Mesenchymal stem cells (MSCs) from adult human bone marrow can differentiate into bone, cartilage, fat, and muscle (1,2) and support hematopoiesis (3,4). Previous studies have shown that MSCs inhibit lymphocyte proliferation in mixed lymphocyte cultures (MLCs) (5–9). They express intermediate human leukocyte antigen (HLA) class I and can be stimulated to express HLA class II and should therefore be recognized by alloreactive T cells. However, MSCs do not induce proliferation of allogeneic lymphocytes, instead they reduce alloreactivity in vitro and prolong skin graft survival in vivo (5). This may indicate that MSCs escape alloreactive recognition.
Preliminary results also indicate that MSCs can reduce the incidence and severity of graft-versus-host disease (GVHD) if cotransplanted with hematopoietic stem cells to treat leukemia (10). No data show that MSCs also inhibit the graft-versus-leukemia (GVL) effect, which is necessary to avoid minimal residual disease and leukemic relapse.
This study determines whether MSCs affect cytotoxic T lymphocytes (CTLs) and natural killer (NK) cells in vitro. Transwell MLCs were also performed to evaluate whether the inhibitory effect of MSCs was mediated by a soluble factor. Finally, we examined whether MSCs were lysed by allogeneic CTLs and NK cells.
MATERIALS AND METHODS
Both fresh and frozen MSCs were used in these trials. Briefly, heparinized bone marrow from healthy persons was washed in phosphate-buffered saline (PBS) and centrifuged at 900 g for 10 min at room temperature. The cells were resuspended in PBS to a final density of 1×108 cells/mL. Five-milliliter aliquots were layered over 1,073 g/mL Percoll solution (Sigma, St. Louis, MO) and centrifuged at 900 g for 30 min. Mononuclear cells were collected from the interface, resuspended in PBS, and centrifuged at 900 g for 10 min. The cells were then grown in Dulbecco’s Modified Eagles Medium, low glucose (Life Technologies, Gaithersburg, MD) supplemented with 10% fetal bovine serum (Sigma), and 1% antibiotic-antimycotic solution (Life Technologies). Cells were cultured at 37°C in a humidified atmosphere containing 5% CO2. When the cultures became near confluence, the cells were detached with 0.05% trypsin and 0.02% EDTA solution and kept in Roswell Park Memorial Institute (RPMI) 1640 (GIBCO, BRL, Paisley, UK) supplemented with human antibody serum, 2 mM l-glutamine, 100 IU/mL penicillin, and 100 μg/mL streptomycin (complete media) pending use. Cryopreserved cells in 10% DMSO were thawed, washed in complete RPMI media, and used directly in experiments. The MSCs used in the present study differentiated into bone, fat, and cartilage under certain culture conditions as previously described (7). By flow cytometry, the cells were positive for CD166, CD105, CD44, CD29, SH-3, and SH-4, but negative for CD34, CD14, and CD45.
Mixed Lymphocyte Culture
Lymphocytes were prepared from two healthy volunteers by the use of Ficoll Hypaque separation (1,077 g/cm3, Axis-Shield PoC AS, Oslo, Norway). Briefly, 5×106 fresh responder cells (A) were cocultured with 5×106 irradiated (20 Gy) stimulator cells (Bx) in 8 mL complete RPMI medium in Falcon culture flasks (Becton Dickinson Labware, Franklin Lakes, NJ) for 5 days at 37°C, 5% CO2 in humidified air. Third-party irradiated MSCs were added on day 0 or day 3. In addition, 10×106 stimulator cells were cultured separately until day 6. On day 3, phytohemagglutinin (PHA, 10 μg/mL medium, Wellcome, Dartford, UK) was added to stimulate blast formation for better uptake of chromium. In the transwell system, 1.25×106 responder cells were cocultured with an equal number of irradiated stimulator cells in 2 mL of medium in a 12-well plate (Falcon, BD). We added 0.125×106 MSCs (10% of responder cells) to the bottom of the wells or in the 0.4-μm transwell inserts (Falcon, BD) and incubated as described.
On day 6, the target cells were washed in serum-free medium at 200 g for 5 min. An excess of 51Cr (100 μCi, sodium chromate in sodium chloride solution, Amersham Pharmacia Biotech UK Limited, Amersham, UK) in 100 to 200 μL serum-free medium was added to 1 to 5 million cells and incorporated for 1 hr at 37°C and shaken at 10-min intervals. After two washes in complete media and PBS, the cell count was adjusted to 0.1×106/mL. The MLCs were washed, counted, and plated in triplicate with 104 target cells per well in a 96-well round-bottomed plate (Nunclone surface, NUNC Brand Products, Roskilde, Denmark) at a 50:1 effector-to-target (ET) ratio and MSC 10%, 1%, or 0.1% of effector cells. Titration also gave 25:1, 12.5:1, and 6.25:1 in a final volume of 200 μL per well. Target cells were mixed with complete media to obtain spontaneous lysis and with 5% triton-X for maximum lysis (discussed later). The plate was centrifuged at 130 g for 4 min and then incubated 4 hr at 37°C, 5% CO2 in humidified air; 100 μL supernatant from each well was collected and counted in a gamma-counter (Wallac, Turku, Finland). The percent specific release was calculated using Brynner’s equation:EQUATION
To determine whether the CTLs could lyse allogeneic MSCs, MLCs were performed as described, and on day 6, PHA-stimulated peripheral blood lymphocytes (PBLs) and MSCs autologous with the stimulator PBLs were 51Cr-labeled and used as targets.
Natural Killer Lysis of K562
Purified NK cells were obtained by negative selection with magnetic beads (Dynal Biotech, Oslo, Norway). The degree of purity was measured with fluorescence-activated cell sorter analysis with labeled antibodies against CD56 (Becton Dickinson, Brussels, Belgium) and CD3 (DAKO, Glostrup, Denmark). We used K562, a well-characterized chronic myeloic lymphoma cell line known for its NK cell sensitivity caused by a lack of MHC class I molecules, as target cells. The cells were labeled, prepared, and plated as described in specific ET ratios. Third-party MSCs were added to wells at concentrations of 10% and 0.1% of NK cells. Controls were used to determine whether the NK cells lysed MSCs by labeling MSCs as described.
Human Leukocyte Antigen Typing
DNA was prepared from fresh blood samples or frozen cells by using the “salting out” method. Erythrocytes in blood were lysed in DNA buffer A (sucrose 0.32 M, Tris-HCl 0.01 M, MgCl2-6H2O 5 mM, Triton X-100 1%, Tris Base 2 mM, and distilled water) and removed by centrifugation for 15 min at 970 g. The procedure was repeated once on the pellet. In the case of frozen cells, the media was removed by washing twice in PBS. The cells were lysed overnight with slow shaking at 37°C in 40 μL proteinase K (10 mg/mL), 40 μL 20% sodium dodecyl sulfate, 180 μL DNA buffer B, pH 8.0 (NaCl 0.375 M, EDTA 0.12 M, and distilled water) and 340 μL distilled water. We added 200 μL saturated NaCl (6M) to precipitate proteins, which were removed by centrifugation for 10 min at 16,060 g. Centrifugation was repeated once for purification. Cold ethanol (99.5%) was used to precipitate DNA and collected by centrifugation for 25 min at 16,060 g. The pellet was washed in cold ethanol (70%), centrifuged as described, and then dried before the DNA was resuspended in distilled water. The concentration and purity of DNA were measured spectrophotometrically (Pharmacia Biotech, Cambridge, UK) at 260 nm and then kept frozen at −20°C. Low- and high-resolution HLA typing was performed for HLA-A, B, and C by polymerase chain reaction with sequence-specific primers, using the kit from Olerup SSP (Saltsjöbaden, Sweden). DNA (30 μg/mL), Mastermix with Taq polymerase, and distilled water were mixed and added to tubes containing various primers. After amplification, the DNA was loaded onto 2% agarose gels (Biowhittaker Molecular Applications, Rockland, ME) containing ethidium bromide (0.375 μg/mL, Olerup SSP) in TBE buffer containing 45 mM Trisbase (Tris[hydroxymethyl]-aminomethane), 89 mM boric acid, and 2 mM EDTA. Detection of an allele is based on the presence or absence of a polymerase chain reaction product visualized on the gel. Alloreactivity was evaluated by comparing various ligands for inhibitory receptors on NK cells and MSCs (11,12).
The P value was calculated using the Student t test, and values less than 0.05 were considered statistically significant.
Mesenchymal Stem Cells Inhibit the Formation of Cytotoxic T Lymphocytes in a Time- and Dose-Dependent Manner
The CTL-mediated lysis was inhibited by MSCs in a time- and dose-dependent manner. When added on day 0 of the 6-day MLC, MSCs significantly inhibited the lysis of target blasts (Fig. 1). When added to the MLC on day 3, MSCs only slightly reduced CTL lysis. When MSCs were added in the cytotoxic phase, however, no significant inhibition of CTL-mediated cytotoxicity was seen. In most CTL experiments, the maximum specific lysis for A+Bx in a 50:1 ET cell ratio was between 55% and 70%, and spontaneous release was always less than 20% of maximum lysis. The lysis of autologous cells, used as negative controls, was always too low to calculate any specific lysis.
Mesenchymal Stem Cells Are Not Lysed by Cytotoxic T Lymphocytes
To determine whether CTLs could lyse allogeneic MSCs, PBLs were cocultured with allogeneic lymphocytes and lysis was studied in a chromium-release assay. As shown in Figure 2, the CTLs caused marked lysis of the PHA blasts derived from the same donor as the stimulating PBL but did not lyse that donor’s MSCs.
Mesenchymal Stem Cells Inhibit Cytotoxic T Lymphocyte-Mediated Lysis Trough Soluble Factor(s)
The use of a transwell system in which responder and irradiated stimulator PBLs were cocultured with MSCs placed in transwell inserts allowed us to study whether the inhibition from MSCs was mediated by direct contact or a soluble factor. We found that the addition of 10% MSCs on day 0 to the MLC, in inserts or in contact with the MLC, induced similar inhibition, indicating that one or more soluble molecules are of importance for the inhibition (Fig. 3).
Mesenchymal Stem Cells Do Not Inhibit Natural Killer-Mediated Lysis of K562 Cells
Negative selection with magnetic beads yielded a median and average of 72% (range 43%–91%) NK cell purity and eliminated more than 98% of all CD3+ cells, as determined by flow cytometry. Specific NK cell lysis of Cr-labeled K562 cells during 4 hr was studied in the presence of various amounts of irradiated MSCs (Fig. 4). The lysis varied to some extent between the NK cell donors. In the 25:1 ET ratio, with no MSCs added, specific NK lysis was 23.0% (median; range 21.1%–41.9%). The addition of 10% or 0.1% MSCs did not affect the lysis, giving a median lysis of 29.8% (28.2%–46.6%) and 25.9% (21.7%–43.5%), respectively.
Mesenchymal Stem Cells Are Not Lysed by Natural Killer Cells
To exclude the possibility that MSCs were unable to affect NK cell lysis, because they were themselves lysed, we studied whether MSCs could be targets for NK cell lysis. Failure to recognize the appropriate killer cell immunoglobulin-like receptor (KIR) ligand on a mismatched cell can trigger NK cell elimination of that target cell (11,13,14). KIR ligand matching, using HLA class I typing, determines the ability of NK cells to lyse targets, and a mismatch is required for alloreactivity. HLA typing of MSC and NK donors confirmed a KIR ligand mismatch (Table 1). Despite this, NK cell-specific lysis of MSCs was approximately 80% less than lysis of K562 cells (Fig. 5), indicating that MSCs escape lysis by alloreactive NK cells.
The findings show that MSCs inhibit T cells in the early activating phase of the allograft reaction, but not in the effector phase. When MSCs were added early in the MLC, they inhibited the cytotoxicity, presumably by preventing the formation of active CTLs. When MSCs were added on day 3 to the 6-day MLC, little effect on cytotoxicity was noted. This would indicate that MSCs inhibit early in the activation process, or that MSCs inhibit a factor that is produced in too high amounts on day 3 for inhibition to occur. The fact that MSCs added to cell-mediated lympholysis failed to inhibit lysis indicates that they do not interfere with the interaction between CTLs and the target cells. The MSCs may secrete some substance that inhibits the cells in an early phase, because MSCs inhibited MLC and then cell-mediated lympholysis in a transwell system without direct contact between MSCs and the MLC. The lysis was slightly greater with transwell inserts, and the closer contact between MSCs and PBLs is necessary for the optimal inhibition. A recent study suggests that the immune modulatory effect of MSCs is caused by their secretion of transforming growth factor-β1 and hepatocyte growth factor, because the use of neutralizing monoclonal antibodies against these factors abolished the suppressive effects on T cells (15). However, we were unable to confirm these findings (K. Le Blanc, M.D., Ph.D., unpublished data, 2003).
A naive cytotoxic T cell is not constitutively cytotoxic; it requires an antigen signal, a costimulatory signal, and interleukin (IL)-2 produced by activated helper T cells to be activated. Formed CTLs then kill through perforin-granzyme–mediated lysis or Fas ligand-induced apoptosis. In contrast, NK cells can lyse cells without prior immunization but with similar mechanisms of killing. IL-2 is needed for the activation of T cells. NK cells are further activated by IL-2 but can also be activated by other cytokines, such as IL-12, IL-15, IL-21, and interferon (IFN)-γ (16). Bartholomew et al. (5) showed that the addition of exogenous IL-2 partly reversed the inhibitory effect MSCs have on MLC. This may indicate that MSCs affect some other signal pathway than that of IL-2. MSCs do not seem to inhibit cells, which are already primed. This indicates that the lytic phase is mediated by cell surface molecules that are probably not affected by MSCs. However, MSCs may reduce the up-regulation of receptors during blast proliferation. This hypothesis would explain why NK cells, being constitutively cytotoxic, show normal lysis even in the presence of MSCs. The addition of MSCs in the lytic phase to NK cells and K562 induced a slight, but not significant stimulation, rather than inhibition, which could be caused by variations between NK-cell donors.
Preliminary clinical data indicate that MSCs can inhibit GVHD (10), and one of the aims of this study was to evaluate the influence of MSCs on the GVL effect associated with GVHD. CTLs cause GVHD (17,18), and by removing T cells in the stem cell graft, GVHD can be prevented. However, T-cell depletion increases the risk of relapse from leukemia after transplantation. Moreover, patients with acute or chronic GVHD experience leukemic relapses less frequently than those with no GVHD (19). However, experimental data support that the elimination of host histocompatibility antigen-specific T cells alone (leaving tumor antigen-specific T cells) prevents GVHD, whereas GVL is preserved (20). Certain cytokines, for example, IL-2, IFN-α (21), IL-11 (22), IL-12 (23), and IFN-γ (24) can induce a GVL effect without generating GVHD. MSCs secrete several such cytokines, including IL-11, IL-12, and IL-15 (3) that may sustain a GVL effect. This could have clinical implications. MSCs should probably be given shortly after transplantation to modify the T cells in the graft when antigen presentation occurs. At the time of acute GVHD, after alloreactive T cells already have developed, MSCs may not be effective. If used to treat GVHD, MSCs should probably be combined with antibody treatment to eliminate the CTLs. NK cells have been believed to be more important for GVL than GVHD (17,18,25). In a mouse model, the addition of NK cells given within 3 days of allogeneic stem-cell infusion induced a graft-versus-tumor effect without causing GVHD. After 3 days, NK cells did not suppress GVHD (26).
MSCs can differentiate into cartilage, bone, and muscle, and in the future they may be used to replace damaged tissue (27,28). We studied whether MSCs are targets for an alloresponse because it is important when choosing autologous or allogeneic MSCs for transplantation. The use of allogeneic MSCs in such cases has practical advantages, because expansion of sufficient numbers of autologous MSCs takes several weeks. Allogeneic MSCs can be stored for immediate use when needed. The finding that MSCs escape recognition by CTLs is surprising because MSCs express HLA class I antigens, which are targets for CTL lysis (7). It is also surprising that MSCs escape NK cell lysis because NK cells lyse allogeneic HLA I expressing cells when there is a KIR ligand mismatch (29). Moreover, in this study we found that MSCs did not inhibit the cytolytic effect of CTLs activated in MLC. However, PBLs, but not MSCs, were lysed by CTLs, although PBLs and MSCs were from the same individual. In vivo experiments are needed to clarify these findings. Our data indicate that MSCs might be transplantable between HLA incompatible persons because they are not destroyed by CTLs or NK cells and do not induce an immune response.
Our observation that MSCs do not inhibit CTLs or NK cells may indicate that these cells will not adversely affect GVL when transplanted together with hematopoietic stem cells. However, MSCs inhibit the development of CTLs, as shown in this study, and alloreactivity, including GVHD. However, clinical studies are needed to determine whether leukemia-free survival can be improved by MSCs.
MSCs inhibited the formation of CTLs in MLC. MSCs did not abrogate CTL- or NK cell-induced lysis. MSCs were not targets for CTL or NK cell destruction. These in vitro findings may have clinical implications for MSC transplantation.
1. Pittenger MF, Mackay AM, Beck SC, et al. Multilineage potential of adult human mesenchymal stem cells. Science 1999; 284: 143.
2. Haynesworth SE, Goshima J, Goldberg VM, et al. Characterization of cells with osteogenic potential from human marrow. Bone 1992; 13: 81.
3. Majumdar MK, Thiede MA, Haynesworth SE, et al. Human marrow-derived mesenchymal stem cells (MSCs) express hematopoietic cytokines and support long-term hematopoiesis when differentiated toward stromal and osteogenic lineages. J Hematother Stem Cell Res 2000; 9: 841.
4. 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.
5. 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.
6. Klyushnenkova E, Mosca JD, McIntosh KR. Human mesenchymal stem cells suppress allogeneic T cell responses in vitro: implications for allogeneic transplantation. Blood 1998; 92: 642a.
7. LeBlanc K, Tammik L, Sundberg B, et al. Mesenchymal stem cells inhibit and stimulate mixed lymphocyte cultures and mitogenic responses independently of the major histocompatibility complex. Scand J Immunol 2003; 57: 11.
8. Krampera M, Glennie S, Dyson J, et al. Bone marrow mesenchymal stem cells inhibit the response of naive and memory antigen-specific T cells to their cognate peptide. Blood 2003; 101: 3722-3729.
9. Tse WT, Pendleton JD, Beyer WM, et al. Suppression of allogeneic T-cell proliferation by human marrow stromal cells: implications in transplantation. Transplantation 2003; 75: 389.
10. Frassoni F, Labopin M, Bacigalupo A, et al. Expanded mesenchymal stem cells (MSC) co-infused with HLA identical hematopoietic stem cell transplants, reduce acute and chronic graft versus host disease: a matched pair analysis. Bone Marrow Transplant 2002; 29: S2.
11. Farag SS, Fehniger TA, Ruggeri L, et al. Natural killer cell receptors: new biology and insights into the graft-versus-leukemia effect. Blood 2002; 100: 1935.
12. Oertel M, Kohlhaw K, Diepolder HM, et al. Alloreactivity of natural killer cells in allogeneic liver transplantation. Transplantation 2001; 72: 116.
13. Davies SM, Ruggieri L, DeFor T, et al. Evaluation of KIR ligand incompatibility in mismatched unrelated donor hematopoietic transplants. Blood 2002; 100: 3825.
14. Ruggeri L, Capanni M, Casucci M, et al. Role of natural killer cell alloreactivity in HLA-mismatched hematopoietic stem cell transplantation. Blood 1999; 94: 333.
15. Di Nicola M, Carlo-Stella C, Magni M, et al. Human bone marrow stromal cells suppress T-lymphocyte proliferation induced by cellular or nonspecific mitogenic stimuli. Blood 2002; 99: 3838.
16. Chiorean EG, Miller JS. The biology of natural killer cells and implications for therapy of human disease. J Hematother Stem Cell Res 2001; 10: 451.
17. Ruggeri L, Capanni M, Urbani E, et al. Effectiveness of donor natural killer cell alloreactivity in mismatched hematopoietic transplants. Science 2002; 295: 2097.
18. Jiang YZ, Barrett AJ, Goldman JM, et al. Association of natural killer cell immune recovery with a graft-versus-leukemia effect independent of graft-versus-host disease following allogeneic bone marrow transplantation. Ann Hematol 1997; 74: 1.
19. Horowitz MM, Gale RP, Sondel PM, et al. Graft-versus-leukemia reactions after bone marrow transplantation. Blood 1990; 75: 555.
20. Chen BJ, Cui X, Liu C, et al. Prevention of graft-versus-host disease while preserving graft-versus-leukemia effect after selective depletion of host-reactive T cells by photodynamic cell purging process. Blood 2002; 99: 3083.
21. Slavin S, Ackerstein A, Kedar E, et al. IL-2 activated cell-mediated immunotherapy: control of minimal residual disease in malignant disorders by allogeneic lymphocytes and IL-2. Bone Marrow Transplant 1990; 6 (Suppl 1): 86.
22. Teshima T, Hill GR, Pan L, et al. IL-11 separates graft-versus-leukemia effects from graft-versus-host disease after bone marrow transplantation. J Clin Invest 1999; 104: 317.
23. Yang YG, Sergio JJ, Pearson DA, et al. Interleukin-12 preserves the graft-versus-leukemia effect of allogeneic CD8 T cells while inhibiting CD4-dependent graft-versus-host disease in mice. Blood 1997; 90: 4651.
24. Yang YG, Qi J, Wang MG, et al. Donor-derived interferon gamma separates graft-versus-leukemia effects and graft-versus-host disease induced by donor CD8 T cells. Blood 2002; 99: 4207.
25. Xun CQ, Thompson JS, Jennings CD, et al. The effect of human IL-2-activated natural killer and T cells on graft-versus-host disease and graft-versus-leukemia in SCID mice bearing human leukemic cells. Transplantation 1995; 60: 821.
26. Murphy WJ, Longo DL. The potential role of NK cells in the separation of graft-versus-tumor effects from graft-versus-host disease after allogeneic bone marrow transplantation. Immunol Rev 1997; 157: 167.
27. Prockop DJ. Marrow stromal cells as stem cells for nonhematopoietic tissues. Science 1997; 276: 71.
28. Fukuda K. Development of regenerative cardiomyocytes from mesenchymal stem cells for cardiovascular tissue engineering. Artif Organs 2001; 25: 187.
29. Ruggeri L, Capanni M, Martelli MF, et al. Cellular therapy: exploiting NK cell alloreactivity in transplantation. Curr Opin Hematol 2001; 8: 355.