., Krause, D.,
Deans, R., Keating, A., Prockop, D. and Horwitz, E. (2006) Minimal criteria for
defining multipotent mesenchymal stromal cells. The International Society for
Cellular Therapy position statement. Cytotherapy 8, 315-317.
Durando, M.M., Zarucco, L., Schaer, T.P., Ross, M. and Reef, V.B. (2006)
Pneumopericardium in a horse secondary to sternal bone marrow aspiration.
Equine vet. Educ. 18, 75-79.
Fortier, L.A. (2005) Stem cells: classifications, controversies, and clinical
applications. Vet. Surg. 34, 415-423.
Fortier, L.A., Nixon, A.J., Williams, J. and Cable, C.S. (1998) Isolation and
chondrocytic differentiation of equine bone marrow-derived mesenchymal stem
cells. Am. J. vet. Res. 59, 1182-1187.
Friedenstein, A.J., Deriglasova, U.F., Kulagina, N.N., Panasuk, A.F., Rudakowa, S.F.,
Luria, E.A. and Ruadkow, I.A. (1974) Precursors for fibroblasts in different
populations of hematopoietic cells as detected by the in vitro colony assay method.
Exp. Hematol. 2, 83-92.
Frisbie, D.D., Kisiday, J.D., Kawcak, C.E., Werpy, N.M. and McIlwraith, C.W. (2009)
Evaluation of adipose-derived stromal vascular fraction or bone marrow-derived
mesenchymal stem cells for treatment of osteoarthritis. J. orthop. Res. 27, 1675-
Georgiou, G.M., Roberton, D.M., Ellis, W.M., Shen, B.J., Ekert, H. and Hosking, C.S.
(1983) CFU-c enrichment from human bone marrow using a discontinuous Percoll
gradient and soybean agglutinin in comparison with Ficoll-paque. Clin. Exp.
Immunol. 53, 491-496.
Guest, D.J., Smith, M.R. and Allen, W.R. (2008) Monitoring the fate of autologous and
allogeneic mesenchymal progenitor cells injected into the superficial digital flexor
tendon of horses: preliminary study. Equine vet. J. 40, 178-181.
Hegewald, A.A., Ringe, J., Bartel, J., Kruger, I., Notter, M., Barnewitz, D., Kaps, C.
and Sittinger, M. (2004) Hyaluronic acid and autologous synovial fluid induce
chondrogenic differentiation of equine mesenchymal stem cells: A preliminary
study. Tissue Cell 36, 431-438.
Herthel, D.J. (2001) Enhanced Suspensory Ligament Healing in 100 Horses by Stem
Cells and Other Bone Marrow Components. Proc. Am. Ass. equine Practnrs. 47,
Houghton, A., Oyajobi, B.O., Foster, G.A., Russell, R.G. and Stringer, B.M. (1998)
Immortalization of human marrow stromal cells by retroviral transduction with a
temperature sensitive oncogene: identification of bipotential precursor cells
capable of directed differentiation to either an osteoblast or adipocyte phenotype.
Bone 22, 7-16.
mesenchymal stem cells as an in vitro model for human adipogenesis. Obes. Res.
Janssens, S., Dubois, C., Bogaert, J., Theunissen, K., Deroose, C., Desmet, W.,
Kalantzi, M., Herbots, L., Sinnaeve, P., Dens, J., Maertens, J., Rademakers, F.,
Dymarkowski, S., Gheysens, O., Van Cleemput, J., Bormans, G., Nuyts, J.,
Belmans, A., Mortelmans, L., Boogaerts, M. and Van de Werf, F. (2006)
Autologous bone marrow-derived stem-cell transfer in patients with ST-segment
elevation myocardial infarction: double-blind, randomised controlled trial. Lancet
Kisiday, J.D., Kopesky, P.W., Evans, C.H., Grodzinsky, A.J., McIlwraith, C.W. and
Frisbie, D.D. (2008) Evaluation of adult equine bone marrow- and adipose-
derived progenitor cell chondrogenesis in hydrogel cultures. J. Orthop. Res. 26,
Koch, T.G., Heerkens, T., Thomsen, P.D. and Betts, D.H. (2007) Isolation of
mesenchymal stem cells from equine umbilical cord blood. BMC Biotechnol. 7,
Koerner, J., Nesic, D., Romero, J.D., Brehm, W., Mainil-Varlet, P. and Grogan, S.P.
(2006) Equine peripheral blood-derived progenitors in comparison to bone
marrow-derived mesenchymal stem cells. Stem. Cells (Dayton Ohio) 24, 1613-
Lennon, D.P. and Caplan, A.I. (2006) Isolation of human marrow-derived
mesenchymal stem cells. Exp. Hematol. 34, 1604-1605.
Mareddy, S., Broadbent, J., Crawford, R. and Xiao, Y. (2009) Proteomic profiling of
distinct clonal populations of bone marrow mesenchymal stem cells. J. Cell
Biochem. 106, 776-786.
Menicanin, D., Bartold, P.M., Zannettino, A.C. and Gronthos, S. (2009) Genomic
profiling of mesenchymal stem cells. Stem. Cell Rev. 5, 36-50.
Nixon, A.J., Dahlgren, L.A., Haupt, J.L., Yeager, A.E. and Ward, D.L. (2008) Effect
of adipose-derived nucleated cell fractions on tendon repair in horses with
collagenase-induced tendinitis. Am. J. vet. Res. 69, 928-937.
Orsini, J.A. and Divers, T.J. (1998) Bone marrow biopsy. In: Manual of Equine
Emergencies: Treatment and Procedures, 1st edn., Ed: J.A. Orsini and T.J. Divers,
W.B. Saunders, Philadelphia. pp 27-28.
Pacini, S., Spinabella, S., Trombi, L., Fazzi, R., Galimberti, S., Dini, F., Carlucci, F.
and Petrini, M. (2007) Suspension of bone marrow-derived undifferentiated
mesenchymal stromal cells for repair of superficial digital flexor tendon in race
horses. Tissue Eng. 13, 2949-2955.
Pittenger, M.F., Mackay, A.M., Beck, S.C., Jaiswal, R.K., Douglas, R., Mosca, J.D.,
Moorman, M.A., Simonetti, D.W., Craig, S. and Marshak, D. (1999) Multilineage
potential of adult human mesenchymal stem cells. Science 284, 143-147.
Pountos, I., Corscadden, D., Emery, P. and Giannoudis, P.V. (2007) Mesenchymal
stem cell tissue engineering: techniques for isolation, expansion and application.
Injury 38 (Suppl 4), S23-S33.
Rangan, G.K. and Tesch, G.H. (2007) Quantification of renal pathology by image
analysis. Nephrol. 12, 553-558.
Reed, S.A. and Johnson, S.E. (2008) Equine umbilical cord blood contains a
population of stem cells that express Oct4 and differentiate into mesodermal and
endodermal cell types. J. Cell Physiol. 215, 329-336.
Rosenbrock, A., Jacobi, R., Jaugstetter, H. and Brehm, W. (2004) Autologous bone
marrow transplantation to stimulate suspensory ligament regeneration in 24 horses
– A clinical case control study. In: 12th ESVOT Congress Proceedings, Eds:
A. Vezzoni and M. Schramme, Munich. pp 264-265.
Schachinger, V., Erbs, S., Elsasser, A., Haberbosch, W., Hambrecht, R.,
Holschermann, H., Yu, J., Corti, R., Mathey, D.G., Hamm, C.W., Suselbeck, T.,
Assmus, B., Tonn, T., Dimmeler, S. and Zeiher, A. (2006) Intracoronary bone
marrow-derived progenitor cells in acute myocardial infarction. N. Engl. J. Med.
Schnabel, L.V., Lynch, M.E., van der Meulen, M.C., Yeager, A.E., Kornatowski,
M.A. and Nixon, A.J. (2009) Mesenchymal stem cells and insulin-like growth
factor-I gene-enhanced mesenchymal stem cells improve structural aspects of
healing in equine flexor digitorum superficialis tendons. J. orthop. Res. 27, 1392-
Seeger, F.H., Tonn, T., Krzossok, N., Zeiher, A.M. and Dimmeler, S. (2007) Cell
isolation procedures matter: a comparison of different isolation protocols of bone
marrow mononuclear cells used for cell therapy in patients with acute myocardial
infarction. Eur. Heart J. 28, 766-772.
© 2010 EVJ Ltd
526 Assessment of equine MSC isolation methods
Smith, R.K. (2008a) Mesenchymal stem cell therapy for equine tendinopathy. Disabil. Download full-text
Rehabil. 30, 1752-1758.
Smith, R.K. (2008b) Use of bone marrow-derived mesenchymal stem cells to enhance
tendon and ligament healing. Proceedings of the 2008 ACVS Veterinary
Symposium: The Sugical Summit 2008, San Diego. pp 172-176.
Smith, R.K., Korda, M., Blunn, G.W. and Goodship, A.E. (2003) Isolation and
implantation of autologous equine mesenchymal stem cells from bone marrow
into the superficial digital flexor tendon as a potential novel treatment. Equine vet.
J. 35, 99-102.
Strober, W. (2001) Trypan blue exclusion test of cell viability. Curr. Protoc. Immunol.
Appendix 3, Appendix 3B.
Massy, M., Mortier, C., Bron, D. and Lagneaux, L. (2005) Mesenchymal
stem cells derived from CD133-positive cells in mobilized peripheral blood
and cord blood: proliferation, Oct4 expression, and plasticity. Stem Cells 23,
T., Meuleman, N.,Delforge, A.,Dejeneffe, M.,Leroy, R.,
Tuan, R.S., Boland, G. and Tuli, R. (2003) Adult mesenchymal stem cells and cell-
based tissue engineering. Arthritis Res. Ther. 5, 32-45.
Vannier, J.P., Monconduit, M. and Piguet, H. (1980) [Comparison between 2 density
gradients for separation of CFU]. Biomedicine 33, 236-239.
Vidal, M.A., Kilroy, G.E., Johnson, J.R., Lopez, M.J., Moore, R.M. and Gimble, J.M.
(2006) Cell growth characteristics and differentiation frequency of adherent
equine bone marrow-derived mesenchymal stromal cells: adipogenic and
osteogenic capacity. Vet. Surg. 35, 601-610.
Vidal, M.A., Kilroy, G.E., Lopez, M.J., Johnson, J.R., Moore, R.M. and Gimble, J.M.
(2007) Characterization of equine adipose tissue-derived stromal cells: adipogenic
and osteogenic capacity and comparison with bone marrow-derived mesenchymal
stromal cells. Vet. Surg. 36, 613-622.
Vidal, M.A., Robinson, S.O., Lopez, M.J., Paulsen, D.B., Borkhsenious, O., Johnson,
J.R., Moore, R.M. and Gimble, J.M. (2008) Comparison of chondrogenic potential
in equine mesenchymal stromal cells derived from adipose tissue and bone
marrow. Vet. Surg. 37, 713-724.
Violini, S., Ramelli, P., Pisani, L.F., Gorni, C. and Mariani, P. (2009) Horse bone
marrow mesenchymal stem cells express embryo stem cell markers and show the
ability for tenogenic differentiation by in vitro exposure to BMP-12. BMC Cell
Biol. 10, 29.
Wagner, W., Wein, F., Seckinger, A., Frankhauser, M., Wirkner, U., Krause, U., Blake,
J., Schwager, C., Eckstein, V., Ansorge, W. and Ho, A.D. (2005) Comparative
characteristics of mesenchymal stem cells from human bone marrow, adip ose
tissue, and umbilical cord blood. Exp. Hematol. 33, 1402-1416.
Wilke, M.M., Nydam, D.V. and Nixon, A.J. (2007) Enhanced early chondrogenesis in
articular defects following arthroscopic mesenchymal stem cell implantation in an
equine model. J. orthop. Res. 25, 913-925.
Wollert, K.C., Meyer, G.P., Lotz, J., Ringes-Lichtenberg, S., Lippolt, P., Breidenbach,
C., Fichtner, S., Korte, T., Hornig, B., Messinger, D., Arseniev, L., Hertenstein,
B., Ganser, A. and Drexler, H. (2004) Intracoronary autologous bone-marrow cell
transfer after myocardial infarction: the BOOST randomised controlled clinical
trial. Lancet 364, 141-148.
Author contributions The initiation, conception and planning of this
study were by S.L., C.B., L.C.S. and J.P.L., the execution and writing
by C.B., S.L., P.V., L.C.S. and statistics by G.B., C.B. and S.L.
© 2010 EVJ Ltd
C. Bourzac et al. 527