Hindawi Publishing Corporation

BioMed Research International

Volume 2014, Article ID 951512, 11 pages
http://dx.doi.org/10.1155/2014/951512

Review Article
Mesenchymal Stem Cells for Regenerative Therapy:
Optimization of Cell Preparation Protocols

Chiho Ikebe and Ken Suzuki

William Harvey Research Institute, Barts and The London School of Medicine and Dentistry, Queen Mary University of London,
Charterhouse Square, London EC1M 6BQ, UK

Correspondence should be addressed to Ken Suzuki; [email protected]

Received 9 October 2013; Accepted 8 December 2013; Published 6 January 2014

Academic Editor: Ryuichi Morishita

Copyright © 2014 C. Ikebe and K. Suzuki. This is an open access article distributed under the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly
cited.

Administration of bone marrow-derived mesenchymal stem cells (MSCs) is an innovative approach for the treatment of a range of
diseases that are not curable by current therapies including heart failure. A number of clinical trials have been completed and many
others are ongoing; more than 2,000 patients worldwide have been administered with culture-expanded allogeneic or autologous
MSCs for the treatment of various diseases, showing feasibility and safety (and some efficacy) of this approach. However, protocols
for isolation and expansion of donor MSCs vary widely between these trials, which could affect the efficacy of the therapy. It is
therefore important to develop international standards of MSC production, which should be evidence-based, regulatory authority-
compliant, of good medical practice grade, cost-effective, and clinically practical, so that this innovative approach becomes an
established widely adopted treatment. This review article summarizes protocols to isolate and expand bone marrow-derived MSCs
in 47 recent clinical trials of MSC-based therapy, which were published after 2007 onwards and provided sufficient methodological
information. Identified issues and possible solutions associated with the MSC production methods, including materials and
protocols for isolation and expansion, are discussed with reference to relevant experimental evidence with aim of future clinical
success of MSC-based therapy.

1. Introduction National Institute of Health (http://www.ClinicalTrial.gov/)

as of July 2013. The cells used are, strictly speaking, mesenchy-
Recent research has extensively shown the potential of bone mal stromal cells, which include MSCs and other cells; but, in
marrow- (BM-) derived mesenchymal stem cells (MSCs) for most cases they are simply referred to as MSCs. Previous pre-
regenerative therapies in various organs including the heart clinical studies and clinical trials have shown feasibility and
[1]. The effects from this approach are dependent on their po- safety of MSC-based therapy; however, the therapeutic effects
ten-cy of secretion of beneficial cytokines and growth fac- observed in clinical trials to date appear to be inconsistent
tors for tissue repair/regeneration and immunomodulation and remain inconclusive [5].
and/or their differentiation for regenerating damaged organs MSCs were first described in 1976 by Friedenstein and
[2]. Since the first clinical trial of BMC injection in 1995 [3], colleagues [6] and are more recently defined by The Inter-
more than 2,000 patients have been administered with al- national Society of Cellular Therapy based on three cellular
logeneic or autologous MSCs for the treatment of vari- properties: (1) adherence to plastic, (2) positive expression of
ous diseases, including graft-versus-host disease, hemato- CD105, CD73, and CD90 and negative expression of CD45,
logic malignancies, cardiovascular diseases, neurologic dis- CD34, CD14 or CD11b, CD79𝛼 or CD19, and HLA class
eases, autoimmune diseases, organ transplantation, refrac- II, and (3) differentiation potential to mesenchymal lin-
tory wounds, and bone/cartilage defects [4]. More than 200 eages including osteocytes, adipocytes, and chondrocyte [7].
clinical trials of MSC-based therapy, completed or ongo- Unfortunately, the frequency of MSCs in BM is low; MSCs
ing, have been listed on the website of the United States represent 0.001–0.01% of BM mononuclear cells or lower [8].
2 BioMed Research International

Although the optimal dosage of MSCs in therapeutic appli- MSCs (Figure 1(c) and Supplementary Table 1). All these
cations is still unclear and will be dependent upon the type of trials successfully supported feasibility to obtain the aimed
therapy, 1.0–2.0 × 106 MSCs per kg body weight is generally number of MSCs, but with using a variety of isolation and
used [8]. Direct collection of such a large number of MSCs expansion protocols. Furthermore, regardless of the protocols
from BM is not practical. Therefore, it is necessary to expand to prepare MSCs and cell number injected, no major safety
isolated MSCs in vitro to obtain a sufficient number for ther- issues that were directly caused by MSC, were reported.
apeutic approaches.
MSCs have a rapid proliferation ability, achieving a thou-
sandfold expansion of cell number in a two- to three- week
period. However, inappropriate expansion may reduce the
quality of MSCs. It is known that extensive in vitro culture 3. Isolation of MSC from BM
induces cellular senescence that is associated with growth
arrest and apoptosis [9]. In addition, particular therapeutic 3.1. BM Preparation for MSC Isolation. Possible techniques
properties of MSCs may be lost during prolonged culture; to isolate MSCs from BM materials include cell adherence-
for example, the cardioprotective effect of passage 5 MSCs is based methods and cell-sorting methods, with the vast
significantly reduced compared to passage 3 MSCs [10]. How- majority of previous clinical trials using the former method.
ever, protocols of MSC preparation used in clinical studies The latter including fluorescence-activated cell sorting and
remain inconsistent and suboptimal. There are surprisingly immune-magnetic bead cell sorting [12] has the advantage
different protocols used in current clinical studies, in terms of collecting a more purified MSC population. However,
of culture materials (flasks, culture media, and supplements), they are hardly used in clinical trials because of the lack
seeding density, passaging, and storage. These factors can in- of appropriately specific simple surface markers for MSCs,
fluence the important properties of MSCs, leading to reduced possible cellular damage, more expensive cost, and more
or unexpected therapeutic results [11]. In addition, such in- demanding labor. For the adherence-based methods, either
consistent protocols make comparison of the results between whole BM cells or BM mononuclear cells separated by density
clinical studies difficult. gradient centrifugation were used. The use of whole BM cells
Establishment of optimal, standardized protocols for is clearly easier and yields higher numbers of adhered cells
MSC isolation and expansion will therefore be a key for MSC- on plastic dishes with reduced loss of MSCs compared to
based therapies to become widespread, generic approaches. density gradient separation methods. However, cells collected
For this aim, understanding of currently used protocols with by an adherence method represent a heterogeneous mixture
their scientific justification is essential. We hereby carefully of cells, including not only MSCs but also hematopoietic cells
searched the protocols used in recent clinical trials of MSC- at different differentiation/commitment stages, endothelial
based therapy by referring PubMed. As a result, a total of 47 cells and endothelial progenitor cells. Although many of these
reports, which sufficiently describe MSC-preparation meth- contaminating cells may be removed during passaging, such
ods, were found, published from January 2007 onwards (see contamination would affect the expansion of MSCs as well
Supplementary Table 1 in Supplementary Material available as the overall effect of the therapy. In order to isolate a more
online at http://dx.doi.org/10.1155/2014/951512). This review homogeneous initial MSC population, BM mononuclear cells
article summarizes the information obtained from these clin- can be separated from whole BM cells by density gradient
ical trials with further referencing to relevant experimental centrifugation using either Ficoll (Paque, Hypaque, or Paque
studies, highlighting issues and solutions associated with Premium) or Percoll (both available from GE Healthcare,
current protocols of MSC isolation and expansion. Uppsala, Sweden). In the current studies we have analyzed,
62% used Ficoll-based density gradient separation, 16% used
whole BM cells without separation, and another 9% employed
2. Background of Clinical Trials of Percoll-based density gradient separation (Figure 2(a)).
MSC-Based Therapy Analyzed in Percoll and Ficoll have usually been used at densities of
This Review 1.073 g/mL [13] and 1.077 g/mL [14], respectively, to isolate
MSCs with high proliferative and differentiative potential.
By literature search using PubMed, we selected 47 reports of Percoll is a suspension of colloidal silica particles (diameter
clinical trials of BM-derived MSC-based therapy published 15–30 nm), which has been widely used for separating cells,
between January 2007 and June 2013, which sufficiently organelles, viruses, and other subcellular particles in basic
describe the methods of isolation and expansion of MSCs science experiments, but it is not produced as a good manu-
(Supplementary Table 1). Most reports provide parts (not facturing practice (GMP) grade reagent. Ficoll, a polymer of
all) of the methodological information of interest to us. sucrose with a high synthetic molecular weight, is generated
The trials aimed to treat a range of diseases, including at GMP grade and has been frequently used for separating
oncological diseases (38%), followed by neurological diseases mononuclear cells and lymphocytes from peripheral blood in
(26%) and cardiovascular diseases (11%) (Figure 1(a)). 66% of clinical practice for several decades, indicating clinical safe-
the studies used autologous MSCs, while the remaining 34% ty of the reagent. Recently Mareschi et al. compared MSCs
used allogeneic MSCs (Figure 1(b)). The number of MSCs collected via Percoll-separated mononuclear cells, Ficoll-
injected ranged from 0.34 to 2.3×106 cells/kg body weight; the separated mononuclear cells, and whole BM cells and found
majority of the reports administered 1-2×106 /kg body weight no significant differences in terms of gross morphology, dif-
BioMed Research International 3

Others
8/47
(17%) Allogeneic
16/47
Knee and Oncological (34%)
4/47
wound 18/47
(9%)
(38%) 31/47
(66%)
Autologous
5/47
Cardiovascular (11%)
12/47
(26%)
Neurological

(a) Disease (b) MSC type

<1 × 106 /kg

3/23
(13%)

>1 × 106 /kg

12/23
1 × 106 /kg (52%)
8/23
(35%)

(c) MSC number

Figure 1: Background of the clinical trials of MSC-therapy reviewed in this article. A total of 47 clinical trials that have reported sufficient
information on the MSC preparation were selected to review in this article. (a) A wide range of diseases were targeted by MSC therapy. (b)
Both autologous and allogeneic MSCs were used for MSC therapy. (c) The number of MSCs administered was 1 × 106 /kg body weight or
more in the majority of clinical trials. Some trials repeated the injection. See Supplementary Table 1 as well.

ferentiation potential, or immunophenotype between the col- adherence-based method including Corning, Falcon, Nunc,
lected cells [15]. However, the whole BM cell method appar- and Greiner. Sotiropoulou et al. compared the effect of these
ently resulted in a greater Colony-Forming Unit-Fibroblast 4 culture flasks to adhere MSCs [16] and indicated that greater
(CFU-F) number and improved cellular growth compared to numbers of MSCs were acquired in Corning flasks followed
gradient-separated cell methods. Given the other advantages by Falcon, Nunc, and Greiner at 7 days after plating (without
in being less demanding in cost and labor, it is proposed that passaging). All these types of flasks are produced from
the whole BM cell method would be the first-choice method polystyrene permanently rendered hydrophilic with corona
for MSC isolation from BM samples. discharge, using high voltage to create a reactive gas plasma
[17]. This process for Falcon flasks takes place in a closed
chamber, thus creating a consistent treatment surface. On the
3.2. Flask for MSCs Isolation. There are many manufacturers other hand, during manufacturing of the flasks from other
that produce plastic flasks suitable for MSC isolation by the companies, the gas is exposed to ambient air and therefore
4 BioMed Research International

Iwaki
Others 1/22
6/45 (5%)
(13%)
Greiner
Percoll 4/22 Corning
4/45 (18%) 6/22
(9%) (27%)

Ficoll
Whole 28/45
BM (62%)
7/45 Nunc
(16%) 5/22
Falcon
(23%) 6/22
(27%)

(a) BM preparation (b) Flask for isolation

Others
5/44
Others 15–20% FBS (11%)
10/47 2/44 (5%)
(21%)

DMEM PL
low glucose 5/44 (11%)
𝛼MEM 25/47
5/47 (53%)
10% FBS
(11%) 32/44
(73%)

DMEM
7/47
(15%)

(c) Expansion medium (d) Serum for expansion

Figure 2: Protocols and materials used for MSC isolation and expansion. Different methods and materials were used in the recent MSC-
therapy clinical trials, in terms of method for bone marrow (BM) preparation for MSC isolation (a), culture flask used for MSC isolation (b),
culture medium used for MSCs expansion (c), and serum used for MSC expansion (d). PL: platelet lysate; FBS: fetal bovine serum.

subjected to day-to-day environmental changes. In the real 2 × 105 BM mononuclear cells/cm2 , the lower initial seeding
world, the most commonly used flask was Corning (27%) and densities achieved increasingly larger numbers of adherent
Falcon (27%) equally, followed by Nunc (23%), Greiner (18%), cells at Passage 0 [16]. Both et al. also reported that MSCs
and Iwaki (5%) in our analysis of current clinical trials (Figure seeded at lower densities had a faster proliferation than
2(b)). those seeded at higher densities, with MSCs plated at 100
cells/cm2 reaching their target of 2 × 108 cells 4 days faster
3.3. Cell Seeding Density for MSC Isolation. Cell seeding den- than cells that were seeded at 5 × 103 cells/cm2 [18]. Further
sity of BM mononuclear cells or whole BM cells is another decrease in the seeding density below 100 cell/cm2 showed
important factor to determine the efficiency of MSC yield a further increase in proliferation rate; however, there is a
as this affects adherence of MSCs, contamination by other lower limit in the plating density in the clinical settings.
cell types, and initial growth of adhered MSCs. Sotiropoulou Given that 1 × 107 –1 × 108 BM mononuclear cells or a
et al. reported that, between the range from 1 × 103 to larger number of whole BM cells are commonly obtained,
BioMed Research International 5

12 on 4 major types of culture flasks has indicated that the

most improved expansion of cultured MSCs was acquired
10 in the flasks from Falcon, followed by those from Corning,
Nunc, and Greiner, although the quality and functions of
Number of reports

8 produced MSCs did not differ among the different types

of flasks examined [16]. In contrast, the most commonly
6 used flask in previous clinical trials we investigated here was
Corning flasks (35%), followed by Nunc flasks (25%), Falcon
4 flasks (20%), and Greiner flasks (20%) (Supplementary Table
1). In the majority of the reports, we found that the same
2 manufacturer’s flasks are preferably used for both isolation
and expansion of MSCs. The use of the same manufacturer’s
0 flasks may be more convenient and economical; however,
it should be noted that the optimal flask surface for initial
<1 × 105

1.5-1.6 × 105

1 × 106
5–8 × 105
2–4 × 105

isolation of MSCs could be different from that for MSC

expansion as the scientific evidence indicates [16].
A wide surface area is required to obtain a sufficient num-
Plating density for MSC isolation (cells/cm2 ) ber of MSCs for clinical application. To reduce the number of
Figure 3: Plating density for MSC isolation. Plating cell densities
culture flasks used, manufacturing companies such as Nunc
of BM mononuclear cells for MSC isolation used in 26 clinical trial and Corning offer large, multilayered culture systems that can
reports are presented. 1.5-1.6 × 105 cells/cm2 was most frequently fit to usual cell culture incubators. Decreasing the number of
used. flasks will improve the microbiological safety and traceability
and also reduce staff workload and cost. The CellStacks
(Corning, USA) and CellFactory (Nunc, Denmark) systems,
which start from a unit surface area of 635 cm2 , offer the pos-
it is not practical to seed such a large number of cells at
sibility of 2, 5, 10, and 40 stages per container. In addi-
below 1 × 104 . Seeding 1 × 108 BM mononuclear cells at tion, these devices can be connected by tubes, allowing for
1 × 103 cells/cm2 would require approximately 600 × 175 cm2 convenient, sterile, GMP-compliant operations (e.g., culture
flasks, which is too high a cost in terms of materials and initiation, medium exchange, and cell harvesting).
labor. As a matter of fact, the cell seeding density used in
current clinical trials is quite high, with extreme variability
ranging from 1.1 × 103 to 1.0 × 106 mononuclear cells/cm2 4.2. Basal Culture Medium. Basal culture medium consists
(Supplementary Table 1). The most commonly used seeding of amino acids, glucose, and ions including calcium, magne-
density of BM mononuclear cells is 1.5-1.6 × 105 cells/cm2 , sium, potassium, sodium, and phosphate. There is no doubt
followed by 1.0×106 cells/cm2 and 2–4×105 cells/cm2 (Figure that the types of culture medium used affect proliferation and
3). For the future clinical application, it is suggested that BM- differentiation of MSCs. There is a preclinical report show-
mononuclear cells should be plated at as a low density as ing that DMEM is preferable to IMDM (Iscove’s modified
far as the cost, facility, and labor allow. There is very limited Dulbecco’s medium) with respect to preservation of MSC
experimental evidence to discuss the optimal plating dose of stemness [19]. It has also been experimentally demonstrated
whole BM cells but a pre-clinical study has shown that 10,000 that 𝛼MEM (minimal essential medium) better preserves
cells/cm2 would be the most advantageous condition [15]. osteogenic properties of MSCs and achieves higher CFU-F
retrieval than DMEM [20]. Figure 2(c) shows that the basal
culture media used for MSC expansion in current clinical
3.4. Medium and Supplement. Optimal medium and culture trial includes DMEM-low glucose (53%), DMEM (15%), and
supplements for MSC isolation remain much less unstudied, 𝛼MEM (11%).
compared to those for MSC expansion (see Section 4 for L-Glutamine is an essential nutrient for energy produc-
detailed information). In the vast majority of current clinical tion as well as protein and nucleic acid synthesis in cell
trials, the same medium and serum/supplement appeared to culture, and thus this is commonly supplemented into culture
be simply used for both MSC isolation and expansion (very media. However, this spontaneously degrades in culture me-
few reports described this particular method). This may be dia, and its chemical breakdown and cellular metabolism lead
convenient and economical in practice; however, the most to ammonia formation, possibly inhibiting cell growth [21].
effective conditions for isolation and expansion of MSCs To solve this issue, Glutamax is recently used as substitute
could be different, requiring further research to elucidate the for L-glutamine, as this is more stable in aqueous solutions
optimal culture medium for MSC isolation. and does not spontaneously degrade. Sotiropoulou et al. sys-
temically compared the expansion efficacy of MSCs among
4. Expansion of MSCs 8 different basal media (IMDM, Optimem, 𝛼MEM with L-
glutamine, 𝛼MEM with Glutamax, DMEM with low glucose
4.1. Flask for MSCs Expansion. A comprehensive laboratory and L-glutamine, DMEM with low glucose and Glutamax,
investigation of the proliferation efficacy of MSCs cultured DMEM with high glucose and L-glutamine, and DMEM
6 BioMed Research International

with high glucose and Glutamax) [16]. The authors have from several healthy donors may be pooled for the use
found significant differences: among the 8 types of medium [26]. Doucet et al. first demonstrated that growth factors
studied, 𝛼MEM containing Glutamax achieved the greatest contained in platelet lysate are able to promote MSC expan-
expansion of cultured MSCs, followed by 𝛼MEM contain- sion in a dose-dependent manner [25]. This was further
ing L-glutamine. Unfortunately many previous clinical trial substantiated by the data showing that a culture medium
papers did not clearly describe the type of glutamine used supplemented with 5% platelet lysate is superior to 10% FBS
(Supplementary Table 1). in clonogenic efficiency and proliferative capacity of MSCs,
therefore providing more efficient expansion, together with
a significant time saving [26]. However, several studies have
4.3. Growth Factor Supplement for MSC Expansion. It is shown limitations of platelet lysate, including reduction of
known that growth factor supplement to culture medium osteogenic or adipogenic differentiation potential [33, 34] and
enhances proliferation with maintenance of important prop- decreased immunosuppressive capacity with altered surface
erties of MSCs. In particular, fibroblast growth factor-2 marker expression [35]. In addition, there is a risk that any
(FGF2) [22], platelet-derived growth factor (PDGF) [23],
allogeneic human product may be contaminated with human
epidermal growth factor (EGF) [24], transforming growth
pathogens that might not be detected by routine screening.
factor (TGF)-𝛽 [23], and insulin-like growth factor (IGF)
Moreover, crude blood derivatives are poorly defined and also
[25, 26] play a role. Previously, fetal bovine serum (FBS)
has been most frequently used (10% FBS in 73% and 15– suffer from batch-to-batch variation, and thus their ability to
20% FBS in 5%) to supply growth factors to MSC culture maintain MSC growth and therapeutic potentials could be
medium (Figure 2(d)), because FBS contains all these factors variable. Further studies are need for platelet lysate to be part
and is relatively readily available at clinical grade. However, of a standard protocol.
it should be noted that FBS shows considerable variation in Issues associated with human products encourage the
growth factor activity from batch to batch, and therefore large use of serum-free and animal component-free MSC culture
amounts of batch-tested FBS will need to be reserved for media. StemPro MSC SFM from Invitrogen is the first FDA-
a clinical application. Furthermore, FBS remains associated approved commercial product of this type. Agata et al.
with safety issues including transmission of prion or viral showed an enhanced effect of StemPro MSC SFM to improve
disease, anaphylatoxic reactions, and production of anti-FBS rapid proliferation at early (<5) passage stages compared to
antibodies [27, 28]. Regulatory authorities in an increasing FBS [36]. Of note, important characteristics of MSCs, in-
number of countries, including Paul-Ehrlich-Institute in cluding surface antigen expression, stemness, and differen-
Germany, now prohibit the clinical use of FBS, while in tiation potential, are different between MSCs cultured with
contrast the Australian Therapeutic Goods Authority allows FBS and MSCs with serum-free medium [36]. Although the
the use of FBS for the production of clinical grade materials formulations of these commercial media are not disclosed, it
as long as it is sourced from cattle in a country free of bovine is important to evaluate each commercial media and select
spongiform encephalitis such as Australia or New Zealand. the most suitable product for each type of treatment.
To avoid such a risk related to the use of animal materials,
the use of human products, including human serum and
platelet lysate, has been proposed. The effect of human 4.4. Direct Addition of Growth Factors for MSC Culture. Al-
allogeneic serum from adult donors to enhance proliferation though ideal growth factor supplements for MSC culture
of MSCs with preservation of important cellular properties are still undefined, administration of several types of growth
is controversial [29, 30]. On the other hand, it will be prob- factors with or without serum or platelet lysate has been
lematic to acquire a large amount of autologous serum tested if it could increase MSC expansion with maintenance
sufficient to generate clinically relevant numbers of MSCs. of important cellular properties. These include at least b-
Moreover, autologous serum from elderly patients may have FGF [22], PDGF [23], TGF-𝛽, EGF [24], and IGF1 [25, 26].
deteriorated capacity to support cell growth. Allogeneic FGF2 induces excellent expansion efficiency of MSCs with
human serum from umbilical cord blood [31] and placenta maintenance of their differentiation potential and has been
[32] has also been proposed as a potential alternative to used for clinical trials of MSC-based therapy [37]. However,
replace FBS because these primitive tissues are a rich source recent data suggests that b-FGF upregulates HLA-DR and
of growth factors. Stro-1 and downregulates CD44 in a dose-dependent fashion
Platelet lysate has recently been a more preferred human [16, 38]. PDGFs were first found in platelets and they might
product; more than 10% of previous clinical trials between be responsible for some of the platelet lysate activity in MSC
2007 and 2013 used Platelet lysate (Figure 2(d)). Platelet growth. PDGFs have a role in osteogenic, adipogenic, and
lysate can be easily obtained from apheresis products [26], chondrogenic differentiation of MSCs; however, the primary
as well as from buffy coats [33] of healthy volunteers. effect is likely to be mitogenic action with inhibition of
Immediately after collection, platelet products are frozen differentiation [39]. The PDGF-BB isoform can activate all
at −80∘ C and subsequently thawed to obtain the release PDGF receptors and therefore may be the best choice as a
of growth factors included in platelets with centrifugation culture supplement. A recent report found that a combina-
to eliminate platelet bodies. The obtained growth factors tion of b-FGF, PDGF, and TGF-𝛽 could replace the serum
include PDGFs, b-FGF, VEGF, IGF-1, and TGF-𝛽 [25, 26], component in cell culture medium to expand human MSCs
which improve proliferative capacity of MSCs. Platelet lysate without compromising differentiation potential, at least up
BioMed Research International 7

10 trypsin-EDTA solution (Supplementary Table 1). Of note,

the concentration of trypsin-EDTA used was widely ranging;
8
0.25, 0.05, and 0.025% trypsin-EDTA was used in 58, 26, and
16% of previous trials, respectively (Figure 5(b)). Excessive
Number of reports

trypsinisation can damage cells, while on the other hand

6 insufficient trypsin-treatment will reduce the yield of cells.
Optimal trypsinisation condition may be different among
4
flask types used and cell density/confluence. Thus, the choice
of trypsinisation conditions (not only concentration but
also duration and temperature) should be carefully decided
2 case-by-case based on scientific evidence. In addition, many
previous trials appeared to utilise porcine-derived trypsin,
0 which should be replaced with human trypsin or other
200– 3,000– 5,000– 8.000
8,000 10,000– alternatives [44, 45] to reduce safety concerns [46].
2,800 4,000 6,000 12,000
Plating density for MSC expansion (cells/cm2 )

Figure 4: Plating density for MSC expansion. Plating cell densities 5.2. Storing of MSCs. Isolated and expanded BM-derived
for MSC expansion used in 25 clinical trial reports are presented. MSCs were sometimes stored until the time of treatment.
3,000–6,000 cells/cm2 was commonly used. In our search 17 out of 49 (35%) trials used cryopreserved
MSCs (Figure 5(c); two out of 47 clinical trials used both
fresh cells and cryopreserved cells, making the total number
49). This allows for great flexibility in the clinical setting,
to 5 passages [23]. Further evidence is required for this but extreme caution is needed on possible adverse effects on
promising approach to become a widespread protocol. MSCs. Although there are many preclinical studies showing
that cryopreservation does not change the biological behavior
of MSCs such as differentiation, growth, and/or surface
4.5. Cell Plating Density for MSCs Expansion. Plating cell marker expression [47, 48], on the other hand, there are
density is influential not only on initial isolation but on reports warning hazardous effects by cryopreservation [49].
expansion of cultured MSCs. Generally, a higher plating Further refinement of the protocol is warrantied. Important
density results in a reduced proliferation ability, probably ingredients in current freezing solution include dimethyl-
due to contact inhibition and/or less availability of nutrients sulfoxide (DMSO) and serum. DMSO has been extensively
per cell [40]. The log phase lasts for a longer duration in used at 5–10% as a cryoprotectant with its high membrane
cells plated at lower densities, and hence more population permeability. However, DMSO can be damaging to cells
doublings occur due to a longer exponential growth phase when used in high concentration, especially during the
[41]. A comparable study showed that, after 10 days in culture, thawing procedure. Also, if it remains in MSC suspension
BM-derived MSCs seeded at 2,500, 250, 25, and 2.5 cells/cm2 for administration, DMSO can cause adverse reactions in
resulted in 2.7 ± 0.5, 4.8 ± 0.4, 6.7 ± 0.5, and 7.6 ± 1.0 patients, including nausea, vomiting, tachycardia, bradycar-
population doublings [42]. This study also showed that the dia, and hypotension. Haack-Sørensen et al. [46, 50] advocate
seeding density does not affect cellular properties of MSCs the use of 5% concentrations of DMSO together with 95%
including cell surface marker expression. FBS. However, the use of animal sera will have a risk in
However, for clinical-scale production (1 × 106 /kg body the use for patients as discussed above. Defined, serum-
weight or more) of MSCs, use of a very low plating density free and animal component-free freezing media, such as
is unrealistic due to demanding cost, facility, and labor; a Cryostor CS10 StemCell Technologies [51] or Plasmalyte-A
plating density of 1,000 cells/cm2 is suggested as a reasonable, [52], may be possible alternatives. Cell concentration during
evidence-based compromise [43]. In the clinical arena, how- cryopreservation was proposed to be optimal with 0.5–1 ×
ever, more compromise for the cost/labor is usually taken; 106 /mL [53]. A controlled rate freezing method (freezing at
over 75% of current clinical trials used plating densities of rate of 1∘ C per minute) will achieve superior outcome than
over 3,000 cells/cm2 (Figure 4). uncontrolled freezing [54].

5. Passaging and Storing of MSC

5.3. Injection Vehicle. It is important to optimise the vehicle
5.1. Dissociation of Adherent MSCs. For the purpose of pas- of MSCs for injection, as this will affect donor cell viability
saging for expansion or collection for administration, and loss before and after injection. In previous clinical trials,
adhered MSCs on plastic flasks need to be dissociated. In 66% used saline and 17% used PBS (Figure 5(d)). There were
our search, administered MSCs were received in less than 1 attempts to supplement human serum albumin to protect
passage in 23%, 1–5 passages in 71%, and over 5 passages in 6% cells from environmental stress and prevent adherence to the
of reported clinical trials (Figure 5(a)). To this end, the major- walls of tubes and needles. Further systematic comparisons
ity of current clinical trials used enzymatic digestion using between injection vehicles are warrantied.
8 BioMed Research International

>P5
2/35
0.025%
(6%) 3/19
P0-P1 (16%)
8/35
(23%)

0.25%
0.05% 11/19
P1–P5
25/35 5/19 (58%)
(71%) (26%)

(a) Passages (b) Trypsin dose

Others
5/29
(17%)

Yes No
17/49 32/49 PBS Saline
(35%) (65%) 5/29 19/29
(17%) (66%)

(c) Cryopreservation (d) Injection vehicle

Figure 5: Preparations of MSCs. (a) Over 70% clinical trials used MSCs that received 1–5 passages. (b) Doses of trypsin used for passaging
were widely varied. (c) Cryopreserved MSCs were used in 35% of clinical trials reviewed. In 47 clinical trials studied, two trials used both
fresh cells and cryopreserved cells, thus the total number of reports shown in the graph is 49 (see Supplementary Table 1). (d) It was common
to use saline as injection vehicle.

6. Conclusion and Future Perspective expansion of MSCs. This review demonstrates how various
the current protocols were. Many protocols lack scientific
Recent advance in basic and medical science and tech- validation and appear to be suboptimal.
nologies has realised the employment of BM-derived MSCs It is now urgently important to solve this issue of the lack
for a variety of therapeutic indications including regenera- of conformity between MSC manufacturing protocols, which
tive therapies. A sufficient sum of initial clinical trials has is considered as potential threat to further development
shown feasibility and safety of this approach at least and of MSC-based therapy. As summarised in this review, a
also suggested the therapeutic effect (though preliminary), range of relevant scientific evidence is available for this
encouraging further study for this approach to become an purpose. Active cooperation between academics, clinicians,
established generic treatment. One of the major hurdles for companies, and regulatory authorities is encouraged in order
this development will be the establishment of optimised to develop international standards for BM-derived MSC
and standardized GMP-compliant protocols for isolation and production, which should be evidence-based, regulatory
BioMed Research International 9

authority-compliant, of good medical practice grade, cost- [12] I. Pountos, D. Corscadden, P. Emery, and P. V. Giannoudis,
effective, and clinically practical, for the future success of “Mesenchymal stem cell tissue engineering: techniques for
MSC-based therapy. isolation, expansion and application,” Injury, vol. 38, no. 4, pp.
S23–S33, 2007.
[13] M. F. Pittenger, A. M. Mackay, S. C. Beck et al., “Multilineage
Conflict of Interests potential of adult human mesenchymal stem cells,” Science, vol.
The authors declare that there is no conflict of interests re- 284, no. 5411, pp. 143–147, 1999.
garding the publication of this paper. [14] A. Muraglia, R. Cancedda, and R. Quarto, “Clonal mesenchy-
mal progenitors from human bone marrow differentiate in vitro
according to a hierarchical model,” Journal of Cell Science, vol.
Acknowledgments 113, no. 7, pp. 1161–1166, 2000.

This work was supported by NIHR-funded Barts Cardiovas- [15] K. Mareschi, D. Rustichelli, R. Calabrese et al., “Multipotent
cular Biomedical Research Unit, UK, British Heart Founda- mesenchymal stromal stem cell expansion by plating whole
bone marrow at a low cellular density: a more advantageous
tion, UK, and Heart Research UK.
method for clinical use,” Stem Cells International, vol. 2012,
Article ID 920581, 10 pages, 2012.
References [16] P. A. Sotiropoulou, S. A. Perez, M. Salagianni, C. N. Baxevanis,
and M. Papamichail, “Characterization of the optimal culture
[1] M. Li and S. Ikehara, “Bone-marrow-derived mesenchymal conditions for clinical scale production of human mesenchymal
stem cells for organ repair,” Stem Cells International, vol. 2013, stem cells,” Stem Cells, vol. 24, no. 2, pp. 462–471, 2006.
Article ID 132642, 8 pages, 2013.
[17] R. S. Sigmons and M. Goldman, “Corona discharge physics and
[2] D. M. Patel, J. Shah, and A. S. Srivastava, “Therapeutic potential
applications,” NATO Advanced Study Institute Series B, vol. 89,
of mesenchymal stem cells in regenerative medicine,” Stem Cells
pp. 1–64, 1978.
International, vol. 2013, Article ID 496218, 15 pages, 2013.
[3] H. M. Lazarus, S. E. Haynesworth, S. L. Gerson, N. S. Rosenthal, [18] S. K. Both, A. J. C. Van Der Muijsenberg, C. A. Van Blitterswijk,
and A. I. Caplan, “Ex vivo expansion and subsequent infusion J. De Boer, and J. D. De Bruijn, “A rapid and efficient method
of human bone marrow-derived stromal progenitor cells (mes- for expansion of human mesenchymal stem cells,” Tissue Engi-
enchymal progenitor cells): implications for therapeutic use,” neering, vol. 13, no. 1, pp. 3–9, 2007.
Bone Marrow Transplantation, vol. 16, no. 4, pp. 557–564, 1995. [19] L. Pieri, S. Urbani, B. Mazzanti et al., “Human mesenchymal
[4] M. M. Lalu, L. McIntyre, C. Pugliese et al., “Safety of cell therapy stromal cells preserve their stem features better when cultured
with mesenchymal stromal cells (SafeCell): a systematic review in the Dulbecco’s modified Eagle medium,” Cytotherapy, vol. 13,
and meta-analysis of clinical trials,” PLoS ONE, vol. 7, no. 10, no. 5, pp. 539–548, 2011.
Article ID e47559, 2012. [20] M. J. Coelho, A. Trigo Cabral, and M. H. Fernandes, “Human
[5] J. Ankrum and J. M. Karp, “Mesenchymal stem cell therapy: two bone cell cultures in biocompatibility testing. Part I: osteoblastic
steps forward, one step back,” Trends in Molecular Medicine, vol. differentiation of serially passaged human bone marrow cells
16, no. 5, pp. 203–209, 2010. cultured in 𝛼-MEM and in DMEM,” Biomaterials, vol. 21, no.
11, pp. 1087–1094, 2000.
[6] A. J. Friedenstein, U. F. Gorskaja, and N. N. Kulagina, “Fibrob-
last precursors in normal and irradiated mouse hematopoietic [21] A. Christie and M. Butler, “Growth and metabolism of a murine
organs,” Experimental Hematology, vol. 4, no. 5, pp. 267–274, hybridoma in cultures containing glutamine-based dipeptides,”
1976. Focus, vol. 16, pp. 9–13, 1994.
[7] M. Dominici, K. Le Blanc, I. Mueller et al., “Minimal crite- [22] G. Bianchi, A. Banfi, M. Mastrogiacomo et al., “Ex vivo enrich-
ria for defining multipotent mesenchymal stromal cells. The ment of mesenchymal cell progenitors by fibroblast growth
International Society for Cellular Therapy position statement,” factor 2,” Experimental Cell Research, vol. 287, no. 1, pp. 98–105,
Cytotherapy, vol. 8, no. 4, pp. 315–317, 2006. 2003.
[8] K. Schallmoser, E. Rohde, A. Reinisch et al., “Rapid large- [23] F. Ng, S. Boucher, S. Koh et al., “PDGF, tgf-2. And FGF signaling
scale expansion of functional mesenchymal stem cells from is important for differentiation and growth of mesenchymal
unmanipulated bone marrow without animal serum,” Tissue stem cells (mscs): transcriptional profiling can identify markers
Engineering C, vol. 14, no. 3, pp. 185–196, 2008. and signaling pathways important in differentiation of MSCs
[9] K. Stenderup, J. Justesen, C. Clausen, and M. Kassem, “Aging into adipogenic, chondrogenic, and osteogenic lineages,” Blood,
is associated with decreased maximal life span and accelerated vol. 112, no. 2, pp. 295–307, 2008.
senescence of bone marrow stromal cells,” Bone, vol. 33, no. 6, [24] K. Tamama, H. Kawasaki, and A. Wells, “Epidermal Growth
pp. 919–926, 2003. Factor (EGF) treatment on Multipotential Stromal Cells
[10] P. R. Crisostomo, M. Wang, G. M. Wairiuko et al., “High passage (MSCs). Possible enhancement of therapeutic potential of
number of stem cells adversely affects stem cell activation and MSC,” Journal of Biomedicine and Biotechnology, vol. 2010,
myocardial protection,” Shock, vol. 26, no. 6, pp. 575–580, 2006. Article ID 795385, 10 pages, 2010.
[11] M. E. Bernardo, F. Locatelli, and W. E. Fibbe, “Mesenchymal [25] C. Doucet, I. Ernou, Y. Zhang et al., “Platelet lysates promote
stromal cells: a novel treatment modality for tissue repair,” mesenchymal stem cell expansion: a safety substitute for animal
Annals of the New York Academy of Sciences, vol. 1176, pp. 101– serum in cell-based therapy applications,” Journal of Cellular
117, 2009. Physiology, vol. 205, no. 2, pp. 228–236, 2005.
10 BioMed Research International

[26] M. E. Bernardo, M. A. Avanzini, C. Perotti et al., “Optimization [39] R. Pytlik, O. Slanar, D. Stehlik, and E. Matejkova, “Production
of in vitro expansion of human multipotent mesenchymal of clinical grade mesenchymal stromal cells,” in Regenerative
stromal cells for cell-therapy approaches: further insights in Medicine and Tissue Engineering—Cells and Biomaterials, pp.
the search for a fatal calf serum substitute,” Journal of Cellular 145–178, InTech, 2011.
Physiology, vol. 211, no. 1, pp. 121–130, 2007.
[40] D. C. Colter, I. Sekiya, and D. J. Prockop, “Identification of
[27] M. Sundin, O. Ringdén, B. Sundberg, S. Nava, C. Götherström, a subpopulation of rapidly self-renewing and multipotential
and K. Le Blanc, “No alloantibodies against mesenchymal adult stem cells in colonies of human marrow stromal cells,”
stromal cells, but presence of anti-fetal calf serum antibodies, Proceedings of the National Academy of Sciences of the United
after transplantation in allogeneic hematopoietic stem cell States of America, vol. 98, no. 14, pp. 7841–7845, 2001.
recipients,” Haematologica, vol. 92, no. 9, pp. 1208–1215, 2007.
[28] A. Mackensen, R. Dräger, M. Schlesier, R. Mertelsmann, and [41] B. Neuhuber, S. A. Swanger, L. Howard, A. Mackay, and I.
A. Lindemann, “Presence of IgE antibodies to bovine serum Fischer, “Effects of plating density and culture time on bone
albumin in a patient developing anaphylaxis after vaccination marrow stromal cell characteristics,” Experimental Hematology,
with human peptide-pulsed dendritic cells,” Cancer Immunol- vol. 36, no. 9, pp. 1176–1185, 2008.
ogy Immunotherapy, vol. 49, no. 3, pp. 152–156, 2000. [42] C. Bartmann, E. Rohde, K. Schallmoser et al., “Two steps to
[29] A. Shahdadfar, K. Frønsdal, T. Haug, F. P. Reinholt, and J. E. functional mesenchymal stromal cells for clinical application,”
Brinchmann, “In vitro expansion of human mesenchymal stem Transfusion, vol. 47, no. 8, pp. 1426–1435, 2007.
cells: choice of serum is a determinant of cell proliferation, [43] I. Sekiya, B. L. Larson, J. R. Smith, R. Pochampally, J.-G. Cui, and
differentiation, gene expression, and transcriptome stability,” D. J. Prockop, “Expansion of human adult stem cells from bone
Stem Cells, vol. 23, no. 9, pp. 1357–1366, 2005. marrow stroma: conditions that maximize the yields of early
[30] K. Tateishi, W. Ando, C. Higuchi et al., “Comparison of human progenitors and evaluate their quality,” Stem Cells, vol. 20, no.
serum with fetal bovine serum for expansion and differentiation 6, pp. 530–541, 2002.
of human synovial MSC: potential feasibility for clinical appli-
cations,” Cell Transplantation, vol. 17, no. 5, pp. 549–557, 2008. [44] G. Brooke, T. Rossetti, R. Pelekanos et al., “Manufacturing of
human placenta-derived mesenchymal stem cells for clinical
[31] J. Jung, N. Moon, J.-Y. Ahn et al., “Mesenchymal stromal cells
trials,” British Journal of Haematology, vol. 144, no. 4, pp. 571–
expanded in human allogenic cord blood serum display higher
579, 2009.
self-renewal and enhanced osteogenic potential,” Stem Cells and
Development, vol. 18, no. 4, pp. 559–571, 2009. [45] P. P. Carvalho, X. Wu, G. Yu et al., “Use of animal protein-
[32] H. Shafaei, A. Esmaeili, M. Mardani et al., “Effects of human free products for passaging adherent human adipose-derived
placental serum on proliferation and morphology of human stromal/stem cells,” Cytotherapy, vol. 13, no. 5, pp. 594–597, 2011.
adipose tissue-derived stem cells,” Bone Marrow Transplanta- [46] M. Haack-Sørensen and J. Kastrup, “Cryopreservation and
tion, vol. 46, no. 11, pp. 1464–1471, 2011. revival of mesenchymal stromal cells,” Methods in Molecular
[33] C. Lange, F. Cakiroglu, A.-N. Spiess, H. Cappallo-Obermann, Biology, vol. 698, pp. 161–174, 2011.
J. Dierlamm, and A. R. Zander, “Accelerated and safe expan-
[47] R. Pal, M. Hanwate, and S. M. Totey, “Effect of holding time,
sion of human mesenchymal stromal cells in animal serum-
temperature and different parenteral solutions on viability and
free medium for transplantation and regenerative medicine,”
functionality of adult bone marrow-derived mesenchymal stem
Journal of Cellular Physiology, vol. 213, no. 1, pp. 18–26, 2007.
cells before transplantation,” Journal of Tissue Engineering and
[34] R. Gruber, F. Karreth, B. Kandler et al., “Platelet-released su- Regenerative Medicine, vol. 2, no. 7, pp. 436–444, 2008.
pernatants increase migration and proliferation, and decrease
osteogenic differentiation of bone marrow-derived mesenchy- [48] A. Nicol, M. Nieda, C. Donaldson et al., “Cryopreserved human
mal progenitor cells under in vitro conditions,” Platelets, vol. 15, bone marrow stroma is fully functional in vitro,” British Journal
no. 1, pp. 29–35, 2004. of Haematology, vol. 94, no. 2, pp. 258–265, 1996.
[35] H. Abdelrazik, G. M. Spaggiari, L. Chiossone, and L. Moretta, [49] M. François, I. B. Copland, S. Yuan, R. Romieu-Mourez, E. K.
“Mesenchymal stem cells expanded in human platelet lysate Waller, and J. Galipeau, “Cryopreserved mesenchymal stromal
display a decreased inhibitory capacity on T- and NK-cell cells display impaired immunosuppressive properties as a result
proliferation and function,” European Journal of Immunology, of heat-shock response and impaired interferon-𝛾 licensing,”
vol. 41, no. 11, pp. 3281–3290, 2011. Cytotherapy, vol. 14, no. 2, pp. 147–152, 2012.
[36] H. Agata, N. Watanabe, Y. Ishii et al., “Feasibility and efficacy [50] M. Haack-Sørensen, L. Bindslev, S. Mortensen, T. Friis, and
of bone tissue engineering using human bone marrow stromal J. Kastrup, “The influence of freezing and storage on the
cells cultivated in serum-free conditions,” Biochemical and characteristics and functions of human mesenchymal stromal
Biophysical Research Communications, vol. 382, no. 2, pp. 353– cells isolated for clinical use,” Cytotherapy, vol. 9, no. 4, pp. 328–
358, 2009. 337, 2007.
[37] N. K. Venkataramana, S. K. V. Kumar, S. Balaraju et al., “Open-
labeled study of unilateral autologous bone-marrow-derived [51] E. J. Woods, B. C. Perry, J. J. Hockema, L. Larson, D. Zhou, and
mesenchymal stem cell transplantation in Parkinson’s disease,” W. S. Goebel, “Optimized cryopreservation method for human
Translational Research, vol. 155, no. 2, pp. 62–70, 2010. dental pulp-derived stem cells and their tissues of origin for
banking and clinical use,” Cryobiology, vol. 59, no. 2, pp. 150–
[38] S. Walsh, C. Jefferiss, K. Stewart, G. R. Jordan, J. Screen, and
157, 2009.
J. N. Beresford, “Expression of the developmental markers
STRO-1 and alkaline phosphatase in cultures of human marrow [52] S. A. Steigman, M. Armant, L. Bayer-Zwirello et al., “Preclinical
stromal cells: regulation by fibroblast growth factor (FGF)-2 and regulatory validation of a 3-stage amniotic mesenchymal stem
relationship to the expression of FGF receptors 1-4,” Bone, vol. cell manufacturing protocol,” Journal of Pediatric Surgery, vol.
27, no. 2, pp. 185–195, 2000. 43, no. 6, pp. 1164–1169, 2008.
BioMed Research International 11

[53] B. C. Goh, S. Thirumala, G. Kilroy, R. V. Devireddy, and

J. M. Gimble, “Cryopreservation characteristics of adipose-
derived stem cells: maintenance of differentiation potential
and viability,” Journal of Tissue Engineering and Regenerative
Medicine, vol. 1, no. 4, pp. 322–324, 2007.
[54] R. Fuller and R. V. Devireddy, “The effect of two different
freezing methods on the immediate post-thaw membrane
integrity of adipose tissue derived stem cells,” International
Journal of Heat and Mass Transfer, vol. 51, no. 23-24, pp. 5650–
5654, 2008.