Male BALB/c (H2Kd) and female C57BL/6 (H2Kb) mice were purchased from Vital River (Charles River, China) at the age of 6–8 weeks. The mice were fed with acidified water containing gentamicin for 7 days post-transplantation. All the animal experiments were performed according to the protocols approved by the Institutional Animal Care and Ethics Committee of Peking University First Hospital.

5-Aza was dissolved in dimethyl sulfoxide (DMSO) to prepare a 50-mM stock solution. DZNep was dissolved in sterilized water to prepare a 50-mM stock solution. Both drugs were diluted in phosphate-buffered saline (PBS) before use, so that the final concentration of DMSO was below 2% in all the experiments.

C57BL/6 (H2Kb) mouse bone marrow cells and splenic mononuclear cells were prepared by Ficoll gradient centrifugation. Cell mixtures of 5 × 10⁶ bone marrow cells and 5 × 10⁶ splenic mononuclear cells per mouse were prepared for intravenous injection into BALB/c (H2Kd) mice at about 6 h after BALB/c mice receiving 7.5 Gy ⁶⁰Coγ of total body irradiation (TBI). The mice were divided into five groups: 1) BM group—mice only received donor bone marrow (BM) cells; 2) BM+splenic cells+vehicle group (aGvHD mouse model)—mice received donor bone marrow cells and splenic cells and were treated with vehicle; 3) 5-Aza group—mice received donor bone marrow cells and splenic cells and were treated with 1 mg/kg 5-Aza; 4) DZNep group—mice received donor bone marrow cells and splenic cells and were treated with 0.1 mg/kg DZNep; and 5) combined treatment group—mice received donor bone marrow cells and splenic cells and were treated with 1 mg/kg 5-Aza and 0.1 mg/kg DZNep. Each treatment was applied on days +3, +5, +7, +9, +11, +13, +15, and +17 intraperitoneally. To explore appropriate drug dosages, we examined DZNep at 0.1 mg/kg with different dosages of 5-Aza at 2 and 1 mg/kg (7, 15). The clinical scores of aGvHD were assessed every 3 days, including weight loss, activity, posture, skin integrity, and fur texture (16). Splenic chimerism, viability, and T-cell differentiation were assessed by flow cytometry on day 8 after transplantation. Histopathology, bone marrow smear, and peripheral blood cell counts were assessed on day 21 after transplantation.

Cells were cultured in the RPMI medium supplemented with l-glutamine (4 mM), penicillin (100 U/ml), and streptomycin (100 μg/ml). Splenic mononuclear cells of C57BL/6 mice were stimulated in the presence of anti-mouse CD3 antibody (2 μg/ml) and CD28 antibody (1 μg/ml). After 12 h stimulation, vehicle (0.001% DMSO-PBS), 0.5 μM 5-Aza, 0.02 μM DZNep, and 0.5 μM 5-Aza combined with 0.02 μM DZNep were added into the culture medium. To determine the proper dosage for the combination treatment of 5-Aza and DZNep that does not show significant cytotoxic effects, we examined 5-Aza at 0.5 μM with different concentrations of DZNep at 0.02, 0.04, and 0.08 μM (17, 18). After another 60 h treatment, the viability and proliferation of splenic T cells were then assessed by flow cytometry. Viability was evaluated using the percentage of cells that were double negative for annexin V and 7-amino actinomycin (7-AAD). Proliferation was evaluated by carboxyfluorescein diacetate succinimidyl ester (CFSE) dilution. T-cell differentiation was examined by flow cytometry. Cells treated with the vehicle were used as the reference control.

On day 21 after transplantation, the liver, lung, and colon tissues of the aGvHD mouse model were fixed in 10% neutral-buffered formalin overnight at room temperature. Fixed tissues were embedded in paraffin, cut into 4- to 6-μm tissue sections, and then stained with hematoxylin and eosin (H&E) using the standard protocols for microscope analysis. The pathological changes were evaluated using a previously published grading scale (19).

On day 8 after transplantation, the spleens of the aGvHD mouse model were crushed through 70-μm screens and the erythrocytes lysed. Single-cell suspensions were incubated in PBS containing the following fluorescently labeled antibodies: CD3 (PE), CD4 (APC/Cy7), and CD8 (BV510) (all from BioLegend, San Diego, CA, USA) at 4°C for 20 min. Apoptosis was assessed using PE Annexin V Apoptosis Detection Kit I (BD Biosciences, Franklin Lakes, NJ, USA). Chimerism was assessed by H2Kb (FITC; BioLegend) positive population. For intracellular cytokine staining, the cells were stimulated with cell stimulation cocktail plus protein transport inhibitors (BioLegend) for 4 h. Cells were then harvested and washed. After surface staining of CD3, CD4, and CD8, the cells were fixed and permeabilized using a fixation and permeabilization kit (BD IntraSure Kit) and stained for the intracellular cytokines IFNγ (PerCP) and IL4 (APC) (both from BioLegend). The CD4⁺CD25⁺Foxp3⁺ population assay was carried out using a mouse regulatory T cell (Treg) staining kit (eBioscience, San Diego, CA, USA). Data were acquired on a FACS Canto II (BD) system and were analyzed using FlowJo software.

Recipient mouse spleens were harvested on day 8 after transplantation. Single-cell suspensions were prepared as described above. Total RNA was extracted using the TRIZOL reagent (Ambion, Austin, TX, USA) according to the manufacturer’s instruction. A complementary DNA (cDNA) library was prepared using 2.5 μg messenger RNA (mRNA) and the RevertAid First-Strand cDNA Synthesis Kit (Thermo Fisher, Waltham, MA, USA). Quantitative PCR (qPCR) was performed using the SYBR Green PCR kit (Sigma) on the ABI Prism 7500PCR system. The C _t values of GAPDH were used as housekeeping controls, and the relative expression of a gene was calculated using the 2^−△△Ct method. The primers used for qPCR are summarized in Table 1 .

Recipient mouse spleens were harvested and processed on day 8 after transplantation, as described above. Total RNA was extracted from the tissues using TRIzol (Invitrogen, Carlsbad, CA, USA) according to the instruction in the manual. About 60 mg of tissues was ground into powder by liquid nitrogen in a 2-ml tube, followed by homogenization for 2 min and resting horizontally for 5 min. The mixture was centrifuged for 5 min at 12,000 × g at 4°C, and then the supernatant was transferred into a new Eppendorf (EP) tube with 0.3 ml chloroform/isoamyl alcohol (24:1). The mixture was shaken vigorously for 15 s and then centrifuged at 12,000 × g for 10 min at 4°C. After centrifugation, the upper aqueous phase where the RNA remained was transferred into a new tube with an equal volume of supernatant of isopropyl alcohol and then centrifuged at 13,600 rpm for 20 min at 4°C. After removing the supernatant, the RNA pellet was washed twice with 1 ml 75% ethanol, and then the mixture was centrifuged at 13,600 rpm for 3 min at 4°C to collect residual ethanol, followed by air drying the pellet for 5–10 min in the biosafety cabinet. Finally, 25–100 µl of diethylpyrocarbonate (DEPC)-treated water was added to dissolve the RNA. Subsequently, total RNA was qualified and quantified using a NanoDrop and Agilent 2100 Bioanalyzer (Thermo Fisher Scientific, Waltham, MA, USA).

Oligo(dT)-attached magnetic beads were used to purify mRNA. Purified mRNA was fragmented into small pieces with a fragment buffer at an appropriate temperature. Then, first-strand cDNA was generated using random hexamer-primed reverse transcription, followed by second-strand cDNA synthesis. Afterwards, A-Tailing Mix and RNA Index Adapters were added by incubating to end repair. The cDNA fragments obtained from the previous step were amplified by PCR. The products were purified using Ampure XP beads and then dissolved in ethidium bromide (EB) solution. The product was validated on the Agilent Technologies 2100 Bioanalyzer for quality control. Double-stranded PCR products from the previous step were heat denatured and circularized by the splint oligo sequence to obtain the final library. The single-strand circle DNA (ssCirDNA) was formatted as the final library. The final library was amplified with phi29 to create DNA nanoballs (DNBs), which had more than 300 copies of one molecule. DNBs were loaded into the patterned nanoarray, and single-end 50-bp reads were generated on the BGIseq500 platform (BGI, Shenzhen, China).

The sequencing data were filtered with SOAPnuke (v1.5.2) (20) by removing 1) reads containing sequencing adapters; 2) reads whose low-quality base ratio (base quality ≤5) was more than 20%; and 3) reads whose unknown base (N′ base) ratio was more than 5%. Thereafter, clean reads were obtained and stored in FASTQ format. The clean reads were mapped to the reference genome using HISAT2 (v2.0.4) (21). Bowtie2 (v2.2.5) (22) was applied to align the clean reads to the reference coding gene set, and then the expression level of a gene was calculated using RSEM (v1.2.12) (23). A heatmap was drawn with pheatmap (v1.0.8) according to the gene expressions in different samples. Essentially, differential expression analysis was performed using DESeq2 (v1.4.5) with a Q value ≤0.05. To obtain insights into the change of phenotype, Gene Ontology (GO; http://www.geneontology.org/) and Kyoto Encyclopedia of Genes and Genomes (KEGG; https://www.kegg.jp/) enrichment analysis of the annotated differentially expressed genes (DEGs) was performed with Phyper (https://en.wikipedia.org/wiki/Hypergeometric_distribution) based on the hypergeometric test. Significant levels of the terms and pathways were corrected using the Q value with a rigorous threshold (Q value ≤0.05) by Bonferroni (24).

Data were reported as the mean ± SD. Survival data were compared using the log-rank (Mantel–Cox) test. Comparisons among various groups were performed using one-way analysis of variance. Statistical analysis was performed using GraphPad Prism, version 5.0. A p < 0.05 was considered significant.

To determine the proper dosage for the combination treatment of 5-Aza and DZNep, we examined the effect of 0.5 μM 5-Aza combined with different concentrations of DZNep (0.02, 0.04, and 0.08 μM) on cell viability and proliferation (see Materials and Methods). 5-Aza at 0.5 μM and 0.04 or 0.08 μM DZNep showed cytotoxic effects on viability or proliferation ( Supplementary Figure S1 ). Only the combination of 0.5 μM 5-Aza and 0.02 μM DZNep exhibited a small impact on both cell viability (85.50 ± 1.84% vs. 92.20 ± 4.75%, p = 0.12) ( Figure 1A ) and proliferation (73.00 ± 5.23% vs. 65.95 ± 5.30%, p = 0.47) ( Figure 1B ) compared with the control groups, indicating no significant cytotoxic effects. Thus, this concentration was then used in subsequent experiments.

T helper 1 (Th1) and T cytotoxic type 1(Tc1) cells are the main effector cells of aGvHD. Th1 is characterized as CD4⁺IFNγ⁺ T cells, while Tc1 is characterized as CD8⁺IFNγ⁺ T cells. As shown in Figures 2A–C , among the four treatment groups (control, 5-Aza, DZNep, and 5-Aza and DZNep), the combined treatment of 5-Aza and DZNep could significantly reduce the proportions of Th1 cells (32.20 ± 2.01% vs. 26.31 ± 2.25%, p = 0.028) and Tc1 (28.37 ± 0.84% vs. 20.03 ± 2.35%, p = 0.004) compared with the control group. Single treatment of 5-Aza or DZNep could decrease the proportions of Th1 and Tc1, although the reduction did not show statistical significance ( Figures 2B, C ).

We also examined the proportions of Th2 and Tc2. As shown in Figures 2D–F , the combination therapy of 5-Aza and DZNep increased the proportions of CD4⁺IL4⁺ Th2 cells (3.18 ± 0.51% vs. 0.72 ± 0.10%, p = 0.045) and CD8⁺IL4⁺ Tc2 cells (2.86 ± 0.13% vs. 1.62 ± 0.04%, p < 0.0001). It is worth noting that a single treatment of 5-Aza or DZNep could also increase the proportion of Th2 cells, although to a lesser extent than the combination treatment (p < 0.05; Figure 2E ). Similar results were observed for Tc2, but were not significant ( Figure 2F ).

The qPCR experiment further confirmed the altered proportions of the differentiated T-cell subsets ( Figures 2G, H ). Compared with the control group, there was a reduction in the expression of IFNγ (p = 0.043; Figure 2G ) and an increase in the expression of IL4 in the two-drug combination group (p = 0.021; Figure 2H ).

To assess the impact of combination therapy on Tregs in vitro, we examined the proportion of CD4⁺CD25⁺Foxp3⁺ cells by flow cytometry ( Figures 2I, J ). The percentage of Tregs in CD4⁺ T cells was found lowest in the control group (3.46 ± 0.55%) and highest in the 5-Aza and DZNep combination treatment group (5.39 ± 0.66%, p = 0.008). The single treatment of 5-Aza or DZNep did not induce high production of Tregs (not significant; Figures 2I, J ), indicating that 5-Aza combined with DZNep synergistically increased the percentage of Tregs in vitro.

In vitro research demonstrated that combination therapy of 5-Aza and DZNep changed the proportions of different T-cell subsets without significant cytotoxic effects. We then examined its effect in vivo using the aGvHD mouse model. The mouse model was established as described in Materials and Methods. To explore appropriate drug dosages, we examined various dosage combinations (see Materials and Methods). We found that mice that received 5-Aza (1 or 2 mg/kg) combined with 0.1 mg/kg DZNep showed significantly improved survival after transplantation ( Supplementary Figure S2A ), while 2 mg/kg 5-Aza combined with 0.1 mg/kg DZNep caused significant cytopenia (white blood cells and platelets; Supplementary Figure S2B ) compared with treatment of 1 mg/kg 5-Aza combined with 0.1 mg/kg DZNep. These results suggest that the combination of 1 mg/kg 5-Aza and 0.1 mg/kg DZNep can effectively prolong the survival of aGvHD mice without causing significant cytopenia. Therefore, these concentrations were then used in subsequent in vivo experiments.

We observed that, after day 17 of transplantation, aGvHD mice developed significant symptoms, such as weight loss, ruffled fur, posture changes, and diarrhea, as indicated by the GvHD score ( Figure 3A ). Single treatment of DZNep did not alleviate the GvHD-related symptoms or survival, single treatment of 5-Aza could mildly improve GvHD survival, although was not significant (p = 0.237), while 5-Aza combined with DZNep significantly alleviated aGvHD symptoms and improved survival (p = 0.0001; Figures 3A, B ). Importantly, most aGvHD mice treated with a combination of 5-Aza and DZNep survived on day 45 after transplantation (survival rate = 70%, p = 0.0002), while no mice survived in the control group. Interestingly, we found that the survival rate of the group that received single treatment of 5-Aza was 40% (p = 0.007, compared with the control), while the group given single treatment of DZNep showed no changes in the survival rate (p = 0; Figure 3B ). Compared with the single treatment of 5-Aza, the combination of 5-Aza and DZNep tended to enhance survival, although the difference was not significant (p = 0.179). These results demonstrate that aGvHD significantly affected the survival of mice and that the combination of 5-Aza and DZNep can significantly improve survival in aGvHD.

Furthermore, histology examination on day 21 after hematopoietic stem cell transplantation (HSCT) showed that the combined treatment of 5-Aza and DZNep significantly alleviated the inflammation in the liver, colon, and lung of recipient mice ( Figure 4 ). Interestingly, single treatment of 5-Aza, but not DZNep, also significantly alleviated the inflammation in the liver and lung ( Figures 4B, C ), but to a lesser extent than the combined treatment. These results suggest that in vivo epigenetic agents, especially the combination of 5-Aza and DZNep, can alleviate the development of aGvHD.

To confirm the successful implantation and the validity of the results, we first evaluated the splenic cell viability and chimeric rate on day +8 by flow cytometry. Viability was assessed using the percentage of 7-AAD and annexin V double-negative cells, and the chimeric rate was assessed using the percentage of H2Kb-positive cells. There was no significant difference in the cell viability for the two-drug combination group and the aGvHD group (75.22 ± 3.69% vs. 82.46 ± 2.89%, respectively, p = 0.16) ( Supplementary Figure S3A ). The chimeric rate of the two-drug treatment group was not significantly different from that of the control group and the single-drug treatment groups ( Supplementary Figure S3B ). Additionally, we evaluated the peripheral blood cell count and the BM on day 21 after transplantation. Peripheral blood cell counts and BM smear showed that each group appeared to have similar neutrophil, platelet, and lymphocyte counts and BM hyperplasia on day 21 after transplantation, indicating successful implantation ( Supplementary Figure S4 ).

During the pathogenesis of aGvHD, the donor-derived T cells can further differentiate into alloreactive effector T cells after transplantation (4). Based on whether they secrete the cytokine IFNγ or IL4, donor-derived T cells can be further subdivided into the Th1and Th2 subtypes and the Tc1 and Tc2 subtypes. We collected the spleens of mice on day 8 after aGvHD induction and examined the T-cell subtypes. Compared with the control group, the proportions of Th1 and Tc1 cells in the 5-Aza and DZNep combination group were significantly decreased (10.24 ± 2.25% vs. 24.73 ± 4.35% and 16.76 ± 2.26% vs. 29.70 ± 5.33%, p = 0.012 and 0.045, respectively), while the decreases in the proportions of Th1 or Tc1 in the groups receiving single-drug treatment of 5-Aza or DZNep were not significant ( Figures 5A–C ). Meanwhile, the proportions of Th2 and Tc2 cells were increased in the combination treatment group (1.02 ± 0.23% vs. 0.46 ± 0.08% and 0.92 ± 0.17% vs. 0.24 ± 0.07%, p = 0.047 and 0.006, respectively). Meanwhile, significant differences were also observed in the single-treatment groups when compared with the control group ( Figures 5D–F ). These results demonstrate that in vivo application of the two-drug combination therapy altered the proportions of the T-cell subsets in a much more significant extent than the single-drug therapy, suggesting the larger impact of combination therapy on the differentiation and/or proliferation of the different T-cell subtypes.

qPCR was performed to further validate the specific effector subsets that were involved in the process. Compared with the aGvHD control group, there was a reduction in the expression of IFNγ and an increase in the expression of IL4 in the two-drug combination group (p = 0.035 and 0.047, respectively), but not in the single-drug treatment groups (p = 0.113 and 0.101, p = 0.198 and 0.999 for 5-Aza and DZNep, respectively) ( Figures 5G, H ).

Hypomethylating agents may induce immune suppression by converting Foxp3⁻ donor T cells to Foxp3⁺ Tregs (25). We then examined the Tregs in the spleen lymphocytes of the aGvHD mouse model on day 8 after transplantation. We found a significant increased in the combined 5-Aza and DZNep group compared to the aGvHD control group (4.29 ± 0731% vs. 0.83 ± 0.19%, p = 0.001) ( Figures 5I, J ). Interestingly, single treatment of 5-Aza, but not DZNep, also increased the proportions of CD4⁺CD25⁺Foxp3⁺ cells (3.75 ± 0.77% vs. 0.83 ± 0.19%, p = 0.007).

To explore the molecular mechanism and biological pathways of the two-drug combination treatment, RNA-seq was performed to identify the DEGs between the two-drug combination treatment group (5-Aza combined with DZNep) and the aGvHD control group. Overal1, 474 DEGs were identified, including 262 upregulated genes and 212 downregulated genes ( Figure 6A ).

GO functional enrichment analysis showed that the DEGs were statistically significantly enriched in 439 molecular function (MF) terms, 457 cellular component (CC) terms, and 445 biological process (BP) terms. The most enriched BP terms were cellular process (n = 355), biological regulation (n = 300), and regulation of biological process (n = 286). The most enriched CC terms included cell (n = 355) and cell part (n = 351). The most enriched MF terms were binding (n = 330), catalytic activity (n = 128), and molecular function (n = 65) ( Figure 6B ).

Furthermore, functional annotation of the 474 DEGs using a major public pathway-related database (KEGG) revealed enriched signaling pathways, including the most enriched cytokine–cytokine receptor interaction signaling pathway [n = 19, enrichment ratio = 0.06, false discovery rate (FDR) = 0.009] ( Figure 6C ). The RNA-seq results showed that the expressions of the Th1 and Tc1-related cytokines, such as IFNγ, IL21, and IL12, were reduced in the group that received 5-Aza combined with DZNep. Meanwhile, the cytokine receptor of IL13 (IL13RA) was upregulated. Genes that regulated the production of Tregs, such as TGF-β and Foxp3, were also upregulated. The chemokine signaling pathway was another enriched pathway (n = 13, enrichment ratio = 0.06, FDR = 0.014). We found that CCR7 and its ligand CCL19 were significantly downregulated. Other enriched pathways included vitamin digestion and absorption, retinol metabolism, etc. ( Figure 6C ).