Mesenchymal Stem Cells Human Mesenchymal Stem Cells Support Unrelated Donor Hematopoietic Stem Cells and Suppress T-Cell Activation

Home , Mesenchymal stem cell

Mesenchymal stem cells Human mesenchymal stem cells support unrelated donor hematopoietic stem cells and suppress T-cell activation

B Maitra1, E Szekely1, K Gjini1, MJ Laughlin1,2, J Dennis3, SE Haynesworth3 andON Koc ¸1,2

1Comprehensive Cancer Center, Case Western Reserve University, University Hospitals of Cleveland, Cleveland, OH, USA; 2Division of Hematology/Oncology, Department of Medicine, Cleveland, OH, USA; and 3Department of Biology, Skeletal Research Center, Case Western Reserve University, Cleveland, OH, USA

Summary: interaction between bone marrow-derived mesenchymal stem cells (MSC) andthe hematopoietic stem cells to Bone marrow-derived mesenchymal stem cells (MSCs) are determine if MSCs could support allogeneic hematopoietic known to interact with hematopoietic stem cells (HSCs) stem cell transplantation by facilitating engraftment with- and immune cells, and represent potential cellular therapy out increasing the risk of GVHD. Several groups have to enhance allogeneic hematopoietic engraftment and described nonhematopoietic, plastic-adherent progenitor prevent graft-versus-host disease (GVHD). We investi- cells derived from adult human bone marrow aspirates, gated the role ofhuman MSCs in NOD-SCID mice which are capable of differentiating into mature mesen- repopulation by unrelated human hematopoietic cells and chymal cells.2–6 It is thought that these progenitors give rise studied the immune interactions between human MSCs to adventitial and other mesenchymal cells in the marrow and unrelated donor blood cells in vitro. When hemato- that fabricate the connective tissue scaffolding, and poietic stem cell numbers were limited, human engraft- produce cytokines, chemokines and extracellular matrix ment ofNOD-SCID mice was observed only after proteins that regulate hematopoietic homing andprolifera- coinfusion of unrelated human MSCs, but not with tion.7–10 Therefore, we hypothesizedthat MSCs could coinfusion of mouse mesenchymal cell line. Unrelated support the survival of hematopoietic stem cells andtested human MSCs did not elicit T-cell activation in vitro and this using an in vivo model of human hematopoietic stem suppressed T-cell activation by Tuberculin and unrelated cell engraftment andsurvival using NOD-SCID mice. allogeneic lymphocytes in a dose-dependent manner. Cell- Furthermore, several groups have recently shown that free MSC culture supernatant, mouse stromal cells and MSCs suppress T-lymphocyte proliferation in response to human dermal fibroblasts did not elicit this effect. These allogeneic stimuli.11–14 This suppressive effect of MSCs preclinical data suggest that unrelated, human bone appears to be partially mediated by soluble factors, marrow-derived, culture-expanded MSCs may improve although there are conflicting data. Therefore, we investi- the outcome ofallogeneic transplantation by promoting gatedthe immune interactions between MSCs and hematopoietic engraftment and limiting GVHD and their unrelatedallogeneic lymphocytes in vitro using Elispot therapeutic potential should be tested in clinic. assay. Bone Marrow Transplantation (2004) 33, 597–604. In experiments described here, we used a human bone doi:10.1038/sj.bmt.1704400 marrow-derived population of adherent cells described by Publishedonline 12 January 2004 Haynesworth et al3 that have extensive proliferative Keywords: stem cell; MSC; stromal cell; GVHD; im- capacity andability to differentiatealong the osteogenic, munosuppression chondrogenic and adipogenic lineages both in vitro and in vivo.15 In unstimulatedcultures, MSCs appear as fusiform fibroblasts with the expression of unique surface proteins Graft failure andgraft-versus-host disease(GVHD) remain (recognizedby monoclonal antibodiesSH2 andSH3) 16 not significant obstacles to a successful outcome in allogeneic foundon hematopoietic precursors. We have previously hematopoietic stem cell transplantation. There is a growing demonstrated the feasibility of culturing large numbers of interest in cotransplantation of allogeneic stromal and these cells ex vivo andsafety of their intravenous infusion hematopoietic progenitors to improve the patient outcome into patients.17,18 Therefore, our results are directly relevant after allogeneic transplantation.1 Here we investigatedthe to acceptable cellular therapy that can be testedin clinic. In this manuscript, we show that coinfusion of human MSCs andunrelatedumbilical cordblood(UCB) cells into NOD- SCID mice results in improvedfrequency anddegreeof Correspondence: Dr ON Koç, Case Western Reserve University, BRB-3 human engraftment when UCB dose is limiting. We also Hematology/Oncology, 10900 EuclidAve., Cleveland,OH 44106, USA; E-mail: [email protected] foundthat allogeneic-unrelatedMSCs donot elicit T-cell Received25 April 2003; accepted07 August 2003 activation in vitro. On the contrary, unrelateddonorMSCs Publishedonline 12 January 2004 inhibitedT-cell activation normally seen in mixedlympho- MSC support HSC B Maitra et al 598 cyte reactions. This effect couldnot be elucidatedwith centrifugation. All cells were usedfresh, without cryopre- MSC culture supernatant alone. servation.

Materials and methods NOD-SCID mice infusion and detection of human engraftment Mesenchymal stem cell procurement Human hematopoietic engraftment was evaluatedin Human MSC cultures were establishedas described irradiated NOD-SCID mice after infusing human UCB previously.17 Briefly, bone marrow was aspirated(10– cells with or without coinfusion of allogeneic human MSCs. 30 ml) under local anesthesia from healthy volunteer Recipient 6–8-week-oldNOD-SCID mice were given donors after consent approved by the IRB of the University sublethal total body irradiation (250 cGy) 24 h prior to Hospitals of Cleveland. Mononuclear cells (MNCs) were infusion using a cesium source. Cohorts of 4–8 mice were isolatedby Percoll ( d 1.073 gm/ml, Sigma, St Louis, MO, injectedintravenously with UCB cells at dosesof 2 to 6 6 USA) gradient centrifugation and plated at a density of 8 Â 10 per mouse, alone or mixedwith 1 Â 10 human 6 1 Â 106 cells per ml in 175 cm2 polystyrene flasks (Falcon, MSCs (unrelateddonorto UCB cells) or 1 Â 10 con- Beckton Dickinson Comp. Franklin Lakes, NJ, USA) in ditionally immortalized mouse mesenchymal progenitor Dulbecco’s minimum essential medium (DMEM) with a cells (clone BMC9).20 Mice were kept in microisolator cages selectedbatch of 10% fetal bovine serum (FBS) (Hyclone, andfedirradiatedfoodandwater containing penicillin. Logan, UT, USA). Cells were allowedto adherefor 72 h Mice were killed6–7 weeks after cell infusion andlong followed by the removal of nonadherent cells and media bones of hindlegs were extracted.Bone marrow cells were changes every 3–4 days. Adherent cells were removed with flushedinto mediumandusedfor flow cytometry or plated 0.05% trypsin-EDTA (Gibco) andreplated(passaged)at a for growth of human MSCs as described above. Human density of 1 Â 106 per 175 cm2. Multiple individual donor cell engraftment was detected by flow cytometry, using preparations were performedandsecondor thirdpassage anti-human CD45 antibody conjugated to PE (Pharmin- MSCs were used for individual experiments. All MSC gen, San Diego, CA, USA) andisotype control antibody, preparations were morphologically identical and uniformly Mouse IgG1-PE (Becton Dickinson, Parsippany, NJ, expressedSH2 andSH3 markers as describedpreviously. 17 USA). Percent human engraftment was calculatedafter subtraction of the backgrounddetectedincontrol mice and 0.5% was designated as the lower threshold for unequi- Dermal fibroblasts vocal human engraftment. Papillary dermal fibroblasts were a generous gift from Dr IA Schafer. Cells were preparedby dermatomingskin at a depth of about 0.3 mm. The epidermal sheet was removed Interferon-g Elispot assay andthe remaining dermalcomponent was cut into small ELISPOT plates (96-well) (Polyfiltronics, RocklandMA, pieces andplacedonto the surface of a plastic culture dish USA) were coatedwith capture antibodyfor human 19 and covered with culture medium as described. High interferon-gamma (2G1, Endogen, Woburn, MA, USA) passage cells, devoid of contaminating keratinocytes and as described elsewhere.21 In total, 150 000 responder vascular endothelial cells, were used in experiments. peripheral blood MNCs were added to each well in 100 ml of medium. The cells were activated in vitro with BMC9 mesenchymal progenitor cell line phytohemaglutinin (PHA 10 mg/ml final concentration; Sigma, St Louis, MO, USA) or unrelated(non-HLA- A conditionally immortalized marrow-derived mesen- matched) allogeneic cells. After 24 h, the plates were chymal cell clone, BMC9, was isolatedfrom a transgenic washed and biotinlated detection antibody (B133.5, Endo- mouse (H-2Kb-tsA58) containing a gene for conditional gen, 4 mg/ml) was added to the wells overnight at 41C. immortality as previously described.20 This cell line exhibits Strepavidin-HPR (Dako, Carpenteria, CA, USA) was then mesenchymal cell phenotypes, chondrocyte, adipocyte, added for 2 h at room temperature. The spots were stromal andosteoblast andsupports LTC-ICs out to 5 developed using 3-amino-9-ethylcarbazole (Pierce, Rock- weeks (unpublisheddata).The expression of large T- ford, IL, USA). The resulting spots were counted on a antigen does not impede the inherent hematopoietic computer-assistedELISPOT image analyzer Immunospot support capabilities of these cells. Nor does large T-antigen (Cellular Technology, Cleveland, OH, USA). expression artificially promote hematopoietic support as other clones of marrow cells show no LTC-IC support. MSC culture-conditioned medium preparation Umbilical cord blood cells Culture-conditioned medium consisted of MSC culture Human UCB was collectedinto sterile bags containing medium (DMEM with selected batch of 10% FBS) placed citrate dextrose after the delivery of full-term neonates over 70–80% confluent human MSCs for 24–48 h. Super- from healthy volunteer mothers. This procedure was natant was obtainedandfilteredtwice using 0.2 mm filters approvedby the IRB of the University Hospitals of to exclude cells. In some experiments, the medium was Cleveland. Red cells were sedimented using 6% dextran conditionedovera mixture of MSCs andunrelateddonor and MNCs were isolated using a density gradient peripheral bloodlymphocytes.

Bone Marrow Transplantation MSC support HSC B Maitra et al 599 Statistical analysis a 100 The results are expressedas mean 7s.e.m. Differences 54% between experimental conditions were analyzed by t test 80 (pairedwhen possible). A P-value of less than 0.05 was considered statistically signiﬁcant. 60

Results 40 Number

Human CD45 cell engraftment in NOD-SCID mice þ 20 To determine whether human MSCs supported engraftment andsurvival of human hematopoietic stem cells in 0 vivo, we infused2, 4 or 8 Â 106 human UCB MNCs with or 101 102 103 104 without 1 Â 106 unrelatedhuman MSCs into sublethally Log PF irradiated NOD-SCID mice. In mice receiving UCB cells alone, high levels of human engraftment were seen with 8 Â 106 UCB cell dose (Figure 1), whereas no engraftment b 6 was seen after infusion of 2 Â 10 UCB cell dose. Only two 80 out of 10 mice haddetectable( 40.5%), but low levels (o1%) of human engraftment after infusion of 4 Â 106 UCB cells. In contrast, eight out of 10 mice hadhuman 60 engraftment after coinfusion of 4 Â 106 UCB cells plus 1 Â 106 unrelatedhuman MSCs ( P ¼ 0.02 by Fisher’s exact 40 test) (Figure 2). None of the four mice hadhuman Number engraftment after coinfusion of 4 Â 106 UCB cells plus 6 0.1% 1 Â 10 immortalizedmouse mesenchymal cells (BMC-9 20 line). This was not statistically signiﬁcantly different comparedto the mice given 4 Â 106 UCB cells alone. Furthermore, the level of human engraftment was 0 1 2 3 4 also signiﬁcantly higher in mice coinfusedwith human 10 10 10 10 MSCs (2.572 vs 0.270.2%, two-tailed P ¼ 0.005, unpaired Log PF t test with Welch correction) comparedto UCB alone (Figure 3). c 70

Immune suppression of human MSCs 60

Human-Interferon-g Elispot assay was usedto determineif 50 human MSCs activatedT-cells in unrelatedhuman blood. Elispot is a highly sensitive assay (limit of detection 5 40 reactive lymphocytes out of 300 000) that is capable of 30 characterizing the frequencies of alloantigen-specific T cells Number in short-term culture. Three pairs of MSCs andperipheral 20 5% blood MNCs derived from six different random adult donors were tested, and we did not detect activated T-cell 10 clones above the backgroundin any pair (Figure 4). Same MNCs gave a high number of spots in mixedlymphocyte 0 1 2 3 4 cultures andwhen stimulatedwith PHA (positive controls). 10 10 10 10 Reactivity against unrelated, allogeneic MSCs was not Log PF observedat 1 : 4, 1 : 1 and4 : 1 target effector ratios (data CD45-PE not shown). Next, the impact of unrelatedhuman MSCs on Figure 1 Flow cytometry detection of human CD45 þ cells in NOD- an ongoing mixedlymphocyte reaction was evaluated.In SCID mice. Representative flow cytometry histograms of mouse bone three independent experiments, a significant reduction of marrow cells after staining with the anti-human CD45-PE antibody: (a) spot formation occurredin the presence of MSCs from a mouse transplantedwith 8 Â 106 UCB cells, (b) from a mouse transplantedwith 4 Â 106 UCB cells alone, (c) from a mouse transplanted (Figure 5). This suppression was also observedwhen the 6 6 activation of T-lymphocytes in the mixedlymphocyte with 4 Â 10 UCB cells and1 Â 10 human MSCs. reaction was enhanced by the addition of human interleukin-2 to the wells. To determine if this inhibition of T- cell activation couldbe achievedby humoral factors contrast to the inhibitory effect of MSCs, the MSC secretedby the MSCs, Elispot reaction was performedin supernatant failedto inhibit T-cell activation, and,on the a cell-free culture supernatant of MSCs (Figure 6a). In contrary, MSC supernatant hada stimulatory effect on

Bone Marrow Transplantation MSC support HSC B Maitra et al 600 100% 100

4/5 75 75% 8/10

50% Number of Spots 25

ns ns

0 25% L MSC L+MSC L+allo L L+PHA Percent Mice with Human Engraftment 2/10 Figure 4 Detection of alloreactive T cells that wouldactivate andsecrete interferon gamma upon encountering unrelatedhuman MSCs by EliSpot assay. A total of 150 000 responder peripheral blood lymphocytes (L) were 0/4 testedalone or mixedwith unrelatedMSCs or unrelatedallogeneic 0% lymphocytes (allo-L). PHA: phytohemaglutinin (10 mg/ml). Each spot corresponds to an activated interferon-g secreting lymphocyte. NS: not UCB UCB UCB UCB signiﬁcantly different (P40.05). Data from three independent experiments × 6 × 6 × 6 × 6 8 10 /m 4 10 /m 4 10 /m 4 10 /m with triplicate wells in each condition. n=10 +MSC +mouse n=10 stroma Figure 2 Frequency of human hematopoietic engraftment in NOD-SCID mice. Following 250 cGy irradiation, NOD-SCID mice were transplanted with 4–8 Â 106 UCB cells per mouse (m) with or without 1 Â 106 human MSCs or 1 Â 106 murine-immortalizedmesenchymal stem cell line BMC-9 500 § cells. Numbers indicate mice with human engraftment/mice transplanted. m ¼ mouse.

400

100.0%

300

200 § 10.0% Number of Spots

100 *

* Human CD45+

% 1.0% 0 L1+L2 L1+PHA L1+L2 +IL2+MSC L1+L2+IL-2 L1+L2+MSC

0.1% Figure 5 Effect of unrelatedMSCs on mixedlymphocyte reaction. Peripheral bloodlymphocytes (150 000/well) from two unrelateddonors UCB UCB UCB UCB (L1 andL2) were mixedin EliSpot plates with or without human 8×106/m 4×106/m 4×106/m 4×106/m interleukin-2 (IL2, 10 U/well). There was a signiﬁcant reduction in n=10 +MSC +mouse activated lymphocytes by addition of 150 000 human MSCs (donor n=10 BMC-9 unrelatedto either blooddonors)both in the absence (*: P ¼ 0.002) and presence of human IL-2 (y: P ¼ 0.009). Data from three independent Figure 3 Percent human hematopoietic cells (CD45 þ ) in an individual experiments with triplicate wells in each condition. mouse shown in Figure 2. Bone marrow MNCs were obtainedfrom each mouse andCD45 þ cell percentage was determined by ﬂow cytometry. m ¼ mouse.

Bone Marrow Transplantation MSC support HSC B Maitra et al 601 a 50 L1+L2

L1+L2 + MSC 40 L1+L2 + MSC sup.

L2 + MSC sup. § L1 + MSC sup. 30

L2 Experiment 1 Experiment 2 L1 20 * Number of Spots 0 100 200 300 400 500 Number of Spots

b 10 L1+L2+ Experiment 3 (MSC+L1sup) *

0 § L1+L2+MSC L L+MSC L+MSC L+PPD L+PPD L+PPD 150,000 300,000 +MSC +MSC 150,000 300,000

L1+L2 Figure 7 Suppression of PPD activation of peripheral bloodlymphocytes by unrelatedMSCs. Peripheral bloodlymphocytes (L) (150 000/well) from 02040 60 80 100 a PPD-positive donor were mixed with 150 000 or 300 000 unrelated donor MSCs andPPD (10 mg/ml) in triplicates in EliSpot wells. *: P ¼ 0.1, y: Number of Spots P ¼ 0.03. Figure 6 Effect of culture supernatants on mixedlymphocyte reaction. (a) Peripheral bloodlymphocytes (150 000/well) from two unrelateddonors (L1 andL2) were mixedin EliSpot plates with unrelatedMSCs or cell-free Number of Spots MSC culture supernatant. There was a reduction in activated lymphocytes 01020304050 by addition of human MSCs, but a signiﬁcant increase by addition of cell free supernatant from the same donor’s MSC cultures. Data from two L1 independent experiments with triplicate wells in each condition. (b)Inan independent experiment, peripheral blood lymphocytes (150 000/well) from L1+Rat MSC two unrelateddonors(L1 andL2) were mixedin EliSpot plates with unrelatedMSCs or cell-free culture supernatant obtainedfrom a mixture of L1+Human MSC MSCs andL1 lymphocytes (*: P ¼ 0.02, y: P ¼ 0.025, both comparedto mixedlymphocytes alone). L1+NIH3T3

L1+Dermal Fibroblast mixedlymphocyte cultures. Surprisingly, the cell-free L1+L2 supernatant obtainedfrom a mixture of MSCs and unrelatedlymphocytes hadan inhibitory effect similar to L1+L2+hMSC that seen with the addition of MSCs (Figure 6b). L1+L2+Dermal Fib To determine if MSCs could also suppress Tuberculin purifiedprotein derivative (PPD, Evans MedicalLim. Figure 8 Elispot analysis of human lymphocytes with adherent cells. Horsham, England)-induced T-cell activation, peripheral Peripheral bloodlymphocytes (150 000/well) from 2 unrelateddonors(L1 MNCs of a PPD skin test positive individual were incubated andL2) were mixedin EliSpot plates with rat MSCs, human MSCs, murine fibroblast cell line NIH-3T3 or human dermal fibroblasts. Both rat and with PPD. The addition of unrelated MSCs to this reaction unrelatedhuman MSCs failedto activate lymphocytes, while NIH-3T3 decreased the number of PPD-stimulated lymphocytes in a cells activatedthem. Mixedlymphocyte reaction (L1 þ L2) was inhibitedby dose-dependent manner (Figure 7). To determine the human MSCs but not by human dermal fibroblasts. specificity of MSC-mediated suppression of T-cell activation, rat MSCs, human dermal fibroblasts and murine Discussion NIH3T3 cells were also testedin the same assay system. While unrelatedhuman MSCs andxenogeneic rat MSCs Using an NOD-SCID repopulation assay we demonstrated, did not elicit human T-cell activation, unrelated human for the first time, that intravenous coinfusion of human dermal fibroblasts and murine NIH-3T3 cells elicited T-cell UCB cells andunrelatedadultdonorMSCs results in activation (Figure 8). Furthermore, human dermal fibro- improvedfrequency andthe level of human engraftment blasts failedto suppress the mixedlymphocyte reaction. when UCB cells infusedare limiting. When adultsare

Bone Marrow Transplantation MSC support HSC B Maitra et al 602 transplantedwith a single unit unrelatedUCB, engraftment cells of unrelateddonors.In the same assay system, failure is approximately 10–15% andhematopoietic recov- unrelated donor dermal fibroblasts and xenogeneic mouse ery is delayed.22 This is likely due to the limiting numbers of fibroblast cell line activatedhuman peripheral blood hematopoietic stem cells foundin a single unit of UCB cells, whereas xenogeneic rat MSCs failedto activate them. comparedto conventional allogeneic graft sources. Surprisingly, addition of third-party MSCs (unrelated Thus, our data suggest that coinfusion of MSCs in adult to either peripheral blooddonors)to a mixed patients undergoing a single donor UCB transplantation lymphocyte reaction demonstrated an inhibitory effect. may improve engraftment rate andkinetics. Similar to This inhibition was even more evident when the reaction the clinical situation, we did not have access to UCB donor was enhancedwith supplemental human IL-2. This effect MSCs andwere forcedto use unrelatedMSCs. Our did not appear to be specific to alloantigens since in vitro studies show that unrelated MSCs were not Tuberculin activation of T-lymphocytes was also inhibited immunogenic, but rather immunosuppressive andsuggest in the presence of MSCs. This is the first use of Elispot that therapeutic MSCs may be obtainedfrom assay, investigating the interaction between MSCs and donors unrelated to the UCB donor in support of UCB immune cells. transplantation. Our results suggest that human MSCs directly inhibit A number of preclinical animal models of stromal cell the activation of alloreactive T-lymphocytes andthis transplantation have been establishedto support hemato- effect may have important therapeutic implications in poietic engraftment.23,24 Both autologous andallogeneic the setting of transplantation across HLA barriers and human marrow stromal cells were shown to improve GVHD. the quantity anddurationof human hematopoietic DiNicola et al13 reportedthat human bone marrow engraftment in a preimmune sheep model.25 Although stromal cells suppress T-lymphocyte proliferation in vitro. stromal cells usedin these experiments containeda mixture The mechanism of this suppression appearedto be of STRO-1 þ cells, macrophages, CD8 þ cells and mediated by soluble factors, including hepatocyte growth rare CD34 þ cells, the effect on human hematopoiesis factor andtransforming growth factor- b. In these experi- was thought to be mediated by the STRO-1 þ stromal ments, monocyte–macrophage depleted long-term bone precursor cells. Noort et al26 reportedthat human fetal marrow culture-derived stromal cells were used. Since these lung-derived MSCs promoted human UCB engraftment in cultures are heterogenous in nature, it is difficult to NOD-SCID mice. Similar to our findings, human engraft- determine the cell population contributing to the observed ment was enhancedonly when transplantedUCB CD34 þ effect. Our results suggest that the homogenous population cell numbers were limitedat 0.03–0.1 Â 106/mouse. In of MSCs may be the predominant effector cells of stromal contrast to adult bone marrow-derived MSCs, the clinical cultures. In contrast to the findings of DiNicola et al,we use of human fetal lung-derived MSCs may be more did not find MSC culture supernatant to have a T-cell difficult due to the limited availability of such tissues inhibitory effect. On the contrary, the addition of MSC andethical concerns with procurement. Angelopoulou conditioned medium to mixed lymphocytes enhanced the et al27 recently reportedthat cotransplantation of lymphocyte activation. However, when we usedcondi- human MSCs enhancedmyelopoiesis andmegakaryopoi- tionedmediumobtainedfrom a mixture of MSCs and esis in NOD-SCID mice transplantedwith the limiting unrelatedlymphocytes, we observeda significant inhibition doses of peripheral blood CD34 þ cells. Our results of the mixedlymphocyte activation. This findingsuggests extendthese observations to umbilical cordblood that MSC-lymphocyte interaction results in the release of cells andsuggest an important supportive role for substances (transferable in medium) either from MSCs or MSCs in the pre-clinical models of hematopoietic lymphocytes that have a direct inhibitory effect. We are transplantation. currently investigating the nature of this substance andits The mechanism of human hematopoietic support in source. Tse et al12 also foundMSCs to suppress mixed NOD-SCID mice by human MSCs is not clear at this time. lymphocyte proliferation; however, this did not appear to Noort et al26 did not find any impact of fetal lung MSCs on be mediated by IL-10, TGF-b1, prostaglandin E2 secretion the homing of a coinfusedCD34 þ cell. It is possible that or tryptophan depletion. Le Blanc et al11 have reportedthat coinfusedMSCs providesecretedhuman cytokines and MSCs suppress T-cell proliferation both by allogeneic hematopoietic niches to support human cells by antigens andmitogens such as PHA andconcanavalin A, direct contact. We have previously determined the wide suggesting a nonspecific antiproliferative effect. More distribution of human MSCs in NOD-SCID mice, recently, Krampera et al31 foundthat mouse bone particularly to bone marrow andpersistence of these marrow-derived adherent cells inhibit naı¨ ve andmemory cells in vivo over 8 weeks.28 Evidence of low-level bone antigen-specific T cells using HY peptide-stimulated marrow engraftment of MSCs was also foundin dogsand splenocytes from transgenic HY-TCRhigh mice. This nonhuman primates after intravenous infusion.29,30 inhibition was contact dependent and not dependent on We cannot exclude the possibility that the major role of APC or CD4 þ /CD25 þ cells. Further research is needed MSCs during cotransplantation is mediated by secreted to understand the mechanism of the inhibitory actions of proteins that have systemic effects andthe exact MSCs on T-lymphocytes. location of MSC engraftment or entrapment may not There is also evidence for in vivo immuosuppression by be relevant. MSCs. Using a baboon skin graft model, Bartholomew Using a sensitive Elispot assay, we foundthat human et al14 showedthat infusion of ex vivo-expanded donor MSCs did not activate alloreactive peripheral blood (baboon) MSCs at a dose of 20 Â 106 MSC/kg recipient

Bone Marrow Transplantation MSC support HSC B Maitra et al 603 weight prolongedtime to rejection of histoincompatible stromal cells selectively expressing the membrane boundand skin grafts. Even ‘third-party’ baboon MSCs, obtained secretedforms of the human homologue of the steel gene from neither recipient nor skin graft donor, appeared to product, stem cell factor. Proc Natl Acad Sci USA 1992; 89: suppress alloreactivity in vivo. These data suggest that 7350–7354. immunologic effects of MSCs observed in vitro may also be 8 Clark B, Keating A. Biology of bone marrow stroma. Ann NY Acad Sci operative in vivo. Since we usedan immunodeficientmice 1995; 770: 70–88. 9 Majumdar M, Thiede M, Haynesworth S et al. Human model in our cotransplantation experiments, we were marrow-derived mesenchymal stem cells (MSCs) express unable to determine the immunosuppressive effects of hematopoietic cytokines andsupport long-term hemato- MSCs on the prevention of hematopoietic stem cell poiesis when differentiated toward stromal and osteogenic rejection in the setting of xenotransplantation. It is possible lineages. that MSC cotransplantation supports allogeneic hemato- J Hematother Stem Cell Res 2001; 9: 841–848. poiesis in vivo, both by direct hematopoietic survival signals 10 Majumdar MK, Thiede MA, Mosca JD et al. Phenotypic and andby inhibition of alloreactive lymphocytes that may lead functional comparison of cultures of marrow-derived me- to rejection. senchymal stem cells (MSCs) andstromal cells. J Cell Physiol In conclusion, our data indicate that human MSCs may 1998; 176: 186–192. et al. have a significant therapeutic potential in the setting of 11 LeBlanc K, Tammik L, Sundberg B Mesenchymal stem cells inhibit andstimulate mixedlymphocyte cultures allogeneic transplantation by improving the engraftment and mitogenic responses independently of the major rate andkinetics andinhibiting T-cell alloreactivity. histocompatibility complex. Scand J Immunol 2003; 57: This effect may be particularly pronouncedin the 11–20. setting of limiteddonorstem cells such as UCB 12 Tse WT, Pendleton JD, Beyer WM et al. Suppression of transplantation andin situations with a high risk of allogeneic T-cell proliferation by human marrow stromal cells: rejection, such as nonmyeloablative preparative regimens implications in transplantation. Transplantation 2003; 75: andT-cell-depletedtransplants across MHC barriers. 389–397. Furthermore, T-cell-suppressive properties of MSCs may 13 Di Nicola M, Carlo-Stella C, Magni M et al. Human bone prove useful in preventing or treating acute andchronic marrow stromal cells suppress T-lymphocyte proliferation Blood GVHD. Hematopoietic support andimmunomodulatory induced by cellular or nonspecific mitogenic stimuli. 2002; 99: 3838. effects of MSCs shouldbe further characterizedin clinical 14 Bartholomew A, Sturgeon C, Siatskas M et al. Mesenchymal trials. stem cells suppress lymphocyte proliferation in vitro and prolong skin graft survival in vivo. Exp Hematol 2002; 72: 1653–1655. Acknowledgements 15 Caplan A. Mesenchymal stem cells. J Orthop Res 1991; 9: 641–650. 16 Haynesworth S, Baber M, Caplan A. Cell surface antigens on We thank Dr Magdalena Tary-Lehmann for assistance human marrow-derived mesenchymal cells are detected by with the Elispot assay, Stem Cell, Athymic Animal, Flow monoclonal antibodies. Bone 1992; 13: 69–80. Cytometry andRadiationResources Core Facilities of the 17 Koç O, Gerson S, Cooper B et al. Rapidhematopoietic Comprehensive Cancer Center of Case Western Reserve recovery after co-infusion of autologous culture-expanded University andUniversity Hospitals of Cleveland(P30 human mesenchymal stem cells (hMSCs) andPBPCs in breast CA43703). This work was supportedby NIH, R01HL62360 cancer patients receiving high dose chemotherapy. J Clin Oncol (PI: ON Koç). 2000; 18: 307–316. 18 Koç O, Day J, Nieder M et al. Allogeneic mesenchymal stem cell infusion for treatment of metachromatic leukodystrophy References (MLD) andhurler syndrome(MPS-IH). Bone Marrow Transplant 2002; 30: 215–222. 1 Koç O, Lazarus H. Mesenchymal stem cells heading to the 19 Schafer I, Pandy M, Ferguson R, Davis B. Comparative clinic. Bone Marrow Transplant 2001; 27: 235–239. observation of fibroblasts derived from the papillary and 2 Bianco P, Constantini M, Dearden L, Bonucci E. reticular dermis of infants and adults: growth kinetics, packing Alkaline phosphatase positive precursors of adipocytes density at confluence and surface morphology. Mech Ageing in the human bone marrow. Br J Haematol 1988; 68: Dev 1985; 31: 275–293. 401–411. 20 Dennis J, Merriam A, Awadallah A et al. A quadripotential 3 Haynesworth S, Goshima J, Goldberg V, Caplan A. Char- mesenchymal progenitor cell isolatedfrom the marrow of an acterization of cells with osteogenic potential from human adult mouse. J Bone Min Res 1999; 14: 700. marrow. Bone 1992; 13: 81–88. 21 Tary-Lehmann M, Hricik D, Justice A et al. Enzyme- 4 Pittinger M, Mackay A, Beck S et al. Multilineage potential of linkedimmunosorbent assay spot detectionof interferon and adult human mesenchymal stem cells. Science 1999; 284: interleukin-5 producing cells as a predictive marker for renal 143–147. allograft failure. Transplantation 1998; 66: 219–224. 5 Jiang Y, Jahagirdar BN, Reinhardt RL et al. Pluripotency of 22 Laughlin MJ, Barker J, Bambach B et al. Hematopoietic mesenchymal stem cells derived from adult marrow. Nature engraftment andsurvival in adultrecipients of umbilical-cord 2002; 418: 41–49. bloodfrom unrelateddonors. N Engl J Med 2001; 344: 6 Reyes M, LundT, Lenvik T et al. Purification and ex vivo 1815–1822. expansion of postnatal human marrow mesodermal progenitor 23 Anklesaria P, Kase K, Glowacki J et al. Engraftment of a cells. Blood 2001; 98: 2615–2625. clonal bone marrow stromal cell line in vivo stimulates 7 Toksoz D, Zsebo K, Smith K et al. Support of human hematopoietic recovery from total body irradiation. Proc Natl hematopoiesis in long term bone marrow cultures by murine Acad Sci USA 1987; 1987: 7681–7685.

Bone Marrow Transplantation MSC support HSC B Maitra et al 604 24 Nolta J, Hanley M, Kohn D. Sustainedhuman hematopoiesis myelopoiesis andmegakaryocytopoiesis in NOD/SCID mice. in immunodeﬁcient mice by cotransplantation of marrow Exp Hematol 2003; 31: 413–420. stroma expressing human interleukin-3: analysis of gene 28 Lee K, Gerson SL, Maitra B, Koc ON. G156A MGMT- transduction of long-lived progenitors. Blood 1994; 83: transduced human mesenchymal stem cells can be selectively 3041–3051. enrichedby O6-benzylguanine andBCNU. J Hematother Stem 25 Almeida-Porada G, Porada C, Tran N, Zanjani E. Cell Res 2001; 10: 691–701. Cotransplantation of human stromal cell progenitors into 29 Sandmaier B, Strob R, Kniley J et al. Evidence of allogeneic preimmune fetal sheep results in early appearance of human stromal engraftment in bone marrow using canine mesenchy- donor cells in circulation and boosts cell levels in bone marrow mal stem cells. Blood 1998; 92: 116a. at later time points after transplantation. Blood 2000; 95: 30 Devine S, Bartholomew A, MahmudN et al. Mesenchymal 3620–3627. stem cells are capable of homing to the bone marrow of non- 26 Noort WA, Kruisselbrink AB, Anker PS et al. Mesenchymal human primates following systemic infusion. Exp Hematol stem cells promote engraftment of human umbilical cord 2001; 29: 244–255. blood-derived CD34(+) cells in NOD/SCID mice. Exp 31 Krampera M, Glennie S, Dyson J et al. Bone marrow Hematol 2002; 30: 870–878. mesenchymal stem cells inhibit the response of naive and 27 Angelopoulou M, Novelli E, Grove J et al. Cotransplantation memory antigen-speciﬁc T cells to their cognate peptide. Blood of human mesenchymal stem cells enhances human 2003; 101: 3722–3729.

Bone Marrow Transplantation