Scintigraphic imaging of MHC class II antigen induction in mouse kidney allografts: a new approach to noninvasive detection of early rejection

Mice with kidney transplants were investigated to determine whether early kidney allograft rejection could be detected by radioimmune scintigraphy targeting major histocompatibility complex (MHC) class II antigens induced on donor organ cells. Allografts from C3H/He (H2^k) donors were transplanted into BALB/c (H2^d) recipients. Each mouse was injected intravenously with 100 Ci of ¹²³I-labeled anti-MHC class II monoclonal antibody (mAb; Y17, anti-IE^k) 16 h before scintigraphy. After imaging, mice were sacrificed for tissue counting and histopathological examination. Radiotracer uptake in the nontreated allografts increased starting on the 3rd day after transplantation, peaked at around the 6th day, and then gradually decreased. Rejecting allografts with only focal perivascular mononuclear cell infiltration could be identified by scintigraphy. However, allografted mice without evidence of rejection and isografted mice did not show an increase in radiotracer uptake. Rejecting BALB/c kidney transplanted into C3H/He mice did not show an increase in Y17 mAb uptake, suggesting that class II antigens induced on donor kidneys are solely responsible for the mAb uptake in positive scintigrams of rejecting allografts. Five allografted mice were treated with anti-CD3 mAb and cyclosporin starting 3–9 days after transplantation. Radiotracer uptake decreased after 4 weeks of treatment and increased 2 weeks after the cessation of immunosuppressive treatment, reflecting suppression and recurrence of rejection, as determined by histological examination. These changes could be followed scintigraphically. We conclude that changes in class II antigen expression can be assessed by the ¹²³I-labeled anti-MHC class II antigen mAb and that it is a sensitive and noninvasive method for detecting kidney allograft rejection.     

Lack of correlation between the induction of donor class I and class II major histocompatibility complex antigens and graft rejection

The induction of donor major histocompatibility complex (MHC) antigens on nonrejected and rejected rat renal allografts was compared at various times after transplantation in two strain combinations, DA-to-PVG and LEW-to-DA. Graft rejection was prevented by preoperative donor-specific blood transfusion (DST). Quantitative absorption analysis and immunohistology were performed using monoclonal antibodies specific for donor class I and class II MHC antigens. A significant increase in the expression of donor MHC antigens, both class I and class II, was demonstrated on nonrejected as well as rejected kidneys after transplantation. A kinetic analysis showed that induction of donor class I antigens was accelerated on the nonrejected grafts, and by day 5 the nonrejected kidneys showed increased expression of class I antigen when compared with the rejected grafts (a 37- vs. a 25-fold increase in expression). Increased expression of donor class I antigens persisted on the nonrejected grafts and was still detectable on long-term-surviving kidneys, 50 days after transplantation. The magnitude of class II antigen induction was similar on both rejected and nonrejected grafts (8-fold by 5 days after transplantation). Immunohistology demonstrated that class I and class II antigens were induced on identical structures in the kidney in both situations. In particular the vessel endothelia, which do not express class II antigens in normal kidney, become strongly positive in both rejected and nonrejected grafts 5 days after transplantation. Although renal allograft rejection is completely suppressed in rats given a single donor-specific blood transfusion before transplantation, graft survival cannot be explained by the lack of induction of donor MHC antigens. Donor MHC antigens are induced on these nonrejected kidney grafts, and therefore they could act as target molecules for the effector cells that mediate graft destruction. Thus the induction of donor MHC antigens on tissue allografts should not be considered as indicative of a rejection response resulting in graft destruction.     

Different patterns of donor MHC antigen induction in rat kidney allografts following active and passive enhancement

We compare the expression of donor class I and class II major histocompatibility complex antigens in DA kidney grafts transplanted to PVG recipients treated by different protocols of donor-specific immunosuppression. MHC expression was evaluated using donor-specific antibodies and assays by immunohistology and quantitative absorption analysis. PVG recipients were either untreated or treated by (A) twice-weekly intravenous injections of 0.5 ml DA blood for 12 weeks; (B) 0.5 ml DA blood intravenously at 7 days pregraft; (C) as for (B), but with the addition or oral cyclosporine at 10 mg/kg/day from the day of grafting; and (D) passive enhancement with DA anti-PVG serum. Grafts were assessed at 3, 5, and 7 days after transplantation. In untreated controls at day 3, there is a periarteriolar leukocyte infiltrate, weak or absent class II induction, but strong class I induction. Class II induction in untreated controls is maximal at day 5. We confirm that active enhancement by blood transfusion, even using the intensive protocol of twice-weekly transfusions for 3 months, results in accelerated leukocyte infiltration and accelerated donor class I and class II MHC induction. At day 3, there is an intense, diffuse leukocyte infiltration and maximal class II induction. Cyclosporine treatment of blood-transfused recipients reduced the leukocyte infiltration and MHC induction to levels seen in untreated controls--i.e., the accelerated MHC induction caused by the transfusion was partially reversible by cyclosporine. In passively enhanced recipients, leukocyte infiltration and class I MHC induction were similar to untreated controls. However, class II induction was much delayed, not being evident until day 7.     

Central role for interferon-gamma receptor in the regulation of renal MHC expression

The role of the interferon-gamma (IFN-gamma) receptor 1 (IFN-gammaR1) was investigated in the regulation of MHC expression in kidney in the basal state, in response to potent inflammatory stimuli, and after renal injury. In this study, MHC regulation in mice lacking IFN-gammaR due to targeted disruption of the IFN-gammaR1 gene (GRKO mice) was compared with regulation in 129Sv/J mice with wild-type IFN-gammaR1 genes. Basal class I expression was reduced by approximately 45% in kidneys of GRKO mice, while basal class II expression was confined to interstitial cells and was not reduced in GRKO kidneys. Recombinant IFN-gamma administration induced widespread expression of class I and II in renal tubules, arterial endothelium, and glomeruli of 129Sv/J mice, but produced no change in kidneys of GRKO mice. Potent systemic inflammatory stimuli (injections of allogeneic cells, skin sensitization with oxazolone, and injection of bacterial lipopolysaccharide) significantly induced both class I and class II expression in 129Sv/J mice, but not in GRKO mice. Acute renal injury increased local expression of class I and II in both 129Sv/J and GRKO mice, but the induction in GRKO mice was reduced compared with 129Sv/J mice. Thus, the IFN-gamma receptor plays a unique and nonredundant role in the regulation of renal MHC in the response to inflammation, in the response to renal injury, and in the basal state.     

Detailed analysis and demonstration of differences in the kinetics of induction of class I and class II major histocompatibility complex antigens in rejecting cardiac and kidney allografts in the rat

In this paper, we analyze in detail donor class I and class II major histocompatibility complex (MHC) antigen induction in heart and kidney allografts in the DA-to-PVG rat strain combination. The immunohistological techniques and quantitative absorption analyses utilize monoclonal antibodies and assay systems specific for donor class I and class II MHC antigens, to enable precise interpretation of the results in terms of the MHC antigens of the graft. Quantitative absorption analyses were performed on homogenates comprising 4-6 allografts pooled at each interval examined (days 1-5 for kidneys, days 3-7 for hearts). In the heart allografts, donor class I antigen induction begins at day 3 after transplantation and proceeds rapidly on the 4th and 5th postoperative days. The maximum level (a 10-fold increase in comparison with normal heart) occurs at day 6, and thereafter the level declines. Donor class II antigen induction in the heart allografts follows a similar pattern. In kidney allografts, it was of particular interest that donor class I induction occurred much more rapidly, being already evident on the first postoperative day, and reaching levels 20-fold greater than normal kidney by day 3. Maximum levels (approximately 30-fold that of normal kidney) of donor class I antigens were reached on days 4 and 5. Donor class II induction, by contrast, developed in kidney grafts with kinetics similar to that seen for class II induction in heart grafts (beginning at day 3 and reaching a maximum of 7-fold over normal kidney at day 5). Immunohistological studies were performed at days 1, 3, 5, and 7 after transplantation. These confirmed the early induction of donor class I antigen in the kidney allografts. In kidney, by the fifth postoperative day, all tubules in the cortex and medulla, and the arteriolar vascular endothelium, were strongly positive for class II antigens. However, the glomerulus, including the glomerular capillary endothelium, remained donor-class-II-negative, except for induction of class II antigens on Bowman's capsule. The endothelium of interstitial capillaries also probably remained class-II-negative. These results have potentially important implications for understanding the development of the rejection response.     

Donor class I and class II major histocompatibility complex antigen expression following liver allografting in rejecting and nonrejecting rat strain combinations

Orthotopic liver allografts in the nonrejecting DA-to-PVG strain combination and in the DA-to-LEW strain combination were studied at various times after transplantation for donor class I and class II MHC expression using immunohistological techniques and quantitative analyses. DA-to-DA isografts were also studied. In the isografts, weak class I induction on hepatocytes and biliary epithelium was noted from day 5, and this persisted to day 15, the last time point examined. In DA-to-PVG allografts, class I induction also appeared on hepatocytes and biliary epithelium from day 5, but was more intense than in the isografts. Nevertheless, the induction was patchy within most grafts, and in some grafts was not prominent. Quantitative absorption analyses demonstrated that the maximum increase in donor class I expression was only 3-fold over the normal liver. In the strong DA-to-LEW combination, class I induction on hepatocytes seemed to appear earlier, beginning at day 3, and was more uniform and intense than in the DA-to-PVG model from day 5. In the isografts, there was no induction of class II antigens on hepatocytes or biliary epithelium at any stage, but from days 5 to 15 there was a marked increase in the number of isolated, class II-positive cells in the hepatic lobule, probably representing class II induction in the Kupffer cells of the isografts. In DA-to-PVG allografts, biliary epithelium became class II-positive from day 5, and this persisted to day 30, the last time point examined. Weak but definite class II induction was seen on some hepatocytes from day 5 through day 30. However, the majority of hepatocytes remained class II-negative. By day 30, there was virtually no donor class II staining the sinusoids, but isolated class II-positive cells of recipient type were seen, the pattern suggesting a replacement of the graft Kupffer cells by recipient Kupffer cells at this stage. By quantitative absorption analysis, donor class II expression in the grafts increased approximately 5-fold. In DA-to-LEW allografts, class II induction was not noticeably different from that seen in the DA-to-PVG model, except that induction of class II antigens on the Kupffer cells possibly appeared earlier in this strain combination.     

The influence of MHC and non-MHC genes on the nature of murine cardiac allograft rejection. I. Kinetic analysis of mononuclear cell infiltrate and MHC-class I/class II expression in donor tissue

While normal cardiac tissue expresses low levels of MHC-class I, undetectable levels of MHC-class II antigens, and no mononuclear cell infiltrates, posttransplantation allogeneic donor cardiac tissue demonstrates dramatic increases of MHC-class I/class II expression coincident with the infiltration of the tissue with mononuclear cells. Results of this study demonstrate that the kinetics of MHC-class I/II antigen expression and the phenotype of mononuclear cell infiltrate are influenced, to a great degree, by the genetic H-2, intra-H-2 and non-H-2 incompatibility between donor and recipient strains of mice. Increases of MHC-class I precede class II expression in cells from donor cardiac tissue from completely allogeneic BALB/c, H-2-disparate B10.D2, B10.BR, and K, I-A and I-E-disparate B10.T (6R) strains of mice implanted in B10 recipients. In contrast, increase in the level of MHC-class II precedes MHC-class I increases in donor cardiac tissue from H-2-identical but non-H-2-incompatible A.By and the I-E + H-2D end-different B10.A(5R) donor tissue. The completely allogeneic, H-2-disparate or K, I-A, I-E-disparate donor cardiac tissue induced the infiltration of predominantly CD8+ T cells, whereas the non H-2 and I-E + H-2D end-different donor cardiac tissue induced the infiltration of predominantly CD4+ T cells. Finally, whereas bm1 donor cardiac tissue is rejected by B6 recipients by day 32, the (bm1 x bm12)F1 allografts are rejected by day 20, and both express MHC-class I antigens followed by MHC-class II antigens, and contain predominantly CD8+ T cells. In contrast, bm12 allografts are not rejected by B6 recipients, express chronic low levels of both MHC-class I and II antigens, and contain predominantly CD4+ T cells. Of interest is our preliminary finding that bm12 allografts placed in one ear of B6 recipients appear to modify the kinetics of MHC antigen expression and the predominant phenotype of mononuclear cell infiltrates in bm1 allografts placed in the opposite ear. Cumulatively, these data suggest that the type of genetic disparity between cardiac donor and recipient greatly influences the quantitative and qualitative host responses.     

Chronic renal allograft rejection: the significance of non-MHC alloantigens

We studied the role of polymorphic endothelial antigens other than MHC in antibody-mediated chronic renal allograft rejection in two models. In the first model, donor Lewis rat kidneys were transplanted into BN recipients that had been made tolerant for donor class I antigens at the B cell (antibody) level. In this setting Lewis kidney grafts were chronically rejected with stable renal function but increasing proteinuria (> 100 mg/24 h). Rejected graft tissue showed mononuclear cell infiltration and the presence of glomerular vasculonecrotic lesions with fibrinoid material, associated with IgG and IgM deposition, but with absent or weak C3 binding. Graft endothelium showed no expression of MHC class II antigens. Serum antibodies were not reactive with donor class I antigens, but did react with endothelial non-MHC alloantigens. In the second model, more direct information on the role of endothelial non-MHC alloantigens in renal allograft rejection was obtained by transplanting Lewis 1 N kidneys into unmodified BN recipients (MHC-matched transplants). Here, similar to the first model, the animals developed severe proteinuria with stable renal function. Histopathological examination showed mononuclear cell infiltration and deposition of IgM and IgG along the glomerular vasculature, but this time in the presence of strong C3 reactivity. However, glomerular vasculonecrotic lesions with intense fibrin deposition were not observed. The data showed that although clinically the two kidney transplantation models used gave similar chronic rejection phenomena, histopathologically some striking differences were observed in the glomeruli. The precise mechanisms effecting chronic rejection of the grafts is still a puzzle. However, immune reactivity against graft (endothelial) non-MHC antigens may play a significant role.     

Increased major histocompatibility complex antigen expression in unilateral ischemic acute tubular necrosis in the mouse

In many studies of renal transplant recipients, acute tubular necrosis has been shown to predispose to a higher rate of graft loss, apparently due to rejection, but the mechanism of this effect is unknown. One possibility is an increased immunogenicity of the graft. To study this possibility, we examined the expression of major histocompatibility complex antigens in kidneys damaged by ischemia, using a mouse model of ischemic ATN. ATN was produced in the left kidney of male CBA mice by temporary clamping of the vascular pedicle for up to 60 min. Class I and II MHC expression was quantified by the extent of binding of monoclonals in radioimmunoassay, after 1 to 35 days in both kidneys. MHC induction was localized by indirect immunoperoxidase staining. Specific steady state mRNA for beta 2 microglobulin and class II were quantified by northern blotting using 32P-labeled probes. Changes in MHC expression were assessed by comparing the ischemically injured left kidney to the control right kidney. By day 1, ATN was evident by histology but there was no change in MHC expression. By day 3, class I was increased in the left kidney by 3-6-fold over the right. In tissue sections, the class I increase was localized to tubular epithelial cells. Starting on day 7 and persisting to day 35, class II was increased by 1.5 to 3 times for the ischemic kidney over the control, primarily in interstitial cells but also in tubular cells. This increase in class II was associated with the appearance of Thy 1.2-positive cells in the interstitial areas. Increased antigen expression was preceded by increased steady state mRNA. Thus unilateral ischemic ATN causes increased MHC expression in tubular cells and the accumulation of an inflammatory infiltrate, both of which may contribute to the increased rate of rejection and graft loss in ischemically injured kidneys.     

The effects of cyclosporine on the induction of donor class I and class II MHC antigens in heart and kidney allografts in the rat

We have previously reported 5-30-fold increases in the expression of class I and class II major histocompatibility complex (MHC) antigens in rejecting heart and kidney allografts in the DA-to-PVG rat strain combination. We examine here the effects of immunosuppression with cyclosporine on the induction of donor class I and class II MHC antigens in heart and kidney allografts in this strain combination. Immunohistological studies and quantitative absorption analyses using monoclonal antibodies and assay systems specific for donor class I and class II MHC antigens were used throughout. Heart allografts in cyclosporine-treated rats were examined on day 3,5,7,9,11, and 14 after transplantation, and kidney allografts in cyclosporine-treated rats were examined at day 7. In addition, untreated heart and kidney isografts were studied at days 1,3,5, and 7 after grafting. Immunohistological studies on frozen sections showed that cyclosporine-treated heart and kidney allografts showed no induction of class II MHC antigens, in contrast to untreated heart and kidney allografts. Class I MHC antigen induction did occur in spite of cyclosporine-therapy, but at levels lower than those seen in untreated allografts. Moreover, the pattern and degree of class I induction in the cyclosporine-treated allografts resembled very closely those seen in isografts, and so this induction was, in all probability, a consequence of the transplantation procedure rather than of specific immune responses. We also noted, in the cyclosporine-treated heart allografts, that all donor interstitial dendritic cells had disappeared and been replaced by recipient interstitial dendritic cells by the end of the second week after grafting. In addition, there was no reduction in the class II antigen content of kidney allografts treated for 7 days with cyclosporine. The absence of class II antigen induction in allografts where rejection is effectively suppressed with cyclosporine might be of clinical value in the differential diagnosis between rejection and cyclosporine toxicity in renal transplantation, and between active and inactive cellular infiltrates in heart transplantation.     

IgM and IgG alloantibody responses to MHC class I and II following rat renal allograft rejection. Effects of transplantectomy and posttransplantation blood transfusion.

Renal allograft rejection in rats and humans is a potent inducer of alloantibody to donor major histocompatibility complex antigens. Alloantibody in such presensitized recipients can cause hyperacute rejection of subsequent renal allografts. In order to characterize alloantibody production in rats presensitized by renal graft rejection, ACI (RT1a) kidneys were transplanted into untreated fully allogeneic PVG (RT1c) recipients and allowed to reject while one native kidney remained in situ for host survival. Serum samples collected at weekly intervals were analyzed by flow cytometry for IgM and IgG antibody binding to ACI lymphoid target cells. The specificity of alloantibody responses was assessed by (1) differential binding to congenic rat strain target cells expressing only donor class I (PVG.R1) versus both donor class I and II (PVG.1A) antigens, (2) differential binding to unseparated donor lymphoid target cells versus lymphoid target cells depleted of class II MHC antigen-expressing cells, and (3) specific blocking of monoclonal antibodies to donor class I (R2/10P, R2/15S) or class II (F17.23.2) epitopes. Alloantibody responses to both donor class I and II MHC antigens were detected. The initial IgM response to donor class I MHC antigens peaked at the time of rejection, followed by a steady decline to relatively low levels by 4 weeks posttransplantation. The IgM response to donor class II MHC antigens was found to be cyclical with apparent peaks at day 7 and 5-6 weeks. The IgG response to donor class I and class II MHC antigens reached maximum by 5-6 weeks before slowly decreasing. IgM and IgG alloantibody specific for class I and class II MHC antigens could be detected through 19 weeks posttransplantation. The effects on circulating alloantibody of two manipulations, posttransplantation donor specific blood transfusion and allograft removal, were examined in this model. The alloantibody responses to class I MHC antigens were not affected by giving DSBT weekly beginning at day 14 after transplantation. However, posttransplantation DSBT eliminated the second peak of IgM alloantibody to class II MHC antigens seen approximately 5-6 weeks posttransplantation and also decreased circulating IgG specific for class II antigens. Transplantectomy at day 5-7 days after transplantation had no apparent effect on circulating IgM or IgG alloantibody through 7 weeks posttransplantation. These data indicate that in a fully allogeneic rat renal allograft model alloantibody responses are elicited to both class I and II MHC donor antigens, but that the kinetics and regulation of the responses to class I differ from those to class II alloantigens.(ABSTRACT TRUNCATED AT 400 WORDS)     

Modulation of angiotensin II and norepinephrine-induced plasminogen activator inhibitor-1 expression by AT1a receptor deficiency

Angiotensin (Ang) II stimulates plasminogen activator inhibitor-1 (PAI-1) expression in many cell types by mechanisms that are cell-type specific. We measured effects of Ang II or norepinephrine on PAI-1 expression in wild type (WT) and Ang type-1a receptor knockout mice (AT(1a)-/-) in the presence or absence of the non-specific AT(1) antagonist losartan. Ang II and norepinephrine increased systolic blood pressure equally, whereas losartan decreased the pressor response of the former but not the latter in WT mice. In AT(1a)-/- mice, baseline systolic blood pressure was lower with no effect of Ang II, norepinephrine, or losartan. Ang II stimulated PAI-1 expression in the heart, aorta, and kidney and markedly in the liver of WT mice. In AT(1a)-/- mice, Ang II-stimulated PAI-1 was significantly attenuated compared with the WT in the heart and aorta but significantly enhanced in the kidney. Losartan decreased the induction in the aorta and liver of WT, and in the kidney and liver of AT(1a)-/- mice. Norepinephrine increased PAI-1 expression in WT heart and aorta, and in AT(1a)-/- heart, kidney, and liver with no effect of losartan. Renal PAI-1 expression correlated with AT(1b) receptor mRNA. We conclude that Ang II stimulates PAI-1 expression in part through the AT(1b) receptor in the kidney and liver. Further, norepinephrine induces PAI-1 expression in vivo with AT(1a) receptor deficiency modulating the effect.     

Cold ischemia augments allogeneic-mediated injury in rat kidney allografts

BACKGROUND: Some clinical studies demonstrate that kidney grafts with prolonged cold ischemia experience early acute rejection more often than those with minimal ischemia. The mechanism, however, is putative. Therefore, the aim of this study was to unravel the impact of ischemia on the immune response in rat kidney allografts compared with that in isografts. METHODS: To induce ischemic injury, donor kidneys were preserved for 24 hours in 4 degrees C University of Wisconsin solution before transplantation. No immunosuppression was administered. The histomorphology according to the BANFF criteria for acute rejection and infiltrating cells were assessed at days 1, 2, 3, 4, 6, and 8 post-transplantation. RESULTS: In allografts, exposure of the kidney to ischemia led to a significantly earlier onset of interstitial cell infiltration and tubulitis compared with nonischemic allografts. The BANFF score of interstitial cell infiltration was 1 +/- 0 vs. 0.25 +/- 0.29 at day 3 and 2 +/- 0 vs. 1.25 +/- 0.25 at day 4. In contrast, in isografts, the effect of ischemia on the histology was not significant. From day 6, the histologic differences between ischemic and nonischemic grafts disappeared. Ischemia led to a more intense expression of P-selectin (day 1), intercellular adhesion molecule-1 (ICAM-1; day 2), and major histocompatibility complex (MHC) class II on endothelium and proximal tubular cells (day 2) in both allografts and isografts. Concurrently with the up-regulated ICAM-1 and MHC expression, significantly more CD4(+) cells and macrophages infiltrated the ischemic allografts at days 2 and 3 and the ischemic isografts at day 4. Importantly, the influx of these cells after ischemia was significantly greater in allografts than in isografts. CONCLUSIONS: Cold ischemia augments allogeneic-mediated cell infiltration in rat kidney allografts. The earlier onset of acute rejection in 24-hour cold preserved allografts may be prevented by better preservation or treatment using tailored immunosuppression.     

Comparison of Abeta(b-/-), H2-DM(-), and CIITA(-/-) in second-set skin allograft rejection.

BACKGROUND: Responses against donor MHC antigens are the major contributor to allograft rejection. Currently, it is unclear whether both direct and indirect recognition pathways are necessary and/or sufficient for allograft rejection. Previously, we found donor MHC class II and H2-DM to have dramatic effects on cardiac allograft survival. METHODS: Here, we used H2-DM(-) mice, which express CLIP-MHC class II complexes, and CIITA(-/-) mice, which lack all class II proteins, to examine the role of direct and indirect recognition on skin allograft rejection. Recipients were primed with donor cultured keratinocytes and later tested for accelerated memory response by challenge with full-thickness tail skin grafts. RESULTS: As previously reported, Abeta(b-/-) grafts survived longer than wild-type grafts, while H2-DM(-) grafts were rejected as rapidly as wild-type grafts. Skin grafts deficient for both beta2m and H2-DM survived longer than grafts lacking only H2-DM, but not as long as Abeta(b-/-) grafts. Additionally, CIITA(-/-) grafts survived as long as Abeta(b-/-) grafts. CONCLUSIONS: The delayed rejection of Abeta(b-/-) compared to H2-DM(-) suggests that indirect recognition of surface-expressed donor MHC class II is sufficient to mediate rapid skin allograft rejection. The equivalent survival of CIITA(-/-) and Abeta(b-/-) grafts suggests that indirect presentation of donor class II molecules (Aalpha or Ebeta) present in Abeta(b-/-) but not CIITA(-/-) mice does not contribute to graft rejection. These results reveal a modest role for surface-expressed donor class II in primed keratinocyte rejection, but also reveal a dramatic contrast to the cardiac allograft system and indicate tissue/organ-specific mechanisms of rejection.