Postoperative administration of donor B cells induces rat kidney allograft acceptance: lack of association with TH2 cytokine expression in long-term accepted grafts¹

Abstract

Background.  

Although donor leukocytes are only thought to prolong survival when administered before transplantation, recent evidence shows that they are effective at transplantation. This study aims to identify the leukocyte subset that is most active in prolonging kidney allograft survival and examine the cytokine expression in long-term acceptance.

Methods.  

PVG rat kidneys were transplanted to completely MHC class I and class II-mismatched DA recipients. Donor B cells or T cells, purified by negative selection, were injected i.v. at the time of transplantation. Expression of interleukin (IL)-2, IL-4, IL-10, interferon (IFN)-γ, and transforming growth factor-β mRNA was measured by quantitative real-time polymerase chain reaction (PCR). Immunohistochemical analysis and terminal deoxynucleotide transferase-mediated dUTP nick-end labelling (TUNEL) staining was used to identify infiltrating cells and apoptotic cells, respectively, in sections of kidney allografts.

Results.  

Median kidney graft survival time (MST) of B cell-treated animals (n=5) was >300 days, compared with 7 days in untreated animals (n=7) (P =0.003), whereas animals treated with the same number of T cells (n=6) had a MST of 17 days (P =0.1 vs. untreated, P =0.03 vs. B cell-treated). Examination of the long-term (>300 days) accepted grafts from B cell-treated recipients showed little evidence of kidney damage but a moderate perivascular infiltrate consisting of T and B cells. This infiltrate seemed to be quiescent because there was no detectable expression of IL-2 receptors or of apoptotic cells. It produced little or no cytokine mRNA, because expression in the long-term accepted grafts was similar to levels in normal kidneys or syngeneic transplants. There was a marked increase of cytokine mRNA early after transplantation in both leukocyte-treated and untreated grafts, with more rapid appearance of IFN-γ and IL-10 in leukocyte-treated grafts.

Conclusions.  

Donor B cells efficiently induce long-term acceptance of transplanted kidneys in a fully MHC-mismatched rat model when administered at transplantation, by a mechanism that seems to be independent of Th2 cytokine expression within the long-term accepted graft.

Administration of donor antigen in the form of donor-specific blood transfusion (DST) ^{(1, 2)} or donor leukocytes ⁽³⁾ prolongs allograft survival in animal models and in clinical transplantation. Early studies showed that this treatment was only effective when administered a week or more before transplantation, which limits its utility for clinical transplantation because it cannot be used for cadaver donors and has the potential to presensitize the recipient ^(4–6, and reviewed in ⁷⁾. Administration at the time of transplantation avoids these problems, and more recent data suggests that donor leukocytes can prolong transplant survival when given at this time. Donor leukocytes, which are transferred with a transplanted liver, are responsible for spontaneous acceptance of the liver in a rat model ^(8,9). Also spleen allografts, which are rich in donor leukocytes, are spontaneously accepted in some rat strains ⁽¹⁰⁾. Furthermore, administration of donor spleen leukocytes at the time of transplantation prolongs transplant survival, although with less potency than pretransplant administration ^(11,12), and we have recently shown that administration of donor spleen leukocytes at completion of the transplant operation leads to long-term acceptance of rat kidney or liver allografts ⁽¹³⁾. Donor leukocytes used in these experiments have been mixed populations, and the identity of the subpopulation that is primarily responsible has yet to be defined.

The role of cytokines in the induction phase of allograft rejection and tolerance remains controversial, because some studies support a role for Th1 cytokines, such as interleukin (IL)-2 and interferon (IFN)-γ, in rejection and for Th2 cytokines, such as IL-4 and IL-10, in acceptance (reviewed in ¹⁴). There is also evidence to the contrary, which shows a requirement for the Th1 cytokines, IL-2 and IFN-γ, in the development of allograft tolerance, although they are not essential for rejection ^(15–17). Studies from our laboratory have shown increased expression of Th1 cytokines in the recipient lymphoid tissues during tolerance that far exceeded the levels in these tissues during rejection. This expression was observed in spontaneous acceptance of transplanted livers ⁽¹⁸⁾ and in donor leukocyte-induced acceptance of transplanted kidneys ⁽¹³⁾. Subsequent to immune activation, apoptosis of activated leukocytes occurred in the recipient lymphoid tissues and graft ^(13,18–20), suggesting that induction of graft acceptance involved early Th1 cytokine-associated activation of recipient T cells followed by their death by apoptosis. The role of Th1 and Th2 cytokines in the maintenance phase of stable allograft tolerance is also controversial.

The aim of the present study was to further investigate the role of donor leukocytes in induction of transplant tolerance by examining the ability of donor B cell and T cell populations to promote acceptance of kidney allografts in a rat model. In addition, the expression of intragraft Th1 and Th2 cytokines during the induction phase of kidney transplant rejection or spleen leukocyte-induced acceptance was examined. The involvement of cytokines in the maintenance phase of kidney transplant acceptance induced by donor B lymphocyte administration was investigated by studying the intragraft expression of Th1 and Th2 cytokines and of transforming growth factor (TGF)-β in long-term surviving kidney allografts.

MATERIALS AND METHODS

Animals

Male PVG (RT1^c) rats, weighing 200–300 g, were used as donors and male DA (RT1^a) rats, weighing the same, were used as recipients. These strains differ at both the class I and class II MHC loci and at multiple minor loci. Animals were purchased from the Animal Resources Centre, Perth, Australia. They were cared for in accordance with the guidelines of the Animal Ethics Committee of the University of Sydney and the Australian code of practice for the care and use of animals for scientific purposes. All surgical procedures were conducted with Fluothane inhalation anesthesia and aseptic technique.

Kidney Transplantation

Orthotopic kidney transplantation was performed as previously described with removal of the contralateral native kidney 3 days after transplantation ⁽¹²⁾. In some experiments, recipients were given donor leukocytes by i.v. injection of cells in 1 ml of phosphate buffered saline via the penile vein. Unseparated spleen leukocytes, at a dose of 6×10⁷ per animal, were injected into the animals that were killed on days 1, 3, and 5 after transplantation for examination of cytokine expression. Negatively selected T or B cell populations were injected as described. Syngeneic kidney transplants consisted of a PVG kidney transplanted to a PVG recipient with subsequent removal of the contralateral native kidney.

Purification of Spleen Cell Subpopulations

Donor (PVG strain) spleen cells were obtained as previously described by dissociation of the spleen and lysis of red blood cells in NH₄Cl buffer ⁽²⁰⁾. For purification of α/β T cells by negative selection, 1.5×10⁸ spleen cells were incubated with an antibody cocktail that labeled B cells (OX12, reactive with rat kappa light chain and MARM-4, reactive with rat immunoglobulin (Ig)M), monocytes/macrophages/dendritic cells (OX42, reactive with CD11b/c), γ/δ T cells (V65/5/7), and natural killer cells (10/78). For negative selection of B cells, 1.5×10⁸ spleen cells were incubated with a cocktail of antibodies that included R73 (reactive with α/β T cells), V65/5/7, 10/78, and OX42. MARM-4 was obtained from Serotec, Oxford, United Kingdom and used at 1:100 dilution. All other antibodies were obtained as tissue culture supernatants from Dr. J. Sedgwick, DNAX, Palo Alto, CA, and were used undiluted. Antibodies were dialyzed extensively to remove sodium azide preservative, stored at −20°C before use, and 100 μl of each was added to the cell pellet. After incubation for 30 min at 4°C the cells were washed with phosphate-buffered saline (PBS) + 1% fetal calf serum (FCS), then incubated with antibody to mouse IgG conjugated to fluorescein isothiocyanate (Sigma, code F2266) at 1:100 concentration in 5% normal rat serum and washed. The labeled cells were incubated with 150 μl of anti-FITC microbeads (Miltenyi Biotec, Gladbach, Germany, code 487-01), washed, and made up to 750 μl in PBS + 1% FCS. Cells were applied slowly to a washed LS separator column (Miltenyi Biotec) and slowly washed with 9 ml of PBS + FCS. The eluate, which contained the negatively selected cells, was collected and an aliquot analyzed for purity by flow cytometry. Cells were resuspended in 1 ml of PBS and injected at the dose described into the penile vein at completion of the kidney transplant operation. Purity of negatively selected T cells, as measured by staining with R73 antibody, was >97% and of negatively selected B cells, with antibody OX12 was >95%. Immunostaining for flow cytometry was performed as previously described ⁽²⁰⁾.

Histological and Immunohistochemical Analysis

Conventional hematoxylin and eosin staining of paraffin sections was performed and analyzed by a trained pathologist (D.M.P.). Frozen sections of 6-μm thickness were cut onto gelatin-coated slides and stained by an indirect immunoperoxidase method as previously described ⁽²¹⁾. Antibodies to rat α/β T cells (R73); activated IL-2R⁺ cells (OX39); CD11b/c⁺ cells (monocytes, macrophages, granulocytes, and dendritic cells) (OX42); donor PVG cells identified with an antibody reactive with PVG MHC class I antigen (OX27); CD45RC-expressing cells (principally B cells as described in reference ²¹) identified by OX22; anti-rat IgM and anti-rat IgD (Serotec, Oxford, UK); and a negative control antibody (MOPC-21) were used as previously described ⁽²¹⁾. The secondary antibody was rabbit anti-mouse immunoglobulin conjugated to horseradish peroxidase (Dako, Copenhagen, Denmark). The positive immunostaining cells were counted at ×400 magnification with the aid of a microscope eyepiece graticule. Ten random fields were observed in each of the kidney perivascular or interstitial areas. All labeled cells within the graticule area were tallied, and the sum of the cells from 10 fields was multiplied by a factor of 1.6 to yield the cell count/mm².

Apoptosis Assay

Apoptotic cells were identified in frozen sections by terminal deoxynucleotide transferase-mediated dUTP nick end labeling (TUNEL) assay using a commercial kit (in situ cell death detection kit POD; Boehringer-Mannheim, Germany) according to the manufacturer’s instructions. Apoptotic cells were counted as described above.

Measurement of Cytokine mRNA

Isolation of total RNA from kidney allografts and reverse transcription was by techniques previously described ⁽²²⁾. cDNA was stored in aliquots, and all samples were measured by quantitative real-time PCR in a single run for each cytokine to minimize run-to-run variation. A cDNA standard curve was prepared for each cytokine. This consisted of cDNA from a sample with high expression diluted in a 10-fold series from neat to 1:1000. Duplicate cDNA standards were included for each cytokine for each run. The cDNA standard curve gave similar results to those obtained using standards derived from purified cytokine PCR product as previously described ⁽²³⁾. Quantitative real-time PCR for rat cytokines was performed in a 25 μl reaction volume in Universal master-mix (PE Applied Biosystems, Foster City CA, code 4304437) as previously described ⁽²³⁾. The sequences for amplification primers and fluorogenic probes for measurement of rat glyceraldehyde phosphate dehydrogenase (GAPDH), IL-2, IFN-γ, and TGF-β have been previously published ⁽²³⁾. The sequences for rat IL-4 were: IL-4FOR, CAG AAA AAG GGA CTC CAT GCA; IL-4REV, GCT CGT TCT CCG TGG TGT TC; IL-4PRROBE, 6FAM-AGA TGT TTG TAC CAG ACG TCC TTA CGG CAA-TAMRA. For rat IL-10 the sequences were: IL-10FOR, CGA CGC TGT CAT CGA TTT CTC; IL-10REV, TCT TGG AGC TTA TTA AAA TCA TTC TTC A; IL-10PROBE, 6FAM-TGT GAG AAT AAA AGC AAG GCA GTG GAG CA-TAMRA. The length of the IL-4 amplicon was 78 base pairs and the IL-10 was 83 base pairs. Reactions were amplified for 45 cycles on a Model 7700 sequence detector (PE Applied Biosystems), and the level of expression was calculated from the standard curve whose values were given as the inverse of the standard cDNA dilution. For example, the 1:1000 dilution of the standard cDNA was assigned a value of 1 U, whereas the undiluted standard cDNA had a value of 1000 U. Expression of cytokine was normalized to expression of the housekeeping gene, glyceraldehyde-3-phosphate dehydrogenase (GAPDH), by dividing the value for the cytokine by the corresponding value for GAPDH for that sample and multiplying by 100.

Statistical Analysis

Kidney graft survival was analyzed with SPIDA Survival Statistical Software (Macquarie University, Sydney, Australia) using log-rank analysis of the product-limit estimate of Kaplan and Meier. Analysis of quantitative data from cytokine mRNA analysis and immunohistochemical staining was by unpaired t test.

RESULTS

Effect of Treatment with Donor Leukocyte Subsets on Renal Allograft Survival

Kidney allografts in untreated animals in the PVG→DA strain combination had a median survival time (MST) of 7 days (Table 1). Treatment with 6×10⁷ unseparated PVG spleen leukocytes at completion of the transplant operation led to prolonged survival of the transplanted kidneys, with a MST of >300 days (Table 1) (data obtained from our published study, reference ¹³). Treatment with 6×10⁷ purified donor B lymphocytes prolonged survival compared with untreated (P =0.003), with a MST of >300 days. The same number of purified donor T cells (6×10⁷ cells) did not significantly prolong kidney allograft survival, with a MST of 17 days (P =0.10 compared with untreated). Thus, B cells were significantly (P =0.03) more effective than the same number of T cells in their ability to prolong renal allograft survival. Donor T cells at a lower dose (2×10⁷ cells/animal) did not significantly prolong graft survival, with a MST of 7 days (P =0.40 compared with untreated). Treatment with donor T cells was not completely ineffective, however, because there was 1 of 8 animals that survived >300 days in the higher dose T cell treatment and 1 of 6 in the lower dose treatment (Table 1).

Table 1: Induction of long-term acceptance of PVG→DA kidney allografts with donor spleen cells or spleen cell subpopulations

Rejection, not GVHD, is the Cause of Reduced Survival in T Cell-Treated Animals

Animals treated with purified donor T cells had reduced survival compared with animals treated with donor B cells. This might have been due either to kidney rejection resulting from the inability of T cells to induce acceptance or to donor T cell-dependent graft-versus-host disease (GVHD) resulting from expansion of the allogeneic donor T cells. Several lines of evidence indicate that rejection and not GVHD was responsible for the poor survival. Examination of the transplanted kidney showed that there was an extensive infiltrate in the interstitial areas between tubules that consisted of recipient, not donor, cells (Fig. 1A). There was also strong expression of donor MHC class I on kidney tubular cells, consistent with rejection (Fig. 1A), and extensive deposition of IgG1 on kidney tubules and a heavy infiltrate of IL-2 receptor-expressing (IL-2R⁺) cells (data not shown). Donor cells were absent from the recipient spleen of T cell-treated animals (Fig. 1B), showing that there was no GVHD-associated proliferation of donor T cells at that site. The overall appearance of the animal was consistent with rejection because there was no sign of hair loss or reddening of the skin on paws and ears, which are hallmarks of GVHD. Mortality from acute renal allograft rejection tends to be faster than from GVHD and is consistent with the short survival of most of the T cell-treated animals (MST of the pooled T cell-treated groups=7 days).

Figure 1: (A, B) Transplanted kidney (A) and recipient spleen (B) obtained 7 days after transplantation from a kidney transplant treated with 6×10⁷ donor T cells. (A) Strong expression of donor MHC class I antigen, identified with OX27 antibody, on tubular cells of the graft but not the infiltrate, showing that graft-versus-host disease is an unlikely cause of kidney dysfunction. (B) Recipient spleen, showing complete absence of injected donor T cells, identified with OX27⁺ antibody. (C–F) Long-term surviving (>340 days) kidneys from donor B cell-treated kidney transplants. (C) Hematoxylin and eosin staining of formalin-fixed tissue, showing little evidence of damage despite a moderate perivascular infiltrate. (D–F) Serial sections, showing immunohistochemical characterization of the perivascular infiltrate from long-term accepted grafts identified large numbers of α/β T cells (D) and B cells, stained with anti-rat IgD (E). No IL-2 receptor-expressing cells were present in the perivascular infiltrates (F).

Microscopic Analysis of Long-Term Accepted Kidney Allografts

B cell-treated kidney graft recipients survived for >300 days and showed no evidence of graft dysfunction. Animals were killed with a functioning graft and conventional hematoxylin and eosin staining of formalin-fixed kidney tissue revealed scant evidence of chronic rejection, although all grafts had been in place for between 340 and 370 days (Fig. 1C). There was little or no sign of intimal thickening of vessels in the long-term functioning grafts, although around occasional vessels and in some peritubular areas there were considerable numbers of infiltrating leukocytes. The T cell-treated animals, most of which rejected in 6–10 days, showed histological evidence of rejection. One T cell-treated animal that survived for 280 days and two that survived >300 days showed no evidence of rejection. Examination of kidney allografts from untreated animals, which were rapidly rejected, showed evidence of acute rejection with a heavy infiltrate, tubular and glomerular damage, and vascular occlusion.

Immunohistochemical staining of long-term accepted kidneys identified a mixed infiltrate of T cells (Fig. 1D) and OX22⁺ cells. The latter co-localized with cells expressing membrane IgM or IgD (Fig. 1E) in serial sections and confirmed that the OX22⁺ cells were predominantly B cells. These infiltrates were located in perivascular and peritubular areas. The extent of T cell and OX22⁺ cell infiltrate in long-term tolerant grafts is compared with their infiltration early after transplantation in Figure 2. Perivascular infiltrates of T cells in the long-term accepted kidneys (634±213 cells/mm²) were intermediate between the level in normal kidney (14±4) and in the day 5 grafts that rejected (2476±437) or that had been treated with donor spleen cells and were in the process of being accepted (1735±155) (Fig. 2A). The extent of T cell interstitial infiltration in the long-term accepted kidneys was significantly less (P =0.0004) compared with the interstitial infiltrate early after transplantation (Fig. 2C). Similarly, there was a moderate perivascular infiltrate of OX22⁺ cells in the long-term accepted grafts compared with the infiltrate early after transplantation (Fig. 2B) but no evidence of an interstitial infiltrate of these cells in the long-term grafts (Fig. 2D).

Figure 2: Characterization of the infiltrate in PVG→DA kidney allografts and normal PVG kidneys. Kidney allografts were from untreated PVG→DA (Rej) recipients on day (d) 5, from donor spleen cell-treated recipients that ultimately developed tolerance to their graft (Tol) on d 5 ⁽¹³⁾, or from donor B cell-treated recipients that survived long-term and were killed with a functioning graft at >340 days (long-term). Infiltrate was measured by immunohistochemical staining of kidney tissue with R73 antibody to α/β T cells or with OX22 antibody to CD45RC. Cells with a perivascular location (A, B) or interstitial location (C, D) were counted. Results show the mean±SD for at least three different animals.

Despite the observation of moderate numbers of leukocytes infiltrating the long-term accepted grafts with a predominantly perivascular disposition, there was no evidence of IL-2R⁺ cells in the infiltrate (Fig. 1F). This was not an artefact, because the same antibody which identified large numbers of IL-2R⁺ cells in rejecting kidneys was used for these long-term grafts. Also, there was little or no evidence of leukocyte apoptosis in these infiltrates as TUNEL staining of the sections revealed 8±5 cells/mm² in the perivascular areas compared with 4±5 cells/mm² in the interstitial areas and 4±1 cells/mm² in the perivascular and interstitial areas of normal PVG kidney. There was no significant difference in apoptosis between these three results.

Cytokine mRNA Expression in Kidney Allografts

Expression of mRNA of the Th1 cytokines, IL-2 and IFN-γ, is shown in Figure 3. The results are normalized to the expression of the endogenous housekeeping gene GAPDH; however, the pattern of expression and the significance of the differences was the same regardless of normalization. Interestingly, there was a marked increase of these cytokines early after transplantation in both rejection and leukocyte-induced acceptance. There was no significant difference between kidneys from untreated or treated recipients on day 5 for IL-2 (P =0.40) or IFN-γ (P =0.47). The most notable difference between kidney acceptance and rejection was the more rapid appearance of IFN-γ within the accepted grafts (Fig. 3) in which there were 313±117 U/GAPDH×100 on day 3 compared with 56±32 in rejection (P =0.01). In the long-term accepted kidney allografts, there was little expression of IL-2 or IFN-γ compared with their levels early after transplantation (Fig. 3). Detailed analysis of expression of cytokines in the long-term accepted transplants compared with normal or syngeneic grafts is shown in Table 2. There was no significant difference between long-term accepted kidney allografts and normal kidneys or syngeneic transplants for expression of IFN-γ. IL-2 expression in long-term accepted kidneys was slightly higher than in normal kidneys (P =0.04) but not significantly different to syngeneic transplants (P =0.06).

Figure 3: Expression of the Th1 cytokines, IL-2 and IFN-γ, of the Th2 cytokines, IL-4 and IL-10, and of TGF-β in kidney allografts, normal PVG kidneys, and PVG→PVG kidney grafts (Syn PVG). Cytokine mRNA expression was measured by quantitative real-time reverse-transcriptase PCR. Kidney allografts were from untreated PVG→DA recipients (Rej) shown as black bars, from donor spleen cell-treated recipients that ultimately developed tolerance to their graft (Tol) ⁽¹³⁾, or from donor B cell-treated recipients that survived long-term and were killed with a functioning graft at >340 days (long-term), shown as white bars. Syngeneic PVG→PVG kidney grafts (Syn PVG), analyzed on day (d) 3, are shown as grey bars. Results show the mean±SD for at least three different animals at each time point. *Significant difference between Tol and Rej on day 3 (P =0.01).

Table 2: Expression of cytokine mRNA in long-term accepted kidney compared with normal kidney or PVG→PVG kidney transplants

Expression of the Th2 cytokines, IL-4 and IL-10, was increased in rejected kidneys as well as in those treated with donor leukocytes, which were accepted long-term (Fig. 3). This pattern of expression was similar to the Th1 cytokines. There was no significant difference between acceptance and rejection in the expression of IL-4 mRNA; however, there was more rapid expression of IL-10 in kidneys from leukocyte-treated animals. In these there were 419±125 U/GAPDH×100 on day 3 compared with 12±6 in untreated, rejecting grafts (P =0.03). Thereafter, the expression of IL-10 in the rejecting grafts rapidly increased, resulting in similar levels in accepted and rejecting grafts on day 5. In the long-term accepted grafts, the levels of IL-4 and IL-10 mRNA declined to the levels observed in normal kidneys or in syngeneic kidney grafts (Fig. 3, Table 2).

Expression of TGF-β increased slightly in both accepted and rejecting kidneys after transplantation (Fig. 3). The level in rejecting animals was increased compared with normal kidneys on days 3 and 5 after transplantation, and in the accepted kidneys it was increased on days 1, 3, and 5 after transplantation. The significance of the increase in allogeneic kidney transplants is unclear, because there was also an increase of expression of TGF-β in the syngeneic transplants, suggesting that raised TGF-β mRNA is a consequence of the transplant operation, rather than a result of the immune response. TGF-β in long-term accepted grafts was not significantly different to the levels in normal kidney but less than the level in syngeneic grafts on day 3 (Table 2).

DISCUSSION

These studies show that when they are administered at completion of the transplant operation, donor B cells more effectively induced long-term acceptance of completely mismatched rat kidney allografts than donor T cells. Four of five animals treated with donor B cells survived for >300 days and were killed with a functioning kidney. In contrast, only one of eight animals treated with the same dose of donor T cells survived >300 days. Reduced survival in the T cell-treated animals was not due to donor T cell-induced GVHD, because there was no evidence of GVHD in the recipient and histopathological analysis of kidneys from T cell-treated animals showed rejection. Examination of long-term accepted kidneys from B cell-treated allograft recipients showed a moderate perivascular and peritubular infiltrate of T cells and CD45RC⁺ cells. These latter seemed to be surface IgM- or IgD-expressing B cells. Cells in these infiltrates seemed to be quiescent, because there was no evidence of IL-2 receptor expression or of cell apoptosis in areas of infiltrate. Cytokine expression in the long-term accepted grafts was similar to the levels found in normal or syngeneic kidneys and was much reduced compared with the levels found early after transplantation in rejection or in grafts from recipients treated with donor spleen cells that were accepted long-term.

There have been numerous studies of the effectiveness of donor T and B cells, of bone marrow cells, and of other donor cell populations, such as red cells, platelets, hepatocytes, immature dendritic cells, and fibroblasts for pretreatment of transplant recipients. All of these cell populations are capable of prolonging survival of transplanted organs when administered before transplantation ^(4,24–30). There is little information, however, regarding the effectiveness of donor cell populations administered at the time of transplantation, despite evidence that donor leukocytes can prolong graft survival in some circumstances, even though they have not been present before transplantation ^(8–10).

This study is the first to show that B cells administered at completion of the transplant operation induce long-term allograft acceptance. It extends our previous findings that unseparated donor spleen cells induce renal allograft tolerance in this model ⁽¹³⁾. This is consistent with the ability of donor B cells and of bone marrow, which is rich in B cells, to induce alloimmune unresponsiveness when administered before transplantation ^(24,25). Donor B cell induction of T cell unresponsiveness is also consistent with the constitutive expression of MHC class II on B cells. This allows them to directly interact with recipient CD4 T cells, the cells that are responsible for allograft rejection in this rat strain ^(31,32) and that are rapidly activated during leukocyte-induced tolerance ⁽¹³⁾. Interaction is likely to occur in the recipient lymphoid tissues, as we have previously demonstrated that spleen cells, which contain approximately 40% B cells, migrate rapidly to these tissues in donor leukocyte-induced kidney transplant tolerance ⁽¹³⁾. In contrast, rat T cells do not normally express MHC class II, which means that donor T cell interactions are limited to recipient CD8 T cells, which are not necessary for rejection in this strain ⁽³¹⁾.

Previous studies of donor leukocyte-induced graft acceptance have proposed that a limited GVH reaction might be responsible for the ability of these cells to promote tolerance ⁽³³⁾. Consistent with this hypothesis, several studies, including one from this laboratory, have indicated that donor T cells might play a role in graft acceptance, raising the possibility of a limited GVH response to recipient alloreactive T cells ^(11,12). In both of these studies, however, the effect of donor T cells was weak, with only slight prolongation of survival noted in response to donor cells ⁽¹¹⁾ or only slight reduction in survival when donor T cells were removed ⁽¹²⁾. The studies reported here show no evidence for a limited GVH reaction in donor T cell-treated recipients, with grafts being lost due to rejection, not GVHD, even with the higher dose of donor T cells.

Examination of the kidney allografts that were accepted long-term after postoperative administration of donor B cells showed that there was a moderate perivascular infiltrate. Despite the presence of this infiltrate, there was no visible damage to the grafts, which showed almost normal histology, although they had been in place for more than 300 days without immunosuppression. The infiltrate consisted mainly of T cells and B cells that did not appear to be activated, because IL-2 receptor expression was virtually absent. Also there was no evidence of apoptosis within these infiltrates. Thus, the infiltrates observed during the maintenance phase of long-term tolerant kidney transplants differed greatly from those during the induction phase early after transplantation, which expressed high levels of cytokines. Infiltrates during the induction phase of tolerance also contained many IL-2 receptor-expressing cells and apoptotic cells ⁽¹³⁾.

Perivascular infiltrates in long-term accepted grafts have been reported in a number of models of transplant tolerance ^(34–36). This raises the possibility that the infiltrate contains suppressor T cells, which maintain immunological unresponsiveness to the graft. Suppressor cells are thought to produce Th2 cytokines, because expression of IL-4 and IL-10 mRNA within the graft has been associated with long-term graft acceptance ⁽³⁷⁾. In addition, T cell clones isolated from tolerant animals have a Th2 phenotype and can suppress Th1 clones and delayed-type hypersensitivity responses ⁽³⁸⁾. Some models of transplant tolerance also show a requirement for IL-4 ⁽³⁹⁾ and an involvement of IL-10 in transferable transplant tolerance mediated by regulatory T cells ⁽⁴⁰⁾.

Our inability to detect increase of mRNA for the Th2 cytokines, IL-4 and IL-10, or of TGF-β in the long-term accepted grafts is at variance with the above findings. It is not due to the current methodology, because the quantitative PCR used here has been shown to be extremely sensitive and reproducible and capable of resolving small differences in cytokine message expression ⁽²³⁾. The difference may be a reflection of differences between the models, and several studies of long-term graft acceptance have shown results similar to those reported here, with little increase in either Th1 or Th2 cytokines observed within the graft ^(41,42). If Th2 cytokines are involved in maintenance of long-term tolerant allografts, then perhaps they might inhibit potentially alloreactive T cells at sites outside the graft, possibly within the recipient lymphoid tissues.

Our findings show that during the induction phase of tolerance, within the first week after transplantation, there are more rapid increases of IFN-γ and IL-10 cytokine mRNA expression in donor leukocyte-treated kidney allografts that are accepted compared with rejection. These results are consistent with previous findings that showed that there are rapid increases of IL-2, IFN-γ, and IL-10 mRNA in the recipient spleen or lymph nodes during donor leukocyte-dependent kidney or liver allograft acceptance, which far exceeds the level in rejecting animals ^(13,18,43). The results here, showing more rapid increases of IFN-γ and IL-10 within the transplanted kidneys of animals that are accepting the graft, suggest that the cytokine-expressing cells migrate from the recipient lymphoid tissues to the transplanted kidney. They also accord with our previously reported findings of more rapid increase of intragraft Th1 cytokine expression in spontaneous acceptance of transplanted liver, compared with rejection of untreated kidney allografts ⁽⁴⁴⁾ or skin transplants ⁽⁴³⁾. These results, showing a rapid increase of Th1 cytokines in transplant acceptance, are also consistent with the requirement for Th1 cytokines, IL-2 and IFN-γ, in some forms of transplant tolerance ^(15–17).

The T cells observed in perivascular aggregates in the long-term accepted kidney transplants did not appear to express IL-2 receptors. This is consistent with a quiescent state and supported by lack of cytokine mRNA expression, although it is not in accord with the finding that CD4⁺ T cells that express the IL-2 receptor are the main cell population responsible for suppression of rejection of long-term tolerant grafts ^(45,46). There is also evidence from nontransplant models that CD4⁺IL-2R⁺ T cells are responsible for suppression ⁽⁴⁷⁾. The difference might be due to the site of action of these suppressor cells, and the role of activated CD4 T cells within the graft and within the recipient lymphoid tissues in the maintenance of graft acceptance is yet to be resolved.

In conclusion, donor B cells injected at the time of transplantation led to long-term acceptance of kidney allografts that were otherwise rapidly rejected in the absence of such treatment. Donor T cells were much less potent than B cells in their ability to induce tolerance but did not cause GVHD. This suggests that treatment of graft recipients at the time of transplantation with donor lymphocytes may be beneficial to graft survival. Kidney graft recipients so treated maintained the graft long-term, with little evidence of damage. The long-term accepted grafts had perivascular infiltrates of T and B cells that appeared to be quiescent, because they did not express IL-2 receptor and were not associated with consistently increased expression of mRNA for any of the cytokines tested. Donor leukocyte-induced kidney graft acceptance was, however, associated with rapid intragraft increases of Th1 cytokines soon after transplantation, suggesting that the selection of treatments to synergize with donor leukocyte-induced acceptance should take into account the possibility that these might interfere with the immune activation that accompanies this form of acceptance.