Suppression of allogeneic T-cell proliferation by human marrow stromal cells: implications in transplantation : Transplantation

Journal Logo


Suppression of allogeneic T-cell proliferation by human marrow stromal cells: implications in transplantation

Tse, William T.1 2; Pendleton, John D.1; Beyer, Wendy M.1; Egalka, Matthew C.1; Guinan, Eva C.1

Author Information
Transplantation 75(3):p 389-397, February 15, 2003. | DOI: 10.1097/01.TP.0000045055.63901.A9
  • Free


Bone marrow (BM) stroma contains developmentally primitive and potentially self-renewing cells that can differentiate into multiple mesenchymal cell types. These cells have been termed marrow stromal cells (MSC), mesenchymal progenitor cells, or mesenchymal stem cells (1,2). Many groups have isolated and expanded MSC from various species and demonstrated their capacity to develop into different cell types, including osteocytes, chondrocytes, adipocytes, myocytes, and cardiomyocytes (3,4). MSC have also been shown to be able to produce a microenvironment that supports hematopoietic development (5).

The capacity of MSC to differentiate into multiple mesenchymal tissues and to constitute a hematopoiesis-supporting stroma has prompted consideration of a role for MSC in clinical transplantation. Animal studies have shown that transplanted MSC can engraft in appropriate tissues after intravenous infusion (6,7). Pilot clinical studies have demonstrated that autologous, ex vivo expanded MSC can be infused into patients without adverse effects (8,9). Allogeneic MSC isolated from donors have been used as supplementary cellular therapy in attempts to replace defective tissues in osteogenesis imperfecta (10). Coinfusion of MSC in hematopoietic stem-cell transplantation has been evaluated as a way to facilitate posttransplantation hematopoietic recovery (11).

To explore the potential use of MSC in clinical transplantation, we set out to determine the in vitro ability of human MSC to function as alloantigen presenting cells (APC). We observed that human MSC did not induce allogeneic T cells to proliferate, even when their major histocompatibility complex (MHC) class II antigen was up-regulated and a costimulatory signal provided by an anti-CD28 antibody. Significantly, MSC suppressed T-cell proliferation induced by direct stimulation, apparently by way of a novel inhibitory mechanism. These observations indicate that human MSC act as nonprofessional APC with a unique immunomodulating property. This behavior might allow allogeneic MSC transplantation to be performed with a reduced need for host immunosuppression.


Derivation of BM-Derived MSC and Dendritic Cells

Human MSC were isolated and expanded from BM as described previously (4). The purified cells stained positive for antigens known to be present on MSC, such as SH2, SH3, and SH4, and were able to differentiate into adipocytes and osteoblasts under appropriate culture conditions (4). Early passage cells were used in all experiments. Derivation of dendritic cells (DC) from BM was performed as published (12) in culture medium supplemented with interleukin (IL)-4 and granulocyte macrophage colony stimulating factor (R&D Systems, Inc., Minneapolis, MN). The resulting cells stained positive with a monoclonal antibody (mAb) that recognizes the CD1a antigen typically expressed on DC (data not shown).

Allogeneic Peripheral Blood Mononuclear Cells-MSC Mixed Cell Cultures

In mixed cell cultures, peripheral blood mononuclear cells (PBMC) from healthy volunteers were fractionated on a Ficoll-Paque column and used as responder cells. In regular mixed cell cultures, allogeneic PBMC isolated from different volunteers were used as stimulator cells. In modified mixed cell cultures, allogeneic MSC or DC were used as stimulator cells. Stimulator cells were irradiated with 3,300 cGy of γ-radiation and cultured in 96-well tissue culture plates at 105 PBMC or 1.5x104 MSC per well. Stimulator MSC were preplated in wells to form adherent monolayers 16 to 18 hours before addition of responder PBMC. Responder PBMC were cultured in the same wells at 105 cells per well. All cultures were done in triplicate in Roswell Park Memorial Institute (RPMI) medium supplemented with 10% heat-inactivated fetal calf serum (FCS), 2 mM glutamine, and 55 μM β-mercaptoethanol.

In experiments to enhance MSC immunogenicity, stimulator MSC were pretreated with 200 units/mL of recombinant human γ-interferon (R&D Systems) for 4 days before used in mixed cell cultures with allogeneic PBMC in medium without γ-interferon. In experiments to provide a costimulatory second signal, an anti-CD28 activating mAb (clone 15E8, Research Diagnostics, Inc., Flanders, NJ) was added to the culture medium of the PBMC-MSC coculture to a concentration of 200 ng/mL.

In the mixed cell culture, responder and stimulator cells were cocultured for 7 days before proliferation of the responder cells was measured by 3H-thymidine incorporation assay. One day before the assay, 3H-thymidine radiolabel (NEN Life Science Products, Inc., Boston, MA) was added to the cell culture at 1.6 μCi per well. Eighteen hours later, the cells were harvested, and radiolabel incorporation into the cells was measured using a Wallac 1400 Beta Counter (Gaithersburg, MD).

Immunophenotyping of MSC

Immunophenotyping of MSC was performed by fluorescent staining with cell-surface antigen-specific mAbs and subsequent flow cytometry analysis. Antibodies (human leukocyte antigen [HLA]-A, B, C, clone G46-2.6; HLA-DR, DP, DQ, clone Tü39; CD54/intercellular adhesion molecule [ICAM]-1, clone HA58; CD58/lymphocyte function-associated antigen [LFA]-3, clone 1C3; CD80/B7-1, clone L307.4; CD86/B7-2, clone 2331; CD40, clone 5C3) were purchased from PharMingen (San Diego, CA). Mouse hybridomas producing the SH2, SH3, and SH4 mAbs were obtained from American Type Culture Collection (Manassas, VA).

For flow cytometry analysis, adherent MSC were detached by treating with 0.05% trypsin-EDTA for 3 minutes at 37°, neutralized with FCS-containing culture medium, and disaggregated into single cells by pipetting. The cells were stained with fluorescein isothiocyanate-conjugated MAbs and analyzed using a FACSCalibur flow cytometer (Becton-Dickinson Immunocytometry Systems, San Jose, CA).

Apoptosis Analysis of PBMC after Coculture with MSC

Apoptosis of cells was measured by an annexin V binding and propidium iodide uptake assay (TACS annexin V-FITC apoptosis detection kit; R&D Systems). 6x106 PBMC were cocultured with preplated subconfluent MSC in a 25-cm2 flask for 4 days. The PBMC were then recovered, stained with fluorescein isothiocyanate-conjugated annexin V and propidium iodide, and analyzed by flow cytometry.

Suppression of T-cell Proliferation by MSC

To assess the ability of MSC to suppress T-cell proliferation, responder PBMC were cocultured with 105 third-party allogeneic stimulator PBMC in 96-well tissue culture plates at 105 cells per well in the presence or absence of 1.5x104 preplated MSC per well. Both stimulator PBMC and MSC were irradiated before the coculture. 3H-thymidine incorporation by the responder PBMC was measured after 7 days of coculture. Alternatively, responder PBMC were treated for 5 days in culture with anti-CD3 and anti-CD28 activating mAbs (clones 1XE and 15E8, Research Diagnostics) at concentrations of 100 ng/mL and 200 ng/mL, respectively, to stimulate T cells to clonally expand. Proliferative response of T-cells was then measured by 3H-thymidine incorporation.

In other experiments, proliferative response of enriched T cells was measured. CD3+ and CD4+ T-cell subsets were enriched from PBMC using the RosetteSep system (StemCell Technologies, Vancouver, BC) and then used in the anti-CD3/anti-CD28 mAb stimulation experiments. The purity of the enriched CD3+ and CD4+ T cells was estimated to be 95% and 90%, respectively. The number of unfractionated PBMC, CD3+ T cells, or CD4+ T cells used in each well was adjusted such that the absolute T-cell content across all wells was equivalent.

In experiments to assess the ability of MSC to suppress T-cell proliferation after initiation of T-cell activation, PBMC were first stimulated with anti-CD3/anti-CD28 mAbs. At 0, 4, and 24 hours after the commencement of mAb stimulation, the PBMC along with their mAb-containing culture media were transferred to separate wells that had been preplated with MSC. PBMC were then cultured in the presence of mAb±MSC for a total of 5 days, after which 3H-thymidine incorporation was measured.

Transwell and Conditioned-Medium Analysis of MSC-Associated Suppressive Activity

In Transwell experiments, responder PBMC were cultured in an upper chamber of the Transwell insert (Costar, Corning Inc., Corning, NY) either alone or with irradiated MSC±irradiated third-party PBMC cultured in a lower chamber. MSC were preplated 16 to 18 hours before addition of the responder PBMC to produce adherent monolayers. The two cell types were separated by a semipermeable membrane with a pore size of 0.4 μm. Anti-CD3/anti-CD28 mAbs were added to the medium at the start of the culture to induce proliferation of T cells. 3H-thymidine incorporation assay was performed after 5 days of culture.

To produce conditioned media, MSC were cultured in 25-cm2 flasks at 70 to 90% confluence for 4 days, with or without coculture of 5x106 third-party allogeneic PBMC. The conditioned media from the cultures were then recovered and filtered through 0.2 μm sterilization filters and used immediately. Responder PBMC were cultured in the MSC-conditioned media and anti-CD3/anti-CD28 activating mAbs added at the start of the culture. On day 4 of culture in the conditioned media, proliferation of the responder PBMC was measured by 3H-thymidine incorporation.

Analysis of MSC Production of Cytokines and Prostaglandin E2

Levels of IL-10 and transforming growth factor (TGF)-β1 in MSC culture medium were measured by colorimetric sandwich enzyme-linked immunoassays (Quantikine ELISA; R&D Systems) after 5 days of culture. In TGF-β1 neutralizing experiments, a neutralizing anti-human TGF-β1 mAb (clone 9016.2, R&D Systems) was added to a culture of responder PBMC±MSC with anti-CD3/anti-CD28 activating mAbs, at increasing concentrations of up to 5 μg/mL. 3H-thymidine incorporation of the responder PBMC was then measured after 5 days of culture.

Levels of prostaglandin (PG)E2 in MSC culture medium were measured by a colorimetric competitive enzyme-linked immunosorbent assay (ELISA, R&D Systems) after 6 days of culture. In PGE2 inhibition experiments, indomethacin (Sigma, St. Louis, MO) was added to a culture of PBMC±MSC with anti-CD3/anti-CD28 activating mAbs, at final concentrations of up to 40 μM. 3H-thymidine incorporation was determined after 5 days of culture.

Analysis of Potential Tryptophan Depletion by MSC

PBMC were cultured in the presence and absence of MSC in standard RPMI medium, which contains 50 μM of tryptophan. A 5-fold excess of additional tryptophan (Life Technologies, Rockville, MD) was added to each well at the start of a culture of PBMC±MSC with anti-CD3/anti-CD28 activating mAbs. Addition of the same amount of tryptophan was repeated twice on day 4 at 3 hours before the addition of the 3H-thymidine label and again at the time of the addition of the label. Eighteen hours later, on day 5 of culture, the PBMC were harvested for measurement of 3H-thymidine incorporation. In some experiments, potential tryptophan catabolism by the enzyme indoleamine 2,3-dioxygenase (IDO) was inhibited by the addition to the medium at the start of the culture of 1 mM 1-methyltryptophan (Aldrich Chemical Co., Milwaukee, WI), a specific inhibitor of the enzyme. Responder PBMC proliferation was assayed by 3H-thymidine incorporation after 5 days of culture.


MSC Do Not Elicit an Allogeneic Proliferative Response Even in the Presence of CD28 Costimulation

To determine whether MSC could function as APC, primary human MSC were isolated and expanded from BM samples. Irradiated PBMC or MSC stimulators were cocultured with allogeneic PBMC responders in mixed cell cultures, and proliferation of the responders was assessed after 7 days. Allogeneic PBMC, as expected, were highly efficient APC, whereas MSC, in contrast, did not induce significant proliferation (Fig. 1A). DC were able to induce a brisk primary proliferative response in allogeneic PBMC while MSC isolated from the same donor did not (Fig. 1B). The failure of MSC to induce proliferation of allogeneic T cells was consistently observed with MSC and PBMC isolated from multiple donors and was observed over the entire 3 to 7 day time course of the primary culture (data not shown).

Figure 1:
Lack of proliferation of peripheral blood mononuclear cells (PBMC) when cocultured with allogeneic marrow stromal cells (MSC) in mixed cell cultures. (A) After 7 days of culture alone, responder PBMC exhibited only baseline proliferation as measured by 3H-thymidine incorporation (counts per minute) (solid bar). Responder PBMC cocultured with allogeneic PBMC exhibited marked increase in proliferation (hatched bar). Responder PBMC cocultured with two different MSC isolates exhibited only baseline proliferation (open bars). (B) Marrow-derived DC elicited a robust proliferative response from allogeneic responder PBMC (hatched bar), in contrast with MSC derived from the same BM donor (open bar). (C) Lack of responder PBMC proliferation persisted regardless of whether MSC were pretreated with γ-interferon. (D) PBMC cocultured with γ-interferon treated MSC did not proliferate even when an anti-CD28 antibody was added in the culture to provide costimulatory signal. DC, dendritic cells.

One reason for the inability of MSC to function as APC might be a failure to express cell-surface molecules necessary for alloantigen-induced proliferation. MSC were therefore examined for the expression of MHC, adhesion, and costimulatory molecules. MSC constitutively expressed MHC class I antigens but were negative for class II antigens (Fig. 2, left). MSC also expressed, at a lower level, the adhesion molecule LFA-3 (CD58). In contrast, MSC failed to express the adhesion molecule ICAM-1 (CD54) or the costimulatory molecules B7-1 (CD80), B7-2 (CD86), or CD40. Pretreatment of MSC for 4 days with γ-interferon resulted in enhanced expression of MHC class I antigens and the induction of MHC class II antigens as well as ICAM-1. There was no increase in expression of LFA-3, and expression of B7-1, B7-2, and CD40 was not induced (Fig. 2, right). Despite induction of both class II and ICAM-1, the change in surface-antigen expression in MSC subsequent to γ-interferon pretreatment nevertheless did not improve the capacity of MSC to function as APC (Fig. 1C).

Figure 2:
Expression of immunologically important cell-surface molecules on MSC. MSC constitutionally express major histocompatibility complex (MHC) class I antigens and lymphocyte function-associated antigen (LFA)-3 molecules but are negative for MHC class II antigens, and intracellular adhesion molecule (ICAM)-1, B7-1, B7-2, and CD40 molecules, as shown by flow cytometry analysis (left). After treatment with γ-interferon, expression of MHC class I and class II antigens and ICAM-1 molecules are up-regulated but expression of the other molecules remains negative (right). (solid curves) fluorescence profiles after staining with antigen-specific antibodies; (dotted curves) isotype-matched negative controls.

Direct engagement of the CD28 molecule on T cells with an activating mAb (13,14) was undertaken as an alternative approach to deliver the B7-mediated costimulatory signal. However, no increase in responder PBMC proliferation was observed even in the presence of anti-CD28 mAb when PBMC were cocultured with γ-interferon-pretreated MSC (Fig. 1D).

MSC Do Not Induce Apoptosis in PBMC

Another possible explanation for the failure of allogeneic responders to proliferate in the presence of MSC was MSC-induced apoptosis of responder cells. To address this, PBMC were cultured for 4 days in the presence or absence of MSC and their apoptosis and cell death analyzed by annexin V binding. We observed no increase in apoptosis and cell death (Fig. 3A). Additional evidence for viability of responder cells in the PBMC-MSC coculture was provided by the observation that recovered PBMC stimulated directly with mitogenic antibodies against CD3 and CD28 proliferated equally well whether they had been recovered from the MSC preculture or from medium alone (Fig. 3B).

Figure 3:
Apoptosis analysis of PBMC after coculture MSC. (A) Flow cytometry fluorescent profiles of FITC-annexin V binding and propidium iodide uptake by PBMC are shown as contour plots. The number in each quadrant of the plot represents the percentage of PBMC that fall within the quadrant. Lower left quadrant of plot contains viable cells that do not bind annexin V and exclude propidium iodide. (upper plot) PBMC cultured alone; (lower plot) PBMC cocultured with MSC. (B) PBMC were precultured either with MSC or in medium alone and tested for ability to proliferate upon subsequent stimulation with anti-CD3/anti-CD28 antibodies. (Solid Bars) no stimulation; (hatched bars) stimulation with antibodies.

MSC Actively Suppress Proliferation of Stimulated T Cells

Despite γ-interferon-induced expression of class II and ICAM-1, MSC were unable to stimulate T cells. Furthermore, unlike observations with other class II positive cells (13,14), addition of direct costimulation by way of CD28 to cultures of PBMC and γ-interferon-pretreated MSC did not lead to proliferation of responders, suggesting that MSC were either incapable of providing an MHC-restricted first signal to T cells or that they actively suppressed alloantigen-induced T-cell proliferation. To address the question of suppression, we examined whether MSC could inhibit a standard mixed lymphocyte reaction (MLR) between responder PBMC and third-party allogeneic stimulator PBMC. The presence of MSC during the primary MLR completely abolished the alloantigen-stimulated proliferative response (Fig. 4A). We also stimulated PBMC directly with mitogenic antibodies to CD3 and CD28. The presence of MSC markedly reduced proliferation of PBMC to this stimulus (Fig. 4B). For both stimuli, failure of PBMC to proliferate in the presence of MSC was observed over a 3 to 7 day time course (data not shown). To determine which subpopulations in the PBMC were suppressed by MSC, CD3+, and CD4+, T cells enriched from a PBMC sample were stimulated with anti-CD3/anti-CD28 mAbs. The presence of MSC in the culture suppressed the proliferation of the T-cell subsets to the same degree as the unfractionated PBMC sample (Fig. 4C). To show whether the presence of MSC at the initiation of T-cell activation is required for the suppressive effect to take place, PBMC were stimulated with anti-CD3/anti-CD28 MAbs before being transferred at different time points to wells preplated with MSC (Fig. 4D). Suppression of PBMC proliferation was still seen even when the PBMC were stimulated 24 hours before being exposed to MSC, but the MSC-mediated suppressive effect was less prominent in that case.

Figure 4:
MSC suppression of PBMC proliferation upon stimulation. (A) PBMC were induced to proliferate when cocultured with third-party allogeneic PBMC (hatched bar) as compared with PBMC without allogeneic stimulation (solid bar). Proliferative response was abolished when MSC were present in the coculture (open bar). (B) PBMC were stimulated to proliferate when treated with anti-CD3/anti-CD28 antibodies (hatched bar). Proliferative response was also abolished when MSC were present in the culture (open bar). (C) Proliferation of responder cells stimulated with anti-CD3/anti-CD28 antibodies (hatched bars) suppressed in the presence of MSC coculture (open bars), independent of whether the responder cells were PBMC, enriched CD3+ T-cells, or enriched CD4+ T cells. (D) PBMC were stimulated with anti-CD3/anti-CD28 antibodies and subsequently placed into coculture with MSC at 0, 4, or 24 hours after initiation of the stimulation and cultured for a total of 5 days before 3H-thymidine incorporation was measured.

MSC Suppression of T-Cell Proliferation is Mediated by Soluble Factors

To determine whether cell-cell contact between MSC and PBMC was required for suppression, responder PBMC stimulated with anti-CD3/anti-CD28 mAbs were cultured in the upper chamber of a Transwell separated by a semipermeable membrane from irradiated MSC cultured in the lower chamber. MSC markedly suppressed PBMC proliferation across the Transwell (Fig. 5A), demonstrating that a soluble factor was operative. MSC-conditioned medium similarly suppressed proliferation of PBMC activated with anti-CD3/anti-CD28 mAbs (Fig. 5B), but the suppressive effect was more variable and less profound than that observed in the Transwell experiments. In both cases, suppression of the responders was enhanced when irradiated third-party PBMC were added to the MSC culture, indicating that physical interaction of PBMC and MSC increased production of the suppressive activity (Fig. 5).

Figure 5:
Suppressive activity of MSC as soluble factors. (A) Anti-CD3/anti-CD28 antibodies induced proliferation of PBMC in upper Transwell chamber (open bar). The proliferation was suppressed when MSC were cultured in lower Transwell chamber (hatched bar) or when MSC plus third-party PBMC were cultured in lower Transwell chamber (solid bar). (B) Proliferation of PBMC induced with anti-CD3/anti-CD28 antibodies (open bar) was partially suppressed when culture medium was previously conditioned with MSC (hatched bar) or with MSC plus third-party PBMC (solid bar).

Candidate T-Cell Inhibitory Factors Do Not Account for the Suppressive Activity of MSC

IL-10 and TGF-β1 are two well-known inhibitors of T-cell activation and proliferation. The concentration of IL-10 in culture media conditioned by MSC alone was below the level of detection (<7.8 pg/mL), suggesting that IL-10 production by MSC was unlikely to account for the observed suppression. ELISA measurements of TGF-β1 levels in conditioned media from cultures of MSC alone, PBMC alone, or PBMC and MSC together were 0.8 ng/mL, 1.2 ng/mL, and 2.1 ng/mL, respectively, above the basal amount present in the FCS-containing culture medium. Because the amount produced by PBMC alone was similar to that produced by MSC, it is unlikely that secretion of TGF-β1 by MSC accounts for the MSC-mediated suppression of PBMC proliferation. To confirm this, we tested whether MSC could suppress PBMC proliferation in the presence of a neutralizing antibody against TGF-β1. The presence of increasing amounts of neutralizing anti-TGF-β1 antibody did not significantly reverse the suppressive effect (Fig. 6A).

Figure 6:
Exclusion of potential mechanisms of MSC suppression. (A) Proliferation of PBMC stimulated by anti-CD3/anti-CD28 antibodies (hatched bars) was suppressed in the presence of MSC coculture (open bars). MSC-associated suppression was not reversed by presence of increasing concentrations of a neutralizing anti-transforming growth factor (TGF)-β1 antibody in the culture medium. (B) Suppression was largely unaffected by presence of increasing concentrations of indomethacin in the culture medium. (C) Suppression was not substantially reversed by addition of either extra tryptophan or indoleamine 2,3-dioxygenase (IDO) inhibitor to culture medium.

PGE2 suppresses T-cell activation by interfering with signal transduction events following stimulation. To examine whether generation of PGE2 by MSC might explain their suppressive effects, we measured the amount of PGE2 in the media from cultures of PBMC alone, MSC alone, or PBMC and MSC together. The level of PGE2 in these three conditions was 80 pg/mL, 160 pg/mL, and 215 pg/mL, respectively. MSC therefore secrete a measurable amount of PGE2 that is increased when PBMC are cocultured with MSC. To determine whether PGE2 produced by MSC could account for the observed suppressive activity, we examined the effect of indomethacin, an inhibitor of PGE2 biosynthesis. PBMC alone, or PBMC with MSC, were cultured in the presence of anti-CD3/anti-CD28 mAbs and increasing concentrations of indomethacin. PGE2 production, as measured by ELISA, decreased in inverse proportion to increasing amounts of indomethacin in the cultures (data not shown). The suppressive effect of MSC on PBMC proliferation, however, remained largely unchanged (Fig. 6B). This was true even if MSC were pretreated with indomethacin for 3 days before addition of PBMC and anti-CD3/anti-CD28 mAbs.

A new mechanism of immune-response regulation has been recently described that involves tryptophan catabolism by IDO (15). In this proposed mechanism, IDO in the regulatory cells depletes tryptophan in the tissue or culture medium. Because uptake of tryptophan is essential for cell proliferation, its depletion leads to cessation of proliferation. We asked whether such a mechanism might explain the MSC-associated suppressive activity. PBMC alone, or PBMC with MSC, were cultured in the presence of anti-CD3/anti-CD28 mAbs and then were given extra tryptophan both at the start of the coculture and on the day of 3H-thymidine addition, reasoning that the supplementary tryptophan might rescue the responder cells from any proliferative arrest mediated by tryptophan depletion. The MSC-mediated suppression of PBMC proliferation was not substantially reversed by either the addition of supplementary tryptophan or the addition of 1-methyltryptophan, a specific IDO inhibitor (Fig. 6C).


In the present report, we show that human MSC do not induce an alloantigen-specific, primary T-cell proliferative response, even after γ-interferon-induced up-regulation of MHC class II and ICAM-1 on the MSC coupled with direct T-cell costimulation by way of an anti-CD28 antibody. Indeed, MSC actively suppress T-cell proliferation induced by either alloantigens or direct engagement of the T-cell receptor (TCR) and CD28 with activating antibodies. The suppression is observed even when the MSC are separated from the target cells by Transwell membranes, indicating the suppression is probably mediated by a soluble factor.

In contrast with professional APC that are of a hematopoietic origin, such as DC or macrophages, tissue-derived nonprofessional APC have a diminished capability to activate naive, unprimed allogeneic T cells, mostly because of their inability to provide an adequate costimulatory signal. This defective antigen-presenting capability can be reversed in many cases either by cytokine pretreatment to alter surface molecules or by provision of an exogenous costimulatory signal. For example, in a well-characterized model system, transfection-induced expression of class II molecules in conjunction with either ICAM-1 or B7 is sufficient to permit NIH-3T3 cells to provide costimulation in a primary mixed cell culture (16). Pretreatment with γ-interferon induces class II and ICAM-1 as well as B7 on many cell types and similarly results in nonprofessional APC acquiring antigen-presenting function (17). Where costimulation remains inadequate, provision of exogenous costimulation has resulted in T-cell proliferation to alloantigen. Human renal tubular epithelial cells, dermal fibroblasts, and myoblasts, for instance, have all been shown to acquire the ability to present alloantigen in the presence of costimulatory anti-CD28 antibody (13,14,18). In contrast, in the case of human MSC, provision of an additional costimulatory signal by an anti-CD28 antibody in the PBMC-MSC coculture fails to bring about an allogeneic T-cell response. Addition of both an anti-CD3 and an anti-CD28 antibody to the coculture, which would have provided the requisite signals for T-cell activation, also does not result in any proliferative response. These differences suggest that nonprofessional APC derived from various tissues might exhibit divergent antigen presenting behaviors. Further investigation of the molecular mechanisms that account for the different capabilities of various tissue-derived nonprofessional APC to affect allogeneic T-cell responses may be revealing.

The suppressive activity associated with MSC might have masked any signs of MSC-induced T-cell activation. Therefore, it remains unclear whether MSC can actually deliver a first signal to stimulate allogeneic T cells. Unlike other MHC class II-positive, B7 family-negative cells, which induce anergy by engaging TCR without simultaneously providing a costimulatory signal (16), γ-interferon-pretreated, HLA-DR7-positive MSC failed to induce alloantigen-specific unresponsiveness (anergy) in PBMC after the PBMC were cocultured with the MSC (unpublished data). The PBMC so treated proliferated upon subsequent restimulation with HLA-DR7-positive B cells. This observation suggests a potentially unique antigen-presenting behavior of MSC but also raises the question of whether MSC can process and present alloantigen at all. Resolution of this question will be facilitated by the use of an alloantigen T-cell clonal system (16) that will allow analysis of the intracellular signaling events occurring in allogeneic T cells when they are exposed to MSC. Specifically, one can examine the expression and phosphorylation patterns of critical molecules in T-cell signaling pathways, such as TCR-ζ, ZAP-70, fyn, cbl, Rap1, p27kip1, and Tob, which have been shown to be differentially regulated in T cells that are either activated or anergized by exogenous signals (19,20). This analysis will help determine whether the molecular events associated with MSC-induced suppression of T-cell proliferation overlap with those encountered in anergy induction.

Further characterization of the MSC-associated suppressive activity might reveal a new mechanism of immunomodulation. Some other nonprofessional APC have also been shown to have immunomodulatory activities, but the suppressive activity associated with MSC appears to be distinct from those that have been described. Retinal pigment epithelial cells, for instance, inhibit allogeneic T-cell proliferation by inducing apoptosis and production of PGE2 (21,22). Human dermal keratinocytes inhibit proliferation of unprimed T cells by production of both TGF-β and PGE2 (23). Mammalian concepti suppress allogeneic T cells by catabolizing tryptophan (15). We have examined each of these mechanisms as an explanation of the MSC-associated suppressive activity and have found them insufficient to account for this activity. It may be mediated instead by a novel T-cell inhibitory factor produced by MSC.

This presumed T-cell inhibitory factor appears to be relatively labile because its effect was maximal when MSC were continually present in the suppression assay and suboptimal when MSC-conditioned medium alone was used in the assay. Actual physical contact between MSC and PBMC is not required for the action of the inhibitory factor. However, “cross-talk” between PBMC and MSC enhances its production, as demonstrated by the increased suppression seen when third-party PBMC were added to the coculture, or when supernatants from PBMC-MSC cocultures were used.

The immune suppressive effect of MSC demonstrated in our studies is consistent with data seen in MSC-transplantation experiments. Persistence of human MSC has been observed in multiple tissues of newborn sheep that had received human MSC infusions in utero even after immunologic competence has been established in the fetus (24). A baboon that was transplanted with allogeneic baboon MSC had persistence of donor cells even without the use of immunosuppression (25). The lack of graft rejection observed in these studies suggests that MSC might not elicit significant alloimmune responses in vivo, in agreement with our in vitro results. Whether the absence of rejection of allogeneic MSC in these animal studies is because of active suppression of host T cells by the transplanted MSC has not been examined, but these results support consideration of the use of this cell type in clinical transplantation.

Some investigations have suggested that simultaneous transplantation of MSC might enhance hematopoietic engraftment after hematopoietic stem-cell transplantation (26,27). Coinfusion of MSC or bone progenitor cells might facilitate hematopoietic engraftment across an allogeneic barrier, decrease the incidence of graft-versus-host phenomenon, and prolong skin graft survival (28–30). Transplantation of donor-derived or even third-party MSC might therefore form the basis for new strategies in clinical transplantation, and the observed immune suppressive behavior of MSC supports the possibility that transplantation of human MSC might be accomplished with minimal or no host immunosuppression.


The authors thank Drs. Samuel Lux, David Nathan, Alan D’Andrea, Lee Nadler, Vassiliki Boussiotis, Angelo Cardoso and Francisco Bonilla for generous support and useful discussions.


1. Deans RJ, Moseley AB. Mesenchymal stem cells: biology and potential clinical uses. Exp Hematol 2000; 28: 875.
2. Bianco P, Riminucci M, Gronthos S, et al. Bone marrow stromal stem cells: nature, biology, and potential applications. Stem Cells 2001; 19: 180.
3. Friedenstein AJ, Chailakhyan RK, Latsinik NV, et al. Stromal cells responsible for transferring the microenvironment of the hemopoietic tissues. Cloning in vitro and retransplantation in vivo. Transplantation 1974; 17: 331.
4. Pittenger MF, Mackay AM, Beck SC, et al. Multilineage potential of adult human mesenchymal stem cells. Science 1999; 284: 143.
5. Kuznetsov SA, Krebsbach PH, Satomura K, et al. Single-colony derived strains of human marrow stromal fibroblasts form bone after transplantation in vivo. J Bone Miner Res 1997; 12: 1335.
6. Nilsson SK, Dooner MS, Weier HU, et al. Cells capable of bone production engraft from whole bone marrow transplants in nonablated mice. J Exp Med 1999; 189: 729.
7. Pereira RF, Halford KW, O’Hara MD, et al. Cultured adherent cells from marrow can serve as long-lasting precursor cells for bone, cartilage, and lung in irradiated mice. Proc Natl Acad Sci U S A 1995; 92: 4857.
8. Lazarus HM, Haynesworth SE, Gerson SL, et al. Ex vivo expansion and subsequent infusion of human bone marrow-derived stromal progenitor cells (mesenchymal progenitor cells): implications for therapeutic use. Bone Marrow Transplant 1995; 16: 557.
9. Keating A, Berkahn L, Filshie R. A phase I study of the transplantation of genetically marked autologous bone marrow stromal cells. Hum Gene Ther 1998; 9: 591.
10. Horwitz EM, Prockop DJ, Gordon PL, et al. Clinical responses to bone marrow transplantation in children with severe osteogenesis imperfecta. Blood 2001; 97: 1227.
11. Koc ON, Gerson SL, Cooper BW, et al. Rapid hematopoietic recovery after coinfusion of autologous-blood stem cells and culture-expanded marrow mesenchymal stem cells in advanced breast cancer patients receiving high-dose chemotherapy. J Clin Oncol 2000; 18: 307.
12. Sallusto F, Lanzavecchia A. Efficient presentation of soluble antigen by cultured human dendritic cells is maintained by granulocyte/macrophage colony-stimulating factor plus interleukin 4 and downregulated by tumor necrosis factor α. J Exp Med 1994; 179: 1109.
13. Wilson JL, Proud G, Forsythe JL, et al. Renal allograft rejection. Tubular epithelial cells present alloantigen in the presence of costimulatory CD28 antibody. Transplantation 1995; 59: 91.
14. Laning JC, Deluca JE, Isaacs CM, et al. In vitro analysis of CD40-CD154 and CD28-CD80/86 interactions in the primary T-cell response to allogeneic “nonprofessional” antigen presenting cells. Transplantation 2001; 71: 1467.
15. Munn DH, Zhou M, Attwood JT, et al. Prevention of allogeneic fetal rejection by tryptophan catabolism. Science 1998; 281: 1191.
16. Boussiotis VA, Freeman GJ, Gray G, et al. B7 but not intercellular adhesion molecule-1 costimulation prevents the induction of human alloantigen-specific tolerance. J Exp Med 1993; 178: 1753.
17. Pober JS. Immunobiology of human vascular endothelium. Immunol Res 1999; 19: 225.
18. Curnow SJ, Willcox N, Vincent A. Induction of primary immune responses by allogeneic human myoblasts: dissection of the cell types required for proliferation, IFNγ secretion and cytotoxicity. J Neuroimmunol 1998; 86: 53.
19. Boussiotis VA, Freeman GJ, Taylor PA, et al. p27kip1 functions as an anergy factor inhibiting interleukin 2 transcription and clonal expansion of alloreactive human and mouse helper T lymphocytes. Nat Med 2000; 6: 290.
20. Tzachanis D, Freeman GJ, Hirano N, et al. Tob is a negative regulator of activation that is expressed in anergic and quiescent T cells. Nat Immunol 2001; 2: 1174.
21. Rezai KA, Semnani RT, Farrokh-Siar L, et al. Human fetal retinal pigment epithelial cells induce apoptosis in allogenic T-cells in a Fas ligand and PGE2 independent pathway. Curr Eye Res 1999; 18: 430.
22. Liversidge J, McKay D, Mullen G, et al. Retinal pigment epithelial cells modulate lymphocyte function at the blood-retina barrier by autocrine PGE2 and membrane-bound mechanisms. Cell Immunol 1993; 149: 315.
23. Laning JC, Isaacs CM, Hardin-Young J. Normal human keratinocytes inhibit the proliferation of unprimed T cells by TGFβ and PGE2, but not IL-10. Cell Immunol 1997; 175: 16.
24. Liechty KW, MacKenzie TC, Shaaban AF, et al. Human mesenchymal stem cells engraft and demonstrate site-specific differentiation after in utero transplantation in sheep. Nat Med 2000; 6: 1282.
25. Devine SM, Bartholomew AM, Mahmud N, et al. Mesenchymal stem cells are capable of homing to the bone marrow of non-human primates following systemic infusion. Exp Hematol 2001; 29: 244.
26. Almeida-Porada G, Porada CD, Tran N, et al. Cotransplantation of human stromal cell progenitors into preimmune fetal sheep results in early appearance of human donor cells in circulation and boosts cell levels in bone marrow at later time points after transplantation. Blood 2000; 95: 3620.
27. Lukomska B, Janczewska S, Interewicz B, et al. Engraftment of donor-derived stromal cells stimulates fast hematopoietic repopulation of vascularized bone marrow transplant recipients. Transplant Proc 2001; 33: 1757.
28. Fan TX, Hisha H, Jin TN, et al. Successful allogeneic bone marrow transplantation (BMT) by injection of bone marrow cells via portal vein: stromal cells as BMT-facilitating cells. Stem Cells 2001; 19: 144.
29. Gurevitch O, Prigozhina TB, Pugatsch T, et al. Transplantation of allogeneic or xenogeneic bone marrow within the donor stromal microenvironment. Transplantation 1999; 68: 1362.
30. Bartholomew A, Sturgeon C, Siatskas M, et al. Mesenchymal stem cells suppress lymphocyte proliferation in vitro and prolong skin graft survival in vivo. Exp Hematol 2002; 30: 42.
© 2003 Lippincott Williams & Wilkins, Inc.