Critical Role of Proinflammatory Cytokine IL-6 in Allograft Rejection and Tolerance


  • X. Zhao,

    1. Transplantation Research Center, Brigham and Women's Hospital and Children's Hospital Boston, Harvard Medical School, Boston, MA
    2. Department of Urology, Nanjing Drum Tower Hospital, Nanjing, Jiangsu Province, P. R. China
    Search for more papers by this author
    • X.Z. and O.B. contributed equally to this work.

  • O. Boenisch,

    1. Transplantation Research Center, Brigham and Women's Hospital and Children's Hospital Boston, Harvard Medical School, Boston, MA
    Search for more papers by this author
    • X.Z. and O.B. contributed equally to this work.

  • M. Yeung,

    1. Transplantation Research Center, Brigham and Women's Hospital and Children's Hospital Boston, Harvard Medical School, Boston, MA
    Search for more papers by this author
  • B. Mfarrej,

    1. Transplantation Research Center, Brigham and Women's Hospital and Children's Hospital Boston, Harvard Medical School, Boston, MA
    Search for more papers by this author
  • Sunmi Yang,

    1. Transplantation Research Center, Brigham and Women's Hospital and Children's Hospital Boston, Harvard Medical School, Boston, MA
    Search for more papers by this author
  • L. A. Turka,

    1. Transplant Institute and Division of Transplant Immunology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA
    Search for more papers by this author
  • M. H. Sayegh,

    1. Transplantation Research Center, Brigham and Women's Hospital and Children's Hospital Boston, Harvard Medical School, Boston, MA
    Search for more papers by this author
  • J. Iacomini,

    Corresponding author
    1. Transplantation Research Center, Brigham and Women's Hospital and Children's Hospital Boston, Harvard Medical School, Boston, MA
    Search for more papers by this author
  • X. Yuan

    Corresponding author
    1. Transplantation Research Center, Brigham and Women's Hospital and Children's Hospital Boston, Harvard Medical School, Boston, MA
    Search for more papers by this author
    • Presently at Beijing University Wujieping Urology Center and Beijing University Shougang Hospital, Beijing University, Beijing, China.

Xueli Yuan, or John Iacomini, PhD,


The proinflammatory cytokine IL-6 plays an important role in controlling T-cell differentiation, especially the development of Th17 and regulatory T cells. To determine the function of IL-6 in regulating allograft rejection and tolerance, BALB/c cardiac grafts were transplanted into wild-type or IL-6-deficient C57BL/6 mice. We observed that production of IL-6 and IFN-γ was upregulated during allograft rejection in untreated wild-type mice. In IL-6-deficient mice, IFN-γ production was greater than that observed in wild-type controls, suggesting that IL-6 production affects Th1/Th2 balance during allograft rejection. CD28-B7 blockade by CTLA4-Ig inhibited IFN-γ production in C57BL/6 recipients, but had no effect on the production of IL-6. Although wild-type C57BL/6 recipients treated with CTLA4-Ig rejected fully MHC-mismatched BALB/c heart transplants, treatment of IL-6-deficient mice with CTLA4-Ig resulted in graft acceptance. Allograft acceptance appeared to result from the combined effect of costimulatory molecule blockade and IL-6-deficiency, which limited the differentiation of effector cells and promoted the migration of regulatory T cells into the grafts. These data suggest that the blockade of IL-6, or its signaling pathway, when combined with strategies that inhibit Th1 responses, has a synergistic effect on the promotion of allograft acceptance. Thus, targeting the effects of IL-6 production may represent an important part of costimulation blockade-based strategies to promote allograft acceptance and tolerance.


antigen-presenting cells




cytotoxic T lymphocyte antigen 4


fluorescein isothiocyanate


forkhead box P3


interferon gamma




mixed lymphocyte reaction


median survival time


peripheral blood mononuclear cell




standard deviation


T helper


tumor necrosis factor α




T cells are key initiators, effectors and regulators of graft rejection. The importance of the interaction between the innate and adaptive immune systems in the allograft response has recently been recognized. Many proinflammatory cytokines, such as IL-17 and IL-6, seem to play a key role in T-cell immunity by connecting innate and adaptive immunity (1–4). IL-6 is a pleiotropic cytokine produced by antigen-presenting cells (APCs), such as dendritic cells, macrophages and B cells, as well as by a variety of nonhematopoietic cells, such as fibroblasts, epithelial cells, keratinocytes and astrocytes (5). This multifunctional cytokine was first identified as a B-cell growth factor that promotes the differentiation of B cells into plasma cells (6). Later, IL-6 was shown to prolong T-cell survival in vitro via maintenance of Bcl-2 expression and downregulation of Fas ligand expression (7–10), and to promote the activation and proliferation of antigen-specific T cells (7,11).

It is well established that, upon activation, T helper (Th) cells differentiate into Th1, Th2 and the recently elucidated Th17 phenotypes. These phenotypes are defined according to the conditions and cytokines required for their differentiation into distinct T-cell subsets, the cytokines they produce and their immunological functions (12,13). The interactions between various cytokines involved in the differentiation of Th1, Th2 and Th17 cells are complex. Cytokines can induce or suppress their own synthesis, as well as the expression of other cytokines and their receptors, leading to antagonism or synergy in many different and often redundant ways (12,13). IL-6 is critical for Th17 development. It prevents undifferentiated naïve T cells from developing into a regulatory population (12), and may increase the effector/memory T-cell population (11).

To date, the role of IL-6 in graft rejection has received limited attention. Liang et al. demonstrated that heart allografts from IL-6-deficient (IL-6−/−) donors survive longer than those from wild-type (WT) donors (14). Using a mouse skin transplantation model, Shen and Goldstein demonstrated that IL-6 synergizes with tumor necrosis factor-α (TNF-α) to promote T-cell alloimmune responses both in vitro and in vivo, and impairs the ability of regulatory T cells (Tregs) to suppress effector T-cell alloimmune responses (15). However, the exact role of IL-6 in determining the cytokine milieu during alloimmune responses has yet to be clearly defined.

Here, using a fully major histocompatibility complex (MHC)-mismatched mouse model of cardiac transplantation, we found that, along with Interferon-γ (IFN-γ), IL-6 is the major cytokine upregulated during allograft rejection in untreated WT recipients. CD28-B7 blockade significantly inhibited IFN-γ expression, but did not affect IL-6. When combined with CD28-B7 blockade, either systemic deficiency or neutralization of IL-6 in the recipients facilitated allograft tolerance by limiting the differentiation of effector cells and by promoting the migration of Tregs into the grafts. These studies strongly suggest that targeting IL-6 or its signaling pathway will prove to be a useful clinical strategy to facilitate tolerance induction.

Materials and Methods

Mice heart transplantation model

BALB/c (H-2d), WT and IL-6- knockout (IL-6−/−) C57BL/6 (B6, H-2b) were purchased from the Jackson Laboratories (Bar Harbor, ME, USA). Animals were maintained in accordance with institutional guidelines. Cardiac allografts from BALB/c donor mice were transplanted into WT and IL-6−/− mice. Vascularized heart engraftment was performed using sterile microsurgical techniques, as described previously (4). Graft function was assessed by daily palpation of the abdomen. Rejection was defined as complete cessation of cardiac contractility, confirmed by direct visualization. Graft survival is shown as median survival time (MST) in days.

Treatment protocols

A single dose (0.25 mg) of CTLA4Ig, a fusion protein blocking the CD28-B7 pathway, was given intraperitoneally (i.p.) on the day of transplantation. Neutralizing IL-6 mAb (R&D Systems, Minneapolis, MN, USA) was administered i.p. at a daily dose of 0.1 mg from day 0 to 3, and then every other day until day 13.


Heart allografts were recovered either at the time of rejection, or at defined time points (such as 7, 14, 21 or 100 days after transplantation). Samples were fixed in 10% formaldehyde, embedded with paraffin, and cut into 5 μm sections. The tissue sections were then stained with hematoxylin and eosin (H&E) for evaluation of cellular infiltration using the standard protocol of the Harvard Rodent Histology Core. The sections were also stained with anti-CD3 and anti-Foxp3 antibodies (BD Biosciences, San Jose, CA, USA) for assessment of cell infiltration and distribution of Tregs within the grafts, as previously described (16). Images were captured using Nikon C1 Plus Confocal Laser Scanning microscope.

Mixed lymphocyte reaction (MLR) proliferation assay

We used the MLR, a well-accepted in vitro assay for the assessment of direct allorecognition, to measure the impact of IL-6 on lymphocyte proliferation and T-cell differentiation. Responder cells (freshly isolated B6 splenocytes, 5 × 105 cells/well) and stimulator cells (irradiated BALB/c splenocytes [3300 rad], 5 × 105 cells/well) were cocultured in U-bottom 96-well plates in a total volume of 200 μL culture medium. After 66 h, 3H-TdR (1 μCi/well) was added for the final 6 h of culture to assess cell proliferation. Cells were lysed with distilled water and collected onto nitrocellulose filter paper using a Filtermate 96 cell harvester (Canberra Pacard, Frankfurt, Germany), followed by measurement of 3H-TdR incorporation, using a Matrix 96 Direct Beta Counter. On foot of the MLR experiments, suppression assays were performed to evaluate Treg function. CD4+CD25+ Tregs were isolated from spleens and lymph nodes of IL-6−/− and WT animals; 2 × 105 Tregs were cultured with responder lymphocytes recovered from skin-sensitized B6 mice and irradiated BALB/c stimulators, resulting in a Treg:Responder ratio of 1:1. After 48 h, the IFN-γ concentration in the cell culture supernatants was measured to calculate the percentage of suppression.

Flow cytometry

Lymphocytes isolated from spleens and lymph nodes of transplanted animals, freshly isolated peripheral blood mononuclear cells (PBMCs) from naïve or transplanted B6 mice, or cells recovered after an MLR experiment (performed as described earlier without the addition of 3H-TdR), were analyzed using flow cytometry to assess the CD8 effector and CD4 regulatory T-cell populations. Cells were washed twice in PBS containing 1% FCS and incubated with either isotype control or fluorescein isothiocyanate (FITC) anti-CD8, phycoerythrin (PE)-conjugated anti-CD44 and allophycocyanin-conjugated anti-CD62L antibodies to evaluate the number of CD8 effectors. To assess Tregs, cells were stained with FITC-labeled anti-CD4, and PE-conjugated anti-CD25 (both from Pharmingen, San Diego, CA, USA) for 45 min at 4°C. Cells were then washed twice. Intracellular Foxp3 staining was performed using an APC-conjugated anti-Foxp3 antibody according to the manufacturer's protocol. Cells were analyzed by flow cytometry using a FACSCalibur (Becton Dickinson, San Jose, CA, USA) and Flowjo software.

Cytokine analysis by ELISPOT and LUMINEX assays

The frequencies of lymphocytes producing selected cytokines were measured by the ELISPOT assay, and the absolute amount of cytokines in cell culture supernatants quantified by LUMINEX. The ELISPOT assay has been previously described (3,4). Briefly, 0.5 × 106 splenocytes recovered from heart allograft recipients at defined time points posttransplantation were cocultured with 0.5 × 106 irradiated donor splenocytes in Immunospot plates (Cellular Technology Ltd., Shaker Heights, OH, USA). These plates were coated overnight with 4 μg/mL of rat antimouse IFN-γ mAb (R4–6A2), rat antimouse IL-4 mAb (BVD4–1D11) or rat antimouse IL-10 mAb (JES5–2A5) (all from Pharmingen) in sterile PBS. After 24 or 48 h of culture at 37°C in 5% CO2, cytokine producing cells were detected with biotinylated rat antimouse IFN-γ, IL-4 or IL-10 mAb (all from Pharmingen) and horseradish peroxidase–conjugated streptavidin (Dako, Carpenteria, CA, USA). Plates were developed using 3-amino-9-ethylcarbazole (Sigma-Aldrich, St. Louis, MO, USA). Spots were counted on a computer-assisted enzyme-linked immunospot image analyzer (Cellular Technology Ltd.), and frequencies were expressed as number of spots per million splenocytes. For LUMINEX assays, splenocytes were recovered from heart allograft recipients and re-stimulated in vitro with irradiated allogenic splenocytes. The cell-free supernatants of individual wells were removed after 72 h and analyzed using a preconfigured 21-plex cytokine bead-based immunoassay according to manufacturer's instructions and as previously described (3,4).


For graft survival analysis, Kaplan–Meier graphs were constructed and log-rank comparison of the groups was used to calculate p-values. For cytokine levels as measured by ELISPOT and LUMINEX, data are presented as mean ± standard deviation (SD), and comparisons between the values were performed using the two-tailed Student's t-test. For all statistical analyses, the level of significance was set at a probability of p < 0.05. All experiments were performed at least three times.


The absence of IL-6 inhibits lymphocyte proliferation and promotes CD4 T-cell differentiation into a regulatory phenotypein vitro

It has been reported that IL-6 enhances T-cell proliferation and activation, promotes the differentiation of CD4 T cells into Th17 cells and impairs Treg differentiation (12). To ensure that IL-6-deficient mice do not phenotypically have fewer effector T cells and more Tregs than WT mice, we first enumerated the CD25+Foxp3+ CD4 T cells and CD44hiCD62Llow CD8 effector/memory T cells in peripheral blood, lymph nodes and spleen of IL-6−/− and WT animals. The percentages of CD4 Tregs and CD8 effector T cells were comparable between IL-6−/− and WT animals in all compartments (p > 0.05 for each comparison of all compartments; Figure 1).

Figure 1.

Percentages of CD4 regulatory and CD8 effector T cells in peripheral blood, lymph nodes and spleens of naïve IL-6-deficient mice are comparable to those of naïve wild-type animals. Naïve IL-6−/− mice demonstrated similar numbers of CD25+Foxp3+ regulatory CD4 cells in the peripheral blood, lymph nodes and spleens (11.4 ± 1.5%, 15.7 ± 2.2% and 10.8 ± 1.7%[IL-6−/−] vs. 12.3 ± 2.5%, 17.1 ± 3% and 12.4 ± 1.8%[WT]), and similar frequencies of CD44hiCD62Llow CD8 effector T cells (2.4 ± 0.3%, 6.3 ± 0.5% and 2.8 ± 0.8% for IL-6−/− mice vs. 2.3 ± 0.2%, 6.5 ± 0.19% and 3.3 ± 0.35% for WT, respectively) to WT mice (p > 0.05 for the comparisons of all compartments). Right panel: Bar graphs show the data collected from three animals in each group (the experiment was performed three times); left panel: representative flow plots of the T-cell populations in each compartment of both mouse strains.

To determine the function of IL-6 in alloimmune responses, we stimulated IL-6−/− B6 lymphocytes with irradiated BALB/c cells, and analyzed T-cell proliferation and differentiation. As expected, the proliferation of lymphocytes from IL6−/- mice was significantly less than those from WT animals (Figure 2A; H-TdR incorporation: 13 346 ± 1758 vs. 382 250 ± 9236 cpm; p < 0.01). This result is consistent with previous data (7,11) and indicates that IL-6 promotes lymphocyte proliferation in an in vitro model of alloimmune responses.

Figure 2.

Effect of IL-6 deficiency on lymphocyte proliferation, T-cell differentiation and function in vitro. Lymphocytes were isolated from naïve IL-6−/− and WT B6 mice, and then stimulated with irradiated BALB/c splenocytes. Data are shown as mean ± SD, p < 0.05 is considered significant. Each panel is representative of at least three independent experiments. (A) Decreased lymphocyte proliferation in the absence of IL-6. (B) CD4 T cells from IL-6-deficient mice differentiated into CD25+Foxp3+ regulatory CD4 T cells more efficiently than CD4 T cells from WT mice. (C) Enhanced regulatory function of regulatory T cells from IL-6−/− animals, when compared to regulatory T cells from WT animals. (D) IL-6 deficiency does not affect CD8 effector differentiation.

We also examined the effect of IL-6 deficiency on the differentiation of regulatory and effector T cells when stimulated with fully allogeneic BALB/c splenocytes. After 72 h, the percentage of CD25+Foxp3+ CD4 Tregs was significantly higher in cell cultures containing IL-6−/− responder cells than in those containing WT responder cells (Figure 2B; 6.92 ± 1.41 vs. 3.74 ± 0.58%; p < 0.01). To determine whether IL-6−/− Tregs have intact suppressive functions, we isolated CD4+CD25+ cells from spleens of either naïve IL-6−/− or naïve WT mice, and cocultured them with responder lymphocytes from WT B6 recipients and irradiated BALB/c stimulators. IL-6−/− Tregs suppressed IFN-γ production to a greater degree than WT Tregs (p < 0.01; Figure 2C). In contrast, the percentages of CD8 T cells that differentiated into CD44hiCD62Llow effector/memory phenotype were not significantly different between IL-6−/− and WT cells (1.54 ± 0.72 vs. 2.47 ± 1.1%; p > 0.05; Figure 2D).

Systemic deficiency of IL-6 alone does not prolong allograft survival, but alters the cytokine profile in the recipients

To study the role of IL-6 in alloimmunity in vivo, BALB/c hearts were transplanted into WT and IL-6−/− B6 mice. Both WT and IL-6−/− mice rejected their grafts within 1 week of transplantation (Figure 3A; MST: 7 ± 1 vs. 7 ± 1 days; p > 0.05). When compared to WT recipients, FACS analysis showed decreased percentages of CD8 effector/memory T cells (CD44hiCD62Llow) in the peripheral blood of IL-6−/− recipients (9.7 ± 1.7 vs. 15.1 ± 2.8%; p < 0.05), but no significant difference in the numbers of circulating CD25+Foxp3+ CD4 Tregs (13.1 ± 5.1 vs. 17.2 ± 2.1%; p > 0.05) (Figure 3B). This result differs from our in vitro observations (Figures 2B and D), and suggests that the impact on T-cell differentiation varies depending on whether the environment is in vivo or in vitro.

Figure 3.

Impact of IL-6 deficiency on the in vivo alloimmune response in untreated B6 recipients of fully MHC-mismatched BALB/c cardiac allografts. (A) No effect of IL-6 deficiency on the survival of vascularized cardiac allografts (n = 5–7 animals per group). (B) At the time of rejection (7 days after transplantation), IL-6−/− recipients demonstrated lower frequencies of CD44hiCD62Llow CD8 effector T cells and similar frequencies of CD25+Foxp3+ regulatory CD4 cells in the peripheral blood, compared to WT recipients. Bottom panel shows representative flow plots of each group. (C) IL-6-deficient recipients show lower frequencies of IFN-γ-producing alloreactive cells, when compared to WT recipients (ELISPOT). (D) Splenocytes from IL-6-deficient recipients show lower IFN-γ-production upon re-stimulation with alloantigen, when compared to WT recipients (LUMINEX).

In WT mice, IFN-γ and IL-6 were the two major cytokines upregulated at the time of rejection. The concentrations of IFN-γ and IL-6 in cell culture supernatants from splenocytes of WT recipients re-stimulated with alloantigen in vitro were 1865 ± 616.7 and 565.5 ± 100.3 pg/mL, respectively. None of the other cytokines analyzed (IL-1a, IL-1b, IL-2, IL-4, IL-5, IL-9, IL-12p40, IL-12p70, IL-13, IL-15 and IL-17) exceeded concentrations of 15 pg/mL, except IL-10 (33.4 ± 6.6 pg/mL). Interestingly, splenocytes from IL-6−/− recipient mice produced greater amounts of IFN-γ at the time of rejection than WT mice (9065 ± 5471.2 vs. 1865 ± 616.7 pg/mL; p < 0.01), suggesting that IL-6 deficiency may favor a Th1-type response (Figure 3C). This was also confirmed by ELISPOT assays in which the frequencies of IFN-γ-producing cells were higher in IL-6−/− recipients than in WT recipients (650.3 ± 50.9 vs. 278.3 ± 186.3 spots/0.5 × 106 cells; p < 0.01; Figure 3D). No significant IL-17 production was observed in either IL-6−/− or WT animals (0 ± 0 and 2.4 ± 1.4 pg/mL, respectively).

IL-6 deficiency synergizes with costimulatory blockade to facilitate tolerance in recipients treated with CTLA4Ig

CTLA4Ig is a fusion protein that binds to B7 molecules, and therefore interferes with costimulatory signals through CD28/CTLA4-B7 pathway. CTLA4Ig has been widely tested in animal transplantation models, and more recently in clinical trials (17). In a fully MHC-mismatched mouse model of heart transplantation (BALB/c into B6), administration of a single dose of 0.25mg CTLA4Ig on the day of transplantation significantly prolongs allograft survival in WT mice (Figure 4A). However, BALB/c heart transplants were eventually rejected with an MST of 22 days (Figure 3A). In contrast, CTLA4Ig treatment resulted in long-term survival (> 100 days) of BALB/c hearts in IL-6−/− recipients (Figure 3A). Histological examination of the allografts 3 weeks posttransplantation showed a significant lymphocyte infiltration in WT recipients, which was absent in the grafts of IL-6−/− recipients (Figure 4B).

Figure 4.

Effect of IL-6 deficiency on allograft tolerance. IL-6−/− and WT B6 recipients of fully MHC-mismatched BALB/c hearts were treated with a single dose of CTLA4Ig (0.25mg) on the day of transplantation. (A) IL-6 deficiency facilitates allograft tolerance. Survival of BALB/c cardiac allografts is significantly prolonged in IL-6−/− recipients, when compared to WT controls (n = 4–7 per group). (B) Pathology of allografts at 3 weeks after transplantation. Note the relatively normal myocardium and vasculature of allograft of IL-6−/− recipient. (C) Bar graphs illustrate IFN-γ production by splenocytes of WT and IL-6−/− recipients assessed by the LUMINEX assay (left-hand side panel), and frequency of IFN-γ-producing cells measured with ELISPOT assay (right-hand side panel). Presented data (mean ± SD, three to four samples in each group) are representative of three independent experiments. (D) IL-6 production by splenocytes of CTLA4Ig-treated wild-type recipients was assessed with the LUMINEX assay. Data are presented as mean ± SD, and are representative of three independent experiments. IL-6 production by splenocytes of CTLA4Ig-treated WT recipient is similar to untreated controls. (E) IFN-γ production by different subset of T cells and effect of CTLA4Ig assessed with intracellular staining and FACS analysis. Four groups of transplanted animals (Untreated WT and IL-6−/−, CTLA4Ig-treated WT and IL-6−/−) were set up. Splenocytes were all recovered at day 7, and intracellular IFN-γ staining and surface staining for CD4 and CD8 were performed. Data were analyzed with FACSCalibur (Becton Dickinson) and Flowjo software. (n = 4 for all groups).

Since IFN-γ, the hallmark Th1 cytokine, was upregulated in IL-6−/− recipients without CTLA4Ig treatment, we examined whether CTLA4Ig-mediated tolerance in IL-6−/− recipients was associated with a loss of the Th1 response. At 21 days posttransplantation, when WT recipients rejected their grafts, net IFN-γ production and the frequencies of IFN-γ-producing cells were significantly lower in CTLA4Ig-treated WT and IL-6−/− than untreated controls (IFN-γ concentration: WT + CTLA4Ig: 161.5 ± 73.6 pg/mL, IL-6−/−+ CTLA4Ig: 502.8 ± 194.7 pg/mL vs. 865 ± 616.7 pg/mL and 9065 ± 5471.2 pg/mL for WT and IL-6−/− controls, respectively. Frequencies of IFN-γ-producing cells (ELISPOT): 36.8 ± 5.3 and 64.4 ± 10.2 spots/0.5 × 106 responder cells for CTLA4Ig-treated WT and IL-6−/− mice, compared to 314 ± 31.8 and 849 ± 22.5 spots/0.5 × 106 responder cells in untreated WT and IL-6−/− controls respectively; p < 0.01 for both groups compared to untreated controls; Figure 4C). However, despite the striking prolongation of graft survival in IL-6−/− mice and the observed inhibition of the Th1 response, IFN-γ production and frequencies of IFN-γ-producing cells were still greater in IL-6−/− mice when compared CTLA4Ig-treated WT recipients. Interestingly, IL-6 production in WT recipients was not affected by CTLA4Ig administration, when compared to untreated controls (1355 ± 975.6 vs. 1305.6 ± 631.3 pg/mL; p > 0.05; Figure 4D). Similar to untreated controls, cells from CTLA4Ig-treated WT and IL6−/- recipients did not produce significant amounts of IL-17 and other cytokines.

In order to determine which T-cell subsets were producing IFN-γ, to what extent they could be inhibited by CTLA4Ig and the difference between IL-6−/− and WT recipients, we set up four groups of transplanted animals (untreated WT and IL-6−/−, CTLA4Ig- treated WT and IL-6−/−). Splenocytes from all recipients were recovered at day 7 posttransplantation, and the intracellular expression of IFN-γ was measured by flow cytometry. Both CD4 and, in particular, CD8 T cells produced IFN-γ with no significant difference between untreated WT and IL-6−/− recipients (6.8 ± 0.9% and 7.4 ± 0.3% of CD4, 28.6 ± 4.1% and 32.2 ± 1.6% of CD8 T cells for WT and IL-6−/− recipients, respectively; p > 0.05 for both comparisons; n = 4 for each group, Figure 4E). Some IFN-γ was also produced by non-T cells (data not shown). IFN-γ production by both WT and, in particular, IL-6−/− CD4 T cells was significantly inhibited by CTLA4Ig (2.6 ± 0.3% and 1.7 ± 0.2% of CD4; p < 0.01 for both CTLA4Ig-treated groups when compared to untreated groups). CTLA4Ig did not affect IFN-γ production by WT CD8 T cells (24.8 ± 3.5% vs. 28.6 ± 4.1% (untreated group), n = 4 for each group; Figure 4E), but inhibited IFN-γ production by IL-6−/− CD8 T cells (13.6 ± 1.6% vs. 32.2 ± 1.6% (untreated group); p < 0.01).

Effects of IL-6 deficiency on Tregsin vivo

Given that IL-6 deficiency appears to result in higher percentages of Tregs after in vitro stimulation, we next examined Tregs in CTLA4Ig-treated IL-6−/− recipients that accepted BALB/c heart transplants, assuming that IL-6 deficiency might lead to increased Treg frequencies. Early after transplantation (day 7), analysis of T-cell subsets from the peripheral blood of transplanted animals revealed that the Treg frequencies in the peripheral blood of CTLA4-Ig-treated IL-6−/− recipients was lower than that observed in similarly treated WT recipients (2.08 ± 0.23% vs. 4.12 ± 0.15%, p < 0.01, Figure 5A). At day 14 and day 21 after transplantation, this difference was no longer detectable (4.05 ± 0.74% and 7.26 ± 0.74% for IL-6−/− mice vs. 3.78 ± 0.56 and 7.11 ± 0.28% for WT recipients at day 14 and day 21, respectively; p > 0.05 for the comparison at both time points (Figure 5A). Consistent with previous data suggesting a role for IL-6 in increasing effector/memory T-cell population (11), a reduction in CD44hiCD62Llow CD8 effector cells on day 7, 14 and 21 was observed in the peripheral blood of IL-6−/− recipients. (2.83 ± 0.65%, 2.67 ± 0.8% and 2.24 ± 0.23% vs. 7.5 ± 0.89%, 6.87 ± 1.68% and 6.32 ± 2.18% in WT controls; p < 0.01; Figure 5B).

Figure 5.

Effect of IL-6 deficiency on regulatory CD4 T cells and CD8 effector cells in CTLA4Ig-treated recipients. (A) Percentage of CD25+Foxp3+ regulatory CD4 T cells in the peripheral blood of IL-6−/− recipients was lower than in wild-type recipients at day 7, but was similar to WT at day 14 and day 21 after transplantation (n = 3 in all groups, data are representative of three independent experiments). Lower panel shows representative flow plots of each group. (B) IL-6−/− recipients exhibit lower percentages of CD44hiCD62Llow CD8 effector T cells in peripheral blood at 7, 14 and 21 days after transplantation, when compared to WT recipients. Lower panel shows representative flow plots of each group (n = 3 in all groups, data are representative of three independent experiments). (C) Representative immunohistochemistry staining of cardiac allografts at day 7 posttransplantation. Grafts from IL-6−/− recipients show significantly more infiltrating Foxp3+ T cells when compared to WT recipients. Foxp3 was stained in red and CD3 in brown. Right panel bar graph shows the number of Foxp3+ CD3 T cells per high power field under optical microscope. Each panel is representative of at least three independent experiments.

Although Tregs were not found to be increased in the peripheral blood of IL-6−/− recipients, we found significant changes within the graft itself. As determined by immunohistochemical staining of the grafts, we observed a significant increase in the number of Foxp3+ CD3 T cells in the grafts recovered from IL-6−/− recipients when compared to grafts from WT controls (Figure 5C; 11.2 ± 5.3 per high power field vs. 0.5 ± 1.6 for WT recipients; p < 0.01).

Neutralization of IL-6 promoted graft acceptance in the CTLA4Ig-treated recipients

From the results of the in vitro and in vivo experiments described earlier, it seems that IL-6 plays an important role in alloimmune responses, and that IL-6 deficiency promotes tolerance when combined with CTLA4Ig. Therefore, targeting IL-6 may be an effective way to promote transplant tolerance by limiting effector T-cell expansion and by promoting Treg function and distribution within the allograft. To verify this hypothesis, we administered IL-6-neutralizing monoclonal antibody to WT mice that were simultaneously treated with CTLA4Ig. IL-6 neutralization in combination with CTLA4Ig resulted in long-term survival of fully allogeneic BALB/c hearts (MST: >100 vs. 22 days (CTLA4Ig only); p < 0.01; Figure 6). Taken together, these data suggest that inhibition of IL-6 is critically important to induce tolerance in this model.

Figure 6.

In vivo neutralization of IL-6 inhibits rejection and facilitates allograft survival with costimulation blockade in wild-type recipients. Survival of fully MHC-mismatched BALB/c cardiac allografts in WT C57BL/6 recipients treated with CTLA4Ig and either anti-IL-6 mAb (n = 5) or control IgG (n = 8). Data are presented as a Kaplan–Meier curve.


The fate of a transplanted organ is determined by the balance between pathogenic and proinflammatory mechanisms that promote rejection, and regulatory or anti-inflammatory mechanisms that facilitate allograft tolerance or acceptance. Cytokines play an important role in regulating alloimmune responses (18). Before the discovery of IL-17-producing T helper (Th17) cells, naïve Th cells were considered to differentiate into two distinct populations, Th1 and Th2 (18). Th1 cells are associated with cell-mediated immunity, including delayed-type hypersensitivity responses, and produce IFN-γ and IL-2. Th2 cells produce IL-4, IL-5, IL-10 and IL-13, protect against extracellular pathogens and provide help to B-cell function (19). In the context of autoimmunity and in some transplantation models, Th1 cells were believed to be pathogenic, while Th2 cells were thought to be protective. However, the simplicity of this paradigm has been challenged by subsequent studies (20). Recently, another subset of Th cells, termed Th17, was identified. The newly discovered Th17 cells play a critical role in many autoimmune diseases that were traditionally thought to be mediated by Th1 immunity (21). Growing evidence indicate that each type of Th cells, including Th1, Th2 and Th17 cells, is capable of mediating allograft rejection (3,22,23). In our previous studies using T-bet−/− recipients, we found that, in the absence of Th1 responses, CD4 Th17 cells mediate aggressive proinflammatory responses that lead to accelerated rejection and severe vasculopathy in a model of chronic cardiac allograft rejection (3). Later, we demonstrated that IL-17-producing CD8 T cells are resistant toward the induction of tolerance by combined blockade of the CD28-B7 and CD40–CD154 pathways in a fully MHC-mismatched model using the same T-bet−/− recipients (4). All these reports point to an important role of inflammatory cytokines in alloimmune responses.

IL-6 production is closely associated with an inflammatory response. Due to its strong anti-apoptotic properties, IL-6 can increase the effector/memory T-cell population (11). IL-6 has also been shown to skew Th cell differentiation toward a Th17 phenotype and to prevent the development of Tregs (12). Our in vitro data presented here indicate that IL-6 deficiency promotes Treg generation and limits T-cell proliferation. However, in vivo data from transplanted animals showed only a significant reduction in effector/memory population in IL-6−/− recipients, but no significant increase of Tregs. This might be due to differences in antigen presentation, or to the niches in which the antigen was encountered.

Using T-bet−/− mice, which lack Th1 responses, our group previously demonstrated a pathological role of IL-17 in allograft rejection (3,4). The role of IL-17 under normal conditions, in which the Th1 response dominates, remains unclear. In this study, IFN-γ and IL-6 were the two major cytokines that were expressed at high levels during allograft rejection. However, despite high IL-6 production, no notable IL-17 production of IL-17 was detectable in WT B6 recipients. IL-6 deficiency induced upregulation of the Th1 response, but did not affect IL-17 production. It has been reported by other groups that IL-6 modulates the Th1/Th2 balance toward Th2 by promoting autocrine IL-4 production, which further enhances Th2 differentiation (24). These findings indicate that IL-6 plays a very important role in alloimmune responses, independent of IL-17.

T-cell activation results from the combined effect of various signals provided by antigen, costimulatory molecules and cytokines. Costimulatory blockade has been shown to be effective in tolerance induction in many different animal models. A mutant version of CTLA4Ig, belatacept (two amino acid mutations in the extracellular domain), has recently been investigated in human clinical trials to prevent transplant rejection and to treat rheumatoid arthritis (25). In our in vivo experiments, we found that CTLA4Ig can significantly inhibit IFN-γ production by both CD4 and CD8 T cells, particularly in IL-6−/− recipients (especially IFN-γ produced by CD8 T cells), but has no significant effect on IL-6 production. Although systemic IL-6 deficiency induced slight elevation of IFN-γ production in untreated animals, it still contributed to the achievement of long-term graft acceptance in CTLA4Ig-treated mice.

Neutralization of IL-6 with an anti-IL-6 monoclonal antibody can be effective by abolishing the effects of IL-6 produced by both the recipient and by donor cells within the graft. This was confirmed by the neutralization experiment. IL-6 is a very important mediator of inflammation during allograft rejection. In renal transplant patients, delayed graft function was associated with high urine levels of IL-6, which declined once graft function recovered. Steroids are strong anti-inflammatory reagents, and inhibition of IL-6 production could be one of the mechanisms by which steroids prevent allograft rejection (26). In the immunosuppressive protocol for heart or heart–lung transplantation (27), earlier steroid administration is preferable, as it might prevent the production of inflammatory cytokines including IL-6, but is unable to inhibit the detrimental effects of already secreted cytokines. Neutralization of these cytokines by monoclonal antibodies could be a useful strategy in this scenario. Based on our findings, the main effects of IL-6 deficiency on alloimmune responses are the limitation of effector T-cell expansion and the promotion of Treg function as well as their distribution within the graft.


This work was supported by grants from the National Institutes of Health: R01 AI070601 (J.I.), R01AI043619 (J.I.), RO1AI-051559 (M.H.S.), R01AI-37691 (L.A.T. and M.H.S.) and PO1-AI41521 (L.A.T. and M.H.S.). O.B. was funded by Research Grants from the German Research Foundation and the American Society of Transplantation. X.Y. was recipient of American Society of Transplantation Basic Science Faculty Development Grant Award.


The authors of this manuscript have no conflicts of interest to disclose as described by American Journal of Transplantation.