cryopreservation of human bone marrow-derived mesenchymal stem cells with reduced dimethylsulfoxide...

Cryopreservation of Human Bone Marrow-Derived Mesenchymal Stem Cells with

Reduced Dimethylsulfoxide and Well-Defined Freezing Solutions

Yang LiuDalian R&D Center for Stem Cell and Tissue Engineering, School of Chemical Engineering, Dalian University of Technology,Dalian 116024, P.R. China

Institute of Biomedical Engineering, Dept. of Engineering Science, University of Oxford, Oxford, UK

Xia XuInstitute of Biomedical Engineering, Dept. of Engineering Science, University of Oxford, Oxford, UK

Xuehu MaDalian R&D Center for Stem Cell and Tissue Engineering, School of Chemical Engineering, Dalian University of Technology,Dalian 116024, P.R. China

Enca Martin-Rendon and Suzanne WattStem Cell Research Laboratory, National Blood Service, Oxford Centre, The John Radcliffe Hospital, Oxford, UK

Nuffield Dept. of Clinical and Laboratory Sciences, University of Oxford, Oxford, UK

Zhanfeng CuiInstitute of Biomedical Engineering, Department of Engineering Science, University of Oxford, Oxford, UK

DOI 10.1002/btpr.464Published online June 22, 2010 in Wiley Online Library (wileyonlinelibrary.com).

The aim of this study is to investigate the feasibility of using well defined, serum-freefreezing solutions with a reduced level of dimethylsulfoxide (DMSO) of 7.5, 5, and 2.5% (v/v) in the combination with polyethylene glycol (PEG) or trehalose to cryopreserve humanbone marrow-derived mesenchymal stem cells (hBMSCs), a main source of stem cells forcell therapy and tissue engineering. The standard laboratory freezing protocol of around1�C/min was used in the experiments. The efficiency of 1,2-propandiol on cryopreservationof hBMSCs was explored. We measured the post-thawing cell viability and early apoptoticbehaviors, cell metabolic activities, and growth dynamics. Cell morphology and osteogenic,adipogenic and chondrogenic differentiation capability were also tested after cryopreserva-tion. The results showed that post-thawing viability of hBMSCs in 7.5% DMSO (v/v), 2.5%PEG (w/v), and 2% bovine serum albumin (BSA) (w/v) was comparable with that obtainedin conventional 10% DMSO, that is, 82.9 � 4.3% and 82.7 � 3.7%, respectively. In addi-tion, 5% DMSO (v/v) with 5% PEG (w/v) and 7.5% 1,2-propandiol (v/v) with 2.5% PEG(w/v) can provide good protection to hBMSCs when 2% albumin (w/v) is present. Enhancedcell viability was observed with the addition of albumin to all tested freezing solutions.VVC 2010 American Institute of Chemical Engineers Biotechnol. Prog., 26: 1635–1643, 2010Keywords: cryopreservation, mesenchymal stem cells, well defined freezing solutions, DMSO

Introduction

Recent development in stem cell therapy and tissue engi-neering has increased the need for large numbers of appro-priate multipotent or pluripotent stem cells. Bone marrow-derived mesenchymal stem cells (BMSCs) have becomegood candidates for cell therapy and tissue-engineeringapplications as they have the capacity for self-renewal andmultilineage differentiation into mesenchymal tissues, suchas osteoblasts, adipocytes, and chondrocytes under appropri-ate culture conditions.1

Cryopreservation plays an important role in obtaining off-the-shelf availability for a variety of tissues and cells. Cryo-preservation decouples cell culture and implantation andmakes possible for long distance transportation and for longterm storage of cells. Cryopreserved cells are likely to be themain cell sources for tissue engineering and stem cell ther-apy, the important aspects of regenerative medicine.2 Theavailability in term of stem cell source and quantity makescryopreservation important. If stem cells can be cryopre-served for long-term and still retain a high level of viabilityand potential to differentiate into tissue-specific cells, theirclinical applications can be greatly simplified. This is notonly for allogeneic cell therapy (i.e., the stem cell medicine)but also important for autologous usage to avoid

Correspondence concerning this article should be addressed to Z. Cuiat [email protected].

VVC 2010 American Institute of Chemical Engineers 1635

transplantation immunity.3 However, an essential prerequisitefor successful applications of BMSCs is to cryopreserve cellsunder well defined conditions to satisfy requirements in clin-ical applications.

Currently, there are two available approaches of cryopre-servation of cells: (i) conventional slow freezing and (ii) vit-rification. Slow freezing is usually achieved by equilibriumin the presence of permeating cryoprotecting agents (CPAs)and the cells are frozen with a controlled cooling rate, nor-mally at 1�C/min. Vitrification is defined as glass-like solidi-fication and/or complete avoidance of ice crystal formation4

involving the use of high concentration of CPAs and fastfreezing rate. Slow freezing has the advantage of using lowconcentrations of CPAs, which are associated with chemicaltoxicity and osmotic shock.5 More importantly slow freezingcan handle large quantity of cells, which makes it more clin-ically relevant. However, as the freezing rate does not varymuch, the cell survival largely depends on the CPAs.

Ever since the discovery in 1949 that glycerol affordedprotection to sperm6 during cryopreservation, it has becomecommon practice to add one or several CPAs to the freezingmedia. Because of its high-membrane permeability, dime-thylsulfoxide (DMSO) has been used extensively as a CPAfor cryopreservation of various cells and tissues. In commoncell culture in laboratories, fetal bovine serum (FBS) is oftenused to provide growth factors and protein source. FBS isalso found to protect cells during freezing as it stabilizes bio-membranes and adjusts osmotic pressure acting as a CPA.Hence in laboratories, the most common CPA formulationconsists of 10% (v/v) DMSO and up to 90% (v/v) FBS andthis had been widely used for various kinds of stem cells,7–12

including hBMSCs.2,13,14 However, both DMSO and FBS havepresented a number of disadvantages.

DMSO, which can stabilize cellular proteins by increasingthe freeze energy of unfolding via the preferential exclusionmechanism15 and stabilizes plasma membranes through inter-acting electrostatically with phospholipids bilayer,16 at ahigher concentration, such as 10% (v/v), has been associatedwith neurological toxicity,17 particularly in the developingbrain.18 DMSO also interacts with hydrophobic residues ofproteins, leading to denaturation and destruction of pro-teins.19 Cryopreserved cells treated with DMSO may causeother adverse effect20 including gene mutation. Therefore, itis encouraging to reduce the levels of DMSO in freezing sol-utions even if complete elimination is not possible. Use of areduced level of DMSO, if in the tolerant range of humanbody, can also eliminate the need for CPA removal and thethawed cells can be directly implanted.

FBS is an undesirable additive to cells for therapeutic pur-poses in humans because it is not defined in composition andalso the use of FBS carries the risk of transmitting viral andprion diseases and proteins that may initiate xenogeneicimmune responses.21 The use of autologous serum, instead ofFBS, may solve the ethical and the safety problems. However,production of autologous serum is a time-consuming andpainstaking procedure and the amount of autologous serumneeded is extremely limited.22 Hence, for the clinical applica-tion of cryopreserved cells, the use of well defined CPA solu-tions containing only authority approved ingredients and theelimination of animal-derived biological ingredients and allpossible sources of infectious diseases are necessary.

The objective of this study is to investigate the possibilityof reducing DMSO concentrations and replacing serum in thefreezing solutions for the cryopreservation and long-term stor-age of hBMSCs. Reduction of DMSO may require the additionof other clinically approved chemicals, such as polyethyleneglycol (PEG), trehalose, and 1,2-propanediol. After cryopreser-vation, cell viability and apoptosis were determined. Cells met-abolic activities were assessed during the first 24-h period afterthawing. Then cell growth characteristics were measured up to11 days after cryopreservation. Differentiation potentials wereassessed as well. The effects of DMSO concentration and pres-ence of albumin on hBMSCs viability and functions after cryo-preservation have been demonstrated.

Materials and Methods

Chemicals

All chemicals used in this study were purchased from theSigma (UK) and culture medium from Invitrogen (UK)unless otherwise stated.

Cell culture

Human bone marrow-derived mesenchymal stem cells(hBMSCs) were obtained from ScienCell (UK) and Lonza (UK)and expanded in our laboratory according to recommended pro-tocols. The cells were cultured in 75 cm2 flasks with alpha-mini-mal essential medium (a-MEM) supplemented with 10% FBS,200 units/mL penicillin and streptomycin. The culture mediumwas changed every 3 days. When the cells became confluent,they were trypsinized and centrifuged and cell pellets were har-vested for further expansion or cryopreservation studies.

Cryopreservation of hBMSCs

All freezing solutions (fs) were prepared in a-MEM me-dium according to the formula in Table 1. The most

Table 1. Formula of Freezing Solutions

No. Solution name FBS (% v/v) DMSO (% v/v) PEG (% w/v) Trehalose (% w/v) BSA (% w/v) 1.2-Propanediol (% v/v)

C1 D10 0 10 0 0 0 0C2 D10S10 10 10 0 0 0 0C3 D10S90 90 10 0 0 0 0fs1 D7,5Peg2.5 0 7.5 2.5 0 0 0fs1A D7.5Peg2.5A 0 7.5 2.5 0 2 0fs2 D5Peg5 0 5 5 0 0 0fs2A D5Peg5A 0 5 5 0 2 0fs3 D5Peg2T 0 5 2 3 0 0fs3A D5Peg2TA 0 5 2 3 2 0fs4 D2.5Peg7.5 0 2.5 7.5 0 0 0fs4A D2.5Peg7.5A 0 2.5 7.5 0 2 0fspA P10A 0 0 0 0 2 10fsppA P7.5Peg2.5A 0 0 2.5 0 2 7.5

D, DMSO; S, serum; P, 1,2-propanediol; A, albumin; T, trehalose

1636 Biotechnol. Prog., 2010, Vol. 26, No. 6

routinely used freezing solutions were chosen as controls:C1 (D10, 10% DMSO), C2 (D10S10, 10% DMSO þ 10%FBS), C3 (D10S90, 10% DMSO þ 90% FBS); whilst otherfreezing solution formulas were regarded as experimentalgroups. Concentrated cells were directly resuspended in thefreezing solutions at a concentration of 1 � 106 cells/mLand then transferred into cryovials in 1 mL aliquots, andequilibrated with the freezing solutions for 10 min at 4�C.Then the cryovials with cells were placed in an isopropanolfreezing box (Nalgene cryo 1C/min freezing container, Nal-gene, USA) within a �80�C freezer (New Brunswick Scien-tific) overnight, and stored in a liquid nitrogen tank (Taylor-Wharton, UK) for at least 1 week. Before use, the cells werethawed by rapidly immersing the cryovials in a 37�C waterbath with gentle shaking. After 2 min, aliquots in each cryo-vial were slowly transferred into 4 mL prewarmed a-MEMin 15 mL centrifuge tubes for 5 min and pelleted by centrifu-gation at 1000 rpm for 5 min. The cells were resuspended inculture medium for the following assays.

Cell viability and apoptosis assay

To analyze cell viability and apoptosis, thawed cells wereprocessed using the Annexin V-fluorescein isothiocyanate(FITC) apoptosis kit (Invitrogen,UK) with a modified proto-col. Briefly, thawed cells were washed with phosphate-buf-fered saline (PBS) twice and then resuspended in 200 lL 1xbinding buffer. Annexin V-FITC (0.5 lL) was added andincubated for 15–20 min at room temperature in the dark.Then propidium iodide (PI) buffer was added at a final con-centration of 5 lg/mL. The cells were placed on ice beforebeing analyzed by FACSCalibur flow cytometer (BD Scien-ces, UK) equipped with Cell Quest software. Viable and ap-optotic cell populations were analyzed with WinMDIsoftware.

Cell metabolic activity and proliferation

Cells from Lonza were cryopreserved with C1, C2, fs1,and fs1A by slow freezing method as described above. Toevaluate freezing solutions cytotoxicity, the metabolic activ-ity of cryopreserved hBMSCs normalized to noncryopre-served cells during 24-h after thawing was determined byalamarBlue assay. After thawing, a same number of viablecells were seeded in 96-well plates. Five percentage ala-marBlue was added into each well at 0, 2, 4, 6, and 24 h af-ter seeding and incubated for 2 h in a CO2 incubator.Noncryopreserved cells plated in 96-well plates were used asa positive control. The absorbance was measured spectropho-tometrically at 570 and 600 nm using a standard microplatereader (TECAN GENios, Australia). Data were processedwith XFluor4 software.

To calculate the normalized alamarBlue reduction betweencryopreserved and noncryopreserved cells, the followingequation was used.

% normalizedAB reduction

¼ eOXk2ð Þ Ak1ð Þ � eOXk1ð Þ Ak2ð Þ½ �cryopreservedeOXk2ð Þ A0k1ð Þ � eOXk1ð Þ A0k2ð Þ½ �noncryopreserved

� 100

In the equation, eOX is the constant representing the molarextinction coefficient of alamarBlue oxidized form. Ak1 andAk2 are the absorbance of test wells at 570 and 600 nm,

respectively. A0k1 and A0k2 are the absorbance of noncryo-preserved positive growth control wells at 570 and 600 nm,respectively.

To assess the effect of cryopreservation on cell prolifera-tion, the growth rate was monitored during 11 days of cul-ture. hBMSCs were plated in 96-well plate at aconcentration of 8 � 105 cells/mL and subsequently dilutedto serial 1:2 dilutions to get a standard curve relating cellnumber and alamarBlue reduction rate. Cells in 5% ala-marBlue solution were incubated at 37�C, in a humidifiedatmosphere with 5% CO2 for 2 h. As negative control, ala-marBlue solution was added to the medium without cells.The absorbance of the test and control wells was measuredspectrophotometrically at 570 and 600 nm using a microplatereader as described above.

To determine the growth rate, the noncryopreserved andcryopreserved hBMSCs were seeded in 96-well plates atdensity of 1 � 104 cells/cm2. The number of viable cellswas determined at days 1, 3, 5, 7, 9, and 11 after plating. Toeliminate differences due to medium color, the experimentwas performed using culture medium without phenol red. Allproliferation assays were performed in quadruplicate in atleast three separate experiments.

The number of viable cells correlate with the magnitudeof dye reduction and is expressed as the percentage of ala-marBlue reduction rate. The calculation of the percentage ofalamarBlue reduction rate is as follows according to themanufacture’s protocol:

% alamar Blue reduction

¼ eOXk2ð Þ Ak1ð Þ � eOXk1ð Þ Ak2ð ÞeREDk1ð Þ A0k2ð Þ � eREDk2ð Þ A0k1ð Þ � 100

In the equation, eOX and eRED are constants representingthe molar extinction coefficient of alamarBlue oxidized andreduced form, respectively. Ak1 and Ak2 are the absorbanceof test wells at 570 and 600 nm, respectively. A0k1 and A0k2are the absorbance of negative control wells at 570 and 600nm, respectively.

Cell morphology, intracellular pH, and mitochondria

To study possible morphological and intracellular changes,noncryopreserved and post-thawed hBMSCs were cultured ina complete culture medium on sterile coverslips (U13 mm,VWR) for 3 days and then incubated with calcein AM(1 lM)/Hoechst 33342 (1 lg/mL)/PI (0.1 lg/mL) mixture,acridine orange (10 lg/mL), and MitoTracker orange(200 nM) (Molecular Probes, Invitrogen) to evaluate the cellviability and biological properties. Stained cells were visual-ized using a fluorescence microscope (Nikon, 80i, Japan).All fluorescent images were taken with an automatic digitalcamera (Hamamatsu ORCA-ER, C4742-80) loaded withHamamatsu ORCA software.

Differentiation of hBMSCs in vitro

Noncryopreserved and cryopreserved cells were plated at1 � 104 cells/cm2 (for osteogenic and adipogenic differentia-tion) in 24-well plates (Nunclon TM). The cells cultured in

Biotechnol. Prog., 2010, Vol. 26, No. 6 1637

standard culture medium were used as negative controlgroups.

Osteogenic differentiation was induced by culturing theconfluent hBMSCs for 2 weeks in the culture medium sup-plemented with 10�8 M dexamethasone, 10 mM b-glycero-phosphate (Rose Chemicals, UK), and 50 lg/mL ascorbicacid. The differentiation medium was changed every 3 days.Osteogenic differentiation was revealed by alkaline phospha-tase (ALP) expression by histochemical staining with BCIP/NBT Liquid Substrate System for 30 min. After incubation,the cells were washed with PBS and viewed under lightmicroscope.

To induce adipogenic differentiation, the confluent humanMSCs were cultured in 1:1 DMEM/F12 (with 2.5 mML-Glutamine, 15 mM HEPES, Pyridoxine HCl, 55 mg/L So-dium Pyruvate; SAFC Bioscienses, UK) supplemented with10% rabbit serum (Invitrogen, UK), 10�8 M dexamethasone,5 lg/mL insulin for 2 weeks. The medium was changed ev-ery 3 days. To confirm the adipogenic differentiation, thecells were rinsed twice with PBS, fixed with 4% paraformal-dehyde in PBS for 15 min at 4�C and then washed with 70%ethanol. Then, the cells were stained with 2% filtered OilRed O for 30 min at room temperature as an indicator of in-tracellular lipid accumulation. Excess stain was removed bywashing with 70% ethanol followed by several steps of dis-tilled water.

To induce chondrogenic differentiation, hBMSCs wereseeded on 24-well plates at a higher seeding density. Briefly,20 lL of a concentrated cell suspension (2 � 106 cells/mL)were plated into the center of each well in 24-well platesand precultured at 37�C for about 2 h to allow cell attach.Then, the culture medium was gently added to each well soas not to detach the cell nodules. After 3 days, the cellswere cultured with chondrogenic differentiation medium,which was constituted of DMEM-HG, 10�8 M dexametha-sone, 100 lg/mL sodium pyruvate, 10 ng/mL transforminggrowth factor (TGF)-b3, 40 lg/mL L-proline, 50 lg/mLascorbic acid, 50 lg/mL ITS þ 1. The medium was changedevery 3 days for 3 weeks. Chondrocytes were detected usingtoluidine blue staining. After washing once with PBS, thecells were fixed in 4% paraformaldehyde for 15 min at 4�C,washed with distilled water and stained by 1% toluidine bluefor 2–4 h at room temperature, and then washed with 95%ethanol several times to remove the excess stain.

The differentiation staining was observed with an invertedmicroscope (Nikon, TES200, Japan). All images were takenwith a digital camera (Nikon, Coolpix 990) mounted to theinverted microscope. Differentiation potential was expressedas percentage of staining positive area in the total culturearea. To evaluate the differentiation potential, three separatestaining were performed. At least 10 images were taken ran-domly from each staining, and the positive areas of the dif-ferentiation staining were semiquantitively analyzed with theImage-Pro Plus software.

Statistical analysis

Data were presented as mean � standard deviation (SD)of at least three separate experiments. In each experiment,data were taken in triplicate. A standard curve of alamarBluereduction rate was plotted and linear regression analysis wascarried out to calculate R2 for the slope. The EXCEL 2007software was used for the statistical tests. Probabilitiesof significant differences were performed using Student

t-test comparison, with P \ 0.05 considered statisticallysignificant.

Results

Cell viability and apoptosis after cryopreservation

Different combinations of CPAs were prepared accordingto Table 1. Cell viability and apoptosis was assessed by thecombination of PI and Annexin V staining as described inMaterials and Methods. Figure 1a shows the effects of differ-ent types and concentrations of CPAs on post-thawinghBMSCs viability. The viabilities of hBMSCs cryopreservedin controls were 82.7 � 3.7% (10% (v/v) DMSO), 83.8 �2.9% (10% DMSO þ 10% (v/v) FBS), and 90.2 � 1.4%(10% DMSO þ 90% FBS), respectively. Post-thawing via-bility of hBMSCs frozen in C3 was the highest and signifi-cantly different from all other freezing solutions (P \ 0.05).

Figure 1. Viability and apoptosis analysis of hBMSCs aftercryopreservation with different freezing solutions.

(a) Cell viability immediately after thawing. (b) Percentage ofearly apoptotic cells immediately (0 h) and 2 h after thawing.Cell viability and apoptosis was determined by Annexin V-FITC/PI kit with a flow cytometer. Cells negative for bothannexin V and PI are considered as viable cells. Cells positiveonly for annexin V are considered to be in early apoptosis. Forstatistical analysis, x and y represents significant difference(P \ 0.05) compared with group C1 and C2, respectively. x0and y0 represents significant difference (P \ 0.05) comparedwith group C2 at 0 h and 2 h, respectively.


However, there was no significant difference among C1, C2,and fs1A. Compared with fs2 and fs3, the presence of albu-min in fs2A and fs3A improved cell viability (P ¼ 0.045and P ¼ 0.014, respectively). However, the addition of tre-halose in the freezing solutions did not increase cell viabil-ity. The combination of 1,2-propanediol and PEG (fsppA)had better protection effects than 1,2-propanediol (fspA)alone (P ¼ 0.012).

To evaluate cell apoptosis and determine whether there isan immediate or gradual loss of cell viability upon freeze-thawing of hBMSCs, apoptosis assay was performed imme-diately (0 h) and 2 h after thawing with the freezing solu-tions, which gave better results in terms of cell viability.Results in Figure 1b shows a significant increase in apoptosisrate 2 h after thawing, however, the total percentage wasvery low, 1.5–4.0%. The percentage of early apoptotic cellscryopreserved in fsppA in the absence of DMSO was thelowest among other freezing solutions both at 0 h and 2 h af-ter thawing and it was significantly different from all otherformula (P \ 0.05). This may be explained by the effect oninduction of apoptosis that DMSO exerts.23 Compared withC2, a significant decrease in apoptosis cells cryopreserved byfs1A was observed at 2 h after thawing.

Cell metabolic activity and proliferation

The metabolic activity of hBMSCs after cryopreservationwas determined by alamarBlue reduction at 0, 2, 4, 6, and24 h after thawing (Figure 2a). Immediately after thawing,the metabolic rates of the cells after cryopreservation werearound 50% of noncryopreserved cells. After 2 h plating, thecell metabolic activities increased for all tested conditionsbut a slight decrease was observed after 4 h seeding, then itreached to the highest rate 6 h after thawing. At 24 h afterthawing, there was no significant difference in alamarBluereduction (i.e., metabolic rates) in all cryopreserved cells.

To quantify cell proliferation rates after cryopreservation,a standard curve based on the relationship between viablecell number and alamarBlue reduction rate was developed(R2 ¼ 0.99336, data not shown). Cell numbers in the culturewere determined every 2 days. Noncryopreserved hBMSCsshows a higher proliferation rate in comparison with the cry-opreserved cells in Figure 2b. Noncryopreserved cell culturesreached the stationary growth period after day 9, whereasnoncryopreserved cells reached at day 7. There is no signifi-cant difference in cell growth kinetics among freezing solu-tions tested.

Cell morphology and intracellular pHand mitochondria distribution

It is well known that hBMSCs had a spindle-shaped,fibroblastic morphology. Figure 3 shows cell morphology, in-tracellular pH, and mitochondria distribution after cryopre-servation with C1, C2, C3, fs1, and fs1A. We used a calceinAM/Hoechst 33342/PI mixed dye for simultaneous stainingof viable and dead cells. Acridine orange was used to detectintracellular pH gradients, and MitoTracker orange to iden-tify mitochondria distribution. It can be noted that cryopre-served hBMSCs showed a similar pattern of cellularmorphology, intracellular pH value, and mitochondria distri-bution as the noncryopreserved cells.

Cell differentiation

Functional hBMSCs should readily differentiate into osteo-blasts, adipocytes, and chondrocytes when incubated in theappropriate differentiation medium. To induce osteogenic dif-ferentiation, subconfluent cells were cultured in osteogenic me-dium for 2 weeks. The activity of ALP, which is apronounced marker of bone differentiation, was revealed byNBT staining. ALP positive cells were widely distributed inthe culture plates (see Figure 4). By contrast, the negative con-trol cells that were incubated in normal (nondifferentiation)culture medium did not show ALP activity (data not shown).Cells cryopreserved with fs1 showed significantly less evi-dence of osteogenic differentiation capacity than the otherfreezing solutions and noncryopreserved cells (Figure 5a).

After induction with adipogenic medium for 2 weeks, OilRed O staining showed the lipid vacuoles with orange redcolor (Figure 4) and the noninduced control cells showed no

Figure 2. hBMSCs metabolic activity and proliferation aftercryopreservation.

(a) Metabolic activity of hBMSCs after cryopreservation during24-h period. (b) Proliferation of hBMSCs after cryopreservationin 11-day culture. Cell metabolic activity and proliferation wasassessed by alamarBlue assay. Cells were incubated in 5% ala-marBlue in a-MEM medium (without phenol red) for 2 h inthe CO2 incubator, and the absorbance was measured spectro-photometrically at 570 nm and 600 nm using a microplatereader. Absorbance values were converted into absolute cellnumbers based on an established standard curve. x and y repre-sents significant difference (P \ 0.05) compared with groupC1 and C2, respectively. * represents significant differencebetween group fs1 and fs1A (P\ 0.05).


detectable lipid vacuoles (data not shown). After 2 weeks,more than 10% of noncryopreserved hBMSCs differentiatedinto adipocytes that stained with Oil Red O. Surprisingly,more cells differentiated to adipocytes after cryopreservationthan noncryopreserved cells as shown in Figure 5b. hBMSCscryopreserved with fs1A has no significant difference com-pared with other freezing solutions in terms of adipogenicdifferentiation capability.

The potential of hMSCs to differentiate before and aftercryopreservation along the chondrogenic lineage was charac-terized after 21 days in culture at high density in chondro-genic medium. The intensive positive toluidine blue stainingindicated the accumulation of extracellular cartilage matrixproteoglycans (Figure 4). In contrast, no positive stainingwas detected in the negative control cells (data not shown).Cells cryopreserved with the fs1 showed significantly lowerchondrogenic potential than noncryopreserved cells and cellscryopreserved with C1 and C3 (Figure 5c).

Discussion

In this study, hBMSCs were cryopreserved with a simpleand practical method of slow freezing and rapid thawingwith various compositions in the freezing solutions. Wefound that a serum-free and reduced DMSO freezing solu-tion, comprised of 7.5% DMSO, 2.5% PEG, and 2% albumin(fs1A), can obtain comparable results with commonly used10% DMSO (C1) in terms of cell viability, apoptotic per-centage, metabolic activity, proliferation, and differentiationcapacities.

Although DMSO can protect cells by reducing the risk ofintracellular ice formation and the rise in electrolyte concen-tration during freezing, several side effects during the infu-sion of DMSO into patients, including sedation, headache,nausea, vomiting, hypertension, bradycardia, hypotension oranaphylactic shock,24 and at cellular level have been demon-strated.20 Therefore, it is of great importance to reduce con-centration of DMSO.

In our experiments, hBMSCs viability after cryopreservedwith fs1A was satisfactory, comparable with cells cryopre-served in C1. With the decrease of DMSO concentration,there was a loss in cell viability, as shown in Figure 1a. Inour case, we can come to a conclusion that to protecthBMSCs properly in the process of cryopreservation, theconcentration of DMSO has to be higher than 5% if it is theonly permeating CPA. That is to say that DMSO cannot becompletely replaced by other nonpermeating CPAs (e.g., tre-halose or PEG). This may also explain that poor cell viabil-ity obtained with freezing solutions fs4 and fs4A and theseviability outcomes have not been improved by albumin sup-plement (P [ 0.05). It can be concluded that a certainamount of permeating CPA is required to protect cells fromfreezing damages. Partly, replacement of PEG by trehalosedid not improve hBMSCs post-thawing viability (fs2 vs. fs3and fs2A vs. fs3A) while at the same concentration ofDMSO. Another permeating CPA, 1,2-propanediol can beused to replace DMSO to an extent and PEG may also beable to lower its required concentrations. Although trehaloseprovided enhanced cryoprotective effects intracellularly,25

Figure 3. hBMSCs morphology, intracellular pH, and mito-chondria distribution.

Noncryopreserved hBMSCs and cells that had undergone cryo-preservation with C1, C2, C3, fs1, and fs1A had been labelledwith calcein AM (green color), Hoechst 33342 (blue color) andPI (red color) mixture, acridine orange, and MitoTracker or-ange (orange color). Scale bar: 100 lm.

Figure 4. Representative images of noncryopreserved and cry-opreserved hBMSCs under osteogenic, adipogenicand chondrogenic differentiation conditions.

Osteoblast differentiation was indicated by ALP staining foralkaline phosphatase positive after 2 weeks induction. Adipo-cyte differentiation was stained with Oil Red O for lipid drop-let formation. Chondrocytes were stained positive withtoluidine blue after 3 weeks differentiation. Scale bar: 200 lm.


extracellulary26 or both,27 it did not exhibit superior protec-tion towards hBMSCs in our experiments.

Albumin contributes to prevention of viability loss ofMSCs during freezing and thawing (Figure 1a). Bovine se-rum albumin (BSA) is used in the experiment for its easinessto obtain. We found BSA effectively increased the hBMSCsviability after thawing, which is consistent with previous

studies.28,29 BSA may play a significant role in the mem-brane structure stabilization and reduced damage caused bycold shock or osmotic stress as a lipid peroxidation inhibi-tor30 or an emulsifying substance31 to protect lipids againstoxidation. Although we used bovine serum albumin as aCPA supplement, good results obtained by using human se-rum albumin (HSA) are expected.

CPA mixtures may have some advantages over solutionscontaining only one permeable CPA. High molecular weightmacromolecules, such as polyvinyl-pyrrolidone (PVP) andhydroxyethylstarch (HES) have been applied in cell cryopre-servation and good results have been demonstrated.32–34 Thehigh molecular weight agents tend to have increasingly highviscosities at low temperature, prohibiting water moleculesto form ice crystals.35 PEG is a polymer of ethylene glycoland it is nontoxic. It has been approved by the FDA for sev-eral medical and food industry applications and broadly usedin cell and tissue engineering,36,37 drug delivery systems,38

and cryopreservation.39–41 PEG is also known to depress thefreezing point of solution and will promote cell dehydrationbecause of its nonpermeating property. Our data showed thatPEG is an effective CPA to hBMSCs but the benefit islimited.

During the cryopreservation, cell death could result fromboth necrosis and apoptosis. It has been demonstrated thatthe major reason for the low viability of human embryonicstem (hES) cells after slow freezing is apoptosis rather thancellular necrosis caused by cryoinjury and the experimentaldata showed cell viability lose dramatically during incubat-ing at 37�C over 90-min period.42–44 In our study, we inves-tigated apoptotic cell death of hBMSCs experienced slowfreezing immediately after thawing and 2 h after thawing(Figure 1b) and found that apoptosis is not the major mecha-nism of the loss of hBMSCs post-thawing viability.

Cell metabolic activity and proliferation after cryopreser-vation is an index how cryopreservation affects cell func-tions. In the present study, we employed alamarBlue assayfor determination of the cell metabolic activity and prolifera-tion. The alamarBlue reduction rates of hBMSCs after cryo-preservation normalized to noncryopreserved cells werehigher at 2 and 6 h after thawing than other time points (Fig-ure 2a). Accordingly, hBMSCs started adhesion at 2 h andfinished spreading at 6 h (data not shown). This active me-tabolism may associate with cell adhesion and expansionbehaviors as cell adhesion involves the participation of cellu-lar metabolic processes.45 After plating for 24 h, metabolicactivity of cryopreserved cells was only 69.3–74.5% of non-cryopreserved cells, indicating that it takes time for cells re-covery from cryopreservation process. This is consistent withcell numbers at day 1 in the growth curves (Figure 2b). Cry-opreserved cells had a similar growth pattern, whereas non-cryopreserved hBMSCs showed higher proliferation rate.Results showed that the cryopreservation process does notappear to alter the proliferation potential with slow freezingprotocol and use of appropriate CPAs.

Intracellular pH is an important factor to maintain cellfunctions.46,47 In our experiments, after having been culturedfor 3 days postcryopreservation, the cells maintained the in-tracellular pH at normal level, which indicated that theintracellular pH regulation system is fundamentally preservedafter cryopreservation (Figure 3). A perinuclear arrange-ment of mitochondria might be a cellular marker for‘‘stemness’’48,49 and alternations in this pattern may accom-pany differentiation or loss of pluripotency.50 After

Figure 5. Semiquantitative analysis of hBMSCs differentiationto osteoblast (a), adipocytes (b) and chondrocytes (c)before and after cryopreserved with various freezingsolutions.

At least 10 visual fields were selected randomly for each testof three separate experiments. The positive areas of the differ-entiated cells were analyzed semiquantitatively with the Image-Pro Plus software. # represents significant difference comparedwith other groups (P \ 0.05). & represents significant differ-ence compared with noncryopreserved cells. * represents sig-nificant difference between compared groups (P\ 0.05).


cryopreservation, perinuclear mitochondria distribution wasremained similar to the noncryopreserved cells. This mightindicate cells retained differentiation competence.

To further confirm whether the presence of PEG and BSAand the absence of FBS with reduced DMSO during cryopre-servation affect hBMSCs post-thawing differentiationcapacity into osteoblasts, adipocytes, and chondrocytes,51 weused ALP staining, Oil Red O and toluidine blue to stain theundifferentiated hBMSCs, osteoblasts, adipocytes, and chon-drocyte, respectively. Cell differentiation patterns were iden-tified by histochemical staining of the cells and the matrixthat they produced (Figure 4) and the differentiation ratiowas calculated based on the image analysis (Figure 5).Osteogenic differentiation was demonstrated by positiveNBT staining of ALP. ALP is a differentiation marker that isactive during matrix production and maturation by cells ofthe osteogenic lineage. The ALP activity of hBMSCs underosteogenic differentiation condition would reach maximumon day 14 and then decreased.52 Hence, we inducedhBMSCs into osteoblasts for 2 weeks and performed ALPstaining. Adipogenic differentiation was assessed by positiveRed Oil O staining of cytoplasmic lipid accumulation. Chon-drogenic differentiation was demonstrated by positive tolu-dine blue staining for proteoglycans accumulated in thematrix. It is inconclusive on the effects of different CPAs onhBMSCs differentiation potential, apart from that the adipo-genic differentiation was significantly enhanced by cryopre-servation. Further study is desirable on tissue specific geneexpressions.

Conclusions

In this study, the effect of cryopreservation on hBMSCsviability, metabolic activities, intracellular pH, mitochondriadistribution, and differentiation potentials has been assessed.We demonstrated the possibility of developing a welldefined, serum-free and reduced DMSO freezing solution forhBMSCs cryopreservation. With DMSO as the sole permeat-ing CPA, we found that the minimum DMSO concentrationis around 5% (v/v) to achieve [73% cell survival. DMSOcan be replaced with another permeating CPA, such as 1,2-propanediol although the cell survival is slightly lower. FBScan be replaced with albumin and 2% (w/v) albumin in CPAsolution can significantly enhance the cell survival. This is asignificant outcome as clinically approved human albumincan easily be obtained. In addition the growth kinetics andosteogenic and chondrogenic differentiation potentials of thecryopreserved hBMSC cells are quite similar to those ofnoncryopreserved.

Acknowledgments

This work was sponsored by the UK Biotechnology and Bio-logical Science Research Council (Grant Reference BBSRCBB/D014751/1). Yang Liu is grateful to China ScholarshipCouncil for partly support to conduct her studies at the Univer-sity of Oxford as a visiting PhD student.

Literature Cited

1. Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R,Mosca JD, Moorman MA, Simonetti DW, Craig S, MarshakDR. Multilineage potential of adult human mesenchymal stemcells. Science. 1999;284:143–147.

2. Hirose M, Kotobuki N, Machida H, Kitamura S, Ohgushi H,Tateishi T. Osteogenic potential of cryopreserved human bonemarrow-derived mesenchymal cells after thawing in culture.Mater Sci Eng C. 2004;24:355–359.

3. Kotobuki N, Hirose M, Machida H, Katou Y, Nuraki A, Taka-kura Y, Ohgushi H. Viability and osteogenic potential of cryo-preserved human bone marrow-derived mesenchymal cells.Tissue Eng. 2005;11:663–673.

4. Kuleshova LL, Gouk SS, Hutmacher DW. Vitrification as aprospect for cryopreservation of tissue-engineered constructs.Biomaterials. 2007;28:1585–1596.

5. Luciano AM, Chigioni S, Lodde V, Franciosi F, Luvoni GC,Modina SC. Effect of different cryopreservation protocols oncytoskeleton and gap junction mediated communication integrityin feline germinal vesicle stage oocytes. Cryobiology. 2009;59:90–95.

6. Polge C, Smith AU, Parkes AS. Revival of spermatozoa aftervitrification and dehydration at low temperature. Nature.1949;164:666–666.

7. Nie Y, Bergendahl V, Hei DJ, Jones JM, Palecek SP. Scalableculture and cryopreservation of human embryonic stem cells onmicrocarriers. Biotechnol Prog. 2009;25:20–31.

8. Zhao J, Hao HN, Thomas RL, Lyman WD. An efficient methodfor the cryopreservation of fetal human liver hematopoietic pro-genitor cells. Stem Cells. 2001;19:212–218.

9. Yang H, Zhao H, Acker JP, Liu JZ, Akabuta J, McGann LE.Effect of dimethyl sulfoxide on post-thaw viability assessmentof CD45þ and CD34þ cells of umbilical cord blood and mobi-lized peripheral blood. Cryobiology. 2005;51:165–175.

10. Milosevic J, Storch A, Schwarz J. Cryopreservation does notaffect proliferation and multipotency of murine neural precursorcells. Stem Cells. 2005;23:681–688.

11. Liu GP, Zhou H, Li YL, Li G, Cui L, Liu W, Cao YL. Evalua-tion of the viability and osteogenic differentiation of cryopre-served human adipose-derived stem cells. Cryobiology. 2008;57:18–24.

12. Yeh HJ, Yao CL, Chen HI, Cheng HC, Hwang SM. Cryopreser-vation of human limbal stem cells ex vivo expanded on amni-otic membrane. Cornea. 2008;27:327–333.

13. Liu GP, Shu CF, Cui L, Liu W, Cao YL. Tissue-engineeredbone formation with cryopreserved human bone mesenchymalstem cells. Cryobiology. 2008;56:209–215.

14. Xiang Y, Zheng Q, Jia BB, Huang GP, Xie GG, Pan JJ, WangJF. Ex vivo expansion, adipogenesis and neurogenesis of cryo-preserved human bone marrow mesenchymal stem cells. CellBio Int. 2007;31:444–450.

15. Buchanan SS, Gross SA, Acker JP, Toner M, Carpenter JF,Pyatt DW. Cryopreservation of stem cells using trehalose: eval-uation of the method using a human hematopoietic cell line.Stem Cells Dev. 2004;13:29–305.

16. Anchordoguy TJ, Cecchini CA, Crowe JH, Crowe LM. Insightsinto the cryoprotective mechanism of dimethyl sulfoxide forphospholipid bilayers. Cryobiology. 1991;28:467–473.

17. Junior AM, Arrais CA, Saboya R, Velasques RD, Junqueira PL,Dulley FL. Neurotoxicity associated with dimethylsulfoxide-pre-served hematopoietic progenitor cell infusion. Bone MarrowTranspl. 2008;41:95–96.

18. Windrum P, Morris TCM. Severe neutotoxicity because of di-methyl sulphoxide following peripheral blood stem cell trans-plantation. Bone Marrow Transpl. 2003;31:315–315.

19. Arakawa T, Carpenter JF, Kita YA, Crowe JH. The basis fortoxicity of certain cryoprotectants: a hypothesis. Cryobiology.1990;27:401–415.

20. Rodrigues JP, Paraguassu-Braga FH, Carvalho L, Abdelhay E,Bouzas LF, Porto LC. Evaluation of trehalose and sucrose ascryoprotectants for hematopoietic stem cells of umbilical cordblood. Cryobiology. 2008;56:144–151.

21. Shahdadfar A, Frønsdal K, Haug T, Reinholt FP, BrinchmannaJE. In vitro expansion of human mesenchymal stem cells:choice of serum is a determinant of cell proliferation, differen-tiation, gene expression, and transcriptome stability. Stem Cells.2005;23:1357–1366.

22. Lange C, Cakiroglu F, Spiess AN, Cappallo-Obermann H, Dier-lamm J, Zander AR. Accelerated and safe expansion of human


mesenchymal stromal cell in animal serum-free medium fortransplantation and regenerative medicine. J Cell Physiol. 2007;213:18–26.

23. Hanslick JL, Lau K, Noguchi KK, Olney JW, Zorumski CF,Mennerick S, Farber NB. Dimethyl sulfoxide (DMSO) produceswidespread apoptosis in the developing central nervous system.Neurobiol Dis. 2009;34:1–10.

24. Scheinkonig C, Kappicht S, Kolb HJ, Schleuning M. Adoptionof long-term cultures to evaluate the cryoprotective potential oftrehalose for freezing hematopoietic stem cells. Bone MarrowTranspl. 2004;34:531–536.

25. Ma W, O’Shaughnessy T, Chang E. Cryopreservation of adher-ent neuronal networks. Neurosci Lett. 2006;403:84–89.

26. Rosa AD, Francesco FD, Tirino GA, Ferraro GA, Desiderio V,Paino F, Pirozzi G, D’Andrea F, Papaccio G. A new method forcryopreserving adipose-derived stem cells: an attractive andsuitable large-scale and long-term cell banking technology. Tis-sue Eng C. 2009;15:1–9.

27. Buchanan SS, Menze MA, Hand SC, Pyatt DW, Carpenter JF.Cryopreservation of human hematopoietic stem and progenitorcells loaded with trehalose: transient permeabilization via theadenosine triphosphate dependent P2Z receptor channel. CellPreserv Technol. 2005;3:212–222.

28. Pugh PA, Aankersmit AE, McGowan LT, Rervit HR. Cryopre-servation of in vitro-produced bovine embryos: effects of pro-tein type and concentration during freezing or of liposomesduring culture on post-thaw survival. Theriogenology. 1998;50:495–506.

29. Leeuw FE, Leeuw AM, Daas JH, Colenbrander B, Verkleij AJ.Effects of various cryoprotective agents and membrane-stabiliz-ing compound on bull sperm membrane integrity after coolingand freezing. Cryobiology. 1993;30:32–44.

30. Alvarez JG, Storey BT. Evidence that membrane stress contrib-utes more than lipid peroxidation to sublethal cryodamage incryopreserved human sperm: glycerol and other polyols as solecryoprotectants. J Androl. 1993;14:199–209.

31. Cabrita E, Anel L, Herraez MP. Effect of external cryoprotec-tants as membrane stabilizers on cryopreserved rainbow troutsperm. Theriogenology. 2001;56:623–635.

32. Bricka M, Bessis M. Sur la conservation des erythrocytes parcongelation a basse temperature en presence de polyvinyl-pyrro-lidone et de dextran. C R Seances Soc Biol Fil. 1955;149:875–877.

33. Luo KH, Wu GQ, Wang QL, Sun Y, Liu HB. Effect of dime-thylsulfoxide and hydroxyethyl starch in the preservation offractionated human marrow cells. Cryobiology. 1994;31:349–354.

34. Clapisson G, Salinas C, Malacher P, Michallet M, Philip I, Phi-lip T. Cryopreservation with hydroxyethylstarch (HES) þ dime-thylsulfoxide (DMSO) gives better results than DMSO alone.Electroc J Oncol. 1999;1:97–102.

35. Fuller BJ. Cryoprotectants: the essential antifreezes to protectlife in the frozen state. CryoLetters. 2004;25:375–388.

36. Deible C, Petrosko P, Johnson P, Beckman E, Russel A andWagner W. Molecular barriers to biomaterial thrombosis bymodification of surface proteins with polyethylene glycol. Bio-materials. 1999;20:101–109.

37. Lu HX, Guo LK, Kawazoe N, Tateishi T, Chen GP. Effects ofpoly (L-lysine), poly (acrylic acid) and poly (ethylene glycol) onthe adhesion, proliferation and chondrogenic differentiation ofhuman mesenchymal stem cells. J Biomat Sci. 2009;20:577–589.

38. Fu J, Fiegel J, Krauland E, Hanes J. New polymeric carriers forcontrolled drug delivery following inhalation or injection. Bio-materials. 2002;23:4425–4433.

39. Monroy B, Honiger J, Darquy S, Reach G. Use of polyethylene-glycol for porcine islet cryopreservation. Cell Transplant. 1997;6:613–621.

40. Haggman HM, Ryynanen LA, Aronen TS and Krajnakova J.Cryopreservation of embryogenic cultures of Scots pine. PlantCell Tissue Organ Cult. 1998;54:45–53.

41. Davies MG, Huynh TT, Fulton GJ, Svendsen E, Brockbank FG,Hagen PO. Controlling transplant vasculopathy in cryopreservedvein grafts with polyethylene glycol and glutathione duringtransport. Eur J Vas and Endovasc Surg. 1999;17:493–500.

42. Heng BC, Ye CP, Liu H, Toh WS, Rufaihah AJ, Yang Z, BayBH, Ge Z, Ouyang HW, Lee EH, Cao T. Loss of viability dur-ing freeze-thaw of intact and adherent human embryonic stemcells with conventional slow-cooling protocols is predominantlydue to apoptosis rather than cellular necrosis. J Biomed Sci.2006;13:433–445.

43. Xu X, Cowley S, Flaim CJ, James W, Cui ZF. Enhancement ofcell recovery rate of dissociated human embryonic stem cells af-ter cryopreservation. Biotechnol Prog. 2010;26:781–788.

44. Xu X, Cowley S, Flaim CJ, James W, Seymour L, Cui ZF. Theroles of apoptotic pathways in the low recovery rate after cryo-preservation of dissociated human embryonic stem cells. Bio-technol Prog. 2010;26:827–837.

45. Yamato M, Okuhara M, Karikusa F, Kikuchi A, Sakurai Y,Okano T. Signal transduction and cytoskeletal reorganizationare required for cell detachment from cell culture surfacesgrafted with a temperature-responsive polymer. J Biomed MaterRes. 1999;44:44–52.

46. Xu X, Cui ZF, Wilkins RJ, Urban JP. Intracellular pH changesin isolated bovine articular chondrocytes during the loading andremoval of cryoprotective agents. Cryobiology. 2003;46:161–173.

47. Hayashi T, Hayashi I, Shinohara T, Morishita Y, Nagamura H,Kusunoki Y, Kyoizumi S, Seyama T, Nakachi K. Radiation-induced apoptosis of stem/progenitor cells in human umbilicalcord blood is associated with alterations in reactive oxygen andintracellular pH. Mutat Res. 2004;556:83–91.

48. Lonergan T, Bavister B, Brenner C. Mitochondria in stem cells.Mitochondrion. 2007;7:289–296.

49. Squirrell JM, Schramm RD, Paprocki AM, Wokosin DL, Bavis-ter BD. Imaging mitochondrial organization in living primateoocytes and embryos using multiphoton microscopy. MicroscMicroanal. 2003;9:190–201.

50. Lonergan T, Brenner C, Bavister B. Differentiation-relatedchanges in mitochondrial properties as indicators of stem cellcompetence. J Cell Physiol. 2006;208:149–153.

51. Dominici M, Blanc KL, Mueller I, Slaper-Cortenbach I, MariniF, Krause D, Deans R, Keating A, Prockop D, Horwitz E. Mini-mal criteria for defining multipotent mesenchymal stromal cells.The International Society for Cellular Therapy position state-ment. Cytotherapy. 2006;8:315–317.

52. Sun XJ, Gan YK, Tang TT, Zhang XL, Dai KR. In vitro prolif-eration and differentiation of human mesenchymal stem cellscultured in autologous plasma derived from bone marrow. Tis-sue Eng A. 2008;14:391–400.

Manuscript received Feb. 3, 2010, and revision receivedMay 4, 2010.


cryopreservation of human bone marrow-derived mesenchymal stem cells with reduced dimethylsulfoxide...

Documents