Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Immunol. Author manuscript; available in PMC 2009 July 15.
Published in final edited form as:
PMCID: PMC2593473

Induction of Tolerance to Cardiac Allografts Using Donor Splenocytes Engineered to Display on Their Surface an Exogenous FasL Protein1


The critical role played by FasL in immune homeostasis renders this molecule as an attractive target for immunomodulation to achieve tolerance to auto and transplantation antigens. Immunomodulation with genetically modified cells expressing FasL was shown to induce tolerance to alloantigens. However, genetic modification of primary cells in a rapid, efficient, and clinically applicable manner proved challenging. Therefore, we tested the efficacy of donor splenocytes rapidly and efficiently engineered to display on their surface a chimeric form of FasL protein (SA-FasL) for tolerance induction to cardiac allografts. Intraperitoneal injection of ACI rats with WF splenocytes displaying SA-FasL on their surface resulted in tolerance to donor, but not F344 third party, cardiac allografts. Tolerance was associated with apoptosis of donor reactive T effector cells and induction/expansion of CD4+CD25+FoxP3+ T regulatory (Treg) cells. Treg cells played a critical role in the observed tolerance as adoptive transfer of sorted Treg cells from long-term graft recipients into naïve unmanipulated ACI rats resulted in indefinite survival of secondary WF grafts. Immunomodulation with allogeneic cells rapidly and efficiently engineered to display on their surface SA-FasL protein provides an effective and clinically applicable means of cell-based therapy with potential application to regenerative medicine, transplantation, and autoimmunity.

Keywords: T regulatory cells, Allograft tolerance, FasL, Apoptosis


Transplantation of foreign cellular, tissue, and organ grafts represents an important therapeutic modality for the treatment of various inherited and acquired disorders, such as end-organ failure, autoimmunity, malignancies, and congenital enzyme deficiencies. However, rejection of foreign grafts by immunocompetent recipients presents a major hurdle for the routine application of transplantation in the clinic (1,2). Although general immunosuppression is presently used to control foreign graft rejection, the chronic use of nonspecific immunosuppression is not only inefficient in preventing graft rejection, but is also associated with various complications, including but not limited to infections, malignancies, and organ toxicity (1,2). Induction of specific tolerance to foreign grafts has the potential to overcome these complications, and as such has been the subject of intense studies since the first successful organ transplantation in 1954 (3). Irrespective of extensive efforts, the routine and consistent induction of transplantation tolerance in the clinic remains to be realized (4).

Modulation of immune responses using donor cells genetically modified to express immunologic molecules that play key roles in the regulation of the immune system has the potential to induce transplantation tolerance (5). However, using gene therapy to express immunomodulatory molecules has various complications, such as safety, inefficient targeting, and low levels of expression of the therapeutic protein (6). In particular, in settings of organ, tissue, and primary cell transplantation where rapid and robust expression of immunological proteins are a prerequisite for therapeutic efficacy, gene therapy has severe limitations. Inasmuch as cell surface receptor ligand interactions are critical to immune decision making and these interactions do not need to be extensive in duration (7), we sought direct display of exogenous immunological proteins on the cell surface as a practical alternative to gene therapy for immunomodulation and developed the ProtEx technology (8). ProtEx involves; i) generation of recombinant proteins that contain extracellular domains of immunological ligands fused to a modified form of core streptavidin (SA); ii) modification of cell membrane with biotin; and iii) display of chimeric proteins on the modified surfaces (8).

We tested the immunomodulatory potential of our ProtEx technology using Fas-ligand (FasL) as an apoptotic molecule to specifically eliminate pathogenic lymphocytes in settings of autoimmunity and allograft transplantation. The choice of FasL as an immunomodulatory molecule is because of its critical role in activation induced cell death (AICD) and tolerance to self antigens (9,10). As such, there has been significant interest in using FasL as an immunomodulatory molecule to induce transplantation tolerance with conflicting observations (11). While some studies reported therapeutic effect of FasL, others demonstrated its contribution to the pathogenesis of the disease (1220). Mechanisms responsible for these opposing effects are complex, and may be regulated by levels of FasL and its receptor, Fas, on target tissues and the sensitivity of these tissues to FasL-mediated apoptosis, the cytokine milieu, the tissue microenvironment, and/or the differential effects of membrane-bound and soluble forms of FasL (reviewed in ref. 11). FasL is initially expressed on the cell surface as a membrane-bound type II protein that has potent apoptotic activity on Fas bearing cells (9). However, membrane-bound FasL is cleaved from the cell surface by matrix metalloproteinases into a soluble form (21), which does not have apoptotic activity and may interfere with apoptosis by competing with the membranous form for binding to Fas on target cells (21,22). Furthermore, soluble FasL has chemotactic activity on neutrophils (23,24) that may be responsible for the rejection of grafts ectopically expressing FasL (13,19).

Therefore, a chimeric molecule containing the extracellular functional domain of FasL lacking the metalloproteinase cleavage sites fused to streptavidin (SA-FasL) was generated (8). We previously demonstrated that the direct display of SA-FasL on heart graft vasculature resulted in the prevention of acute rejection following transplantation into allogeneic recipients (25). However, this approach of direct manipulation of the donor graft resulted in moderate prolongation of graft survival without achieving tolerance. The lack of tolerance in this model might have been due to the inability of the transient, local display of SA-FasL on the graft vasculature to deplete a significant pool of alloreactive T effector cells. Hence, we herein tested whether recipient immunomodulation by systemic infusion of donor splenocytes engineered to display SA-FasL on their surface induces tolerance to cardiac grafts in a totally allogeneic rat strain combination. Infusion of ACI graft recipients with WF donor splenocytes displaying SA-FasL on their surface resulted in robust tolerance to cardiac allografts. Tolerance was donor-specific and maintained by CD4+CD25+FoxP3+ T regulatory (Treg) cells. The rapid and efficient display of exogenous proteins on the cell membrane, therefore, represents a practical and effective cell based therapy with potential application to autoimmunity, transplantation, and regenerative medicine.

Materials and Methods


Eight to 12-week old male Wistar-Furth (WF, RT1u), ACI (RT1a), and Fisher (RT1lv) rats were purchased from Harlan Sprague-Dawley (Indianapolis, IN) and 2C mice were generously provided by Dr. J. Connolly, Washington University School of Medicine, St. Louise, MO. All animals were maintained under specific SPF conditions in our barrier facility and used according to the University of Louisville and NIH institutional guidelines.

Construction of SA-FasL and protein expression in Drosophila S2 cells

The construction of SA-FasL gene was reported previously (8). Briefly, DNA segment encoding core streptavidin was amplified using specific primers (forward 5′-AGATCTCATCATCACCATCACCATATCACCGGCACC and reverse 5′-GAATTCGGAGGCGGCGGACGGCT) and total genomic DNA from Streptomyces avidinii (ATCC) as template for PCR. The 5′-primer was designed to include a BglII restriction site and 6 His residues to allow cloning in frame with the Drosophila secretion signal (BiP) in pMT/BiP/V5-His vector for expression in S2 cells as a secreted protein and purification using Chelating Sepharose Fast Flow Columns. The cDNA encoding the extracellular domain of rat FasL without the metalloproteinase site (nucleotides 428–998) was cloned using total RNA from ConA-activated splenocytes as a template and specific primers (5′-primer; nucleotide 428–453 and 3′-primer; nucleotide 977–998). These primers were engineered to include EcoRI sites for cloning in frame with SA in the pMT/BiP/V5-His vector. The chimeric gene in pMT/BiP/V5-His vector was used to stably transfect S2 cells (Invitrogen). Stable transfectants were induced by 1.0 mM CuS04 and secreted chimeric protein in culture medium was collected 1–4 days after induction. The supernatant was either used immediately or precipitated with 50% ammonium persulfate, dialyzed against PBS, and purified using Chelating Sepharose Fast Flow Columns (Amersham). The concentration of purified SA-FasL or protein in culture supernatant was determined by the Western blot or ELISA using known amounts of commercially available streptavidin as standard.

Engineering donor splenocytes to display SA-FasL on their surface

Spleens were harvested from WF rats, processed into single cell suspension, and red blood cells were lyzed using ACK solution. Cells were biotinylated by incubation in 5 μM biotin solution (Pierce, Rockford, IL) in PBS at room temperature for 30 min. Cells were washed twice with PBS, and then incubated with ~ 200 ng SA-FasL protein/1 ×106 cells in PBS for 30 min by constant rocking in a cold room. After washing twice, cells were resuspended in PBS, and counted as previously described (8). A portion of the cells were stained with streptavidin-APC, FITC-labeled anti FasL, or FITC-labeled anti-streptavidin Abs to assess the level of biotinylation and SA-FasL-decoration, respectively, using flow cytometry.

Splenocyte treatment and transplantation

ACI graft recipients received 13 × 106 unmodified WF donor splenocytes by i.p. injection 6 days prior to heart transplantation. Grafts recipients were then treated on days −3, −1, +1, +3 and +5 with respect to heart transplantation on day 0 with the same number of unmodified (n = 8) or SA-FasL (n = 23) or SA (n = 20) engineered donor splenocytes. Recipients without splenocyte treatment (n = 10) served as control to determine the normal rate of graft rejection. Another group of treated ACI animals (n = 4) were transplanted with F344 third party heart as antigen-specificity controls. Intra-abdominal heterotopic heart transplantation was performed as previously described (26). Ventricular contractions were assessed by daily palpation and rejection was defined as cessation of heartbeat verified by autopsy and pathology.

T cell tracking using 2C model

Twenty million of total spleen cells labeled with CFSE were injected into the tail vain of C57BL/6. SJL mice. These animals were then immunized one day later by i.v. injection of 20 million of allogeneic BALB/c splenocytes either left unmodified or engineered with SA-FasL or control protein SA. Animals were euthanized 3 days after immunization and their spleens and mesenteric lymph nodes harvested for single cell preparation. Cells were stained with antibodies to CD45.2, Vb8.2, and CD8, and analyzed using flow cytometry by gating on CD45.2+CD8+Vb8.2+ T cells. Cell division was assessed using FlowJo software (Ashland, OR).

Mixed lymphocytes reaction (MLR)

Spleens were processed into single cell suspension, labeled with 2.5 μM CFSE, resuspended in DMEM, and 50 × 106 CFSE labeled splenocytes were plated on a petri dish for 45 minutes at 37°C to enrich lymphocytes. After 45 minutes non-adherent cells were collected, washed, and incubated (1 × 105 cells) with irradiated (2000 cGy; 1 × 105 cells) naïve ACI, WF or Fisher splenocytes in 96-well U-bottom titer plates in MLR medium. Irradiated syngeneic cells and Fisher third party splenocytes were used as controls. After 5-days cells were stained with fluorochrome-labeled antibodies against rat CD4 and CD8, and analyzed by flow cytometry. Data were analyzed using FlowJo (Tree Star Inc., San Carlos, CA) software. All proliferation assays were performed in triplicates and are representative of a minimum of 3 animals per group.

Intracellular cytokine staining

Splenocytes were resuspended in MLR medium at a concentration of 2 × 106 cells/ml, stimulated with PMA (5 ng/ml, Sigma) and ionomycin (500 ng/ml, Sigma) for 2 h at 37°C in a 5% CO2 incubator. Two hrs later, cultures were supplemented with Golgi Plug (1 μl/ml, BD Bioscience) and incubated for an additional 2.5 h. Cells were then stained with fluorescent conjugated antibodies against rat CD4, CD8, and IFN-γ, and analyzed by flow cytometry as previously described (27).

T cell phenotyping

Spleen and mesenteric lymph nodes were processed into single cells suspensions and stained with CD4-APC, CD8-PerCP, and CD25-PE mAbs or isotype controls at 4°C for 30 min. Cells were then washed twice with FACS buffer and analyzed by flow cytometry. For FoxP3 staining, cells were stained with antibodies against CD4, CD8, and CD25, washed twice with FACS buffer, fixed, and then stained for intracellular FoxP3 according to the manufacturer’s protocol (eBioscience, San Diego, CA).

Adoptive transfer studies

Spleens were harvested from long-term (> 90 days) graft survivors or acutely rejecting animals at rejection and processed into single cell suspensions. Naïve unirradiated or 450 cGy irradiated ACI rats received 80-95 × 106 cells via i.v. injection 24 h after irradiation. These animals were transplanted with WF hearts 24 h after cell transfer. Graft survival was monitored by abdominal palpation.

In selected experiments, sorted CD25+ T cells were used for adoptive transfer experiments. Briefly, lymphocytes from rats with long-term graft survival or acutely rejecting animals were stained with a saturating dose of anti-rat CD25 Ab (OX-39) conjugated to PE for 30 minutes at 4°C. After extensive washing to remove unbound Ab, cells were incubated with the appropriate amount of anti-PE microbeads (Miltenyi Biotec, Auburn, CA, USA). Magnetic separation was performed according to the manufacturer's instructions. Positively selected CD25+ and lymphocytes depleted of this population were stained with fluorochrome-labeled Abs to FoxP3 and CD4 and analyzed using multiparameter flow cytometry. CD4+ FoxP3+ and CD4+ FoxP3 populations were found to be over 90% pure.


Cardiac graft samples fixed in 10% formalin were embedded in paraffin, sectioned, and stained with hematoxylin and eosin for evaluation of cellular infiltration. Elastin staining was performed using a kit from Sigma Diagnostics (St. Louis, MO) according to the manufacturer’s instructions to demonstrate the vascular lesions and myocardial fibrosis. Briefly, tissue sections were stained in hematoxylin-iodine-ferric chloride solution and differentiated by use of a dilute ferric chloride solution. Van Gieson solution was used as a counterstain to stain collagen fibers red and myocardium yellow. Chronic rejection was assessed by light microscopy. The severity of chronic rejection was graded according to the percentage luminal occlusion by intimal thickening as previously described (28).


Data were analyzed using Student’s t test, nonparametric Mann-Whitney U test, Kaplan-Meier’s log-rank, and Tarone-Ware tests as appropriate. P < 0.05 was considered significant. Statistical analysis was performed using SPSS 13.0 software.


The display of SA-FasL on the surface of splenocytes does not alter the distribution of endogenous cell membrane proteins

The transplant experiments were preceded by a series of preliminary studies to test whether the display of SA-FasL on the surface of splenocytes interferes with the normal distribution of endogenous cell surface proteins, including molecules involved in antigen presentation. WF rat splenocytes were labeled with biotin (5 μM) and engineered with SA-FasL (~ 200 ng/106 cells). The levels of biotin and SA-FasL on the cell surface were assessed by fluorochrome-conjugated streptavidin and an antibody against streptavidin or FasL, respectively, using flow cytometry. Almost all targeted cells were positive for the cells surface biotin and SA-FasL (Fig. 1A). Biotinylated and SA-FasL-engineered cells were then analyzed for the presence of a series of endogenous cell surface molecules, such as CD3, CD4, CD8, class I MHC, and CD80 molecules, involved in the immune response using their respective antibodies in flow cytometry. As shown in Fig. 1B, neither biotinylation nor engineering with SA-FasL significantly altered the cell surface expression patterns of these molecules. Taken together, this data demonstrates that primary cells, such as splenocytes, can be engineered to display on their surface SA-FasL in a rapid and efficient manner without significantly altering the cell surface distribution of endogenous proteins.

Display of SA-FasL on donor splenocytes does not alter the distribution of endogenous cell surface proteins. A, Engineering WF rat splenocytes with SA-FasL protein. Splenocytes were modified with biotin (5 μM) and engineered with chimeric SA-FasL ...

Systemic immunomodulation with SA-FasL-engineered donor splenocytes induces allograft tolerance

T cells become sensitive to Fas/FasL-mediated apoptosis following activation, several rounds of proliferation, and re-encounter with the antigen (10). We, therefore, pretreated ACI recipients of WF donor heart allografts with 13 × 106 unmodified donor splenocytes i.p. to mobilize the alloreactive T cell pool and sensitize them to Fas/FasL-mediated apoptosis. These animals were then injected with several doses of SA-FasL-engineered donor splenocytes pre- and post-cardiac allograft transplantation as shown in Fig. 2A. Seventy percent of ACI rats treated with SA-FasL modified donor splenocytes accepted WF heart grafts over the 100-day observation period while the remaining 30% rejected their allografts in a median survival time (MST) of 12 ± 1.2 (Fig. 2B). In contrast, all untreated animals (n = 10) rejected their grafts in a MST of 9 ± 0.4 days (Fig. 2B). Treatment with splenocytes engineered with control SA protein (n = 20) resulted in only ~25% graft survival, which was similar to that achieved using unmodified splenocytes (n = 8). The induced tolerance was donor-specific as ACI rats (n = 4) treated with SA-FasL-engineered WF splenocytes rejected F344 third party heart allografts in a normal tempo (MST = 8 ± 0 days; Fig. 2B). Importantly long-term grafts (100 days) in SA-FasL treatment group had no inflammatory infiltrates and showed minimal neointimal proliferation (Fig. 2C; panels a,b). In marked contrast, grafts that survived long-term (100 days) as a result of immunomodulation with SA-engineered donor splenocytes had inflammatory infiltrates and showed significant incidences and levels of neointimal proliferation (Fig. 2C; panels c,d).

Systemic immunomodulation using SA-FasL-engineered donor splenocytes prevents cardiac allograft rejection in a fully mismatched WF-to-ACI rat model. A, Schematic diagram detailing the treatment regimen for ACI recipients. Unmodified WF splenocytes (13 ...

To rigorously test the ability of donor splenocytes to induce tolerance in a clinically relevant setting, where the heart and splenocytes are simultaneously available from cadaveric donors, the treatment commenced post-transplantation (Fig. 2D). Inasmuch as allogeneic heart grafts are efficient in activating alloantigen-specific T cells and as such sensitizing them to Fas-mediated apoptosis (29,30), all infusions contained donor splenocytes engineered with SA-FasL. As shown in Fig. 2E, over 62% of ACI recipients (n = 8) treated with SA-FasL engineered WF splenocytes did not reject WF heart allografts whereas all animals (n = 6) treated with control SA-engineered splenocytes rejected their grafts in a MST of 11 ±1.2 days. Taken together, this data demonstrates that systemic immunomodulation with SA-FasL-engineered splenocytes was effective in inducing tolerance to cardiac allografts without the use of immunosuppression.

Systemic immunomodulation with SA-FasL-engineered allogeneic splenocytes results in reduced proliferation and accumulation of alloreactive T cells

To determine if systemic infusion of SA-FasL-engineered splenocytes physically eliminates alloreactive cells in the host, we took advantage of the TCR transgenic 2C mouse model whose CD8+ T cells are specific for H-2Ld class I alloantigen (31). C57BL/6 (CD45.1) mice were adoptively transferred with congenic CFSE-labeled 2C T cells (C57BL/6; CD45.2) followed by immunization with BALB/c (H-2d) splenocytes engineered with SA-FasL or SA control protein. Three days later, splenocytes and lymph node cells from the immunized animals were harvested and analyzed using flow cytometry for the proliferation of 2C cells. Immunization with SA-engineered BALB/c splenocytes resulted in robust proliferation (> 8 cycles) of 2C cells and accumulation of increasing numbers of daughter cells per generation (Fig. 3A). In marked contrast, mice immunized with SA-FasL-engineered splenocytes had moderate proliferation (~ 4 cycles) of 2C cells. Importantly, there was a steady decrease in the number of daughter cells per generation plausibly due to apoptosis. This notion is consistent with published literature that naïve T cells require few cycles of proliferation in response to antigen before becoming sensitive to Fas/FasL-mediated apoptosis (10).

Immunomodulation with SA-FasL-engineered splenocytes results in apoptosis of alloreactive T cells. A, C57B/6.SJL (CD45.1) mice were adoptively transferred with 20 × 106 2C (C57B/6; CD45.2) splenocytes. One day later the animals were immunized ...

To provide direct evidence that alloreactive T cells undergo apoptosis following antigen recognition in the context of FasL on donor cells, we performed in vitro stimulation assays where BALB/c splenocytes engineered with SA-FasL or SA control protein were used as stimulators for 2C cells. There was a significant reduction in the number of proliferating 2C cells responding to SA-FasL-engineered donor cells as compared with control (Fig. 3B; upper panel). The observed reduction in proliferation was due to significant apoptosis of responding 2C T cells (Fig. 3B; lower panel).

Allograft tolerance is associated with donor-specific T cell hypoproliferation and reduced expression of IFN-γ

To further elucidate the mechanisms responsible for the observed tolerance to WF hearts, ACI graft recipients were evaluated for evidence of donor specific proliferation using CFSE-based mixed lymphocyte reactions (32). As shown in Fig. 4A, lymphocytes from SA-FasL-engineered cells treated long-term graft recipients (SA-FasL LT) responded poorly to donor cells but showed a normal response to third party cells, indicating antigen specificity. In marked contrast, lymphocytes from SA-engineered cells treated animals with acute graft rejection (SA Rej) and naïve control animals proliferated vigorously to both donor and third party cells. Moreover, tolerant rats had lower percentages of CD8+ T cells expressing IFN-γ, a signature cytokine for graft rejection (33,34), (Fig. 4B) as compared with SA-engineered cells treated rats with acute graft rejection (7.4 ± 3.6% vs. 28.5 ± 3.0% in CD8+ T cell gate; Fig. 4C). In marked contrast to CD8+ T cells, only a small percentage of CD4+ T cells expressed IFN-γ in all groups irrespective to the treatment regimen (Fig. 4C; gated on CD4+ cells). There was a slight increase in the percentage of CD4+ T cells expressing IFN-γ in FasL-treated animals, which was not statistically significant.

Long-term SA-FasL treated animals are hyporesponsive to donor antigens and have reduced percentages of T cells expressing IFN-γ. A, Splenocytes from long-term (> 90 days) graft survivors treated with SA-FasL-engineered donor splenocytes ...

Tolerance is associated with increased percentage of peripheral CD4+CD25+FoxP3+ T regulatory cells

Although we demonstrated that systemic immunomodulation with SA-FasL-engineered splenocytes induce apoptosis in activated alloreactive T cells (Fig. 3), this effect may be short lived due to the transient present of SA-FasL on the cells surface. Therefore, physical elimination of alloreactive T cells early in transplantation may not be sufficient for long-term allograft survival and may require induced peripheral immunoregulatory mechanisms. We focused on CD4+CD25+FoxP3+ T regulatory cells because of their capacity to induce and maintain tolerance in various transplantation settings (35). Lymphocytes were isolated from the spleen of long-term graft recipients (> 90 days) and phenotyped for CD4 and CD25 expression using flow cytometry (Fig. 5A). There was a significant (P < 0.05) increase in the percentage of CD4+CD25+ T cells (8.2 ± 1.3% of total lymphocytes) in long-term SA-FasL treated graft recipients as compared with naïve rats (5.1 ± 0.4%). In contrast, rats with acute graft rejection had similar percentages of CD4+CD25+ T cells as naïve animals whether they were treated with SA (5.7 ± 1.5 %) or SA-FasL-engineered (5.1 ± 1.2%) splenocytes (Fig. 5B). The increased percentage of CD4+CD25+ Treg cells in long-term graft survivors was further confirmed with an Ab to the signature transcriptional factor FoxP3 (Fig. 5C, gated on CD4+ T cells).

Long-term SA-FasL treated animals have high percentages of peripheral CD4+CD25+ T regulatory cells. Splenocytes from long-term (> 90 days) graft survivors treated with SA-FasL-engineered donor splenocytes (SA-FasL LT), acutely rejecting animals ...

Importantly, sorted CD4+CD25+ T cells from long-term graft survivors inhibited the proliferative responses of naïve T effector cells to donor alloantigens when used at various ratios in a CFSE-based in vitro proliferation assay (Fig. 5D). This finding indicates that CD4+CD25+ T cells from long-term graft survivors are T regulatory cells, rather than newly activated T effector cells expressing CD25. Taken together, these data demonstrate that systemic immunomodulation with SA-FasL-engineered donor splenocytes induces long-term graft survival that is associated with increased percentages of CD4+CD25+FoxP3+ Treg cells in the periphery.

Sorted CD4+CD25+FoxP3+ T regulatory cells from tolerant animals prevent the rejection of donor hearts in secondary naïve recipients

To provide direct evidence for the role of peripheral immunoregulatory mechanisms in the observed tolerance, 80-95 × 106 splenocytes harvested from primary graft recipients were adoptively transferred into a second cohort of naïve rats that had been subjected to 450 cGy total body irradiation one day earlier. All rats (n = 6) that received splenocytes from SA-FasL-engineered cells treated long-term primary graft survivors indefinitely accepted WF donor heart allografts (Fig. 6A). In marked contrast, animals (n = 4) receiving splenocytes from SA-engineered cells treated primary graft recipients with acute rejection rejected donor heart grafts in a MST of 9 ± 0.4 days. We also tested the efficacy of splenocytes from SA-FasL-treatment group that had delayed rejection in prolonging the survival of secondary grafts. Interestingly, adoptive transfer of these cells into secondary graft recipients resulted in prolonged survival of all grafts (MST = 15 ± 10.4 days) with one out of 6 grafts not rejecting during the 100-day observation period. These data suggest that SA-FasL treatment in this group had induced immunoregulatory cells, but these cells were not sufficient to prevent graft rejection in primary or secondary recipients after adoptive transfer. This notion is consistent with the reduced IFN-γ secretion in T cells from this group as compared with the SA-treatment group (Fig. 4C).

Tolerance is maintained by CD4+CD25+ Treg cells. A, Adoptive transfer experiments into irradiated secondary graft recipients. Splenocytes (80-95 × 106) harvested from long-term (> 90 days; SA-FasL) or acutely rejecting (SA or SA-FasL) ...

The presence of immunoregulatory cells in primary long-term graft survivors was further confirmed in a group of nonirradiated secondary graft recipients. Adoptive transfer of splenocytes from SA-engineered cells treated ACI primary graft recipients with acute rejection into secondary unmanipulated naïve ACI recipients resulted in accelerated rejection of all WF donor hearts (Fig. 6B; n = 4, MST = 9 ± 1 days). In marked contrast, all WF grafts (n = 6) transplanted into secondary ACI recipients adoptively transferred with splenocytes from long-term SA-FasL-engineered cells treated ACI primary recipients showed prolonged survival with fifty percent of the grafts not rejecting during the 100-day observation period. Long-term (>100 days) graft survivors rejected third party F344 hearts in a normal tempo (n =4; MST = 7.5±0.5 days) without any effect on the survival of primary WF grafts.

To provide direct evidence for the contribution of Treg cells in the observed tolerance, CD4+CD25+ Treg cells were positively sorted from tolerant splenocytes using Miltenyi beads. Sorted Treg cells (5-9 × 106 cells/recipient) as well as Treg-depleted splenocytes (50-90 × 106 cells/recipient) were adoptively transferred into naïve ACI recipients of WF hearts one day before transplantation. All recipients (n= 4) adoptively transferred with Treg cells accepted their grafts (Fig. 6C). In marked contrast, all grafts in the group receiving Treg cell-depleted splenocytes underwent rejection with a MST of 23.4±3.4 days. However, this rejection time was significantly delayed as compared with normal controls (MST = 9 ± 0.4 days), suggesting either the presence of other immunoregulatory cells or contamination with Treg cells. Importantly, Treg cells sorted from acutely rejecting animals (SA-splenocytes treated or normal controls) did not prevent the rejection of donor grafts following adoptive transfer into naïve secondary recipients under similar conditions as Treg cell sorted from FasL-treated long-term graft survivors. Taken together, these data demonstrate that Treg cells are critical to the observed donor-specific peripheral tolerance achieved by systemic immunomodulation using SA-FasL-engineered splenocytes.


Immunomodulation with genetically engineered cells expressing FasL has been extensively tested for the induction of tolerance to auto and alloantigens with reported efficacy in various preclinical settings (1418,20,3640). However, the translation of this approach to the clinic remains to be realized primarily due to difficulties associated with efficient and rapid manipulation of donor primary cells under clinically applicable conditions to reproducibly express FasL and safety concerns of the gene therapy. Furthermore, FasL has pleiotropic effects on immune and nonimmune cells, and as such long-term stable expression of this molecule using gene therapy might have detrimental consequences. In the present study, we overcame these difficulties by transient display of a recombinant form of FasL protein with potent apoptotic activity (25) on donor splenocytes in a rapid (~ 2 hours ) and efficient manner (~ 100% of the targeted cells). Systemic immunomodulation with FasL-engineered donor cells resulted in peripheral tolerance to cardiac allografts without the use of any additional immunosuppression.

Although, a series of studies using various antigen presenting cells genetically engineered to express FasL for immunomodulation reported alloantigen-specific immune nonresponsiveness, none of these studies demonstrated that such nonresponsiveness eventually leads to long-term allograft tolerance (11,16,37,40). To our knowledge, this is the first study to demonstrate that systemic immunomodulation with donor cells engineered to express FasL on their surface induces long-term peripheral tolerance to cardiac allografts in a totally allogeneic rat strain combination. The mechanistic basis of the induced peripheral tolerance involves physical elimination of alloreactive cells by AICD early after transplantation and maintenance of tolerance by CD4+CD25+FoxP3+ Treg cells. The enhanced immunomodulatory effect of repeated splenocyte infusion reported in the present communication can be attributed to effective elimination of alloreactive immune effector cells at remote sites from the graft, robust induction/expansion of Treg cells, or both. T cells require antigen-specific activation and several rounds of sensitization to acquire sensitivity to Fas/FasL-mediated apoptosis (10). Also, it is well-established that memory immune cells reside within non-lymphoid target tissues, plausibly to generate a rapid and effective secondary immune response to recurrent infection (41). Therefore, systemic and repeated administration of engineered donor splenocytes may have the advantage of tolerazing compartmentalized memory responses defined as heterologous immunity that serves a major barrier for tolerance induction in the clinic (4).

In our view, induction of CD4+CD25+FoxP3+ Treg cells using SA-FasL-engineered donor splenocytes is the most significant finding of the present study. The essential role of Treg cells in tolerance was demonstrated by adoptive transfer experiments where cells sorted from lymphoid organs of long-term graft acceptors transferred tolerance to naïve, unmanipulated secondary recipients (Fig. 6C). Importantly, CD4+CD25+FoxP3+ T cells sorted from acutely rejecting animals did not prevent the rejection of donor grafts in naive secondary recipients following adoptive transfer. This may be because the sorted CD4+CD25+FoxP3+ Treg cells were not donor specific, contaminated with newly activated alloreactive Teff cells expressing CD25, or rat Teff cells transiently express FoxP3 following activation as reported for human Teff cells (42). The presence of T regulatory cells, other than CD4+CD25+FoxP3+ Treg cells, in our model is consistent with our observations that Treg cell-depleted total splenocytes from SA-FasL-treated long-term graft survivors did not prevent the rejection of donor grafts upon adoptive transfer into secondary naïve graft recipients, but caused significant prolongation as compared with controls (Fig. 6C). These regulatory T cells may include CD4CD8TCR+ T cells (43), CD8+FoxP3+ T cells (44), and/or newly described CD8+CD45RClow T cells expressing IFN-γ (45). These cell types are also implicated in allotransplantation achieved using donor-specific transfusion. For example, infusion of allogeneic donor splenocytes into graft recipients was shown to generate CD4CD8TCR+ T regulatory cells that used FasL to physically eliminate alloreactive T cells for tolerance induction (43). Furthermore, tolerance to skin allografts disparate for H-Y antigen has recently been shown to require FasL on donor cells used for infusion and Fas in graft recipients (46). Therefore, Fas/FasL system may serve as a common denominator of mechanisms responsible for tolerance achieved by immunomodulation of graft recipients using donor-specific transfusion in form of donor lymphocytes.

We envision three possible mechanisms for the induction/expansion of Treg cells by SA-FasL-engineered splenocytes. First, FasL interaction with Fas upregulated on the surface of Treg cells in response to donor antigens may transduce a mitogenic signal, leading to their expansion. Although Fas signaling was shown to be involved in the physiological process of T effector cell activation under selected conditions (47,48), it remains to be determined if such a function also applies to Treg cells. Second, Treg cells may be less sensitive to Fas-mediated apoptosis as compared with T effector cells. Immunomodulation with SA-FasL-engineered splenocytes may preferentially eliminate T effector cells, thereby tipping the balance towards the expansion of Treg cells. However, the sensitivity of Treg cells to Fas/FasL-mediated apoptosis has been the subject of few studies with conflicting observations. While some studies reported that Treg cells are more sensitive to FasL-mediated apoptosis than T effector cells (49,50), we (51) and others indicated the exact opposite (52). These conflicting observations may be due to the study designs, such as lack of comparative analysis of Treg and T effector cell sensitivity to FasL-mediated apoptosis in the course of an immune response to antigens under inflammatory conditions. Third, T effector cells undergoing apoptosis may give rise to the generation of adaptive Treg cells or expansion of naturally occurring Treg cells. Consistent with this notion are the observations that apoptotic lymphocytes release two cytokines, IL-10 and TGF-β, that play important roles in the generation and function of Treg cells (53). Furthermore, apoptotic bodies from dying cells were shown to contribute to the generation of Treg cells through mechanisms that involved TGF-β and macrophages (54). Further studies are needed to establish mechanisms that are responsible for the generation of Treg cells in our model and if these cells are of adaptive or natural type.

The immunomodulatory approach presented here has several attractive features with direct relevance to the clinic. First, it allows for rapid, efficient, and transient display of exogenous immunomodulatory proteins on the cell membrane with immediate function without a time lag required for gene transfer-based expression. Second, the transient display of immunomodulatory proteins with pleiotropic effects may minimize the potential undesired effects arising from their stable and long-term expression achieved by gene therapy. Third, several proteins with synergistic functions may be simultaneously displayed on the cell surface to maximize their therapeutic efficacy (Sharma and Shirwan, unpublished data). In particular, our approach may be incorporated into cell-based immunomodulation approaches, such as donor-specific leukocyte transfusion or hematopoietic stem cell transplantation, for tolerance induction in the clinic. Infusion of unmodified donor leukocytes or bone marrow cells into graft recipients for the purpose of immunomodulation has been extensively tested with observed beneficial effects in various preclinical and clinical settings (55,56). The presence of SA-FasL on donor leukocytes and bone marrow cells may further improve their therapeutic efficacy by eliminating alloreactive pathogenic lymphocytes and/or inducing peripheral immunoregulatory mechanisms as shown by the present study. In conclusion, this protein display approach possesses the simplicity, safety, and efficacy required to make it a clinically relevant and practical alternative to gene therapy for immunomodulation with broad application to cell-based therapies, regenerative medicine, autoimmunity, and transplantation.


The authors are grateful to Dr. Suzanne Ildstad for proofreading this manuscript and Mr. Orlando Grimany for technical assistance.

1This work was funded in parts by grants from the NIH (R21 DK61333, R01 AI47864, R21 AI057903, R21 HL080108 to H.S. and E.S.Y.), Juvenile Diabetes Research Foundation (1-2001-328 to H.S.), American Diabetes Association (1-05-JF-56 to E.S.Y. and H.S.), and the Commonwealth of Kentucky Research Challenge Trust Fund.


The ProtEx technology described in this manuscript is licensed out from UofL by ApoImmune, Inc., Louisville, KY, for which Haval Shirwan serves as CSO and Haval Shirwan and Esma S. Yolcu have significant equity interest in the Company.


1. Salama AD, Womer KL, Sayegh MH. Clinical transplantation tolerance: many rivers to cross. J Immunol. 2007;178:5419–5423. [PubMed]
2. Newell KA, Larsen CP. Transplantation tolerance. Semin Nephrol. 2007;27:487–497. [PubMed]
3. Merrill JP, Murray JE, Harrision JH, Guild WR. Successful homotransplantation of the human kidney between identical twins. J Am Med Assoc. 1956;160:277–282. [PubMed]
4. Shirwan H. Allograft tolerance: Is it a journey towards achieving the impossible? Current Opinion Org Transp. 2006;11:351–353.
5. Horn PA, Figueiredo C, Kiem HP. Gene therapy in the transplantation of allogeneic organs and stem cells. Curr Gene Ther. 2007;7:458–468. [PubMed]
6. Anderson WF. Gene therapy. The best of times, the worst of times. Science. 2000;288:627–629. [PubMed]
7. Catron DM, Itano AA, Pape KA, Mueller DL, Jenkins MK. Visualizing the first 50 hr of the primary immune response to a soluble antigen. Immunity. 2004;21:341–347. [PubMed]
8. Yolcu ES, Askenasy N, Singh NP, Cherradi SE, Shirwan H. Cell membrane modification for rapid display of proteins as a novel means of immunomodulation: FasL-decorated cells prevent islet graft rejection. Immunity. 2002;17:795–808. [PubMed]
9. Suda T, Takahashi T, Golstein P, Nagata S. Molecular cloning and expression of the Fas ligand, a novel member of the tumor necrosis factor family. Cell. 1993;75:1169–1178. [PubMed]
10. Refaeli Y, Van Parijs L, London CA, Tschopp J, Abbas AK. Biochemical mechanisms of IL-2-regulated Fas-mediated T cell apoptosis. Immunity. 1998;8:615–623. [PubMed]
11. Askenasy N, Yolcu ES, Yaniv I, Shirwan H. Induction of tolerance using Fas ligand: a double-edged immunomodulator. Blood. 2004;105:1396–1404. [PubMed]
12. Allison J, Georgiou HM, Strasser A, Vaux DL. Transgenic expression of CD95 ligand on islet beta cells induces a granulocytic infiltration but does not confer immune privilege upon islet allografts. Proc Natl Acad Sci U S A. 1997;94:3943–3947. [PubMed]
13. Kang SM, Schneider DB, Lin Z, Hanahan D, Dichek DA, Stock PG, Baekkeskov S. Fas ligand expression in islets of Langerhans does not confer immune privilege and instead targets them for rapid destruction. Nat Med. 1997;3:738–743. [PubMed]
14. Lau HT, Stoeckert CJ. FasL--too much of a good thing? Transplanted grafts of pancreatic islet cells engineered to express Fas ligand are destroyed not protected by the immune system. Nat Med. 1997;3:727–728. [PubMed]
15. Lau HT, Yu M, Fontana A, Stoeckert CJ., Jr Prevention of islet allograft rejection with engineered myoblasts expressing FasL in mice. Science. 1996;273:109–112. [PubMed]
16. Matsue H, Matsue K, Walters M, Okumura K, Yagita H, Takashima A. Induction of antigen-specific immunosuppression by CD95L cDNA-transfected 'killer' dendritic cells. Nat Med. 1999;5:930–937. [PubMed]
17. Min WP, Gorczynski R, Huang XY, Kushida M, Kim P, Obataki M, Lei J, Suri RM, Cattral MS. Dendritic cells genetically engineered to express Fas ligand induce donor-specific hyporesponsiveness and prolong allograft survival. J Immunol. 2000;164:161–167. [PubMed]
18. Swenson KM, Ke B, Wang T, Markowitz JS, Maggard MA, Spear GS, Imagawa DK, Goss JA, Busuttil RW, Seu P. Fas ligand gene transfer to renal allografts in rats: effects on allograft survival. Transplantation. 1998;65:155–160. [PubMed]
19. Takeuchi T, Ueki T, Nishimatsu H, Kajiwara T, Ishida T, Jishage K, Ueda O, Suzuki H, Li B, Moriyama N, Kitamura T. Accelerated rejection of Fas ligand-expressing heart grafts. J Immunol. 1999;162:518–522. [PubMed]
20. Zhang H, Yang Y, Horton JL, Samoilova EB, Judge TA, Turka LA, Wilson JM, Chen Y. Amelioration of collagen-induced arthritis by CD95 (Apo-1/Fas)-ligand gene transfer. J Clin Invest. 1997;100:1951–1957. [PMC free article] [PubMed]
21. Suda T, Hashimoto H, Tanaka M, Ochi T, Nagata S. Membrane Fas ligand kills human peripheral blood T lymphocytes, and soluble Fas ligand blocks the killing. J Exp Med. 1997;186:2045–2050. [PMC free article] [PubMed]
22. Tanaka M, Itai T, Adachi M, Nagata S. Downregulation of Fas ligand by shedding. Nat Med. 1998;4:31–36. [PubMed]
23. Ottonello L, Tortolina G, Amelotti M, Dallegri F. Soluble Fas ligand is chemotactic for human neutrophilic polymorphonuclear leukocytes. J Immunol. 1999;162:3601–3606. [PubMed]
24. Seino K, Iwabuchi K, Kayagaki N, Miyata R, Nagaoka I, Matsuzawa A, Fukao K, Yagita H, Okumura K. Chemotactic activity of soluble Fas ligand against phagocytes. J Immunol. 1998;161:4484–4488. [PubMed]
25. Askenasy N, Yolcu ES, Wang Z, Shirwan H. Display of Fas Ligand protein on cardiac vasculature as a novel means of regulating allograft rejection. Circulation. 2003;107:41–47. [PubMed]
26. Shirwan H, Barwari L, Fuss I, Makowka L, Cramer DV. Structure and repertoire usage of rat TCR α-chain genes in T cells infiltrating heart allografts. J Immunol. 1995;154:1964–1972. [PubMed]
27. Koksoy S, Kakoulidis TP, Shirwan H. Chronic heart allograft rejection in rats demonstrates a dynamic interplay between IFN-gamma and IL-10 producing T cells. Transpl Immunol. 2004;13:201–209. [PubMed]
28. Shirwan H, Mhoyan A, Yolcu ES, Que X, Ibrahim S. Chronic cardiac allograft rejection in a rat model disparate for one single class I MHC molecule is associated with indirect recognition by CD4(+) T cells. Transpl Immunol. 2003;11:179–185. [PubMed]
29. Hosenpud JD, Everett JP, Morris TE, Mauck KA, Shipley GD, Wagner CR. Cardiac allograft vasculopathy. Association with cell-mediated but not humoral alloimmunity to donor-specific vascular endothelium. Circulation. 1995;92:205–211. [PubMed]
30. Seino K, Azuma M, Bashuda H, Fukao K, Yagita H, Okumura K. CD86 (B70/B7-2) on endothelial cells co-stimulates allogeneic CD4+ T cells. Int Immunol. 1995;7:1331–1337. [PubMed]
31. Sha WC, Nelson CA, Newberry RD, Kranz DM, Russell JH, Loh DY. Selective expression of an antigen receptor on CD8-bearing T lymphocytes in transgenic mice. Nature. 1988;335:271–274. [PubMed]
32. Suchin EJ, Langmuir PB, Palmer E, Sayegh MH, Wells AD, Turka LA. Quantifying the frequency of alloreactive T cells in vivo: new answers to an old question. J Immunol. 2001;166:973–981. [PubMed]
33. Diamond AS, Gill RG. An essential contribution by IFN-gamma to CD8+ T cell-mediated rejection of pancreatic islet allografts. J Immunol. 2000;165:247–255. [PubMed]
34. Miura M, Morita K, Kobayashi H, Hamilton TA, Burdick MD, Strieter RM, Fairchild RL. Monokine induced by IFN-gamma is a dominant factor directing T cells into murine cardiac allografts during acute rejection. J Immunol. 2001;167:3494–3504. [PubMed]
35. Wood KJ, Sakaguchi S. Regulatory T cells in transplantation tolerance. Nat Rev Immunol. 2003;3:199–210. [PubMed]
36. Arai H, Chan SY, Bishop DK, Nabel GJ. Inhibition of the alloantibody response by CD95 ligand. Nat Med. 1997;3:843–848. [PubMed]
37. Matsue H, Matsue K, Kusuhara M, Kumamoto T, Okumura K, Yagita H, Takashima A. Immunosuppressive properties of CD95L-transduced "killer" hybrids created by fusing donor- and recipient-derived dendritic cells. Blood. 2001;98:3465–3472. [PubMed]
38. Tourneur L, Malassagne B, Batteux F, Fabre M, Mistou S, Lallemand E, Lores P, Chiocchia G. Transgenic expression of CD95 ligand on thyroid follicular cells confers immune privilege upon thyroid allografts. J Immunol. 2001;167:1338–1346. [PubMed]
39. Whartenby KA, Straley EE, Kim H, Racke F, Tanavde V, Gorski KS, Cheng L, Pardoll DM, Civin CI. Transduction of donor hematopoietic stem-progenitor cells with Fas ligand enhanced short-term engraftment in a murine model of allogeneic bone marrow transplantation. Blood. 2002;100:3147–3154. [PubMed]
40. Zhang HG, Su X, Liu D, Liu W, Yang P, Wang Z, Edwards CK, Bluethmann H, Mountz JD, Zhou T. Induction of specific T cell tolerance by Fas ligand-expressing antigen-presenting cells. J Immunol. 1999;162:1423–1430. [PubMed]
41. Masopust D, Vezys V, Marzo AL, Lefrancois L. Preferential localization of effector memory cells in nonlymphoid tissue. Science. 2001;291:2413–2417. [PubMed]
42. Tran DQ, Ramsey H, Shevach EM. Induction of FOXP3 expression in naive human CD4+FOXP3 T cells by T-cell receptor stimulation is transforming growth factor-beta dependent but does not confer a regulatory phenotype. Blood. 2007;110:2983–2990. [PubMed]
43. Zhang ZX, Yang L, Young KJ, DuTemple B, Zhang L. Identification of a previously unknown antigen-specific regulatory T cell and its mechanism of suppression. Nat Med. 2000;6:782–789. [PubMed]
44. Liu J, Liu Z, Witkowski P, Vlad G, Manavalan JS, Scotto L, Kim-Schulze S, Cortesini R, Hardy MA, Suciu-Foca N. Rat CD8+ FOXP3+ T suppressor cells mediate tolerance to allogeneic heart transplants, inducing PIR-B in APC and rendering the graft invulnerable to rejection. Transpl Immunol. 2004;13:239–247. [PubMed]
45. Guillonneau C, Hill M, Hubert FX, Chiffoleau E, Herve C, Li XL, Heslan M, Usal C, Tesson L, Menoret S, Saoudi A, Le MB, Josien R, Cuturi MC, Anegon I. CD40Ig treatment results in allograft acceptance mediated by CD8CD45RC T cells, IFN-gamma, and indoleamine 2,3-dioxygenase. J Clin Invest. 2007;117:1096–1106. [PMC free article] [PubMed]
46. Minagawa R, Okano S, Tomita Y, Kishihara K, Yamada H, Nomoto K, Shimada M, Maehara Y, Sugimachi K, Yoshikai Y, Nomoto K. The critical role of Fas-Fas ligand interaction in donor-specific transfusion-induced tolerance to H-Y antigen. Transplantation. 2004;78:799–806. [PubMed]
47. Alderson MR, Armitage RJ, Maraskovsky E, Tough TW, Roux E, Schooley K, Ramsdell F, Lynch DH. Fas transduces activation signals in normal human T lymphocytes. J Exp Med. 1993;178:2231–2235. [PMC free article] [PubMed]
48. Peter ME, Budd RC, Desbarats J, Hedrick SM, Hueber AO, Newell MK, Owen LB, Pope RM, Tschopp J, Wajant H, Wallach D, Wiltrout RH, Zornig M, Lynch DH. The CD95 receptor: apoptosis revisited. Cell. 2007;129:447–450. [PubMed]
49. Fritzsching B, Oberle N, Eberhardt N, Quick S, Haas J, Wildemann B, Krammer PH, Suri-Payer E. In contrast to effector T cells, CD4+CD25+FoxP3+ regulatory T cells are highly susceptible to CD95 ligand- but not to TCR-mediated cell death. J Immunol. 2005;175:32–36. [PubMed]
50. Mohamood AS, Trujillo CJ, Zheng D, Jie C, Murillo FM, Schneck JP, Hamad AR. Gld mutation of Fas ligand increases the frequency and up-regulates cell survival genes in CD25+CD4+ TR cells. Int Immunol. 2006;18:1265–1277. [PubMed]
51. Franke DD, Yolcu ES, Alard P, Kosiewicz MM, Shirwan H. A novel multimeric form of FasL modulates the ability of diabetogenic T cells to mediate type 1 diabetes in an adoptive transfer model. Mol Immunol. 2007;44:2884–2892. [PMC free article] [PubMed]
52. Banz A, Pontoux C, Papiernik M. Modulation of Fas-dependent apoptosis: a dynamic process controlling both the persistence and death of CD4 regulatory T cells and effector T cells. J Immunol. 2002;169:750–757. [PubMed]
53. Chen W, Frank ME, Jin W, Wahl SM. TGF-beta released by apoptotic T cells contributes to an immunosuppressive milieu. Immunity. 2001;14:715–725. [PubMed]
54. Kleinclauss F, Perruche S, Masson E, de Carvalho BM, Biichle S, Remy-Martin JP, Ferrand C, Martin M, Bittard H, Chalopin JM, Seilles E, Tiberghien P, Saas P. Intravenous apoptotic spleen cell infusion induces a TGF-beta-dependent regulatory T-cell expansion. Cell Death Differ. 2006;13:41–52. [PMC free article] [PubMed]
55. Siemionow M, Agaoglu G. Role of blood transfusion in transplantation: a review. J Reconstr Microsurg. 2005;21:555–563. [PubMed]
56. Fudaba Y, Spitzer TR, Shaffer J, Kawai T, Fehr T, Delmonico F, Preffer F, Tolkoff-Rubin N, Dey BR, Saidman SL, Kraus A, Bonnefoix T, McAfee S, Power K, Kattleman K, Colvin RB, Sachs DH, Cosimi AB, Sykes M. Myeloma responses and tolerance following combined kidney and nonmyeloablative marrow transplantation: in vivo and in vitro analyses. Am J Transplant. 2006;6:2121–2133. [PubMed]