of June 6, 2018.This information is current as

CellsProtection of Human Mesenchymal Stem The Novel Role of SERPINB9 in Cytotoxic

Philip G. Ashton-Rickardt and Reza AbdiChristopher Ting, Mark Atkinson, Mohamed H. Sayegh,Mounayar, Jamil Azzi, Bechara Mfarrej, Ibrahim Batal, Najib El Haddad, Robert Moore, Dean Heathcote, Marwan

http://www.jimmunol.org/content/187/5/2252doi: 10.4049/jimmunol.1003981July 2011;

2011; 187:2252-2260; Prepublished online 27J Immunol 

Referenceshttp://www.jimmunol.org/content/187/5/2252.full#ref-list-1

, 21 of which you can access for free at: cites 58 articlesThis article

        average*  

4 weeks from acceptance to publicationFast Publication! •    

Every submission reviewed by practicing scientistsNo Triage! •    

from submission to initial decisionRapid Reviews! 30 days* •    

Submit online. ?The JIWhy

Subscriptionhttp://jimmunol.org/subscription

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/About/Publications/JI/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/alertsReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists, Inc. All rights reserved.Copyright © 2011 by The American Association of1451 Rockville Pike, Suite 650, Rockville, MD 20852The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on June 6, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

by guest on June 6, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

The Journal of Immunology

The Novel Role of SERPINB9 in Cytotoxic Protection ofHuman Mesenchymal Stem Cells

Najib El Haddad,*,1 Robert Moore,*,1 Dean Heathcote,† Marwan Mounayar,*

Jamil Azzi,* Bechara Mfarrej,* Ibrahim Batal,* Christopher Ting,* Mark Atkinson,‡,x

Mohamed H. Sayegh,* Philip G. Ashton-Rickardt,† and Reza Abdi*

Clinical trials using allogeneic mesenchymal stem cells (MSCs) are ongoing for the purpose of providing therapeutic benefit for

a variety of human disorders. Pertinent to their clinical use are the accessibility to sufficient quantities of these cells allowing for

repetitive administration, as well as a better understanding of the specific mechanisms by which allogeneic MSCs evade host im-

mune responses that in turn influence their life span following administration. In this report, we sought to characterize and compare

human peripheral blood MSCs (hPB-MSCs) with bone marrow-derived MSCs. hPB-MSCs met the established criteria to char-

acterize this cellular lineage, including capacity for self-renewal, differentiation into tissues of mesodermal origin, and expression of

phenotypic surface markers. In addition, hPB-MSCs suppressed alloreactive proliferation as well as the production of proinflam-

matory cytokines. Examination of the mechanisms by which allogeneic MSCs evade the host immune response, which is crucial for

their therapeutic use, demonstrated that constitutive expression of serine protease inhibitor 9 (PI-9) on hPB-MSCs and bone

marrow-derived MSCs is a major defense mechanism against granzyme B-mediated destruction by NK cells. Similarly, MSCs

treated with small interfering RNA for PI-9 increased MSC cellular death, whereas expression of transgenic PI-9 following

retroviral transduction protected MSCs. These data significantly advance our understanding of the immunomodulatory role

for hPB-MSCs as well as the mechanisms by which they evade host immune responses. These findings contribute to the de-

velopment of MSC-based therapies for diseases. The Journal of Immunology, 2011, 187: 2252–2260.

Due to the plasticity and immunomodulatory function ofmesenchymal stem cells (MSCs), multiple preclinicaland clinical trials in regenerative medicine, graft versus

host disease, inflammatory diseases, and autoimmunity have usedthese cells (1–4). Allogeneic MSCs represent an “off-the-shelf”cell therapy, one allowing for an omission of the normal time spanfor cell isolation and expansion. Allogeneic MSCs, albeit hypo-immunogenic, are subject to rejection, which could compromisetheir therapeutic benefit (5–7).

Although bone marrow-derived MSCs (BM-MSCs) are wellcharacterized in the literature, the procedure for bone marrow ex-traction is invasive, and the number of progenitor cells isolated maybe inadequate for MSC proliferation. In particular, the process isquite cumbersome when the need arises for repetitive adminis-tration of MSCs. Consequently, alternative sources, such as cordblood, adipose tissue, and peripheral blood, have recently becomean increasing focus of research (8–10). Although the concentra-tion of hPB-MSCs is significantly less than that of BM-MSCs intheir respective tissues, contributing to the difficulty in hPB-MSCexpansion from peripheral blood, in pathological conditions inwhich stimuli for their release is present (e.g., patients with burns,cancer, and graft rejection) hPB-MSCs have been isolated (11, 12).Moreover, it is of great value to understand the mechanisms by

which allogeneic MSCs evade host immune responses and, con-versely, how they are rejected in therapeutic settings (13, 14).MSCs that express class I MHCs are killed by GrB-producingCTLs and NK cells (14). In contrast, NK cells constitute a ma-jor component of the innate immune system and do not expressTCRs for recognizing Ags bound to MHC molecules as opposedto CTLs (15). Therefore, NK cells have recently been shown tohave regulatory effects on adaptive responses. As many NK cellfunctions share common features with functions of adaptive CD8+

T cells, such as cytotoxicity and cytokine production, NK cell-activating receptors can stimulate overlapping signaling pathwaysused by the T cell Ag receptor (16). Cell-mediated death can occurthrough either the slow Fas-Fas-L cascade or the rapid GrB-per-forin–dependent cascade (17, 18). MSCs pulsed with peptidesfrom viral Ags stimulated the secretion of IFN-g, which resultedin the GrB-caspase-dependent–mediated apoptosis of MSCs (19).GrB homeostatic regulation is, in turn, mediated through inter-action with inhibitors belonging to the serine protease inhibitor(serpin) superfamily. Proteinase inhibitor 9 (PI-9/SERPINB9) in

*Transplantation Research Center, Brigham and Women’s Hospital, Harvard MedicalSchool, Boston, MA 02115; †Section of Immunobiology, Department of Medicine,Imperial College London, London W12 0NN, United Kingdom; ‡Department ofPathology, University of Florida College of Medicine, Gainesville, FL 32610; andxDepartment of Pediatrics, University of Florida College of Medicine, Gainesville,FL 32610

1N.E.H. and R.M. contributed equally to this work and are considered coauthors forthis publication.

Received for publication December 15, 2010. Accepted for publication June 19,2011.

This work was supported by Juvenile Diabetes Research Foundation Grant 4-2007-1065, Juvenile Diabetes Research Foundation Research and Development Grant 1-2007-713, and National Institute of Allergy and Infectious Diseases Grant R01AI45108 (to P.A-R.) and by a Juvenile Diabetes Research Foundation regular grant(to R.A).

Address correspondence and reprint requests to Dr. Reza Abdi, Brigham and Wom-en’s Hospital, Harvard Medical School, 221 Longwood Avenue, LMRC Building,Room 310, Boston, MA 02115. E-mail address: rabdi@rics.bwh.harvard.edu

Abbreviations used in this article: AM, acetomethoxy; b-FGF, basic fibroblast growthfactor; BM-MSC, bone marrow-derived MSC; GrB, granzyme B; GrB-i, GrB in-hibitor; hPB-MSC, human peripheral blood MSC; MFI, median fluorescence inten-sity; MSC, mesenchymal stem cell; PI-9/SERPINB9, serine protease inhibitor 9;serpin, serine protease inhibitor; siRNA, small interfering RNA; Spi6, serine proteaseinhibitor 6.

Copyright� 2011 by TheAmericanAssociation of Immunologists, Inc. 0022-1767/11/$16.00

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1003981

by guest on June 6, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

humans specifically inactivates GrB in an irreversible manner(20). Serine protease inhibitor 6 (Spi6) is the mouse homologof PI-9 in humans and is required to protect CTLs from GrB-mediated death (21, 22). PI-9 is expressed in the cytoplasm andnuclei of CTLs, immunoprivileged cells (23), and embryonic stemcells (24); PI-9 overexpression allows for the evasion of GrB-mediated cytotoxicity (25, 26). Endogenous GrB inhibitors havebeen characterized in both mice and humans. We have recentlyshown the role of Spi6 in the survival of murine BM-MSCs (5).No data are yet available on the presence and protective role of PI-9 in human MSCs against the GrB machinery. In this article, wedescribe the isolation and characterization of MSCs from pe-ripheral blood and also explore the protective function of PI-9 asa potential escape pathway of MSCs.

Materials and MethodsHuman samples

Human peripheral blood samples were obtained from nine healthy indi-viduals. We were able to isolate MSCs from only six individuals. NK cellswere isolated with a kit (Miltenyi Biotec) according to the manufacturer’sguidelines. BM-MSCs and foreskin-derived fibroblasts were commerciallypurchased (Lonza-Poietics). YT cells were cultured as previously de-scribed (27, 28).

MSCs and fibroblast cell culture

A total of 80 ml peripheral blood was collected from healthy individuals,and PBMCs were isolated using Ficoll-Paque (GE Healthcare). PBMCswere cultured at a concentration of 15 3 106 per 25-cm2 flasks in DMEMcomplete medium (DMEM from Lonza), containing 10% FBS (GeminiBio-Products), 1% penicillin-streptomycin, 1% glutamine (both fromLonza), and supplemented with 6 ng/ml human basic fibroblast growthfactor (b-FGF) (Peprotech). The above conditions were used for thestandardized expansion of PB-MSCs; various parameters, such as theblood volume drawn (40–80 ml), the number of seeded PBMCs for initialculture (5–50 3 106 cells per 25-cm2 flasks), cell confluence in primaryculture (60–90%), and the concentration of b-FGF (2–10 ng/ml) wereadditionally assessed. Cells were incubated at 37˚C and 5% CO2 for 3 d,after which half the medium was changed. Once cell confluence reachedabove 80%, MSCs were trypsinized (0.25% Trypsin-EDTA 13; LifeTechnologies) to a new passage. BM-MSCs and fibroblasts were culturedunder the same conditions. All MSCs, as well as fibroblasts, used in thisstudy were passaged four or five times before use.

MSC characterization and differentiation

hPB-MSCs and BM-MSCs were analyzed for surface Ag median fluo-rescence intensity (MFI) expression with a panel of positive and negativemarkers. Anti-human Abs against CD29, CD44, CD49e, CD73, CD90,CD166, CD105, CD45, CD34, and CD80 were purchased from BDPharmingen. MSC differentiation into chondrocyte, adipocyte, and osteo-cyte cell lineages was induced by the relevant differentiation media (Lonza).Cells were cultured at 37˚C (5% vol/vol CO2) in eight-well culture plates(Lab-Tech Brand Products), and the differentiation medium for each lin-eage was changed twice per week. Cells were analyzed for osteogenic,adipogenic, and chondrogenic differentiation, as previously reported (4).Intracytoplasmic anti-human SERPINB9 (PI-9) Abs were purchased fromAbcam with the relevant IgG isotype provided as a kit. HLA typing ofindividuals was carried out using molecular methodology (29).

Proliferation assay and cytokine profile

Isolated PBMCs were stimulated to proliferate with pure-grade anti-CD3/CD28 Abs (6 mg/well) (eBioscience) and cocultured with six differentconcentrations of irradiated (3000 rad) MSCs for 72 h. PBMC proliferationwas measured by CFSE incorporation (CellTrace CFSE, Molecular Probes,Invitrogen). Isolated PBMCs were incubated with CFSE (1 ml of 15 mM)for 6 min at a cell concentration of 203 106/ml, followed by one wash andreconstitution with RPMI complete medium (RPMI 1640, 13; Cellgro;supplemented with 10% FBS, 1% penicillin-streptomycin, and 1% gluta-mine). Concurrently, increasing concentrations of irradiated hPB-MSCsand BM-MSCs were plated in six replicates in a 96-well flat-bottomplate (Costar) for 1 h prior to the stimulation assay, to ensure MSC at-tachment to the well base of the plates. Negative and positive controlsrepresented unstimulated and stimulated PBMCs, respectively, both of

which were incubated without MSCs. CFSE-labeled PBMCs were addedto all wells. Anti-CD3/CD28 was added to all wells, except the negativecontrol wells, and these cultures were incubated for 72 h at 37˚C and 5%CO2. Cell proliferation was detected by flow cytometry as a percent di-vision from the original population. The supernatant from the wells wascollected to assess the production of proinflammatory cytokines IFN-g andTNF-a, using Luminex (Milliplex Map; Millipore). PBMC proliferationwas assessed with FlowJo software (Tree Star).

Cell killing assay

Human NK cells were isolated with an NK cell isolation kit (MiltenyiBiotec) and were immediately used as effector cells against target MSCsand fibroblasts as a control. hPB-MSC, BM-MSC, and fibroblast killingwas assessed using LIVE/DEAD Viability/Cytotoxicity Kit (MolecularProbes, Invitrogen Detection Technologies) according to the manu-facturer’s recommendations. Briefly, the two-color fluorescence cell vi-ability assay detects live and dead cells, using two probes that measurerecognized parameters for cell viability: intracellular esterase (calceinacetomethoxy [AM]) and plasma membrane integrity (ethidium homo-dimer), respectively (30). MSCs and fibroblasts were incubated withcalcein AM prior to the killing assay, because this intense green fluo-rescent dye is well retained within live cells. Effector NK cells werecoincubated with stained MSCs or fibroblasts at an increasing E:T ratiofor 2 h (1:1, 4:1, 16:1) at 37˚C and 5% CO2 in RPMI complete medium,supplemented with IL-2 (25 ng/ml). At the end of the incubation period,ethidium homodimer, which enters through the damaged membranes ofapoptotic cells and undergoes a 40-fold enhancement of red fluorescenceupon binding to nucleic acids, was added to the wells for 15 min at roomtemperature. Samples were measured using a dual-scanning microplatespectrofluorimeter (Versamax Microplate Reader) (excitation filter: 48569 nm; band-pass filter: 530 6 9 nm). Percent lysis was calculated withthe same formula used for the 51Cr assay and presented as follows: % CellLysis = (Experimental Wells 2 Spontaneous Release)/(Maximum Re-lease 2 Spontaneous Release) (30). Caspase-3 in MSCs, a GrB intracel-lular second messenger, was assessed using the Caspase-3 ColorimetricAssay (R&D Systems) according to the manufacturer’s recommendations.Briefly, lysed MSCs were tested for protease activity by the addition ofa caspase-specific peptide conjugated to the color reporter molecule p-nitroanaline. Cleavage of the peptide by caspase releases the chromo-phore p-nitroanaline, quantified spectrophotometrically at a wavelength of405 nm. The level of caspase enzymatic activity in the cell lysate is di-rectly proportional to the color reaction.

Small interfering RNA knockdown of PI-9 in hPB-MSCs andBM-MSCs

Small interfering RNA (siRNA) against PI-9 was purchased from Dhar-macon (SMARTpool-siRNA and Accell nontargeting pool as negativecontrol from Thermo Scientific). hPB-MSCs and BM-MSCs were trans-fected as directed in the manufacturer’s protocol. Briefly, the MSCs wereplated at 150,000/well in six-well plates. PI-9 and nonspecific scrambledsiRNA were individually mixed with Accell siRNA delivery medium(Thermo Scientific, Dharmacon Accell siRNA delivery media) and addedto MSCs to give a final concentration of 1 mmol/l per well. After 72 hincubation, MSCs were harvested, PI-9 expression was assessed by flowcytometry, and they were subjected to the killing assay as describedabove.

Transfection and infection of hPB-MSCs using the Phoenixamphotrophic packaging line

DNAwas diluted in Tris–EDTA buffer to a final concentration of 1 mg/ml(MIRG1; MIRG1–PI-9). Lipofectamine transduction of cells was per-formed according to the manufacturer’s protocol (Invitrogen). Trans-duction of samples at a final concentration of 1 mg/ml was made in 50 mlDMEM (without FBS) and incubated at room temperature for 5 min.Lipofectamine reagent was diluted to 2 ml/50 ml media. Diluted DNAwasmixed gently with lipofectamine reagent and then incubated for 20 min atroom temperature to allow for DNA–lipofectamine complexes to develop.Following incubation, the DNA–lipofectamine mixture was added to the60–70% confluent Phoenix amphotrophic cells (American Type CultureCollection) cultured in T150 flasks. Transduced Phoenix cells were left toincubate at 37˚C. After 24 h, the medium was changed, and cells wereincubated at 35˚C to induce production of viral packaging proteins con-taining the plasmid. Conditioned medium was collected every 8 h from thePhoenix cell cultures and transferred to human MSCs seeded in T75 flasks(at 70–80% confluence) and replaced with fresh media. Conditioned me-dium was removed every 8 h over a period of 3 d.

The Journal of Immunology 2253

by guest on June 6, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

Statistics

Data were expressed as themean6 SD of at least three separate experimentsperformed in duplicate. Student t test was used for comparison of results.

ResultsMSC characterization and differentiation

Cultured hPB-MSCs and BM-MSCs were characterized at passage4 and/or 5 with a panel of flow cytometry Abs for stem cells. We didnot find a significant difference in the positive MFI expression ofhPB-MSCs and BM-MSCs for various MSCmarkers (Fig. 1A, 1B).Both hPB-MSCs and BM-MSCs showed comparable MFI ex-pression levels for CD29 (200 6 6 versus 190 6 9), CD44 (480 612 versus 466 6 15), CD49e (160 6 6 versus 138 6 6), CD90(926 6 13 versus 948 6 7), CD166 (41 6 3 versus 40 6 6),CD105 (46 6 5 versus 44 6 2), and CD73 (2406 9 versus 238 63), for hPB-MSCs and BM-MSCs, respectively. Hematopoieticlineage markers CD45 and CD34 and costimulatory moleculeCD80 were negative in both MSC types (Fig. 1A–C). Both MSCtypes shared similar fibroblastic characteristic morphology, asobserved under light microscopy (3400 magnification) (Fig. 1D,1E). In addition, hPB-MSCs and BM-MSCs differentiated in-dividually into chondrocytes, osteocytes, and adipocytes uponstimulation with a commercially available differentiation mediumfor each cell lineage (Fig. 1D, 1E).

Proliferation assay and cytokine profile

To examine the immunomodulatory role of hPB-MSCs vis-a-visthat of BM-MSCs, increasing concentrations of both cell types

were added to an anti-CD3/CD28 T cell proliferation assay and

assessed by flow cytometry using CFSE staining. Compared with

the positive control, hPB-MSCs and BM-MSCs similarly and

significantly suppressed T cell proliferation in a dose-dependent

manner (p , 0.02) (Fig. 2A, 2B). We also compared the role of

hPB-MSCs with that of BM-MSCs in suppressing the production

of two proinflammatory cytokines, IFN-g and TNF-a. As shown

with both cytokines, both MSC types exhibited an analogous and

significant suppression in comparison with the positive control

(p , 0.01) (Fig. 2C, 2D).

PI-9 constitutive expression in MSCs and its protective roleagainst NK cell-mediated cytotoxicity

Immunostaining and FACS analysis revealed a constitutive ex-pression of intracellular PI-9 in hPB-MSCs and BM-MSCs, and no

expression in fibroblasts (Fig. 3A, 3B). To assess the protective role

of PI-9 against NK cell-mediated cytotoxicity, freshly isolated NK

cells were incubated with HLA-mismatched MSCs and fibroblasts

for 2 h in a killing assay. The purity of NK cells was .95% (data

not shown). As shown in Fig. 4A, a proportional increase in per-

cent killing of hPB-MSCs and BM-MSCs was observed, with

FIGURE 1. Flow cytometry assessment and differentiation of hPB-MSCs compared with BM-MSCs. Both hPB-MSCs and BM-MSCs showed similar

surface marker expression for CD29, CD44, CD49e, CD73, CD90, CD166, and CD105 markers. MSCs lacked hematopoietic markers CD34 and CD45 and

costimulatorymolecules CD80 (A,B). Data are shown as themeanvalue6 SD of five experiments, n = 6 donors (C). Both cell types showed typical fibroblastic

morphology and differentiated into chondrocytes (Alcian Blue/PAS), osteocytes (Alizarin Red S dye), and adipocytes (Oil red O staining), as shown using

3400 magnification (D, E).

2254 SERPINB9 PROTECTS HUMAN PERIPHERAL BLOOD MSCs

by guest on June 6, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

FIGURE 2. hPB-MSCs and BM-MSCs similarly suppress T cell proliferation in anti-CD3/CD28 stimulation assays. Anti-CD3/CD28 stimulation assay

on PBMCs was performed in the presence of increasing concentration of hPB-MSCs and BM-MSCs. Proliferating cells were stained with CFSE and

analyzed by flow cytometry as percent division (A, B). The histograms show representative proliferation, and the bar graph is the percent division of

proliferating T cells. hPB-MSCs and BM-MSCs similarly suppressed PBMC proliferation in a dose-dependent manner. *p , 0.02. C and D, Both MSC

types significantly suppressed the production of inflammatory cytokines TNF-a and IFN-g. Data represent the average of five experiments, with each

parameter performed in six replicates. *p, 0.01. NEG CTRL, negative control (PBMCs in medium); POS CTRL, positive control (PBMCs cocultured with

anti-CD3/CD28).

The Journal of Immunology 2255

by guest on June 6, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

significant increase attained at an E:T ratio of 16:1 compared with1:1 (376 3.8 versus 2.46 1.6 for hPB-MSCs and 406 4.1 versus1.8 6 0.7 for BM-MSCs, respectively). In contrast, the fibroblastcells revealed a significantly higher percent killing at all E:T ratiosstudied in the killing assay (p , 0.02) (Fig. 4A).To determine the role of GrB in the NK-mediated killing of

MSCs and fibroblasts, we used compound 19, a potent and specificGrB inhibitor (GrB-i) in our killing assay (31). We added threeconcentrations (25 mM, 50 mM, and 75 mM) of GrB-i to freshlyisolated NK cells for 45 min prior to the killing assay. A signifi-cant dose-dependent drop in percent killing occurred in all threecell types at GrB-i concentrations of 50 mM and 75 mM, althougha decreasing trend was evident at 25 mM (Fig. 4B) (p , 0.02).Because GrB induces the activation of caspase-3 in the targetcells, we measured the activity of caspase-3 in MSCs and fibro-blasts, using labeled peptide substrate. When MSCs and fibro-blasts were subjected to NK cells, caspase-3 activity in the threetarget cell types was significantly higher than in the target cellsalone (p , 0.02) (Fig. 4C). In contrast, a significantly highercaspase-3 activity was noted in fibroblasts compared with bothMSC types subjected to NK cells (p , 0.02) (Fig. 4C). NK cellsalone, like MSCs alone, revealed relatively no caspase activity(data not shown). These data imply that PI-9 protects MSCs fromNK-mediated apoptosis by suppressing GrB activity.

Downregulating PI9 using siRNA enhances MSC susceptibilityto death by NK cells

To assess the significance of PI-9 in protecting MSCs from GrB-mediated cytotoxicity, we used specific siRNA for PI-9 to down-regulate PI-9 expression in MSCs. Compared with MSCs trans-fected with nonspecific scrambled siRNA, PI-9 siRNA transfection

resulted in a significant reduction in the expression of PI-9 (Fig.5A). hPB-MSCs and BM-MSCs were individually transfected withPI-9 siRNA and scrambled siRNA for 72 h. The expression of PI-9was assessed by flow cytometry after 72 h of MSC treatment.Subsequently, hPB-MSCs and BM-MSCs were trypsinized andincubated with NK cells at three E:T ratios of 1:1, 4:1, and 16:1 inthe killing assay, as mentioned above. Percent killing was com-pared with that in MSCs transfected with scrambled siRNA.Results revealed a significantly higher percent killing in theknockdown experiments than in MSCs transfected with scrambledsiRNA in all three E:T ratios assessed (Fig. 5B).

Overexpression of PI-9 protects hPB-MSCs from killing

To assess the independent role of PI-9 in protecting hPB-MSCsfrom cytotoxic killing, we transduced our hPB-MSCs with theMIRG1–PI-9 plasmid, to increase PI-9 expression, and challengedthem with YT cells, a human NK cell line. Using flow cytometry,we first assessed the transduced hPB-MSCs for the expressionof GFP with MIRG1 (empty vector) and MIRG1–PI-9, comparedwith nontransduced hPB-MSCs. GFP expression in hPB-MSC–MIRG1–GFP and hPB-MSC–PI-9–GFP was compared with thatin nontransduced MSC (hPB-MSCs) (Fig. 6A). Overall expressionof GFP in these pools of hPB-MSC–MIRG1–GFP and hPB-MSC–PI-9–GFP was 97 6 5% and 91 6 6%, respectively (Fig. 6A).Subsequently, the three cell types were incubated with YT cells toassess the protective role from overexpression of PI-9. It is worthnoting that NK cells and the YT cell line undergo similar mech-anisms of cytotoxicity (32), are stimulated by common surfaceAgs (33), and express similar surface markers (34). Therefore,both cell types are used interchangeably in the literature. hPB-MSC, hPB-MSC–MIRG1, and hPB-MSC–MIRG1–PI-9 were

FIGURE 3. Constitutive intracellular PI-9 detected in hPB-MSCs and BM-MSCs, but not fibroblasts. Immunostaining with PI-9–PE versus nuclear

staining with DAPI is shown at 3200 magnification, compared with its isotype (A). In contrast to fibroblasts, hPB-MSCs and BM-MSCs showed a strong

cytoplasmic staining for PI-9 (solid red), in comparison with the isotype. Flow cytometry revealed a constitutive intracellular PI-9 expression (red line)

versus isotype (blue line) (B).

2256 SERPINB9 PROTECTS HUMAN PERIPHERAL BLOOD MSCs

by guest on June 6, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

loaded with calcein AM and challenged for a standardized 8-hperiod with YT cells. When we subjected them to higher E:Tratios of YT cells, both hPB-MSC and hPB-MSC–MIRG1 cellsshowed increased lysis compared with hPB-MSC–MIRG1–PI-9cells. MSC-MIRG1–PI-9 cells showed a significant inhibition ofpercent killing at the highest ratio (50:1) when compared withhPB-MSC and hPB-MSC–MIRG1 cells (p , 0.01) (Fig. 6B).Therefore, overexpression of PI-9 can protect MSCs from cell-mediated lysis.

DiscussionRecent years have seen an unprecedented increase in the numberof clinical trials in which MSCs are used to treat various immune-mediated diseases andmetabolic and genetic disorders, as well as intissue repair (35). Indeed, to date, .160 clinical trials using MSCshave been listed at www.clinicaltrials.gov.Due to ease of administration, the ability to use controlled or

optimized cells, eliminating the need for dysfunctional autologoussenile MSCs or dysfunctional MSCs with genetic predisposition, aswell as the ability to administer multiple doses, allogeneic MSCshave increasingly been considered for treating various diseases(36). In this regard, we have recently shown that allogeneic MSCs,but not autologous MSCs, reverse autoimmune diabetes in NODmice (4, 37).Because the isolation of abundant MSC numbers from peripheral

blood would set the stage for an improved cell-based therapy, theuse of bone marrow as a source requires further analysis. Currently,bone marrow represents the most abundant source of MSCs forexamining their function experimentally as well as their therapeutic

roles in clinical studies. hPB-MSCs are lower in concentration thanare BM-MSCs in their respective tissues and are rarely detectedin normal individuals (9, 10, 38, 39). CFU-fibroblastic have beendetected in peripheral blood of humans, although they are few innumber and difficult to maintain and culture, compared with bonemarrow CFU-fibroblastic (11, 40). In instances in which stimu-lants such as G-CSF or GM-CSF have been used, or in patho-logical conditions in which intrinsic stimuli have been present(e.g., patients with cancer, burns, allograft rejection, or stroke), en-hanced release and recovery of MSCs were observed. hPB-MSCshave been isolated and expanded with various levels of success (9,12, 41). A total of 500 ml of blood was needed to obtain almost10,000 mesenchymal precursor cells without growth factors,leading to an MSC culture capable of differentiating into the threemesodermal lineages (42).Our data suggest that higher cell confluence, which increased the

likelihood of generating MSCs in primary cultures, was obtainedfrom samples with a greater amount of blood drawn and highernumbers of PBMCs isolated, as well as the addition of b-FGFat a concentration of 6 ng/ml. The supplementation of b-FGF toculture medium enhanced the growth and proliferation of MSCsobtained from low-yield areas, such as dental pulp and synovium-derived stem cells (43, 44). Interestingly, Bian et al. (45) recentlyshowed that peripheral blood of patients with bone sarcomas hadhigher numbers of CD105+ MSCs than did the peripheral blood ofhealthy individuals. This increase was associated with increasedgrowth factors, such as hepatocyte growth factor and vascularendothelial growth factor, in the plasma (45). Therefore, it is ofvalue for future studies to better examine the synergistic effects of

FIGURE 4. Killing assay for hPB-MSCs and BM-MSCs and fibroblasts with NK cells. A, Freshly isolated NK cells were incubated with hPB-MSCs,

BM-MSCs, and fibroblasts for 2 h at three E:T ratios (1:1, 4:1, and 16:1) (n = 3 experiments). Significantly higher percent killing was observed for

fibroblasts versus K562 cells, hPB-MSCs, and BM-MSCs at all three E:T ratios tested (p , 0.02). Percent killing was significantly higher for both hPB-

MSCs and BM-MSCs at an E:T of 16:1 compared with lower E:T ratios of 4:1 and 1:1. *p , 0.02. B, NK cells incubated with three GrB-i concentrations

(25, 50, and 75 mM) for 45 min prior to coincubation with hPB-MSCs, BM-MSCs, and fibroblasts in the killing assay revealed a significant drop in percent

killing with increasing GrB-i concentrations. *p , 0.02. C, Caspase-3 activity in MSCs and fibroblasts alone and coincubated with NK cells (MSC + NK),

compared with MSCs alone, was assessed after 45 min into the killing assay. Caspase-3 in fibroblasts subjected to NK cells revealed significantly higher

activity than was found in both MSC groups subjected to NK cells. *p , 0.02.

The Journal of Immunology 2257

by guest on June 6, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

various growth factors in optimizing hPB-MSC isolation. In ourstandardization process, we also experienced difficulties in iso-lating MSCs from nine donors. We learned that there was notabledonor variability in the numbers of PBMCs isolated from blood. In

three donors of nine, the initial number of MSCs was below op-timal to sustain a culture, and thus we were unable to isolatedMSCs from them in repeated experiments. These individuals mayneed a stimulus, such as G-CSF or GM-CSF, to increase the

FIGURE 5. hPB-MSCs and BM-MSCs

show higher percent killing when PI-9 is

knocked down. A, Both nonspecific scram-

bled siRNA and PI-9 siRNA were added to

BM-MSCs for 72 h prior to the assess-

ment of PI-9 expression by flow cytometry.

Results revealed a significant reduction in

PI-9 protein expression in the PI-9 siRNA-

treated MSCs (right), compared with the

scrambled siRNA-treated MSCs (red line

represents the PI-9 expression and blue line

the isotype control). B, Coincubation of

NK cells with hPB-MSCs and BM-MSCs

knocked down for PI-9 by PI-9 siRNA

revealed a significant increase in percent

killing with increasing E:T ratio (1:1; 4:1;

16:1), compared with their relevant trans-

fected cells with scrambled siRNA. *p ,0.02.

FIGURE 6. Overexpression of PI-9

affords protection against YT killing. hPB-

MSCs were infected with either MIRG1

(empty vector) or MIRG1–PI-9. GFP ex-

pression in hPB-MSC–MIRG1–GFP and

hPB-MSC–MIRG1–PI-9 cells was com-

pared with that in uninfected MSCs (hPB-

MSCs). Overall expression of GFP in these

pools of hPB-MSC–MIRG1–GFP and hPB-

MSC–PI-9–GFP was 97 6 5% and 91 66%, respectively. A, Both hPB-MSC and

hPB-MSC–MIRG1 cells showed increased

lysis over that found in MSC–MIRG1–PI-9

cells (n = 3 experiments). B, MSC–

MIRG1–PI-9 cells showed reduced percent

killing when compared with MSC and

MSC-MIRG1 cells. *p , 0.01.

2258 SERPINB9 PROTECTS HUMAN PERIPHERAL BLOOD MSCs

by guest on June 6, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

shedding of bone marrow progenitors to the periphery. Futurestudies to better comprehend the source of variability in thenumber of progenitors among individuals and innovative tech-nology to improve their isolation from blood will greatly con-tribute to the success of isolating hPB-MSCs. Although muchwork is needed to further optimize the likelihood of recoveringMSCs from peripheral blood, our data indicate that successfullyretrieved MSCs from peripheral blood shared similar surfacemarkers with BM-MSCs, as well as differentiation potential andimmunomodulatory characteristics (46, 47).The immunomodulatory effect of hPB-MSCs was assessed using

the anti-CD3/CD28 T cell stimulation assay. Our results revealedthat hPB-MSCs showed a dose-dependent antiproliferative andanti-inflammatory effect. The data also show that the immuno-suppressive ability of hPB-MSCs is similar to that of BM-MSCsand MSCs (48–50). In addition, as cytokines play key roles inthe pathogenesis of various diseases, we assessed the anti-inflammatory effect of our hPB-MSCs on IFN-g and TNF-a inthese proliferation assays (51). The hPB-MSCs suppressed theproduction of these cytokines in a fashion comparable to the effectof BM-MSCs. These data indicate that immunomodulatory effectsof hPB-MSCs were in concordance with those of BM-MSCs (13).We further investigated the presence of PI-9 in MSCs as a

strategic component for immuno-evasion and a critical factor forthe substantial improvement of cell-based therapy. Recent reportsindicate that, despite being hypoimmunogenic, transplanted allo-geneic MSCs are recognized by the host immune response and areeventually rejected (52). Several defense mechanisms have beenreported as the means whereby MSCs avoid cytotoxicity, but noneof these mechanisms fully explain how they evade the hostimmune response (53–55). Lack of MHC class II expression orcostimulatory molecules could contribute to MSC hypoimmuno-genicity. Nevertheless, Western blotting of cell lysates has shownthat MSCs contain intracellular deposits of class II alloantigens(54, 55). Furthermore, MSCs stimulated by IFN-g mimicking in-flammatory milieu exhibit induction of class II MHC expression(56).In this study, we report the constitutive intracellular expression

of PI-9 in hPB-MSCs and BM-MSCs, but not in fibroblasts. PI-9 isa specific GrB inhibitor and is present in CTLs, NK cells, APCs,endothelial cells, and immunoprivileged sites (57). The over-expression of PI-9 contributes to protection of CTLs from GrB-mediated apoptosis (20). Our data show that PI-9 contributesto the protection of hPB-MSCs and BM-MSCs from NK cell-mediated cytotoxicity, whereas lack of PI-9 was associated withan increased percent killing of fibroblasts. In line with previousreports, our results corroborate the induction of lytic activityagainst MSCs by freshly isolated NK cells at high E:T ratios (58).The specificity of GrB-mediated killing was demonstrated by

blocking the lytic capacity of NK cells with our GrB-i. The sig-nificant decrease in percent lysis of both MSC types and fibroblastsin a dose-dependent manner with GrB-i underscores the essentialrole of GrB in the killing of MSCs, thereby emphasizing thefundamental protective role of PI-9 for MSC survival. Evidence ofGrB activity in MSCs subjected to NK cells was associated withenhanced caspase-3 activity, compared with that in MSCs andfibroblast controls, which signifies a GrB–caspase-3–dependentkilling mechanism. Furthermore, when PI-9 was knocked downby siRNA, percent killing was significantly higher in comparisonwith that in controls. The percent killing of MSCs was inverselyproportional to the expression of PI-9 in MSCs. These data showthat in the absence of the protective effect of PI-9, cells are ren-dered susceptible to GrB-induced apoptosis. Furthermore, theseresults agreed with our recent work on the protective role of Spi6,

the mouse homolog to PI-9, in mouse MSCs (5). In our previouswork, we generated CTL machinery by sensitizing BALB/c micevia a skin allograft from a C57BL/6 donor. This model mimicsa scenario in which allogeneic wild-type MSCs become the targetfor CTL killing following administration. In this report, we haveprimarily used NK cells, which use similar GrB machinery.However, NK cells constitute a major component of the innateimmune system, without engaging TCRs, as opposed to CTLs.We transduced hPB-MSCs with the MIRG1–PI-9 plasmid to

enhance PI-9 production. Overexpression of PI-9 in NK-treatedhPB-MSCs revealed an almost 50% reduction in cell lysis, com-pared with that in untreated MSCs. In contrast, lack of PI-9 ex-pression and increased caspase-3 activity in fibroblasts evidentlyreinforce the protective role of PI-9 against NK cell-mediatedkilling. It is important to note that these data not only are im-portant for the cycles of MSC therapy, but also will establisha potential novel ground for developing anti–PI-9 strategies ininstances in which unwanted growth of tumor occurs in the courseof MSC therapy. These new data advance our understanding ofthe immunomodulatory properties of hPB-MSCs and enable themodulation of a newly discovered pathway in hPB-MSCs toprolong their survival postadministration. These findings will re-duce the need for repetitive injection, while increasing the efficacyof the cells, and will substantially contribute to the improvementof cell therapies using MSCs.

AcknowledgmentsWe thank Zeina Chaptini, the laboratory manager, for dedicated managerial

logistics in support of the project.

DisclosuresThe authors have no financial conflicts of interest.

References1. Tyndall, A., and F. A. Houssiau. 2010. Mesenchymal stem cells in the treatment

of autoimmune diseases. Ann. Rheum. Dis. 69: 1413–1414.2. Gonzalez-Rey, E., M. A. Gonzalez, N. Varela, F. O’Valle, P. Hernandez-Cortes,

L. Rico, D. Buscher, and M. Delgado. 2010. Human adipose-derived mesen-chymal stem cells reduce inflammatory and T cell responses and induce regu-latory T cells in vitro in rheumatoid arthritis. Ann. Rheum. Dis. 69: 241–248.

3. Parekkadan, B., A. W. Tilles, and M. L. Yarmush. 2008. Bone marrow-derivedmesenchymal stem cells ameliorate autoimmune enteropathy independently ofregulatory T cells. Stem Cells 26: 1913–1919.

4. Fiorina, P., M. Jurewicz, A. Augello, A. Vergani, S. Dada, S. La Rosa, M. Selig,J. Godwin, K. Law, C. Placidi, et al. 2009. Immunomodulatory function of bonemarrow-derived mesenchymal stem cells in experimental autoimmune type 1diabetes. J. Immunol. 183: 993–1004.

5. El Haddad, N., D. Heathcote, R. Moore, S. Yang, J. Azzi, B. Mfarrej,M. Atkinson, M. H. Sayegh, J. S. Lee, P. G. Ashton-Rickardt, and R. Abdi. 2011.Mesenchymal stem cells express serine protease inhibitor to evade the hostimmune response. Blood 117: 1176–1183.

6. Spaggiari, G. M., A. Capobianco, S. Becchetti, M. C. Mingari, and L. Moretta.2006. Mesenchymal stem cell-natural killer cell interactions: evidence that ac-tivated NK cells are capable of killing MSCs, whereas MSCs can inhibit IL-2-induced NK-cell proliferation. Blood 107: 1484–1490.

7. Prigione, I., F. Benvenuto, P. Bocca, L. Battistini, A. Uccelli, and V. Pistoia.2009. Reciprocal interactions between human mesenchymal stem cells andgammadelta T cells or invariant natural killer T cells. Stem Cells 27: 693–702.

8. Erices, A., P. Conget, and J. J. Minguell. 2000. Mesenchymal progenitor cells inhuman umbilical cord blood. Br. J. Haematol. 109: 235–242.

9. Kassis, I., L. Zangi, R. Rivkin, L. Levdansky, S. Samuel, G. Marx, andR. Gorodetsky. 2006. Isolation of mesenchymal stem cells from G-CSF-mobilized human peripheral blood using fibrin microbeads. Bone MarrowTransplant. 37: 967–976.

10. Puissant, B., C. Barreau, P. Bourin, C. Clavel, J. Corre, C. Bousquet, C. Taureau,B. Cousin, M. Abbal, P. Laharrague, et al. 2005. Immunomodulatory effect ofhuman adipose tissue-derived adult stem cells: comparison with bone marrowmesenchymal stem cells. Br. J. Haematol. 129: 118–129.

11. Fernandez, M., V. Simon, G. Herrera, C. Cao, H. Del Favero, and J. J. Minguell.1997. Detection of stromal cells in peripheral blood progenitor cell collectionsfrom breast cancer patients. Bone Marrow Transplant. 20: 265–271.

12. Mansilla, E., G. H. Marın, H. Drago, F. Sturla, E. Salas, C. Gardiner, S. Bossi,R. Lamonega, A. Guzman, A. Nunez, et al. 2006. Bloodstream cells pheno-

The Journal of Immunology 2259

by guest on June 6, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

typically identical to human mesenchymal bone marrow stem cells circulatein large amounts under the influence of acute large skin damage: new evidencefor their use in regenerative medicine. Transplant. Proc. 38: 967–969.

13. Ryan, J. M., F. P. Barry, J. M. Murphy, and B. P. Mahon. 2005. Mesenchymalstem cells avoid allogeneic rejection. J. Inflamm. (Lond.) 2: 8.

14. Francois, M., R. Romieu-Mourez, S. Stock-Martineau, M. N. Boivin,J. L. Bramson, and J. Galipeau. 2009. Mesenchymal stromal cells cross-presentsoluble exogenous antigens as part of their antigen-presenting cell properties.Blood 114: 2632–2638.

15. Caligiuri, M. A. 2008. Human natural killer cells. Blood 112: 461–469.16. Biron, C. A. 2010. More things in heaven and earth: defining innate and adaptive

immunity. Nat. Immunol. 11: 1080–1082.17. Kagi, D., F. Vignaux, B. Ledermann, K. Burki, V. Depraetere, S. Nagata,

H. Hengartner, and P. Golstein. 1994. Fas and perforin pathways as majormechanisms of T cell-mediated cytotoxicity. Science 265: 528–530.

18. Millard, P. J., M. P. Henkart, C. W. Reynolds, and P. A. Henkart. 1984. Purifi-cation and properties of cytoplasmic granules from cytotoxic rat LGL tumors. J.Immunol. 132: 3197–3204.

19. Lieberman, J. 2003. The ABCs of granule-mediated cytotoxicity: new weaponsin the arsenal. Nat. Rev. Immunol. 3: 361–370.

20. Hirst, C. E., M. S. Buzza, C. H. Bird, H. S. Warren, P. U. Cameron, M. Zhang,P. G. Ashton-Rickardt, and P. I. Bird. 2003. The intracellular granzyme B in-hibitor, proteinase inhibitor 9, is up-regulated during accessory cell maturationand effector cell degranulation, and its overexpression enhances CTL potency. J.Immunol. 170: 805–815.

21. Zhang, M., S. M. Park, Y. Wang, R. Shah, N. Liu, A. E. Murmann, C. R. Wang,M. E. Peter, and P. G. Ashton-Rickardt. 2006. Serine protease inhibitor 6 protectscytotoxic T cells from self-inflicted injury by ensuring the integrity of cytotoxicgranules. Immunity 24: 451–461.

22. Zhang, M., N. Liu, S. M. Park, Y. Wang, S. Byrne, A. E. Murmann, S. Bahr,M. E. Peter, S. T. Olson, A. Belaaouaj, and P. G. Ashton-Rickardt. 2007. Serineprotease inhibitor 6-deficient mice have increased neutrophil immunity toPseudomonas aeruginosa. J. Immunol. 179: 4390–4396.

23. Bladergroen, B. A., M. C. Strik, N. Bovenschen, O. van Berkum, G. L. Scheffer,C. J. Meijer, C. E. Hack, and J. A. Kummer. 2001. The granzyme B inhibitor,protease inhibitor 9, is mainly expressed by dendritic cells and at immune-privileged sites. J. Immunol. 166: 3218–3225.

24. Abdullah, Z., T. Saric, H. Kashkar, N. Baschuk, B. Yazdanpanah,B. K. Fleischmann, J. Hescheler, M. Kronke, and O. Utermohlen. 2007. Serpin-6expression protects embryonic stem cells from lysis by antigen-specific CTL. J.Immunol. 178: 3390–3399.

25. Sun, J., C. H. Bird, V. Sutton, L. McDonald, P. B. Coughlin, T. A. De Jong,J. A. Trapani, and P. I. Bird. 1996. A cytosolic granzyme B inhibitor related tothe viral apoptotic regulator cytokine response modifier A is present incytotoxic lymphocytes. J. Biol. Chem. 271: 27802–27809.

26. Medema, J. P., J. de Jong, L. T. Peltenburg, E. M. Verdegaal, A. Gorter,S. A. Bres, K. L. Franken, M. Hahne, J. P. Albar, C. J. Melief, and R. Offringa.2001. Blockade of the granzyme B/perforin pathway through overexpression ofthe serine protease inhibitor PI-9/SPI-6 constitutes a mechanism for immuneescape by tumors. Proc. Natl. Acad. Sci. USA 98: 11515–11520.

27. Abdelhaleem, M. 2007. Cytotoxic granule disruption is a late event inchemotherapy-induced apoptosis in natural killer YT cells. Exp. Mol. Pathol. 83:112–114.

28. Pittenger, M. F., and D. M. Helfman. 1992. In vitro and in vivo characterizationof four fibroblast tropomyosins produced in bacteria: TM-2, TM-3, TM-5a, andTM-5b are co-localized in interphase fibroblasts. J. Cell Biol. 118: 841–858.

29. Delaney, M., C. S. Cutler, R. L. Haspel, B. Y. Yeap, S. L. McAfee, B. R. Dey,E. Attar, G. Kao, E. P. Alyea, J. Koreth, et al. 2009. High-resolution HLAmatching in double-umbilical-cord-blood reduced-intensity transplantation inadults. Transfusion 49: 995–1002.

30. Wang, X. M., P. I. Terasaki, G. W. Rankin, Jr., D. Chia, H. P. Zhong, andS. Hardy. 1993. A new microcellular cytotoxicity test based on calcein AMrelease. Hum. Immunol. 37: 264–270.

31. Willoughby, C. A., H. G. Bull, M. Garcia-Calvo, J. Jiang, K. T. Chapman, andN. A. Thornberry. 2002. Discovery of potent, selective human granzyme Binhibitors that inhibit CTL mediated apoptosis. Bioorg. Med. Chem. Lett. 12:2197–2200.

32. Jones, G. J., J. C. Wiseman, K. J. Marr, S. Wei, J. Y. Djeu, and C. H. Mody. 2009.In contrast to anti-tumor activity, YT cell and primary NK cell cytotoxicity forCryptococcus neoformans bypasses LFA-1. Int. Immunol. 21: 423–432.

33. Bauer, S., V. Groh, J. Wu, A. Steinle, J. H. Phillips, L. L. Lanier, and T. Spies.1999. Activation of NK cells and T cells by NKG2D, a receptor for stress-inducible MICA. Science 285: 727–729.

34. Shirakawa, F., Y. Tanaka, S. Eto, H. Suzuki, J. Yodoi, and U. Yamashita. 1986.Effect of interleukin 1 on the expression of interleukin 2 receptor (Tac antigen)on human natural killer cells and natural killer-like cell line (YT cells). J.Immunol. 137: 551–556.

35. Hodgkinson, C. P., J. A. Gomez, M. Mirotsou, and V. J. Dzau. 2010. Geneticengineering of mesenchymal stem cells and its application in human diseasetherapy. Hum. Gene Ther. 21: 1513–1526.

36. Zhang, L. S., Q. F. Liu, K. Huang, Y. Zhang, Z. P. Fan, and S. L. Huang. 2009.[Mesenchymal stem cells for treatment of steroid-resistant chronic graft-versus-host disease]. Zhonghua Nei Ke Za Zhi 48: 542–546.

37. Abdi, R., P. Fiorina, C. N. Adra, M. Atkinson, and M. H. Sayegh. 2008.Immunomodulation by mesenchymal stem cells: a potential therapeutic strategyfor type 1 diabetes. Diabetes 57: 1759–1767.

38. Wexler, S. A., C. Donaldson, P. Denning-Kendall, C. Rice, B. Bradley, andJ. M. Hows. 2003. Adult bone marrow is a rich source of human mesenchymal“stem” cells but umbilical cord and mobilized adult blood are not. Br. J. Hae-matol. 121: 368–374.

39. Tyndall, A., and A. Uccelli. 2009. Multipotent mesenchymal stromal cells forautoimmune diseases: teaching new dogs old tricks. Bone Marrow Transplant.43: 821–828.

40. Tondreau, T., N. Meuleman, A. Delforge, M. Dejeneffe, R. Leroy, M. Massy,C. Mortier, D. Bron, and L. Lagneaux. 2005. Mesenchymal stem cells derivedfrom CD133-positive cells in mobilized peripheral blood and cord blood: pro-liferation, Oct4 expression, and plasticity. Stem Cells 23: 1105–1112.

41. Lund, T. C., J. Tolar, and P. J. Orchard. 2008. Granulocyte colony-stimulatingfactor mobilized CFU-F can be found in the peripheral blood but have limitedexpansion potential. Haematologica 93: 908–912.

42. Zvaifler, N. J., L. Marinova-Mutafchieva, G. Adams, C. J. Edwards, J. Moss,J. A. Burger, and R. N. Maini. 2000. Mesenchymal precursor cells in the bloodof normal individuals. Arthritis Res. 2: 477–488.

43. Kim, J., J. W. Kang, J. H. Park, Y. Choi, K. S. Choi, K. D. Park, D. H. Baek,S. K. Seong, H. K. Min, and H. S. Kim. 2009. Biological characterization oflong-term cultured human mesenchymal stem cells. Arch. Pharm. Res. 32: 117–126.

44. Morito, A., Y. Kida, K. Suzuki, K. Inoue, N. Kuroda, K. Gomi, T. Arai, andT. Sato. 2009. Effects of basic fibroblast growth factor on the development of thestem cell properties of human dental pulp cells. Arch. Histol. Cytol. 72: 51–64.

45. Bian, Z. Y., G. Li, Y. K. Gan, Y. Q. Hao, W. T. Xu, and T. T. Tang. 2009. In-creased number of mesenchymal stem cell-like cells in peripheral blood ofpatients with bone sarcomas. Arch. Med. Res. 40: 163–168.

46. Horwitz, E. M., K. Le Blanc, M. Dominici, I. Mueller, I. Slaper-Cortenbach,F. C. Marini, R. J. Deans, D. S. Krause, A. Keating; International Society forCellular Therapy. 2005. Clarification of the nomenclature for MSC: The In-ternational Society for Cellular Therapy position statement. Cytotherapy 7: 393–395.

47. Prockop, D. J., I. Sekiya, and D. C. Colter. 2001. Isolation and characterizationof rapidly self-renewing stem cells from cultures of human marrow stromal cells.Cytotherapy 3: 393–396.

48. Zhao, S., R. Wehner, M. Bornhauser, R. Wassmuth, M. Bachmann, andM. Schmitz. 2010. Immunomodulatory properties of mesenchymal stromal cellsand their therapeutic consequences for immune-mediated disorders. Stem CellsDev. 19: 607–614.

49. Nauta, A. J., and W. E. Fibbe. 2007. Immunomodulatory properties of mesen-chymal stromal cells. Blood 110: 3499–3506.

50. Najar, M., R. Rouas, G. Raicevic, H. I. Boufker, P. Lewalle, N. Meuleman,D. Bron, M. Toungouz, P. Martiat, and L. Lagneaux. 2009. Mesenchymal stro-mal cells promote or suppress the proliferation of T lymphocytes from cordblood and peripheral blood: the importance of low cell ratio and role of in-terleukin-6. Cytotherapy 11: 570–583.

51. Kunz, M., and S. M. Ibrahim. 2009. Cytokines and cytokine profiles in humanautoimmune diseases and animal models of autoimmunity. Mediators Inflamm.2009: 979258.

52. Chamberlain, G., J. Fox, B. Ashton, and J. Middleton. 2007. Concise review:mesenchymal stem cells: their phenotype, differentiation capacity, immunolog-ical features, and potential for homing. Stem Cells 25: 2739–2749.

53. Le Blanc, K., H. Samuelsson, B. Gustafsson, M. Remberger, B. Sundberg,J. Arvidson, P. Ljungman, H. Lonnies, S. Nava, and O. Ringden. 2007. Trans-plantation of mesenchymal stem cells to enhance engraftment of hematopoieticstem cells. Leukemia 21: 1733–1738.

54. Le Blanc, K., C. Tammik, K. Rosendahl, E. Zetterberg, and O. Ringden. 2003.HLA expression and immunologic properties of differentiated and un-differentiated mesenchymal stem cells. Exp. Hematol. 31: 890–896.

55. Klyushnenkova, E., J. D. Mosca, V. Zernetkina, M. K. Majumdar, K. J. Beggs,D. W. Simonetti, R. J. Deans, and K. R. McIntosh. 2005. T cell responses toallogeneic human mesenchymal stem cells: immunogenicity, tolerance, andsuppression. J. Biomed. Sci. 12: 47–57.

56. Le Blanc, K., and O. Ringden. 2007. Immunomodulation by mesenchymal stemcells and clinical experience. J. Intern. Med. 262: 509–525.

57. Bots, M., E. de Bruin, M. T. Rademaker-Koot, and J. P. Medema. 2007. Pro-teinase inhibitor-9 expression is induced by maturation in dendritic cells via p38MAP kinase. Hum. Immunol. 68: 959–964.

58. Sotiropoulou, P. A., S. A. Perez, A. D. Gritzapis, C. N. Baxevanis, andM. Papamichail. 2006. Interactions between human mesenchymal stem cells andnatural killer cells. Stem Cells 24: 74–85.

2260 SERPINB9 PROTECTS HUMAN PERIPHERAL BLOOD MSCs

by guest on June 6, 2018http://w

ww

.jimm

unol.org/D

ownloaded from