A single infusion of ECP-DL significantly prolonged survival of cardiac allografts via downregulation of T-cell activation, favoring Treg production.
We evaluated the effect of infusing ECP-DL in a heterotopic mouse heart transplant model (27); treatment groups utilized are illustrated in Figs. 1ai-iii. B6 mice were recipients, and BALB/c mice were donors of ECP-DL on day -7 and heart on day 0 (Fig. 1ai). In the absence of IS, this donor-recipient combination results in loss of palpable heart impulse in approximately 10 days. ECP-treated splenocytes from MHC-mismatched third-party mice (C3H) were administered to B6 mice on day -7 before transplant from BALB/c donors in a subset of experiments to test donor specificity (Fig. 1aii). To test the effect of IS, tacrolimus (TAC) or rapamycin (RAPA) was added from day -1 to post-operative day (POD)8 to ECP-DL treatment (Fig. 1aiii). Syngeneic heart transplants served as controls (Iso, Fig. 1b). Pre-transplant infusion of ECP-DL was effective in prolonging allograft survival; animals receiving ECP-DL alone had significantly prolonged graft survival until POD30, whereas allografts receiving no therapy (Untreated), unmanipulated donor-type splenocytes (DL), or ECP-treated third-party cells (ECP-UL) all lost palpable impulse of the graft by POD10 (Fig. 1b). Combined with TAC or RAPA, further prolongation of allograft survival was observed; the combination of ECP-DL and RAPA demonstrated the longest duration of cardiac impulse (Fig. 1c). We also tested administration of ECP-DL on days -3, and -1 and found no statistical difference in graft survival (Supplemental Fig. 1a). Finally, experiments were performed utilizing a reversal of donor and recipient strains (B6 grafts/ECP-DL into BALB/c mice) to correct for known differences in strain combination; results were identical to those reported above (Supplemental Fig. 1b). A subset of animals receiving no treatment and those receiving ECP-DL were sacrificed on POD7 to examine the grafts histologically. Grafts from untreated animals had increased inflammatory infiltrate compared to those from ECP-DL animals. We performed stains to identify T-cells (CD3), monocyte/macrophages (F4/80), and B-cells (B220); untreated grafts had more prominent infiltration of all inflammatory cell types (Fig. 1d).
To evaluate the effect of ECP-DL infusion on T-cell response, transplant recipients receiving either ECP-DL on day -7 or no treatment were sacrificed on POD7. Lymphocytes from the spleen were recovered and subjected to flow cytometry; activation of CD4+ and CD8+ T-cells in the ECP-treated mice was significantly reduced as indicated by decreased CD44+CD62L− effector T-cell populations (Fig. 2a-c). Secretion of IFN-γ from both CD4+ and CD8+ T-cells was lower in the ECP treated group, indicating a reduced inflammatory response (Fig. 2d, e); this was statistically significant in the CD4+ T-cell fraction. The frequency of Tregs was significantly increased in treated mice (13% vs 4.5% in untreated animals, Fig. 2f, g). Together, these data illustrate that in an antigen-specific manner, ECP-DL treatment prolongs graft survival, decreases graft infiltrating inflammatory cells, down-regulates T-cell activation, and promotes Treg production.
A single infusion of ECP-DL decreased production of antigen-specific CD4 + and CD8+ T-cells and suppressed their proliferation through direct antigen presentation.
Subsequently, we utilized an established 2C and 4C transgenic mouse model to evaluate the antigen-specific effects of ECP therapy on alloreactive T-cells prior to transplantation . CD-4C and CD8-2C T-cells were isolated from TCR transgenic B6 mice and labeled with carboxyfluorescein succinimidyl ester (CFSE) to evaluate lymphocyte proliferation rates. CD4-4C and CD8-2C T-cells recognize epitopes presented by donor MHC class II or I molecules, respectively. Labeled T-cells from transgenic B6 mice were injected into standard B6 mice on day 0. On day 1, splenocytes from standard BALB/c mice were isolated as in prior experiments and were injected into the B6 recipient; splenocytes were either ECP-treated or not. On day 3 B6 animals were sacrificed to assess the frequency of T-cells in the spleen; none underwent transplant. These procedures are depicted in Fig. 3a.
Treatment with ECP resulted in a statistically significant depletion in the number of antigen-specific CD4+ T-cells found infiltrating the spleen compared to untreated animals, by 48% (Fig. 3b, c). CFSE concentration within the antigen-specific CD4+ T-cells in ECP-treated animals was higher, indicating reduced proliferation; this was also statistically significant (Fig. 3d, e). Results in the CD8+ compartment were striking: proliferation within treated animals was essentially absent compared to high levels of proliferation seen in control animals (Fig. 3f, g).
Given the remarkable difference between groups in the CD8 compartment, an additional set of experiments were performed to evaluate the fate of CD8 cells after transplant; B6 mice were injected with CFSE-labeled CD8-2C cells on day -8. On day -6, the animals were given either ECP-treated or untreated splenocytes; heart transplant was performed on day 0 and animals were sacrificed on POD5. Flow cytometry was performed on cells isolated from the spleen and graft (Supplemental Fig. 2a). There were substantially fewer infiltrating CD8+ cells in treated animals (Supplemental Fig. 2b). Collectively, the 2C/4C transgenic mouse experiments illustrate a drastic reduction of antigen-specific alloreactive T-cells in ECP-treated animals, which, when present, are less proliferative.
A Single infusion of ECP-DL inhibited production of donor specific antibodies (DSA).
We next asked how ECP-DL infusions affected B-cells by evaluating production of DSA. We performed cardiac transplants in two groups of animals, untreated and those receiving ECP-DL on day -7. DSA were measured at POD5, 7, 14, 21, and 28. Total IgG (Fig. 4a), and subtypes IgG1 (Fig. 4b), IgG2 a/b (Fig. 4c), and IgG3 (Fig. 4d) were assessed. Importantly, minimal DSA were detected in the ECP-DL-treated animals at POD5, suggesting they are not sensitized following ECP-DL treatment. At all timepoints, DSA were lower in the treated group; the differences were statistically significant across all antibody subtypes at most times. This data was reproduced in our liver transplant model (data not shown).
ECP induced apoptosis; most apoptotic cells were phagocytosed in the spleen.
We sought to determine the fate of ECP-DL in non-transplanted animals. BALB/c splenocytes were treated with ECP and subjected to Annexin V staining. Approximately half of ECP-DL became apoptotic prior to infusion (Supplemental Fig. 3a). We subsequently performed experiments in which ECP-DL were labeled with PKH26 dye and injected into B6 mice for tracking (without transplant). Fifteen hours after injection, mice were sacrificed and spleens analyzed; the intensity of PHK26 was tested in monocytes (CD11b+/Ly6C+), granulocytes (CD11b+/Ly6G+), and macrophages (CD11b+/F4/80+). Around 6% of CD45+ leukocytes and up to 66% of macrophages (CD11b+F4/80+) took up PKH dye (Supplemental Fig. 3b). Within the myeloid line, PKH26 signal was detectible primarily in macrophages but not in neutrophils or monocytes, indicating that ECP-DL which had undergone apoptosis had been phagocytosed by macrophages in the spleen (Supplemental Fig. 3c). Together, these data indicate that ECP-DL, once injected, become apoptotic and are taken up by recipient macrophages in the spleen, allowing for donor antigen processing by myeloid cells.
Infusion of ECP-DL led to localization of DCs with tolerogenic phenotype in the allografts.
Given the known regulatory effect of DC on the state of T-cell activation and transplant tolerance, we sought to characterize graft infiltrating DC. Heterotopic heart transplants were performed with or without a single infusion of ECP-DL on day -7, and we evaluated the number and phenotype of infiltrating cells within the graft on POD6 and POD12. Gating strategies utilized to evaluate infiltrating DC are provided in Supplemental Fig. 4. Host DC population was identified within the H-2Db/CD11b gate and is MHC II+/CD11c+. At POD6 there was a trend to increased DC within the grafts retrieved from treated animals compared to control, but this difference was not statistically significant. At POD12, however, statistical significance was achieved (Fig. 5a, b). To better characterize the functional phenotype of these cells, we evaluated the change in co-stimulatory molecule expression CD86, CD80, and MHC Class II (I-Ab, Fig. 5c-f). Interestingly, CD86 expression on dendritic cells is statistically significantly reduced at both POD6 and POD12 in the grafts of ECP treated animals. CD80 expression on graft infiltrating DC were increased at POD6 but then reduced compared to controls at POD12. Additionally, MHC class II expression is higher on the graft infiltrating DC from ECP-treated animals. Taken together, we demonstrate an increase in DC infiltrating the grafts of ECP-treated mice, which exhibit a tolerogenic phenotype.
ECP-DL treatment led to an increase in anti-inflammatory and pro-reparative macrophages infiltrating the graft.
To determine whether ECP-DL influences the recruitment of myeloid cells into the allograft, we enumerated and characterized host myeloid cells (e.g. monocyte/macrophages, granulocytes) in the allografts on POD6 and POD12 by using a Ly6C/Ly6G-based strategy (Supplemental Fig. 5a-c) . We found that infiltrating host (H-2Db+) myeloid cells primarily consist of activated CD11b+ Ly6G+/Ly6Cint cells (granulocytes) and CD11b+Ly6G−/Ly6Chi cells (macrophages) at POD6 and POD12 based on their expression of costimulatory molecules CD80, CD86 and MHC class II (Supplemental Fig. 5d-g). ECP-DL infusion significantly reduced the number of granulocytes in grafts at POD6 and POD12 compared to untreated controls. Interestingly, the number of macrophages infiltrating the grafts of treated animals were reduced compared to control at POD6, while at POD12 there were more; the differences at both time points were statistically significant. CD86 expression on infiltrating macrophages (Supplemental Fig. 5h-k) was significantly downregulated in the ECP treated animals compared to controls. In this cell population, CD80 expression was decreased compared to control animals at both time points, while MHC Class II expression was lower in ECP treated animals at POD6 but higher than controls at POD12. Together, these data indicate ECP-DL treatment not only reduced number of graft infiltrating CD11b+ Ly6G+/Ly6Cint cells and CD11b+Ly6G−/Ly6Chi myeloid cells, but also regulated differentiation of these myeloid cells by down-regulating expression of co-stimulating molecules, particularly on CD11b+Ly6G−/Ly6Chi cells, suggesting that the cells present tended towards an anti-inflammatory phenotype over time.
To further define the monocyte population infiltrating the grafts at POD12, we evaluated the number and phenotype of monocytes expressing MerTK. There was a significantly higher number of MerTK expressing monocytes in the ECP-DL treatment group compared with control; most cells that were MerTK positive in the ECP treated grafts were also Ly6C negative, indicating a less inflammatory phenotype (Fig. 6a, b). Importantly, these monocytes were also phenotypically different in ECP-DL treated vs untreated animals, in that expression of CD80, CCR2, and CXCR1 expression were all significantly lower in the grafts from ECP-DL animals. MHC class II expression was higher in the ECP treated group (Fig. 6c, d). The significant increase in MerTK expression along with the reduction of Ly6C expression indicates the presence of tolerogenic monocytes capable of promoting a favorable cytokine milieu, promote Treg production, negatively regulate nuclear factor kappa B (NF-κb) expression, and promote tissue repair through efferocytosis.
A single infusion of ECP-DL significantly prolonged the survival of liver and kidney allografts, maintained long-term allograft function and histologic appearance.
We subsequently looked to confirm these promising results in orthotopic kidney and liver transplants in rats, which are functional transplants in that entire native kidneys and liver in each model are removed; they are also more immunologically robust than the cardiac model. We first evaluated a rat orthotopic kidney transplant model; Lewis rats were recipients, and MHC-mismatched ACI rats were donors. Terminal rejection will develop in less than two weeks without IS in this high-responder pair [29, 30]. As expected, untreated allogeneic recipients died of renal failure in the first 10 days after transplantation (Untreated, Supplemental Fig. 6a). In stark contrast, ECP-DL pre-conditioning on day -7 significantly prolonged allograft survival to more than 50 days. Moreover, 60% of animals receiving ECP-DL alone survived to day +200. While less than 40% of those receiving TAC alone survive, all animals receiving combination ECP-DL plus TAC survived indefinitely (Supplemental Fig. 6a), demonstrating robust synergy. Creatinine levels were monitored in the blood at pre-determined times. Animals receiving either ECP-DL, TAC, or combination therapy had normal levels of creatinine at POD7 and POD90 (Supplemental Fig. 6b). Grafts were examined histologically at the time of sacrifice (time of renal failure or POD200). Kidneys from rats receiving ECP-DL appear like isografts without significant damage to glomeruli or collecting ducts, while those from untreated animals demonstrated damage consistent with rejection (Supplemental Fig. 6c). Together, these results illustrate that even in a highly immunogenic kidney transplant model, ECP-DL treatment leads to improved graft survival; addition of short-course IS was synergistic and led to indefinite survival.
Similar experiments utilizing ACI rats as cell and liver donors and Lewis rats as recipients were then performed. Syngeneic liver transplants (Lewis/Lewis) served as controls (Iso); as expected, they survived indefinitely (>200 days). Both untreated allograft recipients and allografts receiving ECP-RL (ECP-treated autologous splenocytes) infusion pre-transplant experienced rejection and died within two weeks. Strikingly, 60% of allogeneic animals receiving a single infusion of ECP-DL seven days prior to SOT survived to POD200 without additional therapy; this difference was statistically significant. Adding post-transplant autologous ECP-RL infusions at POD14 and POD28 to ECP-DL pre-conditioning on day -7 further promoted allograft survival, with 80% of ECP-DL/ECP-RL animals surviving to POD200 (Fig. 7a); this improvement in survival was statistically significant. Fig. 7b shows the survival of animals that were treated with TAC +/- ECP-DL infusion. TAC alone prolonged allograft survival with 38% of animals surviving to study endpoint, while allografts receiving pre-transplant ECP-DL infusion plus TAC survived indefinitely, demonstrating a marked synergistic effect. Notably, recipients of ECP-treated unrelated third-party cells (ECP-UL) prior to transplantation were not protected, again illustrating a donor-specific effect (Fig. 7b). Serum was tested at specified intervals to evaluate liver function. Despite an initial rise in alanine aminotransferase (ALT) in the ECP-DL group at week 4, levels approached those of the ECP-DL/ECP-RL and isograft groups by week 24 (Fig. 7c). Histological analysis of the transplanted liver revealed that allografts in the ECP-DL, ECP-DL/RL, and ECP-DL/TAC groups maintained intact architecture and had minimal cellular infiltrates, comparable to isografts, while animals receiving no therapy had disruption of portal triads and cellular infiltration consistent with rejection (Fig. 7d).
Finally, to further confirm that immune hypo-responsiveness is donor-specific, additional animals received ECP-DL seven days prior to liver transplant, and subsequently underwent skin transplant from either ACI or a third-party rat (Brown Norway, BN) at day +100. By day +130 the skin graft from BN had ulcerated (indicating rejection), while donor-derived skin allografts healed well, progressing to a normal appearance by time of sacrifice at day +180 (Fig. 8a). The successful engraftment of donor-origin skin grafts contrasted with rejection of third-party grafts provides further evidence that the effects of ECP-DL pre-conditioning are donor specific. The transplanted liver in animals receiving a single infusion of ECP-DL maintained normal gross appearance at the time of sacrifice (Fig. 8a).
We hypothesized that CD4+/CD25+/FoxP3+Tregs were involved in the hypo-responsiveness induced by ECP-DL; flow cytometry was performed on peripheral blood at specified intervals post-transplant. When comparing untreated animals to those receiving TAC and day -7 ECP-DL (on POD7), the ECP-DL group had a higher number of Tregs (Fig. 8b). Longitudinally, both the ECP-DL and ECP-DL/ECP-RL groups had similar and significantly higher number of Tregs in the peripheral blood at all time points compared to the isograft group (Fig. 8c). Collectively, these experiments reinforce that this is an antigen specific process, and that treated animals have significantly higher levels of circulating Tregs compared to control over time.
Overall, our results have shown that while pre-conditioning with a single infusion of ECP-DL resulted in long-term allograft survival, infusion with ECP-treated autologous cells prior to transplant did not protect the graft, demonstrating that the effect seen in the ECP-DL group was not simply because of the infusion of apoptotic cells. Furthermore, pre-transplant infusion of ECP-DL plus post-transplant infusions of ECP-RL or TAC were synergistic in facilitating immune hypo-responsiveness to both liver and kidney allografts promoting indefinite survival.