BALB/c and C57BL/6 male mice (6–8 weeks old, male) were obtained from Guangdong Medical laboratory Animal Center (Guangdong, China) while Rag1^−/− mice (B6 background) were purchased from Jackson Laboratory (Bar Harbor, ME, USA). All mice were housed under a specific pathogen-free condition. All experiments were performed according to the Chinese national guidelines for the Care and Use of Laboratory Animals and approved by the Institutional Animal Care and Use Committee of Guangdong Provincial Academy of Chinese Medical Sciences (Ref. No: 201).

Skin donors were 7- to 8-week-old wild-type BALB/c mice (male), while skin graft recipients were 7- to 8-week-old-male C57BL/6 mice. Round-shaped full-thickness trunk skin with an approximate size of 10 mm² was transplanted to the dorsal flank area of a recipient mouse and secured with a bandage of Band-Aid (Johnson Johnson, New Brunswick, NJ, USA). The bandage was removed 8 days after transplantation. Skin allograft rejection was monitored daily and defined as graft necrosis greater than 90%, as described in our previous publication (18).

Mice were randomly divided into control groups and groups treated with emodin (10 mg/kg body weight), cyclosporine (CsA: 20 mg/kg body weight), and emodin plus CsA for four consecutive weeks or until graft rejection/sample collection. CsA and emodin (Sigma) were prepared with saline and sodium carboxymethyl cellulose (CMC-Na; Sigma), respectively. Control groups received CMC-Na orally. Our primary data showed that oral administration of CMC-Na did not alter allograft rejection compared to untreated groups (unpublished observation). Emodin was also administered orally while CsA was given through intraperitoneal injection. To deplete Tregs, recipient mice were treated i.p. with anti-CD25 Ab (Clone: PC61, eBioscience) at 0.2 mg on days 0, 3, 6, and 10 or anti-CD122 Ab (Clone: TMβ1, eBioscience) at 0.1 mg on days 0, 7, and 14 posttransplantation as described previously (19, 20).

Allografts were fixed with 4% paraformaldehyde for 24 h and then embedded in paraffin. Tissues were cut into 3-μm thick sections and placed on slides. The sections were then used for H&E and immunohistochemistry staining. For immunohistochemistry, slides with sections were incubated with primary anti-CD3 antibody (1:100, Abcam) at 4°C overnight, then secondary antibody HRP-anti-Rabbit IgG (Maxim), followed by diaminobenzidine color development. For quantitative analyses, slides were imaged at a magnification of 200×. The area of inflammation/infiltration and integrated optical density (IOD) of CD3 were measured using ImagePro plus 6 software.

Draining lymph node (LN) and spleen cells were harvested and stained with anti-CD4-FITC (Clone H129.19)/CD8-FITC (Clone 53-6.7), CD11c-PE (Clone HL3), CD44-V450 (Clone IM7), CD62L-APC (Clone MEL-14), CD80-FITC (Clone 16-10A1), CD86-FITC (Clone GL1), and anti-CD122-PE antibodies (Clone TM-Beta 1) (all from BD Biosciences). To determine intracellular FoxP3 expression, cells were fixed and permeated according to the protocol of Foxp3/Transcription Factor Fixation/Permeabilization Concentrate and Diluent Kit (eBioscience). Then, cells were stained with anti-FoxP3-APC antibody (Clone FJK-16s, eBioscience) and finally analyzed by a flow cytometer (FACSCalibur, BD Biosciences). To purify CD4⁺CD25⁺ and CD8⁺CD122⁺ Tregs for adoptive transfer experiments, spleen cells were stained with anti-CD4-PE (Clone RM4-5) and anti-CD25-FITC (Clone 3C7) or, separately, anti-CD8-PE (Clone 53-6.7) and anti-CD122-FITC (Clone TM-Beta 1) Abs (BD Biosciences). CD4⁺CD25⁺ or CD8⁺CD122⁺ cells were then sorted out by FACSAria III (BD Biosciences). The purity of the sorted cells was typically >95%.

Recipient mice were pulsed intraperitoneally with 0.1 mg of 5-ethynyl-2′-deoxyuridine (EDU, RIBOBIO) in PBS 10 days after transplantation. 24 h later, spleen and LN cells were isolated and single-cell suspensions were prepared for EDU detection using Cell-Light™ Apollo^®488 Stain Kit (RIBOBIO) according to the manufacturer’s instructions. Then, cells were stained for cell surface markers with anti-CD8-APC-Cy7 (Clone 53-6.7), CD44-V450 (Clone IM7), and CD62L-APC Abs (Clone MEL-14) (all from BD Biosciences). To detect cell apoptosis, cells were stained for the same surface markers and then Annexin V-FITC according to the protocol of Annexin V-FITC Apoptosis Detection Kit (BD Pharmingen). Cells finally were analyzed via a flow cytometer (FACSCalibur).

FACS-sorted CD3⁺ T cells (2 × 10⁵/well), derived from B6 mice, were cultured with irradiated Balb/c spleen cells (2 × 10⁵/well) in the absence or presence CsA (2.5 µg/ml) or emodin (50 µM) in 96-well plates (Corning Costar) in complete RPMI 1640 medium (10%FCS, 2 mM glutamine, 100 U/ml penicillin, and 100 µg/ml streptomycin) for 3 and 5 days, respectively. Cells were pulsed with [³H]-thymidine at 0.5 μCi per well for last 8 h. They were finally harvested and analyzed by a Scintillation counter (Perkin Elmer, Wellesley, MA, USA).

Spleen cells derived from BALB/c mice were harvested and their red blood cells were lysed. B cells in splenocytes were first depleted using B220 microbeads (Miltenyi Biotec) via negative selection. Cells (1 × 10⁶/sample) then were stained with diluted serum (1/10) from naïve or transplanted C57BL/6 mice. They were further incubated with PE-anti-mouse IgM or FITC-anti-mouse IgG (Biolegend). The mean fluorescence intensity was determined by a flow cytometer (FACSCalibur, BD Biosciences).

Cultured T cells were lysed in RIPA buffer (50 mM Tris pH 7.5, 150 mM NaCl, 1% Triton X-100, and 5 mM ethylenediaminetetraacetic acid). Cell protein was extracted using RIPA buffer and protein concentration was measured using BCA Kit (Pierce, IL, USA). Samples were run on 10% SDS-PAGE gels and transferred onto a PVDF membrane. TBST with 5% milk was used to block the membrane, which was then incubated with a primary antibody anti-P70S6K or anti-phospho-P70S6K (1:1,000; Cell Signaling Technology) at 4°C overnight. After the incubation, membranes were washed and incubated with a secondary antibody, HRP-conjugated anti-rabbit IgG (1:10,000, Abbkine), for 1 h. GAPDH (1:1,000, Cell Signaling Technology) was also used for loading controls. Finally, signals were detected by an ECL method (Promega) and analyzed by Image J Program software.

Comparisons of the means were performed using Student’s t-test and one-way ANOVA. Data were presented as the mean ± SD and analyzed through GraphPad Prism 6 (GraphPad Software, La Jolla, CA, USA). The analysis of graft survival was conducted using Kaplan–Meier method (log-rank test). A value of P < 0.05 was considered statistically significant.

To study the effects of emodin on allograft rejection, C57BL/6 mice received a skin graft from a donor Balb/C mouse and were then treated with emodin and/or CsA. As shown in Figure 1, we found that emodin significantly prolonged skin allograft survival compared to the control [median survival time (MST) = 24 vs. 13 days, P < 0.05] while allograft survival time of the recipient mice treated with CsA was also longer than that of control group (MST = 25 vs. 13 days, P < 0.05). More importantly, combined treatments with emodin and CsA further extended skin allograft survival compared to the treatment with either CsA or emodin alone (MST = 36 vs. 25 or 24 days, both P < 0.05).

Since emodin suppressed allograft rejection, we asked whether emodin would reduce cellular infiltration in an allograft. Graft-infiltrating cells were analyzed by H&E and immunohistochemical staining 10 days after transplantation. H&E staining showed obvious cellular infiltration in a skin allograft of the recipients without any treatment while much less cellular infiltration was observed in the recipients treated with emodin, CsA or both of them (Figures 2A,C). Similarly, immunohistochemistry revealed an obvious decrease in CD3⁺ T cell infiltration in a skin allograft of the recipients treated with emodin and/or CsA compared to that of control recipients (Figures 2B,D). These data suggest that emodin indeed suppresses allograft rejection and ameliorates alloreactive CD3⁺ T cell infiltration in the skin allografts.

To determine whether emodin would control effector CD8⁺ T cells (Teff) in vivo, draining LN and spleen cells from emodin- or CsA-treated recipient mice were isolated 10 days after transplantation and analyzed via EDU-staining and FACS analysis. As represented by Figure 3A, both emodin and CsA obviously decreased the percentages and absolute numbers of CD8⁺CD44^highCD62L^low effector T cells in both LNs and spleens of the recipients. Furthermore, we found that either emodin or CsA alone suppressed the proliferation of CD8⁺CD44^highCD62L^low T cells in the recipient mice while emodin plus CsA further inhibited their proliferation compared to either emodin or CsA alone (Figure 3B). As a control, we isolated spleen cells from naïve mice without skin transplantation and observed that there was no any difference in CD8⁺CD44^highCD62L^low T cell numbers between untreated and emodin-treated groups (data not shown). On the other hand, emodin did not promote the apoptosis of effector CD8⁺CD44^highCD62L^low and CD4⁺CD44^highCD62L^low T cells 10 days after transplantation (Figures 3C,D). We also demonstrated that emodin did not induce the apoptosis of total CD4⁺ and CD8⁺ T cells 20 days posttransplantation (Figure S1 in Supplementary Material). Thus, our data suggest that emodin hinders effector CD8⁺ T cell expansion/proliferation, but does not induce T cell apoptosis, implying that a treatment with this dose of emodin is not cytotoxic.

Regulatory T cells play an important role in long-term transplant survival or tolerance. Thus, we examined if emodin would suppress allograft rejection by inducing Tregs. Draining LN and spleen cells were isolated 10 days after allogeneic skin transplantation, and CD4⁺Foxp3⁺ Tregs were enumerated by flow cytometry. As shown in Figure 4, emodin significantly increased the percentages and absolute numbers of CD4⁺Foxp3⁺ Tregs in draining LNs while CsA did the opposite. Furthermore, a reduction in LN Treg numbers resulted from CsA treatment was totally reversed in the recipients treated with both emodin and CsA (Figure 4). On the other hand, there was no markedly variance in the frequencies and absolute numbers of splenic Tregs between all groups. Similar findings were also seen 20 days after skin allotransplantation (data not shown). These data suggest that emodin promotes CD4⁺Foxp3⁺ Treg generation mainly in the draining LNs of recipient mice.

Our previous studies demonstrated that CD8⁺CD122⁺ Tregs were more potent in inhibiting T cell proliferation and transplant rejection than CD4⁺CD25⁺ Tregs (21). Hence, we asked whether emodin would also exert its suppressive effects on allograft rejection through inducing CD8⁺CD122⁺ Tregs. We measured the percentages and absolute numbers of CD8⁺CD122⁺ Tregs in the LNs and spleens of recipient mice via flow cytometry 10 days after allogeneic skin transplantation. We demonstrated that both emodin and CsA significantly increased the percentages of CD8⁺CD122⁺ Tregs in both LNs and spleens of the recipients (Figure 5), suggesting that emodin generally promotes the development of CD8⁺CD122⁺ Tregs in vivo. Interestingly, emodin increased the absolute numbers of CD8⁺CD122⁺ Tregs in LNs, but not spleens, of the recipients.

To determine whether emodin also enhances the suppressive function of Tregs, CD4⁺CD25⁺ or CD8⁺CD122⁺ Tregs were isolated from recipient mice treated without or with emodin and/or CsA. These Tregs were then transferred to lymphocyte-deficient Rag1^−/− mice that received syngeneic T cells as well as a skin allograft. As shown in Figure 6A, adoptive transfer of conventional T cells to Rag1^−/− recipients caused their rejection of skin allografts while control Rag1^−/− recipients did not reject the allografts. Transfer of both control CD4⁺CD25⁺ Tregs and T cells extended allograft survival compared to that of T cells alone. However, transfer of CD4⁺CD25⁺ Tregs derived from emodin-treated, but not CsA-treated, recipient mice resulted in longer allograft survival than that of control CD4⁺CD25⁺ Tregs derived from control recipients, suggesting that emodin enhances the suppressive function of CD4⁺CD25⁺ Tregs. Transfer of CD4⁺CD25⁺ Tregs derived from CsA- and emodin-treated recipients also caused longer allograft survival than that of the Tregs from the recipients treated with CsA alone. On the other hand, CD8⁺CD122⁺ Tregs derived from the recipient mice treated with either emodin or CsA prolonged allograft survival, but only as effectively as control CD8⁺CD122⁺ Tregs derived from control recipients (Figure 6B), indicating that emodin or CsA does not alter the suppressive capacity of CD8⁺CD122⁺ Tregs.

Given that emodin induced Tregs in our animal model, we asked whether the effects of emodin on allograft survival were dependent on the Tregs. C57BL/6 mice received a skin graft from a BALB/c mouse and were treated with emodin or CsA. Tregs in the recipients then were depleted using anti-CD25 (PC61) or anti-CD122 (TMβ1) Ab. As shown in Figure 7, depleting CD25⁺ Tregs largely reversed skin allograft survival prolonged by emodin while depleting CD122⁺ Tregs partially abrogated the allograft survival extended by emodin. In contrast, depletion of either CD25⁺ or CD122⁺ Tregs did not significantly shorten allograft survival induced by CsA. As control experiments, isotype control Abs did not alter allograft survival (data not shown). Our results suggest that Tregs contributes to allograft survival induced by emodin.

Given that anti-donor antibodies play a role in allograft rejection, we examined whether emodin also reduced donor-specific antibodies. We then measured IgG and IgM in the serum of B6-recipient mice treated with CsA and/or emodin via FACS analyses using BALB/c splenocytes as target cells. As shown in Figure 8, either emodin or CsA significantly lowered both IgG and IgM levels 2 and 3 weeks after transplantation while administration of both CsA and emodin further decreased IgG and IgM levels compared to the treatment with CsA or emodin alone. Our data indicate that emodin indeed inhibits alloantibody production.

To determine if emodin affects DC maturation posttransplantation, draining LN and spleen cells were isolated 10 days after allogeneic skin transplantation, and CD11c⁺CD80⁺ and CD11c⁺CD86⁺ DCs were enumerated by flow cytometry. As shown in Figure 9, treatment with either CsA or emodin reduced CD11c⁺CD86⁺ cell numbers in LNs of the recipient mice. However, combined treatments with both CsA and emodin, but not treatment with CsA or emodin alone, significantly decreased CD11c⁺CD86⁺ cell numbers in spleens of the mice (Figure 9A). On the other hand, either CsA or emodin reduced CD11c⁺CD80⁺ cell numbers in spleens, but not LNs, of the recipients (Figure 9B). These findings suggest that emodin reduces CD11c⁺CD86⁺ DC numbers in LNs of recipient mice while decreasing CD11c⁺CD80⁺ DC numbers in spleens of the recipients.

Since we found that emodin suppressed allograft rejection and T cell infiltration in an allograft, we then examined whether emodin would also inhibit T cell proliferation in vitro. To determine the effects of emodin on alloreactive T cell proliferation in vitro, one-way MLR was set up using irradiated Balb/C splenocytes as stimulators and B6-derived T cells as responders. As shown in Figure 10, either emodin or CsA inhibited T cell proliferation 3 (Figure 10A) or 5 days (Figure 10B) after MLR culture. Moreover, combined treatments with both emodin and CsA further suppressed T cell proliferation compared to the treatment with emodin or CsA alone.

Previous experiments using rapamycin, an inhibitor of mTOR (22), have proved that mTOR signal transduction suppresses the development and function of Tregs and that rapamycin promotes CD4⁺Foxp3⁺ Treg generation (23). Since our results demonstrated that emodin increased the frequencies of CD4⁺Foxp3⁺ Tregs, we determined whether emodin would also induce Tregs by blocking mTOR signaling. T cells were activated in an MLR in the medium without or with emodin for 48 h, and the phosphorylation of p70S6K protein (P-p70S6K) was measured via western blotting. As shown in Figure 10C, emodin effectively inhibited the phosphorylation of downstream protein p70S6K at Ser411 compared to the control group while CsA only slightly reduced its phosphorylation. These results indicate that emodin likely exerts its effects on Treg generation and T cell proliferation through blocking mTOR signaling pathway.