Prolongation of allograft survival by passenger donor regulatory T cells

Tissue resident lymphocytes are present within many organs, and are presumably transferred at transplantation, but their impact on host immunity is unclear. Here, we examine whether transferred donor natural regulatory CD4 T cells (nT‐regs) inhibit host alloimmunity and prolong allograft survival. Transfer of donor‐strain lymphocytes was first assessed by identifying circulating donor‐derived CD4 T cells in 21 consecutive human lung transplant recipients, with 3 patterns of chimerism apparent: transient, intermediate, and persistent (detectable for up to 6 weeks, 6 months, and beyond 1 year, respectively). The potential for transfer of donor nT‐regs was then confirmed by analysis of leukocyte filters recovered from ex vivo normothermic perfusion circuits of human kidneys retrieved for transplantation. Finally, in a murine model of cardiac allograft vasculopathy, depletion of donor CD4 nT‐regs before organ recovery resulted in markedly accelerated heart allograft rejection and augmented host effector antibody responses. Conversely, adoptive transfer or purified donor‐strain nT‐regs inhibited host humoral immunity and prolonged allograft survival, and more effectively so than following administration of recipient nT‐regs. In summary, following transplantation, passenger donor‐strain nT‐regs can inhibit host adaptive immune responses and prolong allograft survival. Isolated donor‐derived nT‐regs may hold potential as a cellular therapy to improve transplant outcomes.


| INTRODUC TI ON
Although still considered a novel technology, ex vivo perfusion of recovered organs from deceased donors is likely to become widely adopted in the near future. 1,2 Ex vivo perfusion offers the potential to assess the viability of organs before transplantation, and to extend the acceptable period between recovery and implantation. It may also enable targeting of the isolated organs with specific therapies aimed at prolonging allograft survival. 3 One particular focus of such strategies is likely to be donor-derived T cell populations (naïve or memory) that are resident within the graft. 4,5 We have recently reported that passenger T cells are present within human donor organs recovered for transplantation and, using murine transplant models, have demonstrated that donor T effector cells can augment host alloimmune responses directed against the allograft. 6 Thus, although seemingly counterintuitive, these passenger lymphocytes contribute to rejection of the organ. Here, we examine whether donor-derived natural regulatory CD4 T cells (nTregs) can, conversely, prolong allograft survival.

| Identification of circulating donor CD4 T lymphocytes in human lung transplant recipients
Following adult deceased-donor lung or heart plus lung transplantation, blood from consenting recipients was sampled at predetermined time points (initially weekly for the first 2 months after transplantation, and monthly/bimonthly thereafter) and donor CD4 and with the relevant MHC class I HLA-specific biotinylated antibody that were selected to bind exclusively to donor (but not recipient) HLA class I MHC alloantigen (see Table S1; kindly gifted

| Characterization of lymphocyte subsets released during ex vivo normothermic perfusion
Kidneys underwent 1 hour of normothermic machine perfusion, as described previously, 7

| Heterotopic heart transplantation
Vascularized cardiac allografts were transplanted intra-abdominally as described previously. 9,10 Heart graft survival was monitored by daily abdominal palpation, with rejection defined as cessation of a detectable beat. Grafts were excised at predetermined time points after transplantation and stored at −80°C or fixed in 10% buffered formalin. In certain experiments, recipient B6 mice were depleted of CD4 T-regs by treatment with 0.5 mg of anti-CD25 mAb (PC-61, Bio X Cell, West Lebanon, NH), i.p., on day -1 followed by 0.25 mg, i.p., on days 1, 3, 5, and 7, in relation to bm12 heart graft transplantation.
Donor T-reg depletion was achieved by administering 0.5 mg of anti-CD25 mAb (PC-61), i.p., on days -6 and -2 before recovery of heart allograft. Pilot experiments confirmed that this treatment resulted in depletion of typically 85%-90% of FoxP3 +ve splenic CD4 T cells.

| Adoptive transfer of donor/recipient-derived nT-regs
Recipient B6 mice were adoptively transferred by tail-vein intravenous injection with 1 × 10 6 nT-regs derived from B6 or bm12 animals on the first postoperative day after bm12 cardiac transplantation. nT-regs were purified from spleens of naïve B6 or bm12 animals using the CD4 + CD25 + Regulatory T Cell Isolation Kit (Miltenyi Biotec, Auburn, CA) and an autoMACS separator (Miltenyi); cell purity (typically >90% CD25 +ve CD4 +ve ) was analyzed by flow cytometry prior to injection.

| Quantification of humoral autoantibody responses
Antinuclear autoantibody responses were determined by HEp-2 indirect immunofluorescence (The Binding Site, Birmingham, UK) as described previously, 11 by incubating test sera on slides coated with HEp-2 cells and detecting bound antibody with FITC-conjugated goat anti-mouse IgG (STAR 70; Serotec, Oxford, UK). For each test serum, photomicrographs were taken, and the intensity of staining was determined by integrated morphometric analysis using MetaMorph software. The fluorescence value was then derived by comparison with a standard curve obtained for each assay by serial dilutions of a pooled hyperimmune serum that was assigned an arbitrary value of 1000 fluorescence units.

| Histopathology
Cardiac allograft vasculopathy was assessed on elastin van Giesonstained paraffin sections by morphometric analysis as described previously. 11 Luminal stenosis [percentage cross-sectional area luminal stenosis = (area within internal elastic lamina -area of lumen)/area within internal elastic lamina × 100]. All elastin-positive vessels in each section were evaluated, with approximately 10 vessels/heart analyzed.

| Statistics
Data were presented as mean ± standard deviation (SD) where appropriate. Mann-Whitney tests were used for analysis of nonparametric data. Two-way analysis of variance (ANOVA) was employed for comparison of antinuclear and anti-vimentin autoantibody responses. Graft survival was depicted using Kaplan-Meier analysis and groups compared by log-rank (Mantel-Cox) testing. Analysis was conducted using GraphPad 4 (GraphPad Software, San Diego, CA).
Values of P < .05 were considered significant.

| Different CD4 T cell lineages are released from human allografts
Having previously demonstrated the presence of CD4 T effector cells within human organs recovered for transplantation, 6,12 we sought to determine whether donor CD4 T cells, and specifically, donor T-regs, could potentially also be released into the recipient's circulation following transplantation. Human lung transplant recipients (n = 21) were therefore followed for the first year following transplantation, and the presence of circulating donor-derived CD4 T cells determined by surface expression of mismatched HLA donor antigen ( Figure S1). As shown in Figure 1, donor-derived CD4 T cells were detectable immediately following transplantation in all patients, representing between 0.06% and 6% of the total CD4 T cell population detectable in the recipient (mean chimerism at 1 week; 1.54 ± 1.41%). Numbers of cells recovered were too small to defi- The release of donor T-regs was then assessed by analysis of leukocyte filters recovered from human kidneys that had been obtained using standard recovery techniques, but then perfused normothermically ex vivo using leukocyte-depleted blood. 2 Hence leukocytes captured by the filter in the circuit reflect those cells that would be released into the recipient circulation had the organ been transplanted without first being subject to ex vivo perfusion.
CD4 T cells were readily recovered from the filters and represented 6.57 ± 1.30% of the total lymphocyte population ( Figure 1B

| T-reg depletion results in augmented humoral immunity and accelerated allograft rejection
The influence of donor and recipient T-regs on allograft outcomes was then examined using an MHC class II-mismatched murine model of chronic heart allograft rejection. Our previous work has highlighted that chronic allograft vasculopathy (CAV) in this model is associated with the development of effector autoantibody responses that are triggered by graft-versus-host recognition of MHC class II on host B cells by passenger donor CD4 T lymphocytes. 6,12,13 In comparison to unmodified WT C57BL/6 recipients, depletion of the T-reg population by administration of anti-CD25 mAb to C57BL/6 mice at, and following, transplantation with bm12 (B6(C)-H2-Ab1bm12/KhEgJ) heart allografts resulted in much more rapid heart graft rejection, and was associated with markedly augmented host autoantibody responses (Figure 2A,B). This accelerated rejection was nevertheless dependent on adoptive transfer of donor CD4 T cells, because heart  Figure 2A). This suggests that the T-regs were principally influencing the donor T cell/host B cell axis.

| Donor-derived T-regs prolong allograft survival more effectively than recipient T-regs
In the preceding experiments, anti-CD25 treatment of the recipient was continued after transplantation, raising the possibility that F I G U R E 1 Solid organ human transplants contain passenger CD4 T lymphocyte subsets. A, Donor HLA class I mismatched antigens were used as a target for detection of donor CD4 T cell chimerism in lung transplant recipients using flow cytometry.

| D ISCUSS I ON
Our results demonstrate that following solid organ transplantation, donor-derived CD4 T cells are released into the recipient circulation, and, at least following lung transplantation, may persist for some time. Within a larger population of conventional CD4 T effector cells, smaller numbers of regulatory T cells can be identified, and our murine studies confirm that these can inhibit host adaptive immune responses. These findings may hold particular pertinence to ex vivo organ perfusion strategies currently being developed; they highlight that rather than blanket depletion, preservation of select passenger lymphocyte subsets within the allograft may be beneficial.
It is perhaps surprising that donor-derived nT-regs were more effective than recipient-derived nT-regs at blocking host humoral responses. Although the precise target epitopes remain ill-defined, 16,17 nT-regs are thought to recognize specific, self-restricted peptide epitopes (typically autoantigens 18 ). Donor-derived nT-regs therefore presumably recognize intact host MHC class II complexes on recipient cells via the direct pathway, 19 and in which case, do so with a much greater precursor frequency than for a self-restricted response, with approximately 5% of the clonal repertoire responding. 20 We have recently demonstrated that this enables naïve donor T cells to provide promiscuous, "peptide-degenerate" help to all host   donor with recipient alloantigen 6 weeks prior to heart donation) are rejected much more rapidly than hearts from unmodified donors, with greatly augmented autoantibody responses. 6,28 Such a use of third-party T-regs to block host humoral alloimmunity would be distinctly different from proposed strategies that differentiate/expand T-regs with self-restricted specificity for alloantigen from the individual's endogenous T cell population, 25,33 and may offer a particular advantage. T cell help for alloantibody production can only be provided by host CD4 T cells with indirect allospecificity. [34][35][36] Thus, for maximum effectiveness, recipient-derived T-regs would need to recognize the relevant allopeptide epitope presented by host MHC class II. Prediction of these peptides is, however, challenging, not least because the repertoire of presented allopeptide peptides may change with time. 37  This approach may have wider uses beyond transplantation. It could, for example, be refined as a potential treatment for humoral autoimmunity, wherein nT-regs from a third-party donor that have direct-pathway allospecificity for the individual's (recipient's) MHC class II antigens would be expected to block cognate interactions between autoreactive B and T helper cells in the host, thereby inhibiting autoantibody production.

D I SCLOS U R E
The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.