Our previous studies in rats showed that incubation of monocytic dendritic cells (DCs) with the chemotherapeutic drug mitomycin C (MMC) renders the cells immunosuppressive. Donor-derived MMC–DCs injected into the recipient prior to transplantation prolonged heart allograft survival. Although the generation of DCs is labour-intensive and time-consuming, peripheral blood mononuclear cells (PBMCs) can be easily harvested. In the present study, we analyse under which conditions DCs can be replaced by PBMCs and examine their mode of action. When injected into rats, MMC-incubated donor PBMCs (MICs) strongly prolonged heart allograft survival. Removal of monocytes from PBMCs completely abrogated their suppressive effect, indicating that monocytes are the active cell population. Suppression of rejection was donor-specific. The injected MICs migrated into peripheral lymphoid organs and led to an increased number of regulatory T-cells (Tregs) expressing cluster of differentiation (CD) markers CD4 and CD25 and forkhead box protein 3 (FoxP3). Tolerance could be transferred to syngeneic recipients with blood or spleen cells. Depletion of Tregs from tolerogenic cells abrogated their suppressive effect, arguing for mediation of immunosuppression by CD4+CD25+FoxP3+ Tregs. Donor-derived MICs also prolonged kidney allograft survival in pigs. MICs generated from donor monocytes were applied for the first time in humans in a patient suffering from therapy-resistant rejection of a haploidentical stem cell transplant. We describe, in the present paper, a simple method for in vitro generation of suppressor blood cells for potential use in clinical organ transplantation. Although the case report does not allow us to draw any conclusion about their therapeutic effectiveness, it shows that MICs can be easily generated and applied in humans.
Due to its broad suppressive activity, standard immunosuppression after organ transplantation causes side effects such as increased incidence of infections and malignancies. Alternative donor-specific strategies are needed that target only the unwanted immune response against the allograft.
In the present study, we describe a method for induction of donor-specific immunosuppression using donor blood cells incubated with the drug MMC (MICs). Without using additional immunosuppressants, administration of MICs before transplantation led to a strong prolongation of allograft survival in a rat heart and pig kidney transplant model. Immunosuppression could be transferred to syngeneic animals by blood cells and was lost upon removal of CD4+CD25+FoxP3+ Tregs.
A first application of MICs in a patient with therapy-resistant stem cell transplant demonstrated the transferability of this approach to the clinic.
Enormous efforts have been made in organ transplantation for the development of therapies which lead away from general immunosuppression towards donor-specific inhibition of the immune response . Current strategies for induction of clinically relevant immunologic tolerance are based on inhibition or depletion of T- or B-cells, blocking of co-stimulatory signals or cell therapies . The latter include induction of mixed chimaerism by infusion of donor bone marrow into the myeloablated recipient, infusion of ex vivo expanded regulatory T-cells (Tregs), dendritic cells (DCs), macrophages or mesenchymal stroma cells . Even if some of these strategies will prove to be highly effective in humans, they have major side effects or require laborious and time-consuming procedures, factors which limit their application in clinical routine.
Previous work of our laboratory showed that monocyte-derived DCs, which are strongly immunostimulatory, can be converted into suppressive cells by treatment in vitro with mitomycin C (MMC). The latter is a chemotherapeutic drug routinely used in clinical cancer therapy . In a rat heart transplant model, pre-treatment of recipients with MMC-incubated donor DCs induced a strong prolongation of allograft survival without any immunosuppressive therapy . When loaded with myelin basic protein (MBP), an autoantigen of the central nervous system, human MMC-treated DCs specifically suppressed the in vitro response of T-cells derived from patients with multiple sclerosis (MS) . Such cells generated from the patient's own blood monocytes might be used as ‘suppressive bullets’ for controlling the derailed immune response. In an experimental autoimmune encephalomyelitis (EAE) model, the correlate of MS in humans, MBP-loaded MMC-treated DCs were injected into healthy mice, whereupon the animals became resistant to EAE, showing that tolerogenic vaccination protecting an individual from autoimmune disease is possible . In vitro generation of DCs is expensive, time-consuming and yields a limited number of cells. Moreover, the method is difficult to standardize since minimal changes of cell culture conditions may result in different, sometimes even functionally opposing, DCs . Even if rendered suppressive in vitro, when injected into patients DCs may regain stimulatory activity with unpredictable consequences. An elegant alternative would be the use of stable suppressor cells, easily generated from fresh blood cells in sufficient amount and quality for clinical treatment.
In the present paper, we describe the generation of clinical grade suppressor cells [MICs (MMC-incubated blood cells)] by incubation of peripheral blood cells with MMC, demonstrate their suppressive function in an allogeneic rat heart and pig kidney transplant model and analyse their mode of action. Moreover, the safe use of MICs for the treatment of a patient with a haploidentical bone marrow transplant is presented.
MATERIALS AND METHODS
Adult (220–350 g) male Brown-Norway (BN, RT.1n), Dark-Agouti (DA, RT.1av1) and Piebald Virol Glaxo (PVG, RT.1c) rat strains were purchased from Harlan Winkelmann, and Lewis (LEW, RT.1l) rats were purchased from Janvier Labs. The animals were kept at the Interdisciplinary Research Facility of Biomedicine (IBF), University of Heidelberg, Germany. During surgical and scintigraphic imaging procedures the rats were kept under inhalatory anaesthesia using either 2% (v/v) isoflurane (Baxter) or 1% (v/v) sevoflurane (Baxter). Donor and recipient animals were placed on a heating pad or kept under an infrared heating lamp in order to maintain a constant body temperature and to prevent hypothermia. Immediately after organ transplantation, animals received the analgesic carprofen (Rimadyl®, 4.0 mg/kg, Pfizer Pharma).
Male and female common-breed pigs (24–28 kg) were used for the allogeneic kidney transplantation model. The pigs were kept at the animal facility of the Pius Branzeu Center for Laparascopic Surgery and Microsurgery of the Victor Babes University of Medicine and Pharmacy, Timisoara, Romania. All operations were performed under general anaesthesia as previously described .
Allogeneic heart transplantation in rats
Vascularized heterotopic abdominal cardiac transplantation was performed as previously described [8,9]. Donor hearts were implanted into the recipient's abdominal cavity with end-to-side anastomosis of donor aorta to recipient abdominal aorta and donor pulmonary artery to recipient vena cava respectively. Cold ischaemia time was ≤ 30 min. The graft function was monitored daily by palpation of heart contractions through the abdominal wall. Cessation of heart beating was defined as rejection.
Allogeneic kidney transplantation in pigs
Male common-breed pigs were used as donors and females as recipients. This is a model with a high donor–recipient porcine leucocyte antigen (PLA)-incompatibility. Donor operation and surgical preparation of the recipient were performed in parallel, where one donor served two recipients. Kidneys were harvested, perfused with cold (4°C) organ preservation solution (Custodiol HTK solution, Essential Pharmaceuticals) and implanted heterotopically into the abdominal cavity by end-to-side anastomosis of the renal artery to the infrarenal aorta as well as of the renal vein to the vena cava. Rejection was monitored by determination of the peak systolic velocity (PSV) of the renal artery and histopathological evaluation of renal biopsies. Echo-Doppler measurement of PSV of the allograft was done using a Mindray M7 portable colour Doppler device (Mindray Medical International). A PSV of ≥ 20 m/s denoted normal kidney function, PSV of ≤ 10 m/s denoted clinically evolving rejection and PSV of < 5 m/s denoted complete rejection. PSV of the renal artery of the left native kidney served as positive control.
In addition, post-transplant needle biopsies were performed to confirm rejection episodes. Samples were evaluated according to the Banff 97 criteria . After diagnosis of rejection, both the allograft and the native kidneys were excised for gross anatomy inspection.
Isolation of cells
Peripheral blood mononuclear cells (PBMCs) were isolated from heparinized (rat and pig) peripheral blood by density gradient centrifugation on separation medium (Lymphodex, Inno-Train Diagnostik). Cells were washed twice with RPMI-1640 and resuspended in culture medium [RPMI-1640 with 10% FCS (foetal calf serum), 2 mM L-glutamine, 100 units/ml penicillin, 100 μg/ml streptomycin, 1% non-essential amino acids, 1% sodium pyruvate and 50 μM β-mercaptoethanol].
Rat splenocytes (SPCs) were isolated in the same way, following preparation of single cell suspensions by gently forcing harvested spleens through a 70-μm cell strainer mesh and hypotonic red blood cell lysis.
Depletion of monocytes from PBMCs
PBMCs were depleted of monocytes by plastic adherence. PBMCs (2×106/ml) were cultured in 2 ml of culture medium in six-well plates (Greiner Bio-One) for 90 min at 37°C and 5% CO2. Non-adherent cells were retrieved and further processed.
Depletion of CD4+CD25+FoxP3+ Tregs
PBMCs and SPCs from tolerant animals were isolated, as described, pooled and incubated with anti-CD25-phycoerythrin (PE) antibody (Ab, clone OX39, eBioscience) and anti-PE (phycoerythrin) microbeads (Miltenyi Biotec) according to the manufacturers’ instructions. Cells were then loaded on to an LD column (Miltenyi Biotec) and run through via gravity flow in order to capture the labelled CD25+ cells. The unlabelled cells passing through the column were collected, washed twice with PBS and injected into the recipients. Depletion of CD4+CD25+FoxP3+ Tregs was verified by flow cytometric analysis.
Generation of mitomycin-induced cells
PBMCs were treated with MMC (Medac) (1.5×106 PBMCs/40 μg of MMC/ml). After 30 min of incubation at 37°C (5% CO2), with shaking on an oblique roller (30 rpm), the cells were washed twice with RPMI-1640. Supernatants were analysed for traces of MMC by spectrophotometry at 363 nm in a NanoDrop instrument (Thermo Scientific) in order to avoid the transfer of soluble MMC.
PBMCs were suspended in RPMI-1640 (supplemented with 1% non-essential amino acids, 1% L-glutamine and 1% penicillin/streptomycin), incubated with 75 μg/ml MMC in six-well plates (4 ml/well) for 30 min at 37°C (5% CO2), washed once and injected into recipients in 5 ml of PBS.
One millilitre of heparinized blood from donor rats was incubated with various substances in a 2-ml cryotube at 37°C and 5% CO2 on an oblique roller. Except for MMC and brefeldin A (BFA), the concentration per millilitre was deduced from the therapeutic dose per kilogram of body weight or surface (dose per millilitre and incubation time shown in brackets): MMC (100 μg, 30 min, medac), mycophenolate mofetil (MMF, 10 mg/kg, 120 min; CellCept®, Roche), tacrolimus (TACRO, 0.02 mg/kg, 120 min; Prograf®, AstellasPharma), irinotecan (IRT, 350 mg/m2, 120 min; Campto®, Pfizer Pharma), mitoxantrone (MTX, 12 mg/m2, 120 min; Ralenova®, Wyeth Pharma), BFA (14.02 μg, 120 min; Sigma–Aldrich) and cyclosporine A (CsA; 5 mg/kg, 120 min; Sandimmun®, Novartis Pharma). Treated blood was injected intravenously (i.v.) into the recipients 7 days to transplantation.
Induction of apoptosis by UV-C irradiation
PBMCs (3×106/well in 2 ml) were irradiated in six-well culture plates with various doses of UV-C light (254 nm) using a CL-1000 UV cross-linker (UVP). Cells were exposed to 12, 25 and 50 mJ/cm2 respectively.
Rat PBMCs and SPCs were examined for the presence of CD4+CD25+FoxP3+ Tregs. To this end, 106 cells were blocked with mouse-anti-rat CD32 Ab in FACS buffer, stained for CD4 and CD25 with specific fluorochrome-conjugated mAbs (monoclonal antibodies), fixed, permeabilized and stained for intracellular FoxP3 protein. Cell viability was determined with annexin V and 7-aminoactinomycin D (7-AAD).
FACS antibodies, isotype controls, annexin V and 7-AAD were used at concentrations indicated by the manufacturers (BD Biosciences and eBioscience). Data acquisition was performed on a FACSCalibur flow cytometer and analysed using CellQuest software (BD Biosciences).
Cellular treatment of animals
Rat transplant model
A number of 106, 107 or 108 MMC-treated (MICs) or 106, 107 or 108 untreated PBMCs from donor rats (DA) were injected in 250 μl of PBS into PVG recipients via the femoral vein. Seven days later, recipients received a heart allograft from the same donor strain. In additional experiments, recipients received 8.3×107 monocyte-depleted MICs, naïve PBMCs or 108 UV-C-irradiated PBMCs.
In third-party experiments, PVG recipients received 108 DA MICs. One week later, they were transplanted with a BN heart (third party). As controls served untreated PVG rats transplanted with a syngeneic (PVG), allogeneic (DA) or third-party (BN) heart.
For cell transfer experiments, naïve PVG rats were sublethally irradiated (4 Gy) using an ONCOR Linear Accelerator (Siemens Medical Solutions). One day later, they received (9.82±3.67)×106 PBMCs or (4.63±1.35)×107 SPCs, isolated from tolerant (MIC-treated) recipients or (2.46±0.96)×107 PBMCs or (13.88±8.57)×107 SPCs from rejecting animals (non-treated) or (5.53±1.62)×107 Treg-depleted cells from MIC-treated tolerant recipients. On the following day, a DA heart was transplanted.
Pig transplant model
Recipients were pre-conditioned with daily oral gavage of 10 μg/kg CsA starting 14 days before transplantation (day 14). On day 7, recipients received a single dose of either 108 MMC-treated (MICs) or untreated donor-derived PBMCs. CsA serum levels were monitored to reach a concentration of 150–300 μg/ml. On day 3, CsA was withdrawn reaching non-immunosuppressive levels on the day of transplantation (day 0). One donor (two kidneys and cells) served for cellular pre-conditioning of and kidney donation to two recipients.
In vivo tracking of 111indium-labelled MICs and Tregs in allograft recipients
MICs from donor DA rats or Tregs from tolerant PVG rats were obtained as described above. MICs (108) or 3×106 Tregs were washed twice with 50 ml of PBS, resuspended in 2 ml of PBS and labelled with 20 MBq 111indium (In)-oxine (Mallinckrodt Pharmaceuticals) for 15 min at room temperature. After removal of free 111In-oxine by washing once with 50 ml of PBS, the cells were resuspended in 250 μl of PBS and injected into the femoral vein of recipient PVG rats. Subsequently, migration of the cells was followed by planar scintigraphy with a γ-imager-sct (Biospace Lab), performing scans of the thorax and the upper abdomen 10, 20 and 30 min, 2, 4, 24 and 48 h and 5 and 7 days after injection of the labelled cells. Scans were recorded using the γ Vision+ software (Biospace Lab). The activity concentration of the radioactive tracer, correlating with the number of cells gathered at a certain location was visualized via a heat map colour-coding scale [11,12]. For evaluation of bio-distribution of the injected Tregs, the recipients were killed 24 h after injection. Specimens from the indicated organs were collected, weighed and activity was measured in a γ-counter. For each sample, the activity of the In-tracer in 1 g of tissue was calculated in relation to the originally total injected dose (% ID/g).
Immunohistological evaluation of rat allograft tissue sections
Allografts were fixed with 4% phosphate-buffered formalin and embedded in paraffin; 3-μm sections were stained by immunoperoxidase techniques using avidin–biotinylated enzyme complex. Sections were deparaffinized and rehydrated. Endogenous biotin was blocked by 3% H2O2 [0.3% in the case of complement-4d (C4d) staining]. FoxP3 slides were retrieved in citrate buffer (pH6) in the microwave. C4d slides were processed in 1× trilogy buffer (Cell Marque). The tissue was blocked with 10% goat serum and incubated with primary antibodies against FoxP3 (biotinylated mouse IgG2a/κ, eBioscience) or C4d (rabbit polyclonal IgG, courtesy of Dr William Baldwin and Dr Kazunori Murata, Johns Hopkins Medical Institution, Baltimore, MD, U.S.A.) in a wet chamber. C4d-stained slides were incubated with biotinylated anti-rabbit IgG. After addition of streptavidin-conjugated peroxidase, samples were developed with substrate-chromogen 3,3′-diaminobenzidine (Dako). Nuclei were counterstained with haematoxylin. Microscopic evaluation of the heart specimens was performed independently by two observers (magnification 400×). The number of positive cells per view field was counted. Each heart was analysed at three levels: base, middle line and apex. Thus, 3×11 consecutive view fields per heart were counted. The mean value of positive cells per high power view field and heart was calculated.
Treatment of relapsing rejection episodes of haploidentical allogeneic stem cell transplants using mitomycin-treated monocytes
A 6-year-old girl with relapse of acute lymphoblastic leukaemia (ALL) received conditioning with total body irradiation (TBI), fludarabine, etoposide and anti-CD3 mAb muromonab-CD3 (clone OKT3). The patient was transplanted with CD3-/CD19-depleted peripheral blood stem cells (PBSCs) from her haploidentical father. Post-transplant immunosuppression consisted of MMF. The transplant was rejected within 40 days. The patient received a second transplant with CD3-/CD19-depleted PBSC from her haploidentical mother after conditioning with fludarabine, thiotepa, anti-thymocyte globulin (ATG, rabbit, Fresenius) and OKT3 . Post-transplant immunosuppression was performed with MMF. The transplant was also rejected within 40 days. In both cases, the rejection was resistant to therapy with stem cell boosts and donor lymphocyte infusions (DLIs). After conditioning with total lymphoid irradiation, thiotepa, fludarabine, thymoglobulin and OKT3 a third transplantation was performed again with CD3-/CD19-depleted PBSCs of the father. The patient received stem cells boosts and DLI post-transplant. The patient showed engraftment for leucocytes on day +11 and for platelets on day +15. Additionally, alemtuzumab was applied on day +15 and day +16, (20 mg/m2/day). More than 2 months after the third transplantation, the patient developed thrombocytopenia and FACS analysis revealed an increasing percentage of autologous natural killer (NK) cells and CD3+ T-cells, which had been early signs of graft rejection in the previous transplantations. Therefore, a third infusion with alemtuzumab was applied on day +70. Already, before the third transplantation viraemia with cytomegalovirus (CMV) and adenovirus (ADV) were present and required antiviral therapy, particularly against CMV.
On day +75, the patient received a first transfusion of MICs. CD3-/CD19-depleted PBSCs (109 cells) were incubated with MMC (0.1 mg/ml, 30 min, 37°C) in PBS (B. Braun) supplemented with 5% human albumin (HA) Kabi (Octapharma) washed by centrifugation, filtered through a 40-μm cell strainer mesh, resuspended in 50 ml of PBS/5% HA, aseptically transferred with a 50-ml syringe to a blood transfusion bag and infused with a transfusion set through a 200-μm filter. On day +82, a second transfusion of MICs (2×109 cells) followed. Subsequently, haematopoiesis remained stable on a low level and autologous NK cells, B-cells and T-cells nearly disappeared (see Figure 6). No immunosuppressive treatment was given. Starting from day +91, five consecutive stem cell boosts with CD3-/CD19-depleted PBSCs were administered at monthly intervals followed by one donor mesenchymal cell transfusion (day +257) and CMV-specific T-cell transfusion (day +329). Cotrimoxazol prophylaxis, as well as continuous systemic anti-fungal (itraconazole), antiviral (aciclovir, ribavirin) therapy and intravenous IgG substitution were administered. Thirteen months after the third transplant acute, CMV-pneumonia developed. Despite antiviral treatment (ganciclovir/foscarnet) the patient's condition was worsening. Intensive care interventions were unsuccessful and the patient passed away on day +418.
Statistical analysis was done using Prism 5.0 (GraphPad Software) and SPSS predictive analytics software (IBM). If not otherwise mentioned, results are shown as means ± S.E.M. Differences between the groups were determined by paired Student's t test, Wilcoxon test and log-rank test. P<0.05 was considered to be significant (*), P<0.01 to be highly significant (**) and P<0.001 to be very highly significant (***).
Approval for animal and human studies
Animal experiments were approved by the Animal Welfare Board of the Governmental Office (Karlsruhe, Germany), the University of Heidelberg (Heidelberg, Germany) and the Victor Babes University (Timisoara, Romania) Committees for Ethics on Laboratory Animal Experimentation and were performed in compliance with F.E.L.A.S.A. (Federation of European Laboratory Animal Science Associations, Ispwich, U.K.) regulations. The treatment of the patient with MICs was an individual unique emergency use in accordance with German ethical regulations based on the principles of the Declaration of Helsinki and approved by the Ethics Committee of the University of Heidelberg. Informed consent was obtained by the legal guardian of the patient.
Mitomycin-treated donor peripheral blood mononuclear suppress allograft rejection in a rat heart transplant model
We incubated PBMCs of the prospective donor with MMC and injected them into recipients. One week later, recipients received a heart allograft. Figure 1(A) shows that pre-treatment with MICs strongly prolong transplant survival. Notably, the degree of graft protection depended on the number of injected cells. Animals receiving 108 MICs displayed a mean graft survival of 64.8±17 days, in recipients treated with 107 cells graft survival was of 21.3±5 and in those treated with 106 cells 7.5±0.9 days (untreated controls: 8.6±0.3 days). Most importantly, approximately 50% of animals pre-treated with 108 cells were long-term acceptors of allograft (more than 70 days).
MMC-incubated PBMCs (MICs) or blood of the donor prolong allograft survival
As a chemotherapeutic agent, MMC can induce apoptosis. Our findings showed that 24 h after treatment with MMC approximately 60% of PBMCs are dead. Apoptotic cells may induce immunosuppression [14,15]. Therefore, we wanted to find out to what extent apoptosis of donor cells plays a role in induction of immunosuppression in our in vivo system. To this end, donor PBMCs were rendered apoptotic by UV-C irradiation (Supplementary Figure S1) and a similar number of apoptotic cells was injected into prospective recipients. In contrast with MMC-cells, they did not prolong allograft survival; on the contrary, they induced accelerated rejection (Figure 1A, orange curve).
Removal of monocytes from MMC-treated donor PBMCs abrogates their suppressive effect
In previous experiments, monocytic DCs were rendered suppressive by treatment with MMC . If the suppressive cell fractions of MMC-PBMCs are the monocytes, their removal should diminish the tolerogenic effect. In one experiment, 108 donor PBMCs were depleted of monocytes. The remaining cells were treated with MMC and injected into the recipients. No prolongation of graft survival was noted (MICs without monocytes: 9.4±0.7 days; untreated controls: 8.6±0.3 days; Figure 1A).
Mitomycin-treated donor blood suppresses allograft rejection
Protection from heart transplant rejection was also induced when donor blood (instead of PBMCs) treated with MMC was administered (untreated controls: green curve; untreated blood: blue curve; MMC-blood: red curve; Figure 1B). In line with previous observations , donor blood transfusion already prolonged allograft survival (controls: 8.6±0.3; blood: 21±4.2 days); treatment with MMC significantly enhanced this effect (MMC-blood: 34.4±4 days). In parallel experiments, we analysed whether the same effect is obtained when MMC is replaced with other chemotherapeutics (CsA, MMF, TACRO, IRT, MTX, BFA). This was not the case (untreated blood: 21.0±4.2, CsA: 20.9±3.1, MMF: 27.8±1.7, IRT: 20.0±4.7, MTX: 20.9±3.1, BFA: 21.0±6.0 days). MMC induced the strongest suppression (Figure 1B).
Another question was whether the effect of immunosuppressive drugs administered after transplantation is potentiated by concomitant MIC therapy. In one group (n=10) recipients were treated with CsA: 10 mg/kg for 20 days after transplantation. Another group (n=9) received the same treatment plus MIC (108 cells) on day +7, +9 and +11. Whereas untreated controls (n=6) had an allograft survival of 7.3±0.2 days, in the CsA group the transplants survived 53±6 days and in the CsA+MIC group 75±9.6 days. Although the difference between the last two groups was statistically not significant, it shows that MIC given together with CsA does not reduce but probably improves the immunosuppressive effect.
MMC-treated donor PBMCs preferentially migrate into peripheral lymphoid organs of recipients
Donor MICs were labelled with In-oxine and injected into the recipient. Planar scintigraphic γ scans were performed 10, 20 and 30 min, 2, 4, 24 and 48 h as well as 5 and 7 days after administration of cells (Figure 2A). After transient retention of cells in the lung, the majority of MICs accumulated in the spleen within 2 h post injection, where they persisted.
Migration of injected donor-derived MICs and of Tregs in the allograft recipient
MMC-treated donor PBMCs induce Tregs in peripheral lymphoid organs
In contrast with rejecting animals, heart vessels of tolerant animals displayed reduced C4d deposition (Figures 3A–3C). Most importantly, animals treated with MICs showed increased numbers of CD4+CD25+FoxP3+ Tregs in their peripheral blood and spleen (blood: 6.1±1% tolerant compared with 5.5±0.3% rejecting; spleen: 8.3±1.1% tolerant compared with 6.7±0.1% rejecting recipients; P = 0.016; percentage of CD4+ cells), as well as an increased FoxP3+ cell infiltration in the graft (Figures 3D–3F). When comparing the T-cell infiltration of tolerated with that of rejected grafts, it becomes evident that the proportion of FoxP3 cells among CD3 cells is significantly higher in tolerated (55.4±5.1%) than in rejected hearts (25.9±3.4%; mean ± S.E.M., day 7 post-transplant, P = 0.012). This suggests a shift of balance from effector T-cells to Tregs in MIC-treated animals. In one experiment, CD4+CD25+ cells were isolated from the spleen and blood of MIC-treated animals, labelled with In and re-injected into allograft recipients. The cells could be detected in the graft (Figure 2B), indicating migration from the lymphoid organs into the transplanted organ.
Decreased C4d deposition in blood vessels and increased FoxP3+ cell infiltration in heart allografts of tolerant rats
Immunosuppression can be transferred by peripheral blood cells from tolerant to naïve syngeneic recipients
If CD4+CD25+FoxP3+ cells mediate the immunosuppression, it should be possible to transfer tolerance to naïve animals with peripheral blood cells of tolerant animals. We collected PBMCs or spleen cells from rats accepting their graft for at least 70 days and transferred them into syngeneic animals. Recipients received a heart allograft from the same donor strain. Transplants were accepted for a long period of time (Figure 4A), with approximately 50% of recipients developing operational tolerance. In contrast, control recipients receiving cells from non-tolerant animals rapidly rejected their allografts.
MIC-induced tolerance is transferable by peripheral blood cells and is donor-specific
Elimination of CD4+C25+FoxP3+ cells from peripheral blood cells of tolerant animals abrogates their tolerogenic effect
In the next experiment, we depleted the CD4+CD25+FoxP3+ cells before their transfer. This time, they did not transmit tolerance to syngeneic recipients (graft survival with CD4+CD25+FoxP3+: 64.8±17, without: 9.25±0.48; Figure 4A, brown curve).
Suppression of heart allograft rejection is donor-specific
To address the question of donor-specificity, PVG rats pre-conditioned with MMC-treated cells from DA donors were transplanted with a heart from BN rats (Figure 4B). Whereas donor-specific transplants showed prolonged survival, third-party grafts did not, reflecting donor-specific action of MMC-cells.
MMC-treated donor PBMCs (MICs) prolong allograft survival in pigs
In the next step, we studied the effect of MMC-treated donor PBMCs in an allogeneic kidney transplant model in pigs. Recipients received 108 MMC-treated donor PBMCs prior to transplantation. Taking into consideration that tolerance induction was cell-number-dependent and that rats received the same number of cells, 108 PBMCs was an extremely low dose. In order to enhance the effect of suppressive cells, recipients were conditioned from day -14 to -3 with CsA. The animals had non-suppressive serum levels on the day of transplantation and did not receive any CsA after surgery. MIC-pre-treated recipients showed significantly prolonged allograft survival (Figure 5A). In contrast, PBMC-pre-treated recipients rejected their kidney in an accelerated fashion (Figure 5A) with an early decrease in renal artery blood flow (Figure 5B) and a corresponding ischaemic macroscopic image of the allograft (Figure 5C).
Donor-derived MMC-incubated PBMCs (MICs) prolong kidney allograft survival in common-breed pigs
Use of MMC-treated donor monocytes (MICs) for control of rejection in a patient with allogeneic stem cell transplant
A 6-year-old girl with relapse of acute lymphatic leukaemia, who had rejected consecutively the grafts of her haploidentical father and her haploidentical mother, was re-transplanted a third time with stem cells from her father . More than 2 months after the third transplantation signs of rejection occurred, despite intensive immunosuppressive treatment with ATG, OKT3 and alemtuzumab. In this life-threatening situation we administered two transfusions with MICs from her father (109 and 2×109 cells; day +75 and +82; Supplementary Table S1). The patient did not receive specific pre-medication in order to prevent possible side effects. After the treatment, no acute symptoms attributable to MIC treatment occurred. A continuously decreasing percentage of autologous NK cells, T-cells and B-cells was noted, resulting in complete haematopoietic donor chimaerism (Figures 6A–6C). On the background of chronic viral infections (CMV, ADV), cytotoxic antiviral therapy and putative bone marrow stromal cell damaged by previous conditioning therapies, persistent tricytopenia (Figures 6D and 6E) occurred, necessitating repeated donor stem cell boosts and one mesenchymal stem cell transfusion (day +257). Only after therapy with CMV-specific T-cells (day +329), was an increase in mature T-cell numbers observed (Figures 6D and 6E). After MIC-therapy, the girl did not receive immunosuppressants and exhibited stable complete donor chimaerism for about 1 year. On the background of her chronic CMV infection, the patient developed CMV-pneumonia and died on day 418 after transplantation.
Development of cellular chimaerism and distribution of cell populations in a recipient of haploidentical stem cells after injection of donor-derived MICs
Our findings show that not only MMC-treated donor DCs  but also PBMCs and whole blood have an inhibitory action on allograft rejection if injected into prospective transplant recipients. If MICs are administered after transplantation to recipients receiving CsA, they do not counter-act but possibly enhance the immunosuppressive effect. Most importantly, the immunosuppression is donor-specific. Recipients pre-treated with donor MMC-PBMCs accepted heart transplants of the same donor, whereas rejecting hearts from third-party donors. This effect might be particularly interesting in a clinical setting, where reduction in broadly reacting chemical immunosuppression in favour of donor-specific inhibition of the immune response is desirable.
PBMCs are a mix of several cell subpopulations. It was of interest to find out which of them is responsible for the suppressive effect. Previous cell culture studies with monocytes, as well as observations in the rat heart transplant model with negatively selected monocytes (unpublished results from Flavius Sandra-Petrescu and Laura Dittmar) and monocytic DCs  directed our attention towards monocytes. The present data show indeed that deletion of monocytes from MMC-treated PBMCs fully annihilates their suppressive effect.
A feature of MICs was their propensity to apoptosis. Predisposition to apoptosis is of advantage for therapeutic application. Experience shows that living regulatory cells generated in vitro with cytokines and other biomolecules can re-convert into stimulatory cells upon injection into patients. Apoptosis, however, is an irreversible process which precludes return to a stimulatory status. Thus, MICs promise to be a stable therapeutic tool.
Apoptotic cells were found to inhibit the immune response under certain circumstances . Donor-derived leucocytes rendered apoptotic prolonged heart allograft survival in mice when injected into recipients before transplantation (reviewed by ). Bittencourt et al.  showed that injection of apoptotic donor SPCs into mice enhances bone marrow engraftment independently of the origin of cells. But apoptotic cells can also stimulate the immune response [18,19]. Many factors seem to decide whether apoptotic cells exert immunostimulatory or suppressive effects. Early stage apoptotic cells, for example, are more likely to induce immunosuppression than late stage apoptotic cells . The number of injected cells , delivery of danger signals  and interaction with other cells  also influence their immunological action. A detailed analysis  revealed that if apoptotic cells express activating molecules, they acquire immunostimulatory properties and vice versa. The same study showed that UV-C-irradiated cells up-regulate immunoactivating molecules. In our setting, in contrast with MMC-induced apoptosis, UV-C apoptotic donor cells accelerated allograft rejection in vivo. It is well known that donor blood prolongs allograft survival [16,24–26]. When we treated blood cells with MMC, their suppressive effect was significantly increased. However, when treating donor blood cells with IRT, MTX or BFA, chemotherapeutics which also induce apoptosis, they did not prolong allograft survival more than untreated blood cells. Whereas all these findings do not exclude with certainty a role of apoptosis in our system, they indicate that apoptosis cannot be the only explanation for MIC-induced immuno-suppression.
Tolerant animals had an increased number of CD4+CD25+FoxP3+ Tregs in peripheral lymphoid organs and showed infiltration with FoxP3 cells in the allograft. Tolerance could be transferred with spleen or blood cells from tolerant to syngeneic recipients. However, when CD4+CD25+FoxP3+ cells were depleted from the tolerogenic cells, they lost their suppressive action. All these findings show that Tregs are a possible mediator of immunosuppression induced by MIC. It has been shown that some immunosuppressive drugs impede the generation of Tregs . Further studies should clarify whether the suppressive effect of MIC administered before transplantation might be diminished by particular immuno-suppressants given after transplantation. When injected into recipients, donor-derived MICs migrate into peripheral lymphoid organs. The migrational behaviour and the accumulation pattern resembled the properties described for human leucocytes after extracorporeal photophoretic treatment and re-infusion into human patients . We speculate that in the present study they are recognized by donor-reactive effector T-cells, which are converted into Tregs. The suppressive cells then migrate into the transplanted heart and protect it from rejection. This is supported by our findings showing that CD4+CD25+ cells isolated from the spleen and blood of MIC-treated animals can be traced in the allograft. A second hypothetical mechanism is the uptake of MICs by recipient DCs which subsequently induce donor-specific Tregs.
Although kidney allografts in pigs pre-treated with donor MICs showed significantly prolonged survival, the recipients did not become long-term transplant acceptors. This might be due to the relatively low number of MICs used for conditioning of recipients. It is conceivable that at higher cell numbers allograft survival will be considerably longer.
The first clinical application described in the present paper in detail, in the form of an individual unique emergency treatment, shows that our cell therapy can be easily applied in a clinical setting. No acute adverse reactions (including nephrotoxic, hepatotoxic- and inflammatory laboratory parameters) related to the treatment were observed. It is interesting to note that upon injection of donor-derived MICs, autologous T- and NK cells (as an indicator of rejection) virtually disappeared and long-lasting haematopoietic donor chimaerism evolved. Conceivably, this might also have been caused by alemtuzumab administered prior to our suppressive cells. However, the same mAb had previously been administered without controlling the rejection.
We were concerned that the development of CMV-pneumonia might have been the consequence of a general state of immunosuppression induced by MICs. It should be noted that the patient showed several re-activations of CMV infection already before the MIC-based therapy, suggesting that the active infection was not due to MICs. Also, there was a time gap of almost 1 year between the cell therapy and development of pneumonia which makes a causal relationship improbable. We are aware that this case does not prove any effect of MICs in humans. It merely shows that the cells can be applied in a clinical setting and encourages the initiation of controlled clinical trials.
A series of clinical trials using immunosuppressive cells for the control of allograft rejection have been started during the last couple of years (reviewed in ). Induction of haematopoietic chimaerism by administration of donor bone marrow cells to pre-conditioned recipients seems to be one of the most effective strategies of tolerance induction in organ transplantation [29–32]. Whereas in this approach, preparation of donor cells is relatively easy, the intervention remains cumbersome and entails major risks such as the graft-versus-host disease (GVHD).
One focus of present cell therapies is on the use of Tregs whose key role in transplantation tolerance has been demonstrated in many animal experiments. The basic clinical concept is to isolate and expand ex vivo either donor-specific or -unspecific Tregs and inject them into the transplant recipient. Previous results from bone marrow transplant studies show promising safety and possibly therapeutic efficacy [33–35]. However, this therapy needs to overcome some major obstacles (reviewed in ). One problem is the instability and plasticity of in vitro generated Tregs. The function of CD4+CD25+FoxP3+ Tregs, for example, depends on high expression of FoxP3 which in turn is dependent on methylation of a specific non-coding DNA sequence , as well as on post-translational acetylation [38,39]. Both processes are influenced by a series of external factors. For clinical application, it is essential to have a robust Treg manufacturing protocol . Moreover, the generation of Tregs in sufficient number is laborious and time-consuming, making their clinical application cumbersome.
Further cell-based tolerogenic therapies are under investigation . Among others, tolerogenic DCs have been envisaged as a tool for tolerance induction [40,41]. The major obstacles of this approach are the demanding procedure of generation and the risk of conversion into stimulatory cells which may sensitize the transplant recipient.
Mesenchymal stromal cells showed promising results in a phase II clinical trial in steroid-resistant patients with severe acute GVHD . However, a phase III trial failed to show benefit in refractory GVHD [43,44]. In another study regulatory macrophages, called transplant acceptance-inducing cells, generated from donor splenic monocytes were used for specific immunosuppression in kidney transplant recipients. After confirming the safety of cells  in a phase I trial, a second trial was conducted . Although the results did not provide conclusive evidence of a beneficial effect, they suggested that a state of specific unresponsiveness was achieved, which might allow minimization of pharmacological immunosuppression. Apart from the question of their therapeutic efficacy, current cell-based therapies are limited due to high costs, the required technical equipment and issues of standardization .
A general requirement of effective cell therapeutics is a simple and reproducible method of generation with a high yield of clinical grade cells at low cost. These requirements are met by the method described by us. The technique can be easily carried out and relies on the use of PBMCs, a cell population which can be harvested in virtually unlimited amount by cytapheresis. A more demanding alternative would be the use of purified monocytes. Our strategy worked in rats, mice (unpublished data from Elisabeth Mohr and Christian Kleist), pigs and can be transposed to humans. It is in line with similar experiments in other animal models [48–52].
In conclusion, the present study devises a simple method for in vivo generation of suppressor cells for control of allograft rejection in organ transplantation.
cluster of differentiation
donor lymphocyte infusion
experimental autoimmune encephalomyelitis
forkhead box protein 3
total injected dose
myelin basic protein
MMC-incubated blood cell
mean survival time
peripheral blood mononuclear cell
peripheral blood stem cell
peak systolic velocity
Piebald Virol Glaxo
Christian Kleist, Flavius Sandra-Petrescu, Laura Dittmar and Elisabeth Mohr participated in the designing of the study, performing the experiments, data analysis and the completion of the manuscript. Lucian Jiga performed the kidney transplantation and monitoring of pigs. Johann Greil was the supervisor of the clinical application of MICs in the patient with stem cell transplantation. Peter Terness conceived the study, participated in the conception and design of the study, data analysis and the completion of the manuscript. Peter Lang, Luis E. Becker and Walter Mier participated in the performance of experiments, data analysis and the completion of the manuscript and contributed reagents, materials and analysis tools. Gerhard Opelz participated in data analysis and the completion of the manuscript.
We thank Helmut Simon, Jürgen-Heinz Schnotz, Christiane Christ, Monika Hexel and Stephanie Mechler (Department of Transplantation Immunology, Institute for Immunology, Heidelberg, Germany) for their excellent technical assistance. We are grateful to Marie-Luise Gross and Heike Conrad (Institute for Pathology, Heidelberg, Germany) for expert assistance with histology. Many thanks to Janina Jiga and Bogdan Hoinoiu (Department of Vascular Surgery, Victor Babes University of Medicine and Pharmacy, Timisoara, Romania) as well as George Dindelegan (Iuliu Hatieganu University of Medicine and Pharmacy, Cluj-Napoca, Romania) for their expert assistance with the kidney transplantations in pigs. We thank Stefan Rieken and Gerald Major (Department of Radiology, Division of Radiation Oncolocy, University of Heidelberg, Heidelberg, Germany) for performing the whole body irradiation of the rats. We also thank Karin Leotta (Department of Radiology, Division of Nuclear Medicine, University of Heidelberg, Heidelberg, Germany) for excellent support with the acquisition of scintigraphic images for the cell tracking experiments.
This work was supported by the general budget resources of the University Hospital Heidelberg allocated to the Department of Transplantation Immunology.
These authors contributed equally to this work.