PD-1 Signaling Is Essential For The Skin Allograft Protection By Alternatively Activated Dendritic Cell Infusion In Mice

[Background]Alternatively activated DC (aaDC) can prolong allograft survival in the mousemodel. However, the molecular mechanism(s) by which these DCs function toregulate alloreactive-T cell responses remains to be clearly defined.[Methods]1. Preparation of various types of DCs and infusion into recipients miceBMDCs were generated from bone marrow cells as described previously (19). Inbrief, bone marrow cells were obtained by aspirating femurs and tibial bones andthen seeded at 2×10⁶ cells per 100 mm dish in 10 ml RPMI-1640 complete mediumsupplemented with 20 ng/ml rGM-CSF (Pepro Tech Inc, New Jersey, USA).Immature DCs (imDC) were generated by adding 20 ng/ml rmIL-10 (Pepro TechAsia, Rehovot, Israel) at day 6. To obtain alternatively activated DCs (aaDC), 1μg/ml LPS (L4130, Sigma-Aldrich) was added to imDC at day 7. Mature DCs (mDC)were generated by adding 1μg/ml LPS to untreated BMDCs 24 hr before harvesting.To test the in vivo regulatory effects of these DCs originated from BALB/c mice,each C57BL/6 recipient mouse was infused with 2x 10⁶ untreated DC, imDC,aaDC, or mDC via tail vein at 7 days before the transplantation. As a result, 7experimental groups were established (6 mice per group) : group 1: untreated DC ;group 2: imDC ; group 3: aaDC; group 4: aaDC plus anti-PD-1 mAb; group 5:anti-PD-1 mAb alone ; group 6: mDC and group 7: aaDC plus control IgG. Anti-PD-1mAb or control IgG was given intraperitoneally after i.v. infusion of aaDC at 500μg on day 0, followed by 250μg on days 2, 4 and 6.2. Analysis of DC phenotype by flow cytometryFor analysis of DC phenotype, cells were preincubated with anti-CD16/CD32 mAbto block FcγR. After staining with anti-CD11c, the cells were co-stained for CD80,CD86, MHCII, PD-L1 or PD-L2 antibody at 4℃for 30 min, washed twice withPBS, and analyzed on a FACSCalibur^TM flow cytometer (Becton Dickinson).3. Procedures for skin transplantationFull-thickness skin grafts were transplanted using a modified method of Billinghamand Medawar (20). In brief, the recipient mice were anesthetized and shaved aroundthe back. Skins were prepared from the trunk of BALB/c mice. After removing thesubcutaneous fat, 12-mm-diameter circles of full-thickness pieces weresimultaneously transplanted onto the alternate dorsal thorax of the C57BL/6 recipients.The graft bed was slightly larger than the skin allograft. The skin graft was sutured andcovered with a vaseline gauze and adhesive bandage for 7 days. Skin grafts wereinspected daily. Rejection was defined as the complete loss of viable epidermal grafttissue characterized by 50% of the graft surface became raised and necrotic by visualexamination. In selected animals, allograft rejection was further confirmed byhistological analysis.4. HistologyHistological analysis was performed as previously described (21). The skinallografts were excised and stained with hematoxylin/eosin for the analysis ofpathological changes.5. Determination of cytokine production by ELISAIFN-γand IL-10 were measured in supematants harvested from co-cultures ofC57BL/6 splenic T cells with various types of DCs from BALB/c mice, using ELISAkits (Biosource International, Camarillo, CA) according to manufacturer'sinstructions.6. Mixed lymphocyte reaction Stimulator cells were various types of DCs originated from BALB/c mice. The DCswere treated with mitomycin C and seeded into 96-well plates with V-bottom. T cellsfrom recipient mice (C57BL/6) were purified by the nylon wool technique (22). ThoseT cells were stained with CFSE and used as responders. 4×10⁵ responder and 4×10⁵stimulator cells were mixed and incubated at 37℃under 5% CO₂. After culturing for5 days, the cells were harvested and subjected to flow cytometric analysis of CFSEfor T cell proliferation.7. Detection of CD4⁺CD25⁺ Foxp3⁺ Tregs by flow cytometryC57BL/6 T cells after 5 days of coculture with BALB/c DCs were co-stained withanti-CD4-FITC and anti-CD25-PE, fixed and permeabilized, and then stained withAPC-labeled anti-Foxp3 by using commercial kits (Foxp3 Staining Buffer Set,eBioscience) according to the manufacturer's instructions. The cells were analyzed byflow cytometry as described (23). The collected data were analyzed with CellQuestsoftware (BD Biosciences, Mountain View, CA) as instructed.8. Intracellular cytokine determination by flow cytometryFor determination of intracellular cytokine production, single-cell suspensions ofsplenocytes were prepared and stimulated for 4 hr in culture medium containing 50ng/ml PMA, 1μg/ml ionomycin, and 2μM monensin (all Sigma-Aldrich) at 37℃under 5% CO₂. Cells were then stained for surface marker, fixed in 4 % formaldehydein PBS, and permeabilized with 0.5 % saponin plus 1 % BSA PBS, followed bylabeling with specific cytokine antibodies or isotype controls. Cells were then analyzedon a FACSCalibur using CellQuest software.9. Statistical analysisData were analyzed by one-way analysis of variance. Statistical significance wasdefined as P＜0.05. Allograft survival differences among groups were analyzed usingthe method of Kaplan and Meier (24). Comparisons of two means were carried out byone-way anova. Regression curves were fitted using SigmaPlot software (version 8.0;SPSS, Chicago, IL). [Results]1. Partially matured phenotype of aaDCWe first examined the phenotypic profiles of various types of DCs generated fromBALB/c BM progenitors. As shown in Fig. 1, we observed a slight increase of CD80and CD86 as compared with imDC and untreated DC, while a similar level ofMHCII expression was observed between aaDC and imDC or untreated DC. Incontrast, the levels of these molecules on aaDC were much lower than those on mDC.These analysis revealed that aaDC displayed a semi-mature phenotype. Of note, weobserved that aaDC had significantly higher levels of PD-L1 and PD-L2 as comparedto imDC and untreated DC (Fig. 1).2. aaDC displayed a reduced alloreative T cell stimulating capacity that wasassociated with PD-1/PD-L pathwayNext, we performed an MLR to test the stimulatory capacity of aaDC preparedfrom BALB/c donor mice for the proliferation of allogeneic T cells from C57BL/6recipients. As expected, stimulation with mDC resulted in a strong proliferativeresponse (Fig.2). In contrast, stimulation with aaDC or imDC gave rise to a weakT-cell proliferation. Of important note, we observed that the decreased T-cellproliferative response to aaDC was largely restored when PD-1/PD-L pathway wasblocked by anti-PD-1 Ab.3. aaDC enhanced the production of IL-10 while inhibited the secretion of IFN-γin an MLRWe also measured the production of IFN-γand IL-10 in the supernatant harvestedfrom the cocultures by ELISA (Fig.3). aaDC induced a reduced IFN-γand an enhancedIL-10 production as compared with mDC, which were reversed in the presence ofanti-PD-1 blocking Ab. Together, these results suggest that aaDC have an impairedallo-stimulating capacity and, PD-1/PD-L pathway plays an important role in theregulatory capacity of aaDC in this setting.4. Augmented expansion of CD4⁺CD25⁺Foxp3⁺ Tregs coculture with aaDC Given the importance of regulatory DCs in the induction of regulatory T cells(Tregs), we next checked the differences in Tregs induced by various types of DC(imDC, aaDC, mDC). To this end, C57BL/6 T cells were cocultured with varioustypes of DC generated from BALB/c mice for 5 days. Expanded T cells originatedfrom each group were stained for intracellular Foxp3 after co-staining for surface CD4and CD25, and then subjected to flow cytometry analysis. As shown in Fig.4, aaDCinduced an increased expression of Foxp3 in CD4⁺CD25⁺ T-cell population ascompared mDC and untreated DC, and this effect was partially dependent onPD-1/PD-L pathway.5. Prolonged skin allograft survival by aaDC infusion was reversed by blockadeof the PD-1/PD-L pathwayBased on the above results, we next sought to explore the feasibility of aaDC forinduction of antigen-specific allograft tolerance in transplantation. For this purpose,2 x 10⁶ various types of DC (aaDC, imDC, mDC and untreated DC) originated fromBALB/c mice were injected into C57BL/6 recipient mice via tail vein respectively.At 7 days after injection, the mice were transplanted with BALB/c-derived skinallografts. As shown in Fig. 5, the mean graft survival time (MST 22.67±4.08 days)in the recipient mice pre-infused with donor-specific aaDC was significantly longerthan that in the recipients pre-treated with either untreated DCs (MST 11.83±2.40days) , imDC (MST 16.00±3.10 days) or mDC (MST 7.50±1.64 days) (P＜0.01).Notably, pre-treatment of the recipient mice with anti-PD-1 blocking Ab reversed theprolonged survival of skin allografts by aaDC (MST 22.67±4.08 vs 12.33±3.33 ofaaDC + anti-PD-1, P＜0.01), while administration of anti-PD-1 mAb alone showedonly a marginal effect on the allograft survival as compared with that untreated DCgroup ( MST 10.83±2.79 vs 11.83±2.40 days, p＞0.05 ).6. Intravenous infusion of aaDCs attenuated the infiltration of leukocytes in thegrafts.Consistently, histological analysis of allografts at 7 days after transplantation revealed much severe inflammatory infiltrates in the recipient mice pre-injected withmDC and untreated DC as compared to the recipients pre-infused with aaDC or imDC(Fig. 6). Allograft in the recipients treated with aaDC and PD-1 blocking Ab showed asevere infiltration and necrosis (Fig. 6). These results indicate that the PD-1/PD-Lpathway is essential for the aaDC-mediated allograft protection.7. aaDC infusion changed the cytokine profiles produced by splenic T cells fromrecipient miceFinally, we questioned whether infusion of donor-derived aaDCs affects thecytoldne profiles of recipient splenocytes. To address this question, splenocyteswere isolated from each group of recipient mice and then subjected to intracellularcytokine determination by flow cytometry. As shown in Figure 6A, we observed adecreased level for IL-2 (^**p＜0.01) (Fig. 7A) while an increase of IFN-γin thesplenic CD4⁺ T cells from the recipient mice infused with donor-derived aaDCs ascompared to those in the other groups. On the contrary, an increase of IL-10production by CD4⁺ T cells was detected from the recipient mice with aaDCs infusion.Interestingly, we observe a reduction of IFN-γproduction by splenic CD8⁺ T cellsin the recipients treated with aaDCs as compared with other types of DCs (_*^*p＜0.01),whereas IL-2 secretion by CD8⁺ T cells were not affected by infusion of aaDCs (Fig.7B). These regulatory effects of aaDC on the cytokine production by splenic T cellswere abrogated when PD-1/PD-L pathway was blocked by anti-PD-1(Fig. 6).[Conclusions] Our data indicate that the PD-1/PD-L pathway plays an important rolein aaDC-mediated prolongation of skin allograft survival.