Third-party Tolerogenic Dendritic Cells Reduce Allo-reactivity In vitro and Ameliorate the Severity of Acute graft-Versus-host Disease in Allo-bone Marrow Transplantation

Authors


Abstract

Tolerogenic dendritic cells (tDCs) potently induce and maintain tolerance based on their distinct characteristics compared with conventional DCs. Recent reports show that donor or host tDCs promote allograft survival in mice. In this study, the efficacy of third-party tDCs in the prevention of acute graft-versus-host disease (aGVHD) was evaluated. In vitro, tDCs derived from the bone marrow (BM) of D1 mice were induced by GM-CSF, IL-10 and TGF-β1. The phenotypes, expression of cytokines and suppression of tDCs were analysed. In vivo, the effects of adoptive transfer of third-party-tDCs were evaluated in an MHC-mismatched aGVHD mouse model. Survival, body weight, GVHD scoring, histopathological specimens and serum cytokines were analysed in tDC-treated mice and untreated controls. Tolerogenic DCs had low expression of MHC and co-stimulatory molecules, expressed high levels of ‘immunosuppressive’ cytokines and suppressed allo-CD4+T cell proliferation. In the B6→D2 mouse model, all aGVHD mice died within 18 days. Fortunately, third-party tDCs transferred at low doses (104) effectively prolonged survival after allo-BMT. Furthermore, in the mice treated with 104 tDCs, serum levels of IL-10/TGF-β were significantly higher and the percentage of Foxp3+ cells continually increased compared with the mice treated with other doses of tDCs. Third-party tDCs play a crucial role in reducing the severity of aGVHD by modulating the secretion of various cytokines and expanding Foxp3+ regulatory T cells, which suggests the possibility of using third-party tDCs for therapeutic applications. Furthermore, special attention should be paid to the optimal range of tDCs for preventing allograft rejection.

Introduction

Dendritic cells (DCs) are the most potent antigen-presenting cells (APCs), having the unique ability to activate or suppress immune responses depending on maturation status, phenotype and tissue of origin [1, 2]. Due to their inherent plasticity, DCs are regarded as key instigators or regulators of innate and adaptive immunity. ‘Tolerogenic DC’ (tDC) has been assigned as a blanket description of immune regulatory DC that are usually immature [3], plasmacytoid in morphology [4], alternatively activated [5] or a specific subset [6]. Experimental generation of tDCs, propagated from BM precursors in low concentrations of GM-CSF, has been accomplished through treatment with maturation-inhibiting agents [7], blockade of co-stimulatory molecules either with antibodies [8] or antisense oligonucleotides [9], as well as pretreatment with chemical immunosuppressants [10], such as corticosteroids [11], cyclosporine A [12], rapamycin [13] and mycophenolate mofetil [14]. DCs exposed to immunosuppressive cytokines, such as IL-10, TGF-β and IL-6, also exhibit the potential for tolerance induction [15]. Furthermore, apoptotic-cell-treated DCs play an important role in inhibiting autoimmune diseases development, such as MS/EAE [16]. Tolerogenic DCs generated under these conditions typically present low numbers of self-peptide-MHC complexes (signal 1) coupled with limited co-stimulatory molecule expression (signal 2) and pro-inflammatory cytokine production (signal 3), leading to T cell anergy and apoptosis. Thus, tDCs reflect an incomplete or altered status in DC differentiation and are being considered in the design of therapeutic strategies.

Adoptive transfer of tDCs of donor or recipient origin offers the potential to inhibit immune responses [1]. This has been demonstrated in preclinical rodent models of BM, skin, heart and pancreatic islet cell transplantation, using immature, maturation-resistant or alternatively activated DC (AADC, propagated from BM cells in IL-10 and TGF-β followed by stimulation with LPS for 24 h). Evidence has emerged, demonstrating that treatment of mice after allogeneic BM transplantation (BMT) with recipient strain-derived tDCs markedly suppresses lethal graft-versus-host disease (GVHD) [17] or lethal endotoxemia [18]. Furthermore, donor-type tDCs in combination with CTLA4-Ig also induce long-term organ allograft survival [5]. This protective effect indicates that this type of tDCs preferentially secrete IL-10 and generate peripheral alloreactive CD4+CD25+Foxp3+Tregs from donor-derived CD4+CD25Foxp3T cells [19]. While donor-type tDCs can be more easily harvested compared with tDCs from patients undergoing BMT, the potential use of unrelated third-party tDCs could be advantageous as such cells can be prepared in large numbers in advance and can be screened for biological activity, as well as for pathogens and other quality control parameters. Thus, if third-party tDCs ameliorate BM allograft rejection, these cells could eventually become a new pharmaceutical agent rather than an esoteric mode of cell therapy. However, it is still not fully understood whether third-party tDCs exert similar suppressive activity in allogeneic immune responses as donor- or recipient-type tDCs, and which mechanisms of action are involved in inhibiting T cell alloreactivity and inducing tolerance in vitro and in vivo.

In this study, we evaluated, for the first time, the tolerogenic capacity of third-party tDCs propagated from mouse BM in IL-10 and TGF-β1, in alloreactive T cell responses in vitro and in vivo. Our data demonstrated that third-party tDCs exert immunosuppressive effects in allo-MLR assays and in a MHC-mismatched aGVHD mouse model. Furthermore, this effect was associated with various mechanisms, including the induction of allogeneic Tregs and modulation of the secretion of various cytokines. Moreover, the inoculation dose is an important factor in the determination of third-party-tDC-mediated immune reactions.

Materials and methods

Mice

Wild-type male B6 (H-2Kb), D1 (H-2Kq) mice (aged 6–8 weeks) and male D2 (H-2Kd) mice (weight ≥25 g/animal) were purchased from the Shanghai Laboratory Animal Center of the Chinese Academy of Science (SLACCAS, Shanghai, China). Animals were housed in a specific pathogen-free environment and provided with drinking water supplemented with gentamicin.

Ethics statement

This study was carried out in strict accordance with the recommendations in the guidelines of the Institutional Animal Care and Use Committee of the Chinese Association for Laboratory Animal Sciences. The protocol was approved by the Committee on the Ethics of Animal Experiments of Shanghai Blood Center (Permit Number: 09-0022). All surgery was performed under diethyl ether, and all efforts were made to minimize suffering.

Preparation of DCs

DCs were prepared as described previously [18] with some modifications. In brief, BM cells were plated at a density of 5 × 105 cells/ml in RPMI 1640 medium supplemented with 10% foetal bovine serum (PeproTech, Invitrogen, Rocky Hill, NJ, USA). On day 0, recombinant murine GM-CSF (20 ng/ml, PeproTech) was added, and then fresh medium and GM-CSF with additional recombinant murine IL-10 (15 ng/ml, PeproTech) and recombinant human TGF-β1 (15 ng/ml, PeproTech) were supplied on days 4 and 7. After 10 days in culture, non-adherent CD11c+ cells were collected as tolerogenic DCs (tDCs), and mature tDCs (mtDCs) were harvested after stimulation with lipopolysaccharide (LPS, 100 ng/ml, Sigma, St. Louis, MO, USA) for the final 48 h of the culture. Immature DCs (iDCs) were propagated under the same conditions in GM-CSF alone, and mature DCs (mDCs) were generated following exposure to LPS for 48 h.

Flow cytometric analysis

DCs were stained with anti-mouse FITC-CD11c, PE-CD80, -CD86, -IA-IE, -CD40 for phenotypic analysis. Anti-mouse mAbs were purchased from BD Biosciences Pharmingen (San Diego, CA, USA). The incidence of positive cells and MFI was analysed using CellQuest software (Becton-Dickinson, Heidelberg, Germany) by comparison with cells stained with PE-labelled isotype control mAbs.

Quantification of cytokines and function-associated molecules

Real-time quantitative PCR was performed on an ABI Prism 7500 Sequence Detection System (Applied Biosystems, Foster City, CA, USA) using SYBR® Premix DimerEraser™ Kit (TaKaRa, Otsu, Shiga, Japan). PCR conditions were as follows: 5 s at 95°C, 30 s at 55°C and 45 s at 72°C (45 cycles for real-time PCR). Forward and reverse oligonucleotide primers are as follows, in the 5′ to 3′ orientation: IL-12p40, GGAAGCACGGCAGCAGAATA and AACTTGAGGGAGAAGTAGGAATGG3; TGF-β, TTGCTTCAGCTCCACAGAGA and TGGTTGTAGAGGGCAAGGAC; IL-10, CCAAGCCTTATCGGAAATGA and TTTTCACAGGGGAGAAATCG; β-actin, ATCCGTAAAGACCTCTATGC and ACACAGAGTACTTGCGCTCA. Relative gene expression levels were determined using a previously described method [20]. All results were normalized to the expression of the housekeeping gene β-actin in PCRs. Data represented are from six independent experiments.

MLR assay culture

As responder cells, CD4+T cells were isolated from spleen cells (SC) of B6 mice using a CD4+T Cell Isolation Kit (Miltenyi Biotec, Bergisch Gladbach, Germany). Mature DCs derived from D2 mice were induced as described above, as stimulator APCs. A total of 1 × 105 responding CD4+T cells labelled with 5 μg of carboxyfluorescein succinimidyl ester (CFSE; Invitrogen, Germany) were activated by 2 × 104 mDC in 96-well U-bottom plates. The tDCs (suppressors) derived from D1 mice were then added at different ratios to the responder cultures in a final volume of 200 μl. Four days later, the cells were collected, stained with propidium iodide (PI, Sigma) and analysed by flow cytometry. Suppression activity was measured as the percentage of divided CFSE+ cells in the co-culture in the presence or absence of tDCs.

In vitro expansion and functional assessment of CD4+Tregs

To quantify Treg proliferation, CD4+CD25+T and CD4+CD25-T cells (B6) stimulated by D2 mDCs were labelled with CFSE as described above, before co-culture for 4 days with D1 tDCs. CD4+T cells collected at the end of the assay were stained for surface CD4-PE and then stained intracellularly with APC-anti-FoxP3 mAb. Proliferation of CD4+FoxP3+ cells was evaluated from the CFSE dilution profile. To assess their regulatory function, CD4+CD25+ cells from 4-day tDC-induced cultures, or freshly isolated CD4+CD25+T cells from normal B6 mice, were purified based on CD25 expression and tested for their ability to inhibit proliferation of normal CFSE-labelled, syngeneic CD4+T cells in response to stimulation with D2 mDCs.

Acute GVHD model

An acute GVHD model was established as described previously [21]. Briefly, 10- to 12-week-old D2 mice (recipients) were subjected to total-body lethal irradiation (8 Gy) from a γ-ray source and were injected with 3 × 107 BM cells and 1 × 107 SC from B6 mice (donors) via the tail vein after 24 h. Donor cells were resuspended in 0.1 ml phosphate-buffered saline (PBS, GIBCO, Grand Island, NY, USA), and an equivalent volume of PBS was injected as a control.

Adoptive transfer of third-party tDCs

Various numbers (105, 104 and 103) of third-party tDCs (D1) were adoptively transferred via the tail vein into D2 mice treated with irradiation as described above, in combination with equivalent numbers of BM and SC from B6 mice as in the aGVHD model. Each group contained 10 mice.

Monitoring of GVHD and survival

Survival and appearance were monitored on alternate days, and body weight was measured weekly. GVHD was assessed by a scoring system that summed changes in five clinical parameters [22]: weight loss, posture, activity, fur texture and skin integrity (maximum index, 10). Mean body weights of surviving mice in each group were determined during the 63-day observation period.

Histopathology

Histopathological specimens from the liver of surviving host mice were obtained on days 7, 21 and 63 after BMT. The specimens were fixed in 10% formalin and embedded in paraffin blocks. Sections (4–5 μm) were stained with haematoxylin and eosin (H&E) using standard protocols. Microscopic images were obtained using an Eclipse E1000M microscope (Nikon,Tokyo, Japan) and a SPOT RT digital camera and acquisition software (Diagnostic Instruments), with the final magnification (objective, 200 × /0.45 numerical aperture) provided in each figure. Image processing was performed using Photoshop CS (Abode Systems) with standard adjustments of brightness, contrast and colour balance to the entire image. A semi-quantitative scoring system was used to assess the following abnormalities known to be associated with GVHD, as previously described [23]: 0, normal; 0.5, focal and rare; 1.0, focal and mild; 2.0, diffuse and mild; 3.0, diffuse and moderate; and 4.0, diffuse and severe.

Serum cytokine analysis by CBA and ELISA

Serum was obtained from surviving recipient mice on days 7, 21 and 63 post-transplant and stored at −80°C. Levels of IL-4, IL-10, IL-12p70 and IFN-γ were simultaneously determined using the mouse cytokine cytometric bead array (CBA Kit, BD Pharmingen) according to the instructions provided by the manufacturer. Cytokine bead staining was analysed by flow cytometry, and the data were compiled with BD Biosciences CBA software. The concentration of TGF-β in serum was assessed using the Mouse TGF-β1 Platinum ELISA Kit (eBioscience, San Diego, CA, USA) according to the instructions provided by the manufacturer. A total of three samples were analysed in each group.

Induction of Treg subsets in the mouse model

Spleen cells were obtained from surviving recipient mice on days 7, 21 and 63 post-transplant. All of the SC were incubated with anti-mouse CD4-FITC mAb and then fixed, permeabilized and stained with anti-mouse PE-FoxP3 or isotype control mAbs (eBioscience).

Statistical analysis

Results are expressed as the mean ± SEM. Statistical differences in animal survival were analysed using the log-rank test. Differences in the weight of hosts and in serum cytokine levels between transplanted groups were analysed with computer statistical software (STAT 6.0, StatSoft, Inc.,Tulsa, OK, USA). For all tests, a P value of < 0.05 was considered significant.

Results

tDCs maintained the tolerogenic surface phenotype, expressed substantial levels of ‘immunosuppressive’ cytokines and exhibited suppressive activity in allo-CD4+T cell proliferation

It has been previously reported that tDCs derived from mouse BM are induced by GM-CSF with additional IL-10 and TGF-β1, which did not affect DC development from replicating BM progenitors [5]. In our study, low expression of MHC molecules (IA-IE) and co-stimulatory molecules (CD80, CD86 and CD40) was detected on tDCs (Fig. 1A). Following stimulation with LPS for 48 h, the mean fluorescence intensity (MFI) of cells for each of these molecules was consistently lower on the mature tDCs than those on the mature DCs (mDCs). This indicated that tDCs were comparatively resistant to maturation in response to LPS stimulation and retained the tolerogenic surface phenotype. The cytokine expression of tDCs was also analysed by real-time PCR. The results showed that IL-12p40 production by tDCs was negligible, even upon stimulation by LPS, while high levels of ‘immunosuppressive’ cytokines, such as IL-10 and TGF-β, were expressed (Fig. 1B). Surprisingly, TGF-β mRNA expression in mature tolerogenic dendritic cells (mtDCs) was obviously lower than that in tDCs, although there was no statistical difference in the levels of IL-10 produced by tDCs and mtDCs.

Figure 1.

Characteristic profile of tDCs derived from D1 mice BM. iDCs, mDCs, tDCs and mtDCs were induced in vitro as described in ‘Materials and methods’, and CD11c+ cells were harvested. (A) These four types of cells were stained with IA-IE, CD80, CD86, CD40 (thick lines), respectively, or isotype-matched mAbs (thin lines) and analysed by FACS. The MFI representative of 10 experiments is shown in the FACS profile. *< 0.01 in comparison with mDCs versus mtDCs by unpaired t-tests (n = 5). (B) Expression of IL-12p40, IL-10 and TGF-β by these four cell types was examined by real-time PCR. All results were normalized to the expression of the housekeeping gene β-actin and expressed as the Mean ± SEM of six individual experiments. *< 0.05 in comparison with iDCs versus tDCs or mDCs versus mtDCs by unpaired t-tests (n = 5). (C) Freshly isolated CD4+T cells (105) from B6 mice were used as responder cells and activated by allogeneic mDC (2 × 104) from D2 mice alone or in the presence of third-party tDCs from D1 at the different doses. After co-cultivation for 4 days, cells were collected and analysed by FACS. Progressive dilution of CFSE was used as measures of proliferation of responder cells, and the expansion rates of cells were further analysed in a histogram. The results are representative of five experiments. *< 0.05, as compared to no-tDCs group by unpaired t-tests (n = 3).

Next, in order to demonstrate the ability of third-party tDCs to suppress the proliferation of effector T cells, an MHC-mismatched allogeneic MLR assay was utilized as a functional readout. As shown in Fig. 1(C), in the absence of suppressors, the responder cells underwent vigorous proliferation, generating a large number of dividing T cells in the cultures. This intensive proliferation, however, was suppressed by the presence of third-party tDCs (suppressors) at different suppressor-to-responder cell (S: R) ratios. Interestingly, the low dose of tDCs (102; S: R, 1:1000) resulted in the most effective inhibition of allo-CD4+T cells expansion. Furthermore, proliferation of allo-CD4+T cells stimulated by third-party tDCs alone was negligible (data were not shown).

Taken together, these results indicated that tDCs induced by IL-10 and TGF-β1 maintained the tolerogenic surface phenotype even after LPS stimulation, expressed high levels of ‘immunosuppressive’ cytokines and exerted in vitro suppressive activity.

Third-party tDCs induced generation of allogeneic CD4+C25+Foxp3+T cell in vitro

The correlation of the tolerogenic properties of third-party tDCs with specific CD4+Foxp3+Treg interactions was then investigated. As illustrated in the Fig. 2(A), third-party tDCs retained the ability to induce proliferation of naturally existing CD4+Foxp3+Tregs (53.86 ± 8.44%, compared with the control (no tDCs), 59.12 ± 5.36%, > 0.05) and also promoted the conversion of naïve CD4+CD25-Foxp3-T cells to adaptive CD4+Foxp3+ Tregs (15.73 ± 8.94%, compared with the control, 5.31 ± 1.74%, < 0.01).

Figure 2.

Third-party tDCs induced generation and proliferation of allogeneic CD4+Foxp3+T cell in vitro. (A) To quantify Treg proliferation, CD4+CD25+T (Tregs) and CD4+CD25-T cells were isolated from total CD4+T cells (B6). These two subsets (105) stimulated by mDCs (D2, 2 × 104) were co-cultured with tDCs (D1, 102) for 4 days, respectively. Proliferation of CD4+FoxP3+ cells was evaluated from the CFSE dilution profile. The results are shown as dot plots that are representative of three experiments. (B) After co-cultivation for 4 days as described in A, third-party tDC primed (p)/expanded CD4+CD25+T cells (B6) were isolated and used at different doses as inhibitors of naïve CD4+T cells (B6) proliferation in response to stimulation with mDC (D2). The suppressive activity of the freshly isolated CD4+CD25+T cells from normal B6 mice also was examined as the control. The proliferative responses were measured after 4 days. Data are shown in a histogram, and the results are representative of three experiments.

The natural frequency of Foxp3+ cells exceeds 90% of CD4+CD25+ cells in the mouse [24]. Therefore, CD4+CD25+ cells (B6) from 4-day tDC-induced cultures were purified based on CD25 expression. The capacity of these cells to inhibit proliferation of normal CFSE-labelled, syngeneic CD4+CD25-T cells (B6) in response to stimulation with D2 mDCs was then evaluated. Compared with freshly isolated CD4+CD25+ T cells from normal B6 mice, CD4+CD25+T cells induced by third-party tDCs were equally effective on a per cell basis (at Treg/Tresp ratios of 1:1, 1:4, 1:16) in suppressing syngeneic CD4+CD25-T cell (B6) proliferation (Fig. 2B). These results indicated that third-party tDCs induced the generation and proliferation of functional allogeneic CD4+Foxp3+Treg cells in vitro.

Adoptive transfer of third-party tDCs reduced the severity of aGVHD and prolonged survival time of aGVHD mice

The in vivo efficacy of third-party tDCs was further evaluated in an acute GVHD mouse model. In this model, D2 mice (recipients) were subjected to total-body lethal irradiation, and acute GVHD was induced by injection of B6 spleen cells after allo-BMT [21]. Recipient D2 mice displayed typical GVHD symptoms, including diarrhoea, body weight loss, hunched back and hair loss. Histological examination also confirmed GVHD in representative mice. And all of the aGVHD mice died within 18 days. Following co-transfer of B6 SC and BM into D2 mice, third-party tDCs derived from D1 significantly enhanced the survival of aGVHD mice. Compared with aGVHD mice, 60% survival of mice was observed at least 63 days after transfer of 104 third-party tDCs (Fig. 3A). However, when the doses of tDCs were reduced to 103 cells, only 20% of mice survived past day 63 and when increased to 105, all of the mice died within 37 days after allo-BMT. Based on the clinic GVHD scoring and histological examination, it was confirmed that the cause of death was acute GVHD. And the tDC-treated mice lost less weight at earlier time points than untreated aGVHD mice and those that received low doses of tDCs (104/mouse or 103/mouse) maintained relatively stable body weight during the 63-day observation period (Fig. 3B). Furthermore, as demonstrated by summing the changes in five clinical parameters of aGVHD, the GVHD symptoms in mice treated with additional third-party tDCs were rapidly alleviated in the early phase after allo-BMT, while those of the aGVHD mice progressed further. The severity of GVHD was significantly milder after day 21 in mice treated with 104 tDCs than the mice treated with 103 or 105 tDCs (Fig. 3C).

Figure 3.

Adoptive transfer of third-party tDCs reduced the severity of aGVHD and prolonged survival time of aGVHD mice in the dose-dependent manner. D2 mice (recipients) were subjected to total-body lethal irradiation and received transplants of allo-BM cells and SC obtained from B6 mice (donors). Mice were then treated with or without different doses of third-party tDCs derived from D1 mice, designated aGVHD mice (□, N = 10), GVHD+tDC (D1 103) mice (■, N = 10), GVHD+tDC (D1 104) mice (▲, N = 10) and GVHD+tDC (D1 105) mice (◆, N = 10). The survival (A) from groups shown as above was monitored for 63 days after transplantation. *< 0.05 compared with aGVHD mice; +, compared with mice given 103 third-party tDCs; ≠, compared with mice given 105 third-party tDCs (analysed by the log-rank test). The mean and SD of the body weight (B) and clinical GVHD score (C) from groups (N = 10) shown as above were examined during the 63-day observation period after transplantation. +< 0.05, compared with mice given 103 third-party tDCs; ≠, compared with mice given 105 third-party tDCs (analysed by Student's paired t-test).

Liver specimens obtained from surviving recipient mice on days 7, 21 and 63 after allo-BMT were analysed histopathologically (Fig. 4A). Microscopic examination of livers obtained at autopsy on day 7 post-transplant revealed severe changes typical of GVHD in aGVHD mice, including the presence of marked hepatic lymphocyte infiltration surrounding the portal area and local putrescence. In contrast, very little infiltration and no putrescence were detected in mice treated with 104 third-party tDCs during the whole observation period, although partial cloudy swellings were observed in liver cells in the early phase after allo-BMT. Third-party tDCs transferred at higher or lower doses led to severe inflammatory cell infiltration and even some focal necrosis at day 21. Fortunately, GVHD symptoms were reduced at day 63 post-transplant in mice treated with 103 tDCs. GVHD histopathological scores were consistent with microscopic examination (Fig. 4B). Therefore, these results indicated that third-party tDCs, transferred at low doses (104), effectively alleviate symptoms of aGVHD and markedly prolonged survival after allo-BMT in vivo, whereas higher (105) or lower (103) doses of tDCs mediated less effective aGVHD inhibition.

Figure 4.

Histopathological analysis of specimens from the liver of host mice showed adoptive transfer of third-party tDCs reduced symptoms of aGVHD. (A) Microscopic examination of H&E stained sections of liver obtained at autopsy in survived D2 hosts treated with 8 Gy TBI followed by transfer of allo-BM cells, SC and different doses of third-party tDCs transfer at the indicated time points after transplantation. Symbols: → indicates marked hepatic lymphocyte infiltration surrounding the portal area; × indicates local putrescence in the liver specimen. Normal liver sections from normal mice are shown for comparison. Liver specimens are shown at ×200 magnification. (B) Histopathological GVHD scores of liver specimens from recipients treated with or without the different doses of third-party tDCs as described in Fig. 3 are expressed as the Mean ± SEM of three individual experiments. *< 0.05 compared with the indicated groups by Student's paired t-tests (n = 3).

Third-party tDCs suppressed allogeneic immune cell proliferation in aGVHD mice by modulating IL-10 secretion and inducing Foxp3+Tregs expansion in vivo

The ‘cytokine storm’ caused by massive alloreactive T cell is considered to be characteristic of GVHD [25]. To determine whether the attenuation of GVHD in mice after administration of third-party tDCs was a result of their modulating effect on the secretion of various cytokines, the production of IFN-γ, IL-4, IL-10, IL-12p70 and TGF-β in serum of mice treated with different doses of third-party tDCs and untreated aGVHD mice was analysed at the indicated time points (Fig. 5A–B). On Day d post-transplantation, the production of IFN-γ in mice treated with third-party tDCs was lower than that in aGVHD mice, regardless of the dose, although there were no marked differences between third-party-tDC-treated mice and aGVHD mice in the secretion of IL-4 and IL-10. Surprisingly, the level of IL-12p70 in aGVHD mice treated with 105 third-party tDC was markedly higher than at other doses, even in aGVHD mice. Moreover, the secretion of IL-12p70 in serum of mice treated with 105 third-party tDCs increased continuously until day 21. The levels of IL-10 in serum were also noticeably higher in mice treated with 104 tDCs than at other doses, and IFN-γ production was lower than that in the other two tDC-treated groups on day 21 after allo-BMT. Furthermore, during the 63-day observation period, the production of IL-10 remained high, while IL-4 and IL-12p70 secretion was consistently low. The production of TGF-β in serum of mice treated with 104 third-party tDCs increased continuously during the whole observation period. It can be speculated that high-dose third-party tDC treatment cannot effectively suppress aGVHD due to the high levels of IL-12p70 secreted, while low-dose treatment increases IL-10/TGF-β production and further prevents lethal aGVHD in mice.

Figure 5.

Third-party tDCs suppressed allogeneic immune cell proliferation in aGVHD mice by modulating IL-10 secretion and inducing Foxp3+Treg expansion in vivo. (A) Different groups of D2 mice (recipients) were treated as described in Fig. 3. Subsequently, serum was collected from each group of recipient mice at the indicated time points after transplantation. Serum levels of IFN-γ, IL-4, IL-10, IL-12p70 and TGF-β were measured by CBA assays. Error bars represent SD. *< 0.01 compared with the indicated groups by Student's paired t-tests. Results of three replicated experiments were pooled. (B) Changes in serum cytokines levels of the GVHD+tDC (D1 104) mice at different time points after transplantation, including 7 days (□), 21 days (▓) and 63 days (■). *< 0.01, #< 0.05 compared with the indicated groups by Student's paired t-tests. Results of three replicated experiments were pooled. (C) SC cells were obtained from each recipient mouse at the indicated time points after transplantation and stained intracellularly with anti-Foxp3 mAb. The expression of Foxp3 on SC from third-party tDC-treated mice at different doses was analysed by flow cytometry. Data are shown as a histogram in which cells were stained with anti-Foxp3 mAbs (thick lines) or isotype-matched mAbs (thin lines). The rate of positively stained cells and MFI representative of three experiments are shown in the FACS profile.

Changes in in Foxp3 expression in SC from 104tDC-treated mice on the days 7, 21 and 63 after allo-BMT were determined in order to investigate the capacity of third-party tDCs to effectively inhibit aGVHD by inducing FoxP3+Tregs subsets in vivo. The results showed that FoxP3+Tregs were continually induced in vivo and the percentage increased from 2.95 ± 0.63% on day 7 to 17.82 ± 3.58% on day 60 after allo-BMT (Fig. 5C). These results indicated that suitable doses of third-party tDCs played a role in reducing the severity of aGVHD, and this effect was associated with high IL-10/ TGF-β secretion and Foxp3+Tregs expansion in vivo.

Discussion

Over the past decade, detailed studies on both mouse and human tDCs and their relation to regulation of allo-immunity and the outcome of organ transplantation have demonstrated that tDCs, derived either from the donor or the host, play an important role as modulators of allo-reactive and auto-reactive T cell responses [1]. However, the use of third-party tDC-based immunization as a strategy to reduce allogeneic immune responses remains to be elucidated. In this study, we evaluated the tolerogenic capacity of third-party tDCs, propagated from mouse BM in GM-CSF, IL-10 and TGF-β, in alloreactive T cell responses in vitro and in vivo. The suppression of third-party tDCs was observed in allo-MLR assays and in an MHC-mismatched aGVHD mouse model. Mechanistic studies demonstrated that the immunosuppressive effect after immunization with third-party tDCs was critically dependent on (1) the induction of allogeneic Treg generation and proliferation, which further induced effector T cell hyporesponsiveness, and (2) the modulation the secretion of various cytokines, especially IL-10/TGF-β. Interestingly, we observed that the inoculation size (cell number) of third-party tDCs determined the fate of the immune response. Specifically, third-party tDCs were shown to be more immunogenic than tolerogenic at high doses. To the best of our knowledge, this is the first report that third-party tDCs can reduce alloreactive T cell responses in vitro and allogeneic aGVHD mortality and severity in vivo.

It is well known that tDCs could induce an anergic state in CD4+ memory T cells and can be used as a tool to induce proliferation of naturally occurring or adaptive regulatory T cells. In our study, third-party tDCs also successfully induced the generation of allogeneic CD4+Foxp3+ cells (Tregs) in short-term cultures. Moreover, as freshly isolated CD4+CD25+T cells, third-party tDC-expanded CD4+CD25+T cells were also effective in suppressing effector CD4+T cell expansion in response to stimulation with alloantigen. Such rapid expansion of Tregs, within days of allo-stimulation, has been reported in vivo [26]. These observations extend the functional characterization of third-party tDCs generated in IL-10 and TGF-β and strongly suggest their potential to regulate allo-immune reactivity in vivo.

As the mechanisms of action of third-party tDCs in vivo are still poorly understood, it is uncertain whether in vitro suppressor activity predicts in vivo activity. Thus, the allogeneic aGVHD model (B6→D2) was used to evaluate the in vivo activities of third-party tDCs (D1) in mice. We found that a single infusion of third-party tDCs in the aGVHD model displayed potent in vivo effects, inhibiting local inflammation and preventing aGVHD, which suggests that these third-party tDCs inhibited allogeneic T cell activation and expansion in vivo. The ‘cytokine storm’ caused by massive alloreactive T cell responses is considered to be characteristic of GVHD. Therefore, our study further focused on cytokine changes in vivo to determine the relationship between the modulation of cytokine secretion and the attenuation of GVHD in mice after third-party tDC administration. From our results, it was clear that third-party tDCs protected against lethal aGVHD by decreasing the production of inflammatory cytokines and increasing the secretion of immunosuppressive cytokines. IL-10 and TGF-β were shown to play a crucial role in this protective effect. The use of tDCs to generate Tregs has been previously demonstrated by other groups. For example, rapamycin-generated tDCs induce a Treg population that prolonged allograft survival [13]. Furthermore, Lan et al. [5] found that IL-10-producing CD4+T cells and CD4+CD25+Foxp3+ Tregs were increased in a vascularized heart transplant model treated with tDCs propagated with IL-10 and TGF-β and then stimulated with LPS. Thus, intracellular levels of Foxp3, which are related to Tregs, were also examined in spleen cells in 104-third-party-tDC-treated mice at the indicated time points. The percentage of CD4+Foxp3+ cells increased continuously during the entire observation period, indicating that third-party tDCs induce allo-Tregs in vivo and achieve potent transplantation tolerance characterized by reduced aGVHD symptoms and prolonged survival time. Further, Chan et al. provided evidence that indirect CD4 allo-immunity can be futile depending on its target of elimination, thereby leading to differential susceptibility of allogeneic tissues and generating split tolerance in allogenic transplants [27]. They identified a potential new mechanism for allo-tolerance in CD4+T cells and showed that passenger lymphocytes could be immunogenic. Their studies suggested that DC seemed not to the only immunogenic passenger leucocytes within transplants, and identifying cells that provide a source of immunogenic donor (even third-party donor) antigen is important for the future development of tolerance strategies.

Interestingly, the inoculation size (number) of third-party tDCs was shown to determine the fate of immune reactions. In repeated experiments, third-party tDCs exerted potent suppressive activity in aGVHD mice only at low doses (104/mice). Symptoms of aGVHD in the mice vaccinated with 105 third-party tDCs were comparatively accelerated, although high doses inhibited aGVHD transiently in the early phase after BMT. Furthermore, treatment with the lower doses (103/mice) of third-party tDCs did not effectively suppress aGVHD progression. These results were further supported by immunological analysis. IL-10 was induced prominently in the serum acquired from 104-third-party-tDC-vaccinated mice. However, this phenomenon was not observed in the 105-third-party-tDC-treated mice. Levels of IL-12p70, known as a T cell-stimulating factor, increased rapidly. Co-culture experiments in vitro with varying numbers of third-party tDCs and a fixed number of allo-CD4+T cells further demonstrated that these opposing immune phenotypes are dependent on third-party tDC density. Although the mechanism underlying the immunogenic effect of third-party tDC at densities is unknown, we hypothesize that third-party tDCs are further matured via DC–DC interactions at high densities in the presence of alloreactive T cells, thereby resulting immunogenic rather than tolerogenic responses.

Notably, Smyth et al. [28] had made the conflicting observations that despite demonstrating ‘tolerogenic’ function in vitro, DexD3-treated DCs did not prolong skin or heart allograft survival in vivo, but accelerated graft rejection. They suggested that negative function using drug-treated donor DCs confers a major risk of sensitization owing to processing and presentation of alloantigens by endogenous DCs, leading to priming of CD4+T cells with indirect allospecificity. It prompted us again that vaccination with DCs, which could achieve transplantation tolerance successfully, may be depended on many factors, such as the methods used to induce ‘tolerogenic’ DCs, the types of transplantations chosen and the additional therapies applied. And caution should be taken when planning future clinical interventions based on DC-cell therapy because the suppression effect in vitro may not necessarily reflect the in vivo situation.

In conclusion, we have demonstrated that immunization with third-party tDCs can be used as an effective strategy to reduce undesired immune responses, such as aGVHD. The use of third-party tDCs could allow advance preparation of a large bank of tDCs, with all the appropriate quality controls required for cell therapy. However, particular attention should be paid to third-party-tDC-based immunotherapy with regard to the optimal the number of third-party tDCs in the inoculation. Therefore, the importance of the clinical application of these data to the efficacy of third-party tDCs cannot be overestimated, considering the practical advantages offered using tDCs of normal volunteers rather than donor or host cells. Further studies with human tDCs in vitro and in vivo are required to ascertain the potential of third-party cells as a valuable resource for clinical applications.

Acknowledgment

Financial support was from the Science and Technology Commission of Shanghai Municipality (09140903000). We thank the Department of Animal Science in Shanghai Jiaotong University School of Medicine for laboratory animal husbandry.

Conflict of interests

The authors declare no financial interest or commercial conflict of interest.

Ancillary