Anti-CD80/86 antibodies inhibit inflammatory reaction and improve graft survival in a high-risk murine corneal transplantation rejection model

We investigated the effects of anti-CD80/86 antibodies in a murine high-risk corneal transplantation rejection model. A mixed lymphocyte reaction (MLR) assay was conducted with anti-CD80/86 antibodies. Inflammatory cytokine levels in the culture supernatant were measured using an enzyme-linked immunosorbent assay. Interferon (IFN)-γ-producing CD4+ T cell frequencies in the MLR were assessed using flow cytometry. In vivo, high-risk corneal allograft survival and IFN-γ-producing CD4+ T cell frequencies in corneal grafts were assessed with intraperitoneal injection of anti-CD80/86 antibodies compared to phosphate-buffered saline (PBS). RNA-sequencing was performed on corneal grafts 2 weeks post-transplantation. Anti-CD80/86 antibodies significantly decreased T-cell proliferation, IFN-γ+-producing CD4+ T cell frequencies, and IFN-γ, interleukin (IL)-1β, IL-2, IL-10, and tumor necrosis factor-α production in the MLR compared to PBS injection. Intraperitoneal injection of anti-CD80/86 antibodies significantly prolonged corneal graft survival and decreased IFN-γ+-producing CD4+ T cell frequencies compared to PBS injection. Gene set enrichment analysis showed that the gene sets mainly enriched in the control group were related to allograft rejection and inflammatory response compared to PBS injection. Anti-CD80/86 antibodies significantly prolonged corneal graft survival by inhibiting T-cell proliferation and inflammatory response.


Results
Anti-CD80/86 antibodies suppressed CD4 + T cell proliferation. Lymphocyte proliferation in the mixed lymphocyte reaction (MLR) assay was significantly inhibited when anti-CD80/86 antibodies were administrated. Figure 1a shows the representative flow cytometry plots for CD4 + CFSE low T cells in the MLR. On day 3 of incubation, the frequency of CD4 + CFSE low T cells in the MLR was significantly reduced in allogeneic stimulation group with anti-CD80/86 antibodies compared with that in without anti-CD80/86 antibodies group ( Fig. 1b; allogeneic stimulation group, 6.9% ± 0.3%; allogeneic stimulation with anti-CD80/86 antibodies group, 4.4% ± 0.3%; n = 3, P < 0.001). The BrdU assay showed a significant lower opacity density in the allogeneic stimulation with anti-CD80/86 antibodies group compared to without anti-CD80/86 antibodies in allogeneic stimulation and no allogeneic stimulation (Fig. 1c, n = 3; P = 0.007, P = 0.019, respectively). Figure 1d shows the representative flow cytometry plots for IFN-γ-producing CD4 + T cells in MLR.

Anti
(d) Kaplan-Meier survival curves show a significant increase in the survival of grafts with anti-CD80/86 injection compared with that in the control group (log-rank test, n = 6/group, ***P < 0.001). (e) The median graft survival shows that only grafts with anti-CD80/86 injection significantly improved graft survival in high-risk recipients to levels observed in the control group (unpaired t-test, n = 6/group. ***P < 0.001).  Table S1 shows the RNA-Seq read statistics) and gene set enrichment analysis (GSEA) were performed. Figure 6a shows the volcano plot of DEGs in the corneal grafts between the anti-CD80/86-injected and control (PBS injection) groups. Among them, the expression of 672 and 1492 genes was significantly higher in the anti-CD80/86 and control groups, respectively. Figure 6b shows the hierarchical clustering of the identified DEGs between the  IL-2 was downregulated in anti-CD80/86 injection group compared with that in the PBS group (n = 3, **P = 0.003). TNF-α was downregulated in the anti-CD80/86 injection group compared with that in the PBS group (n = 3, ***P < 0.001). There was no difference in IL-10 between the anti-CD80/86 injection group and the PBS group (n = 3, N.S, P = 0.111). TGF-β1 was upregulated in the anti-CD80/86 injection group compared with that in the PBS group (n = 3, **P = 0.003). An unpaired t-test was applied for all statistical analyses. www.nature.com/scientificreports/ . The gene sets with a false discovery rate (FDR) of less than 0.01 were used as a threshold to determine the significance of DEGs. Green dots represent DEGs, and red dots indicate transcripts that did not significantly change between the anti-CD80/86 and PBS injection groups (control). (b) Hierarchical clustering of DEGs between the anti-CD80/86 injection group and the PBS injection group (control). The X axis represents the two compared samples (anti-CD80/86 injection and PBS injection group). The Y axis represents DEGs. The color (from blue to red) represents gene expression intensity from low to high. (c) The bubble plot of Gene Ontology terms. The z-score is assigned to the X axis and the negative logarithm of the P value to the Y axis, as observed in the bar plot (the higher the more significant). The area of the displayed circles is proportional to the number of genes assigned to the terms. Only the immunologically enriched GO terms' labels are displayed. www.nature.com/scientificreports/ anti-CD80/86 injection and control groups. In particular, the expression of Pecam1 (Cd31), Il12a, cd80, Il1β, vascular endothelial growth factor-a (Vegfa), Cd8a, Il2ra, Ctla4, Ifng, Cd4, Ptprc (CD45), Cd86, and Il10 was significantly lower in the anti-CD80/86-injected graft cornea group than in the control group. The Gene Ontology (GO) term enrichment analysis results revealed the significantly enriched biological processes in the anti-CD80/86 injection (Supplementary Table S2) and control groups (Supplementary Table S3). Figure 6c shows a bubble plot of the identified immunological-related GO enrichment terms. We compared the datasets for high-risk corneal grafts between the anti-CD80/86 antibody injection and control groups using the GSEA to identify the activated signaling pathways. Figure 7 shows the GSEA querying hallmark genes (h.all.v6.0.symbols.gmt), showing significant enrichment of Allograft Rejection genes, Inflammatory response genes, Interferon gamma response genes, and TNF-α signaling via NF-κB.

Discussion
In this study, we investigated the suppressive effect of anti-CD80/86 antibodies in the local inflammatory microenvironment in a high-risk corneal rejection model. Our results demonstrated that blocking CD80 and CD86 conferred immunosuppression and prolonged graft survival by inhibiting T-cell proliferation and pro-inflammatory cytokine production. Therefore, anti-CD80/86 antibodies may be useful for suppressing immunological rejection in high-risk cases of corneal transplantation.
The interaction of CD80/CD86 on APCs and CD28 on T lymphocytes is an essential co-stimulatory pathway 20 , which could provide essential signals for T-cell growth, survival, activation, and proliferation 13 . This study showed that the injection of anti-CD80/86 antibodies inhibited T-cell proliferation both in vitro and in vivo. The GO term enrichment results revealed significant enrichment of the biological process T-cell activation and proliferation in the control group compared with that in the anti-CD80/86 injection group. These results suggest that anti-CD80/86 antibodies inhibit immune and inflammatory reactions by inhibiting T-cell proliferation.
CD4 + T cell infiltration and interaction with APCs at the graft site could promote dendritic cell maturation and activate IFN-γ-producing T cells 21 . Higher levels of IFN-γ have been correlated with graft rejection 22 . Ju et al. reported that both CD4 + and CD8 + T cells produce IFN-γ during acute rejection of an allogeneic transplant 23 . This study showed decreased frequencies of IFN-γ-producing CD4 + following treatment with anti-CD80/86 antibodies in the allogeneic stimulation. Antibodies counteracted relative CD80/86 exposure to T cells and, conversely, the activated T cells produced less IFN-γ due to insufficient co-stimulation. Furthermore, the GO term enrichment analysis and GSEA analysis revealed significant enrichment of the biological process IFN-γ response in the control group. Therefore, anti-CD80/86 antibodies could suppress Th1 immune response by reducing IFN-γ responses.
IFN-γ, IL-1β, and TNF-α are pro-inflammatory cytokines produced during immune responses, whereas IL-2 drives T-cell growth and induces Treg differentiation [24][25][26][27] . IL-2 is mainly produced by activated CD4 + , generally in synergy with the level of IFN-γ 28 . We observed a similar elevation of IFN-γ and IL-2 in the supernatant used for MLR assay, and a similar reduction was observed when cells were treated with anti-CD80/86 antibodies. This could be attributed to the comprehensive suppression of T-cell proliferation. IL-1β, a potent proinflammatory cytokine that is crucial for host defense responses 27,29,30 , is mainly secreted by monocytes and macrophages as a part of the innate immune system 31,32 . The level of TNF-α, primarily produced by M1 macrophages 33,34 , is elevated in the rejection of solid organ transplants 35 . The reduced IL-1β and TNF-α expression in this study, achieved using anti-CD80/86 injection, indicates that the blockade of CD80/86 also suppresses the innate immune response. Furthermore, the GO term enrichment analysis also revealed significant enrichment of the biological processes, inflammatory responses, cellular response to cytokine stimulus, and cytokine-mediated signaling pathway in the control group. In the GSEA analysis, inflammatory response and TNF-α signaling were identified as activated pathways. In summary, inflammatory cytokines and relative genes expression were suppressed by the treatment with anti-CD80/86 antibodies in this high-risk rejection model. IL-10 is known for their anti-inflammatory properties, including the ability to inhibit activation and effector function of T cells, shifting T cells towards a tolerogenic phenotype 36,37 . In this study, anti-CD80/86 antibodies decreased the production of IL-10 in vitro in accordance with that findings of a previous study 38 . However, we observed no significant difference in IL-10 mRNA expression in the corneal graft in vivo. On the contrary, as a major IL produced by dendritic cells, IL-12 showed no significant differences in production in the MLR supernatant and mRNA expression in corneal grafts. This discrepancy may be due to the difference in conditions between in vitro and in vivo 39 . The in vitro MLR experiments merely justified the possible mechanism of direct pathway of APCs involved in allogeneic sensitization, but not the indirect pathway. In vivo, the mRNA expression of IL-12 in the grafts of the two groups was similar; this result is consistent with the findings of Torres et al. 40 . Torres et al. assessed IL-12p40 mRNA level in experimental corneal allografts by selectively eliminating dendritic cells, and only observed a transient increase on post-grafting day 3, but not at a later time (days 7-17). Meanwhile, the RNA-Seq results in this study showed a lower level of the DEG Il12a in the grafts with anti-CD80/86 injection compared to PBS injection. These results suggest that IL-12 may play a more complex role in the development of Th1 responses. In summary, in the presence of anti-CD80/86 antibodies, a comprehensive suppression was achieved, and this process combined with the reduction in the level of inflammatory cytokines, such as IFN-γ, IL-1β, TNF-α, IL-2, and IL-10.
This study had some limitations. First, anti-CD80/86 antibodies block the presentation of CD80/86, which inhibits the co-stimulation pathway for T-cell proliferation. However, it remains unclear how this blockade is associated with other mechanisms; for example, we could not clarify the roles and activities APCs in the rejection process, and how the direct and indirect pathways play their respective roles in allograft rejection as the study was based on the one-way MLR in vitro assay. Further studies are needed to investigate the influence on the CD28-CTLA-4 interaction of CD80/86 blockade 41  www.nature.com/scientificreports/ after grafting because we considered the strongest corneal inflammation generally occurred in the early stage of grafting 42 . A longer injection in a previous study showed that very few CD86 + or CD80 + cells were observed in the cornea, cervical lymph node, and spleen from the mice treated with anti-CD80/CD86 antibodies 40 . However, www.nature.com/scientificreports/ their effects on APC migration were not fully investigated, and observations with longer injections should be considered. Third, in this study, we only performed graft mRNA analysis at 2 weeks after transplantation; however, analysis at other time points such as mid-or late-stages of transplantation is necessary. Our study revealed that the inhibitory effects of anti-CD80/86 antibodies prolonged the survival of highrisk rejection corneal grafts. Blockade of the CD80/86 signaling pathway inhibits CD4 + T-cell proliferation, lowers pro-inflammatory cytokine production, and may preserve Treg suppressive function. Therefore, anti-CD80/86 antibodies may be useful for suppressing immunological rejection in high-risk rejection of corneal transplantation. Intraperitoneal anesthetic (ketamine-xylazine solution; 120 mg/kg body weight and 20 mg/kg body weight) was administered to the mice to induce general anesthesia.
Carboxyfluorescein diacetate succinimidyl ester-labeling. RPMI-1640 medium (Lonza, Basel, Switzerland) supplemented with 1 mg/mL streptomycin sulfate (Meiji Seika, Tokyo, Japan) was used as the complete medium. BALB/c mouse lymphocytes (1 × 10 7 cells/mL) were incubated with carboxyfluorescein diacetate succinimidyl ester (CFSE) (eBioscience, San Diego, CA, USA) at a final concentration of 1 μM in pre-warmed PBS for 10 min at 20-22 °C in the dark. The labeling was stopped by adding complete media (5 mL), followed by incubation on ice for 5 min. The cells were then washed thrice with the complete medium. Finally, the CFSElabeled cells were counted with an automated cell counter (Bio-Rad, Hercules, CA, USA) and immediately used in the MLR assay.
MLR assay. In vitro, lymphocytes procured from BALB/c mice were selected as responder cells, and 30-Gy irradiated B6 mouse splenocytes were used as stimulators. The responder and stimulator cells were mixed at a concentration of 1 × 10 6 cells/mL 44 . Proliferation was measured using the 5-bromo-2ʹ-deoxyuridine (BrdU) incorporation assay (EMD Millipore, Billerica, MA, USA) with 2 μg/mL anti-CD3 antibody CD3e (Clone: 145-2c11; BioLegend Inc., San Diego, CA, USA). The cells were co-cultured with and without anti-CD80 and anti-CD86 antibodies (10 μg/mL, respectively) and incubated for 4 days at 37 °C. On day 4, the culture supernatants were collected for an ELISA. High-risk allogeneic corneal transplantation and graft survival. Inflamed, neovascularized, "highrisk" recipient beds were created 7,8,45 by placing three intrastromal sutures into the central cornea 14 days before corneal transplantation. As described previously [7][8][9]46,47 , C57BL/6 corneas were grafted onto BALB/c host beds. The central cornea was excised from donor C57BL/6 mice. The graft bed was prepared by excising the central cornea from BALB/c mice. The donor button was then placed onto the recipient bed and secured with eight interrupted 11-0 nylon sutures, which were removed 7 days after the surgery. Ofloxacin ophthalmic ointment 0.3% (Pharmaceutical Co., Ltd., Osaka, Japan) was applied on the graft cornea immediately after corneal transplantation.
Graft neovascularization score, opacity score, and survival rate were evaluated using a slit lamp biomicroscope 7,45,47 . We used a standardized scoring system to calculate the neovascularization (range 0-8) and opacity (range 0-5+) scores. Corneas with an opacity score of 2 + for two consecutive examinations were considered rejected. Scoring was performed every 7 days for 8 weeks.
Flow cytometry. A single-cell suspension of co-culture cells and grafted corneas was prepared as previously described 7,8,48 . To avoid non-specific staining, the cells were blocked with an anti-FcR-blocking antibody (eBioscience, San Diego, CA, USA). The isolated cells were stained with the following antibodies: FITC-anti-CD4 (RM4-5), and anti-IFN-γ (XMG1.2) antibodies (BioLegend, San Diego, CA, USA). The stained cells were examined using flow cytometry and the data were analyzed using FlowJo software X 10.5.3 (FlowJo LLC, Ashland, OR, USA). www.nature.com/scientificreports/ Corneal RNA extraction and RT-PCR. Two weeks after transplantation, the corneal grafts were excised and immediately submerged in an RNAlater solution (Ambion, Austin, TX, USA). The total RNA from five corneas was isolated for each group using a NucleoSpin RNA isolation kit (Macherey-Nagel GmbH, Duren, Germany). Next, cDNA was prepared using the ReverTra Ace qPCR RT kit (Toyobo, Osaka, Japan). Thereafter, RT-qPCR was performed, with all reactions performed thrice. The results were analyzed using the 2 −ΔΔCt method, and Gapdh was used as an internal control. Specific primer sets were used for Ifnγ, Tgfb1, Il1b, Il2, Il10, Tnfa, Il12a, Il12b, and Gapdh (Supplementary Table S4). Identification and analysis of differentially expressed genes. DEGs were identified using DESeq2

RNA
(version 1.24.0) analysis with a threshold padj < 0.01 and abs (Log2 FC) > 0.5. Gene set enrichment analysis of the RNA-Seq data was performed using GSEA software (version 4.0.3, https:// www. gsea-msigdb. org/ gsea/ index. jsp) and MSigDB v 7.1, focusing on H hallmark gene sets. Log normalized sample counts, following GSEA parameters, were used: norm = meandiv, nperm (number of permutations) = 1000, and permute (permutation type) was set to "permutation on gene set" because of the low number of samples employed. Gene sets with a false discovery rate < 0.25 were considered significant. The raw and processed RNA-Seq data were deposited in the NCBI GEO database (accession number: GSE181893).

Statistical analysis.
Mann-Whitney U test was used to compare means. Data of more than two groups were analyzed using the one-way or two-way analysis of variance (ANOVA) with post-hoc Bonferroni's multiple comparison test. Kaplan-Meier analysis with the log-rank test was used to evaluate graft survival after transplantation. Data are presented as mean or median ± standard error of the mean and considered statistically significant at P < 0.05. All statistical calculations were performed using Prism Version 8.4.3 software (GraphPad, La Jolla, CA, USA).

Data availability
All datasets generated during or analyzed during this study are included in this published article.