Graft-infiltrating host dendritic cells play a key role in organ transplant rejection

Abstract

Successful engraftment of organ transplants has traditionally relied on preventing the activation of recipient (host) T cells. Once T-cell activation has occurred, however, stalling the rejection process becomes increasingly difficult, leading to graft failure. Here we demonstrate that graft-infiltrating, recipient (host) dendritic cells (DCs) play a key role in driving the rejection of transplanted organs by activated (effector) T cells. We show that donor DCs that accompany heart or kidney grafts are rapidly replaced by recipient DCs. The DCs originate from non-classical monocytes and form stable, cognate interactions with effector T cells in the graft. Eliminating recipient DCs reduces the proliferation and survival of graft-infiltrating T cells and abrogates ongoing rejection or rejection mediated by transferred effector T cells. Therefore, host DCs that infiltrate transplanted organs sustain the alloimmune response after T-cell activation has already occurred. Targeting these cells provides a means for preventing or treating rejection.

Figure 1. Replacement of donor DCs by host DCs in heart grafts.

(a) CD45.2 BALB/c (Allo), (B6 x BALB/c) F1 (F1) or B6 (Syn) heart grafts were transplanted to CD45.1 B6 mice. Intragraft recipient (CD45.1) and donor (CD45.2) DCs were analysed by flow cytometry at indicated time points (day 0, day 1, day 3 and day 7). Representative flow plots and line graphs are shown (mean±s.d., 3 or 4 mice per time point). Full gating shown in Supplementary Fig. 1. CSA, Cyclosporine A. NS, not significant. *P<0.01 (one-way analysis of variance (ANOVA)). In the case of donor DC enumeration significant change was detected in DC number over time but not between groups. (b) Representative photomicrograph of a tissue section of one of 3 BALB/c heart allografts harvested on day 7 and stained for CD11c (red) and proliferation marker Ki67 (green). Cell nuclei are stained with DAPI (blue). The CD11c⁺ DCs (in red) were Ki67^neg. The arrow indicates Ki67 expression, as an endogenous control of Ki67 labelling, in the nuclei of CD11c^neg cells shown in detail in the insets. Magnification × 200. Scale bar, 15 μm. (c) Characterization of Lin^negLy6G^neg, recipient (CD45.1⁺) myeloid cells in BALB/c heart allografts on day 7 after transplantation. Cells were analysed by flow cytometry (top panels) and CD11c⁺ and CD11c^neg subsets were sorted and tested for their ability to induce proliferation (centre panels) and IFN-γ production (bar graph) in naive, allogeneic T cells in the direct MLC. Proliferation was measured by CFSE dilution and IFN-γ by ELISA of culture supernatants on day 5 of MLC. One representative experiment out of two is shown. P values were generated by one-way analysis of variance (ANOVA). Bar graphs are mean±s.d. (3 mice per group). (d) IL-12p70 production by sorted CD11c⁺ and CD11c^neg subsets cultured for 24 h. Cells sorted from 2 different BALB/c heart allografts on day 7 were pooled and ELISA was performed in duplicates on the culture supernatant. The experiment was performed twice (4 mice total). Bar graphs are mean±s.d. of 4 measurements (2 experimental × 2 technical replicates). P values were generated by one-way ANOVA. (e) Ability of sorted CD11c⁺ subset to stimulate proliferation of CD4⁺ 1H3.1 and CD8⁺ 2C T cells ex vivo in the absence (left panels) or presence (right panels) of cognate peptide added to the culture (3 mice per group).

Figure 2. Origin of recipient DCs in heart allografts.

(a) BALB/c heart allografts were transplanted to B6 WT or CCR2^KO mice. Grafts and recipient blood were collected on day 7. DC (CD11c⁺) and non-DC (CD11c^neg) subsets were identified and enumerated in the graft (>95% at this time point are recipient-derived as shown in Fig. 1a), and recipient classical (Ly6C^hi) and non-classical (Ly6C^lo) monocyte subsets were enumerated in the blood. As control, monocyte subsets in blood of untransplanted (naive) B6 mice were also analysed. Bars represent mean±s.d. (N=3 or 4 mice per group). P values were generated by one-way analysis of variance (ANOVA). (b) Representative photomicrographs out of three allografts per group depicting graft tissue stained for CD11c (red) and recipient MHC-II (IA^b) (green). Overlay (orange) demonstrates recipient origin of DCs in grafts removed from either WT or CCR2^KO mice. Magnification × 200. Scale bar, 50 μm. (c) Equal numbers of unlabelled, classical (Ly6C^hi) and CFSE-labelled, non-classical (Ly6C^lo) CD45.1⁺ B6 monocytes were co-transferred to CD45.2⁺ B6 mice 5 days after receiving CD45.2⁺ BALB/c heart allografts. Flow analysis of graft-infiltrating leukocytes was performed 2 days later. Flow plots show the differentiation of transferred (CD45.1⁺) monocytes into DCs (CD11c⁺MHC-II⁺) in the graft, the majority of which were CFSE⁺ (derived from non-classical monocytes). Bar graph is mean±s.d. (N=3 mice). P values were calculated with two-sided Student's t-test.

Figure 3. Rapid replacement of donor DCs in kidney allografts by recipient monocyte-derived DCs.

(a) F1 CD11c-YFP kidney allografts (CD45.2) were transplanted to B6 CX3CR1^gfp/+ recipients (CD45.1) and two-photon intravital imaging was performed at indicated time points. Representative volume-rendered images are shown in the top panels and quantitation of DC volume and DC proportions in the graphs below. Panel on far left (Day 0) is an image of a native CD11c-YFP kidney before transplantation. Scale bar, 50 μm. Frame indicates dimensions of the 3D volume. N=14–18 images from 3 or 4 mice per time point. (b) Cells were extracted from allografts at the end of imaging and analysed by flow cytometry. Representative flow plots and line graphs (mean±s.d., N=3 mice per time point) of absolute number and proportion of recipient and donor DCs are shown. (c) F1 kidney allografts not transgenic for a fluorescent protein were transplanted to B6 CX3CR1^gfp/+xCD11c-YFP recipients and two-photon intravital imaging was performed at the indicated time points. Representative volume-rendered images are shown in the top panels. Panel on far left (Day 0) is an image of a native non-transgenic F1 kidney before transplantation. Graphs show proportions of DC subsets by volumetric image analysis (left) and flow analysis (right). In the image analysis, single positive (YFP⁺) cells represented cDCs, and double-positive (YFP⁺GFP⁺) cells mono-DCs. In the flow analysis, cDCs (CD11b^neg) and mono-DCs (CD11b⁺) were distinguished based on CD11b expression on CD45.1⁺Lin^negLy-6G^negCD11c⁺MHC-II⁺ cells. Scale bar, 50 μm. N=7–12 images from 3 mice per time point. In a and c, horizontal lines are mean values and P values were calculated using two-sided Student's t-test.

Figure 4. Interactions between recipient DCs and effector T cells in kidney allografts.

F1.Act-OVA kidneys were transplanted to B6 mixed bone marrow chimeras that harbour equal numbers of H2-K^b−/− (YFP⁺) and H2-K^b+/+ (GFP⁺) DCs. OT-I effector T cells were transferred 6 days after transplantation and intravital two-photon microscopy was performed 1 day later. (a) Representative, volume-rendered image of kidney allograft with inset showing an OT-I cell (red) in close contact with a DC (green). Scale bar, 50 μm. (b) Number of OT-I cells located inside (intravascular) or outside (extravascular) of blood vessels. (c) Proportion of OT-I cells making >2 min contact with either type of DC in the graft. (d) Proportion of OT-I cells in contact with H-2K^b+/+ versus H-2K^b−/− DCs. Lines join data points from same image volumes. (e) Contact time, mean velocity, and arrest coefficient of OT-I cells interacting with H-2K^b+/+ versus H-2K^b−/− DCs during entire time-lapse recording (∼33 min). N=11 image data sets (movies) from five mice. (f) F1.Act-OVA kidney grafts were transplanted to CD11c-DTR or WT bone marrow chimeric B6 mice. DT was administered to both groups of mice on days 7 and 9, OT-I T cells were adoptively transferred on day 7, and imaging performed on day 10 after transplantation. N=9 image data sets (movies) from 4 mice/group. In b–f, horizontal lines are mean values and P values were calculated using two-sided Student's t-test.

Figure 5. Effect of delayed recipient DC depletion on allograft survival and T-cell proliferation and apoptosis in the graft.

(a) BALB/c heart allografts were transplanted to chimeric B6 mice that received either B6 WT or CD11c-DTR bone marrow. Mice were injected with DT or PBS every other day starting on day 5 after transplantation until day of rejection (d 5→). Allograft survival was monitored by daily palpation of heart contractions. Number of mice per group is between parentheses. Kaplan–Meier graft survival plot is shown. P value was calculated by the log-rank method. (b–d) Transplantation and DT treatment were performed as in a except that all grafts were sacrificed on day 7 after transplantation (2 days after a single DT injection) for immunostaining and flow cytometry. Quantitation of intragraft DCs by immunostaining of CD11c is shown in b. The number of intragraft CD4 and CD8 T cells and proportion undergoing proliferation (BrdU⁺) or apoptosis (Annexin-V⁺) is shown in c. T cells undergoing proliferation in situ were detected and quantified by staining for Ki67 or Histone 3 pS10 (d). White arrows point to T cells that are positive for the proliferation markers. Images are representative out of 5–6 fields, at × 200 magnification, analysed on tissue sections of each of four allografts per group. Magnification × 200. Scale bar, 20 μm. Bar graphs are mean±s.d. N=4 mice per group. P values were generated by one-way analysis of variance (ANOVA).

Figure 6. Effect of recipient DC depletion on rejection mediated by effector T cells.

(a) BALB/c heart allografts were transplanted to splenectomized, CD11c-DTR→B6 LTβR^KO bone marrow chimeras that lack all secondary lymphoid tissues. Experimental groups received effector T cells on day 2 after transplantation and were either left untreated or treated with DT starting at time of transplantation (d0→) or 10 days after transplantation (d10→). DT was injected every other day until the day of graft rejection. Control groups included splenectomized CD11c-DTR→B6 LTβR^KO chimeras that did not receive effector T cells (one group treated with DT and one not), and splenectomized WT→B6 LTβR^KO chimeras that received T cells and were treated with DT starting on day of transplantation. Allograft survival was monitored by daily palpation of heart contractions. Kaplan–Meier graft survival plot is shown. P value was determined by log-rank test. (b) Infiltration of heart allografts with recipient DC (CD11c⁺IA^b+ cells), CD4, and CD8 T cells in untreated and DT-treated (+DT) mice that received effector T cells on day 2. Representative photomicrographs out of three fields, at × 200 magnification, analysed on tissue sections of each of 4–6 allografts per group. Inset in photomicrograph is a DC stained positive for both CD11c and IA^b. Horizontal lines in scatter plots represent mean for each group. Magnification × 100. Scale bar, 50 μm. P values were calculated using two-sided Student's t-test.