JOURNAL
OF SURGICAL
53,X2-187
RESEARCH
(19%)
Pretransplant Sensitization with Major Histocompatibility Complex Class I+ Class II- Hepatocytes Leads to Accelerated Skin Graft Rejection P. STEPHEN ALMOND, M.D.,l
GINNY L. BUMGARDNER,
University
of Minnesota,
Department
M.D., SALLY CHEN, B.S., AND ARTHUR J. MATAS, M.D. of Surgery,
Minneapolis,
Minnesota
55455
Submitted for publication April 18, 1991
tor (TCR). This signal can be supplied to CD4+ T cells through binding of their TCR to allogeneic major histocompatibility complex (MHC) class II antigens on donor antigen presenting cells (APCs) which are present within the transplanted organ (direct or allo-recognition). Alternatively, foreign antigens can be processed by recipient APCs and presented with self-class II molecules to recipient T cells (indirect recognition). The APC, either donor or recipient, then supplies a second or costimulatory signal. Currently, there is considerable controversy over which method of antigen recognition (direct or indirect) is more important in allograft rejection. Proponents of direct recognition emphasize the immunogenicity of donor MHC class II+ “ passenger leukocytes” within the allograft, suggesting that rejection could be avoided if these cells were removed pretransplant [2, 31. Attempts to increase graft survival by removing these cells via organ culture, anti class II monoclonal antibodies, and ultraviolet irradiation have met with varying success (4, 5). In addition, several investigators have suggested that any changes in immunogenicity as a result of organ pretreatment may alter antigen expression on the graft parenchymal cells [6]. To study antigen recognition we have used two models--in uitro, the mixed lymphocyte hepatocyte culture (MLHC) [7], and in. uivo, the sponge matrix allograft model [8,9]. When used as a stimulator, the hepatocyte has several advantages over solid organs and other single-cell suspensions. First, Percoll-purified murine hepatocytes are free of contaminating class II+ passenger leukocytes [ 71. Therefore, T cell responses cannot be secondary to direct recognition of allogeneic MHC class II molecules on donor APCs. In addition, by using purified hepatocytes from untreated donors, we have avoided the potential adverse effects of graft pretreatment on hepatocyte antigens. Finally, murine hepatocytes are MHC class I+, class III, so even if they were capable of supplying a second signal, they have no way of supplying the first (i.e., they are MHC class III). Thus, our system
The immunogenicity of major histocompatihility complex (MHC) class I+ class II- hepatocytes is controversial. We studied the effect of pretransplant donorspecific sensitization with either purified hepatocytes (HC) or splenocytes (Spl) on subsequent skin allograft survival. Five million Percoll-purified DBA HC or 10 X lo6 DBA Spl were injected into C57BL/6 recipients either intraperitoneally (ip) or into a sponge matrix allograft. Twelve days later, sensitized mice received a DBA skin graft. On the same day, allogeneic (DBA) and syngeneic (BL/6) skin grafts were placed on naive BL/6 mice. In naive BL/6 mice, allogeneic skin graft survival was 7.8 + 0.5 days (n = 4), and syngeneic survival was indefinite (n = 5). Skin graft survival (mean f SD in days) in recipients sensitized with hepatocytes ip was 6.0 ? 1.2 days (n = 5) compared with 5.6 + 0.5 days in recipients sensitized with splenocytes ip. Similarly, graft survival in recipients that received hepatocytes into a sponge matrix allograft was 5.67 ?Z1 days (n = 6) compared with 5.2 + 1.1 days (n = 8) in those that received splenocytes into the sponge. There was no difference in graft survival between mice sensitized with HC vs Spl, nor between mice injected ip vs with the sponge. All sensitized mice experienced accelerated graft rejection compared with naive controls (P < 0.000). These results demonstrate that purified MHC class I+, class II- murine HCs are immunogenic in vivo. Sensitization with donor-specific HCs led to accelerated rejection of subsequent skin grafts, similar to the accelerated rejection seen after sensitization with MHC ClaSS I+ and ChSS II+ Sph?nOCyteS. 0 1992 Academic Press, Inc.
INTRODUCTION T cell activation by alloantigen is a two-signal process [l]. The first signal is delivered through the T cell recep-
1To whom reprint requests should be addressed at University of Minnesota, 420 Delaware Street SE, Box 328 Mayo, Minneapolis, MN 55455. 0022-4804/92 $4.00 Copyright 0 1992 by Academic Press, All rights of reproduction in any form
182 Inc. reserved.
ALMOND
ET AL.:
TABLE Injection 1. 2. 3. 4. 5. 6.
site
DBA HC SMA DBA SP SMA DBAHCip DBASPip None None
PRETRANSPLANT
SENSITIZATION
1
Skin
No. of mice
Days to rejection
DBA DBA DBA DBA B6 DBA
6 5 5 8 5 4
5.67 + l.O* 5.2 + l.l* 6.0 f 1.2* 5.6 k 0.5* Indefinite 7.8 + 0.5
* P < 0.05 vs line 6.
would allow us to determine if (a) passenger leukocytes are required for allograft rejection and (b) there is evidence for indirect antigen presentation in the MLHC. Previous reports from our laboratory have demonstrated that Percoll-purified MHC class I+ class III murine hepatocytes are immunogenic both in vitro [7] and in uiuo [9]. In primary culture, allogeneic hepatocytes stimulate the proliferation of both CD4+ and CDB+ T cells and the development of CDBf cytotoxic T lymphocytes (CTLs) [ 71. These CTLs are specific for the class I molecule on hepatocytes: they lyse allogeneic, but not syngeneic or third party targets [7, 91. Finally, T cells harvested from primary MLHCs undergo a second set response when restimulated with allogeneic hepatocytes in vitro [9]. To determine if this same phenomenon occurred in uiuo, donor-specific hepatocytes or splenocytes were injected pretransplant. Twelve days later, sensitized mice received an allogeneic skin transplant. All sensitized mice rejected their skin grafts in an accelerated fashion, compared with controls, suggesting that hepatocytes are immunogenic in uivo. MATERIAL
AND
METHODS
Animals. Mice were purchased from the National Institutes of Health (Bethesda, MD) and maintained in our animal facility in accordance with the National Research Council’s guide for animal care. The K, I, D alleles of the mouse strains used were as follows: C57BL/6 (b, b, b) and DBA/S (d, d, d). Hepatocyte harvest. Six-week-old DBA/2 mice were anesthetized with ip Nembutal. The abdomen was prepped with a 70% alcohol solution, shaved, and opened widely. Hepatocytes were harvested using a modification of Seglen’s perfusion technique [lo]. Briefly, a 22-gauge Jelco was inserted into the inferior vena cava, the portal vein was divided, and the suprahepatic vena cava was ligated. The liver was flushed with an ethyleneglycol-bis-N,N,N’,N’-tetraacetic acid (EGTA) calciumfree salt solution warmed to 37°C at a rate of 4 cc/min for 5 min, followed by a 15min perfusion with a 0.05% collagenase (Sigma type I, 300 units/mg) solution. The liver was then removed, minced, and passed over a 40-
AND
SKIN
GRAFT
183
SURVIVAL
pm filter. The resulting suspension was washed with RPM1 and centrifuged at 35g three times. Hepatocyte purification. The hepatocyte suspension was layered onto a 60% Percoll gradient and centrifuged at 140g for 5 min. The supernatant was discarded and the hepatocytes were harvested, counted, and viability was determined by trypan blue exclusion. Preparations with viability > 90% were used for immunizations. Splenocyte preparation. DBA/2 mice were killed by CO, overdose. The spleen was removed aseptically and placed in a petri dish. Then it was crushed with a glass stopper, filtered (40 pm) to remove parts of the capsule, and washed three times with DMEM and centrifuged between washes. Sponge matrix allografts. Our technique has previously been described [B, 111. Briefly, polyurethane sponges (approximately 1 cm3) were implanted subcutaneously on the backs of C57BL/6 mice. Five days later, hepatocytes or splenocytes were injected into the sponge. Immunizations. C57BL/6 mice were injected with either 5 x lo6 DBA hepatocytes or 10 X lo6 DBA splenocytes. These cell numbers were selected based on previous in uivo studies [ 111. Two routes of administration were used: ip and into a sponge matrix allograft. Skin grafts were performed 12 days later. Skin transplants. Donors were killed by CO, overdose. Allo- and syngeneic skin grafts were harvested from DBA and C57BL/6 mice, respectively, and cut into approximately 1 X l-cm sections. Recipients were anesthetized with a 10% Nembutal solution. The flank was prepped and shaved. A 1 X l-cm area of skin was removed and the defect grafted. Grafts were not sewn in place. Sensitized mice received allogeneic skin grafts, and naive animals either allogeneic or syngeneic skin grafts. Grafts were covered with Vaseline gauze and inspected every day beginning on postoperative Day 3. Graft inspections were conducted under anesthesia. Rejection was defined as loss of > 50% of the graft or a “white” graft. At the initiation of the study, there were eight animals in each of the four pretreated groups. In the naive recipients, there were initially six animals in each of the two groups. Animals were removed from the study if anesthetic death occurred (n = 3), if the graft was chewed off (n = 4), or if the graft had “slipped” off (n = 4). Statistics. The Gehan test was used to analyze graft survival data. RESULTS
Pretransplant, Donor-Specific Accelerated Rejection
Sensitization
Leads to
Allogeneic grafts on naive animals survived for 7.8 + 0.5 days (range 7 to 8 days, n = 4) (P = 0.0018 vs spleno-
184
FIG. 1. splenocytes
JOURNAL
Skin grafts on post-transplant (middle right), naive recipient
OF SURGICAL
RESEARCH:
53, NO. 2, AUGUST
1992
Day 5: ip hepatocytes (upper left), ip splenocytes (upper right), SMA hepatocytes (lower left), and syngeneic control (lower right).
cytes and 0.0057 vs hepatocytes). Skin allografts were rejected in an accelerated fashion by all sensitized recipients (Table 1, Fig. l-4). Graft survival in recipients sensitized (either ip or with SMA) with splenocytes was 5.5 +- 0.7 days (range 4 to 6 days, n = 13); with hepatocytes, 5.8 f 1 days (range 4 to 7 days, n = 11). There was no difference in graft survival between mice sensitized with hepatocytes vs splenocytes. Sensitization Occurs after Injection ip or into a SMA
VOL.
of Donor Cells either
Skin allografts were rejected in an equal and accelerated fashion by mice injected (either ip or SMA) with donor cells (hepatocytes or splenocytes) (Table 1, Fig. l-4). Graft survival in mice sensitized ip with spleno-
(middle left), SMA
cytes was 5.6 + 0.5 days (range 5 to 6 days, n = 8); with hepatocytes, 6 -t 1.2 days (range 4 to 7 days, n = 5) (P = 0.26). Graft survival in mice sensitized with SMA with splenocytes was 5.2 + 1.1 days (range 4 to 6 days, n = 5); with hepatocytes, 5.7 ? 1 days (range 4 to 7 days, n = 6) (P = 0.48). DISCUSSION
The diagnosis of acute liver allograft rejection is based on three pathologic findings: endothelialitis, a portal infiltrate, and injury to > 50% of the bile ducts [ 121. Hepatocytes, which form the bulk of the liver parenchyma, are relatively unaffected by the rejection process. Several reasons have been put forth to explain this apparent difference in immunogenicity. The most popular expla-
ALMOND
ET AL.:
PRETRANSPLANT
SENSITIZATION
AND
SKIN
GRAFT
SURVIVAL
185
nge Math
FIG. 2. splenocytes
Skin grafts on post-transplant (middle right), naive recipient
Day 8: ip hepatocytes (upper left), ip splenocytes (upper right), SMA hepatocytes (lower left), and syngeneic control (lower right).
nation is that there are differences in MHC antigen expression between hepatocytes and other cells within the liver and this difference is related to the observed difference in immunogenicity. Endothelial and bile duct epithelial cells, for example, express class II MHC on their cell surface. In contrast, hepatocytes are MHC class I+ and express little, if any, class II antigen [ 131. It has also been suggested that a pure class I stimulus by itself is not immunogenic and may, in fact, be immunosuppressive [ 14,151. Finally, there is evidence of an immunosuppressive factor in hepatocyte cytosol, which may be released at the time of cell death [ 161. The immunogenicity of hepatocytes is a controversial subject, the resolution of which will have a direct impact on two aspects of transplantation: hepatocyte transplantation and possibly the induction of donor-specific hyporesponsiveness. In an effort to bring hepatocyte
(middle left), SMA
transplantation to clinical trials, we have been studying the immunogenicity of Percoll-purified murine hepatocytes. In previous experiments, we [7, 9, 111 and others [ 171, however, have demonstrated that Percoll-purified, allogeneic, MHC class I+, class III murine hepatocytes are capable of eliciting T cell responses in vitro. Murine hepatocytes stimulate the proliferation of allogeneic T cells in a primary MLHC. In addition, responder cells harvested from primary MLHC demonstrate a second set response when restimulated with allogeneic, but not third party hepatocytes, in a standard primed lymphocyte assay. Hepatocytes can also stimulate the development of CDB+ allospecific cytotoxic T lymphocytes (CTLs) in uitro. This process requires the presence of CD4+ T cells and responder APCs; removal of either from culture abrogates the development of cytotoxicity. In fact, at opti-
186
JOURNAL
FIG.
3.
Syngeneic controls,
OF SURGICAL
post-transplant
RESEARCH:
Day 14.
ma1 cell numbers, MHC class If, class II+ hepatocytes appear to be more immunogenic than MHC class I+, class II+ splenocytes [ll]. Interestingly, this difference in immunogenicity correlates well with the increased class I antigen expression on hepatocytes compared with splenocytes. Similar results have been obtained in viva with the sponge matrix allograft model. The injection of allogeneic Percoll-purified murine hepatocytes into subcutaneous sponge matrix allografts (implanted 5 days prior) leads to the accumulation of alloreactive cells within the sponge. Effector cells harvested from the sponge on Day 12 are CDB+ and allospecific; they kill allogeneic but not third party targets. Treating the recipients with either silica (which is taken up by macrophages and interferes with their phagocytic function) or CD4 monoclonal antibody (which depletes CD4+ T cells) significantly decreases cytotoxicity. The value of the sponge matrix allograft model is that graft-infiltrating cells can be easily obtained, analyzed, and tested in functional assays. The trade-off is that the model does not provide a method for measuring graft function or graft survival. That is to say, although there are alloreactive cells within the sponge which are specific for the injected allogeneic cells, it is not known if these cells could mediate the rejection of a functioning allograft. To address this concern, we set up an in vivo correlate of our primed lymphocyte assay. Recipients were sensitized with either donor hepatocytes or splenocytes; subsequent skin allograft survival was compared with allograft survival in unprimed recipients. If our previous in vitro data were correct, skin grafts on mice primed with hepatocytes should undergo accelerated rejection when compared with unprimed recipients.
VOL.
53, NO. 2, AUGUST
1992
The results of this experiment indicate that, in a murine skin allograft model, donor-specific pretreatment with 5 X lo6 hepatocytes or 10 X lo6 splenocytes 12 days prior to transplantation leads to accelerated skin graft rejection, compared with unsensitized allograft controls. In addition, there was no difference in the time to rejection between animals pretreated with hepatocytes vs splenocytes. These data extend our previous observations in two important ways. First, the data suggest that allogeneic MHC class I+, class III hepatocytes can activate responding T cells and that once activated, these cells can and do mediate allograft rejection. Second, the data suggest that indirect antigen presentation is as effective as direct antigen presentation in activating responding T cells. A single-cell suspension of Percoll-purified MHC class I+, class III murine hepatocytes is devoid of contaminating class II+ “passenger leukocytes” and therefore cannot activate alloreactive CD4+ T cells directly. Instead, hepatocyte antigens must be processed and presented by responder APC and then presented to responding CD4+ T cells in the context of self-class II MHC (indirect recognition). In contrast, a bulk splenocyte preparation contains a mixture of cells including class II+ passenger leukocytes. Therefore, this cell preparation should be capable of activating alloreactive T cells (i.e., CD4+ T cells which can recognize allogeneic class II antigens directly without the need for antigen processing). The data from this experiment suggest that either method is equally effective in activating responding T cells.
FIG.
4.
Allogeneic
controls,
post-transplant
Day 14.
ALMOND
ET AL.: PRETRANSPLANT
SENSITIZATION
Other investigators have studied the effect of donorspecific hepatocyte transfusions on allograft survival. In a rat heart allograft model, priming with donor hepatocytes 72 hr pretransplant had no effect on subsequent heart allograft survival [14]. In contrast, rat renal allograft survival was prolonged in rats receiving lo7 donorspecific hepatocytes 7 days pretransplant [El. Similar results were obtained by using purified membranebound class I molecules (isolated from liver cells) in the form of protein micelles [14]. This donor-specific prolongation of allograft survival was also demonstrated in a mouse heart allograft model by pretreating the recipient with 5 X lo6 syngeneic cells transfected with the donor MHC class I or class II gene [14]. Pretreatment with higher or lower doses had less effect or no effect on graft survival. Our findings are not inconsistent with these previously published reports [ 14, 151. Skin grafts, in contrast to heart allografts, are not immediately vascularized and instead require neovascularization for survival. In addition, skin grafts are more immunogenic than vascularized grafts. Allograft survival in unsensitized murine recipients supports this hypothesis, with skin allografts being rejected faster (7.8 +- 0.5 days) than heart (9.1 + 0.7 days) [14] or renal allografts (10 days) [15]. Finally, the prolongation seen in graft survival in these models is the result of long-term survival in a few animals. Taken together with our data, this would suggest that the difference between inducing hyporesponsiveness and sensitization is dependent on the antigenic load, timing of antigen dosing, route of antigen administration, and immunogenicity of the organ transplanted. REFERENCES 1. Goodman, J. W., and Sercarz, E. E. The complex of structures involved in T cell activation. Annu. Reu. Immunol. 1: 475, 1983. 2.
3.
Lafferty, K. F., Prowse, S. J., Simeonovic, C. J., and Warren, H. S. Immunobiology of tissue transplantation: A return to the passenger leukocyte theory. Annu. Reu. Zmmunol. 1: 143, 1983. Gill, R. G. Role of passenger leukocytes in islet allograft immunity. Clin. Transplant. 4: 176, 1990.
AND
SKIN
GRAFT
SURVIVAL
187
4. Gores, P. F., Sutherland,
D. E. R., Platt, J. L., and Bach, F. H. Depletion of donor Ia+ cells before transplantation does not prolong islet allograft survival. J. Immunol. 137: 1482, 1986.
5. Faustman, D. L., Hauptfeld,
V., Lacy, P., and Davie, J. Prolongation of murine islet allograft survival by pretreatment of islets with antibody directed to Ia determinants. Proc. Nutl. Acud. Sci. USA 78: 5156,1981. 6. Hullett, D. A., and Sollinger, H. W. Modification of allograft antigenicity. Transplant. Proc. 22(4): 1926, 1990. 7. Bumgardner, G. L., Chen, S., Hoffman, R., Cahill, D. C., So, S. K., Platt, J., Bach, F. H., and Ascher, N. L. Afferent and efferent pathways in T cell responses to MHC Class I+, II-hepatocytes. Transplantation 47: 163, 1989. 8. Roberts, P. J., and Hayry, P. Effector mechanisms in allograft rejection. Assembly of sponge matrix allografts. Cell. Immunol.
26:160,1976. 9. Bumgardner, G. L., Ascher, N. L., Cahill, D., Chen, S., Payne, W. D., and Matas, A. J. Analysis of the in vitro and in vivo immune response to purified hepatocytes. Transplantation 49(z): 429, 1990. 10. Seglen, P. 0. Preparation of isolated rat liver cells. Methods Cell.
Biol. 13: 29, 1976. 11. Almond, P. Stephen, Bumgardner,
G., Cunningham, T., Chen, S., Payne, W., and Matas, A. Correlation of class I antigen expression with immunogenicity. Curr. Surg. 47: 262, 1990. 12. Snover, D. C. Liver transplantation. In George E. Sale (Ed.), The Pathology of Organ Transplantation. MA: Butterworth, 1990. Pp.
103-132. 13. Lautenschlager,
I., Nyman, N., Vaananen, H., Lehto, V. P., and Hayry, P. Antigenic and immunogenic components in rat liver. Scund. J. Immunol. 17: 61, 1983. 14. Foster, S., Wood, K. J., and Morris, P. J. Comparison of the effect of protein micelles containing purified class I MHC antigen and a cytosolic preparation containing water soluble class I molecules on rat renal allograft survival. Trun.splunt. Proc. 21:
375,1989. 15. Madsen, J. C., Wood, K. J., Superina,
R. A., and Morris, P. J. Induction of immunological unresponsiveness using recipient cells transfected with donor class I or class II MHC genes. Transplant. Proc. 21:477, 1989. 16. Bumgardner, G. L., Dunn, G., Chen, S., Cahill, D., and Ascher, N. L. Immunosuppressive effects of murine liver cytosol in vivo. Surg. Forum 39: 370, 1988. 17. So, S. K., Wilkes, L. M., Platt, J. L., Ascher, N. L., and Simmons, R. L. Purified hepatocytes can stimulate allospecific cytotoxic T lymphocytes in a mixed lymphocyte-hepatocyte culture. Trunsplant. Proc. 19: 251, 1987.
OF SURGICAL
53,X2-187
RESEARCH
(19%)
Pretransplant Sensitization with Major Histocompatibility Complex Class I+ Class II- Hepatocytes Leads to Accelerated Skin Graft Rejection P. STEPHEN ALMOND, M.D.,l
GINNY L. BUMGARDNER,
University
of Minnesota,
Department
M.D., SALLY CHEN, B.S., AND ARTHUR J. MATAS, M.D. of Surgery,
Minneapolis,
Minnesota
55455
Submitted for publication April 18, 1991
tor (TCR). This signal can be supplied to CD4+ T cells through binding of their TCR to allogeneic major histocompatibility complex (MHC) class II antigens on donor antigen presenting cells (APCs) which are present within the transplanted organ (direct or allo-recognition). Alternatively, foreign antigens can be processed by recipient APCs and presented with self-class II molecules to recipient T cells (indirect recognition). The APC, either donor or recipient, then supplies a second or costimulatory signal. Currently, there is considerable controversy over which method of antigen recognition (direct or indirect) is more important in allograft rejection. Proponents of direct recognition emphasize the immunogenicity of donor MHC class II+ “ passenger leukocytes” within the allograft, suggesting that rejection could be avoided if these cells were removed pretransplant [2, 31. Attempts to increase graft survival by removing these cells via organ culture, anti class II monoclonal antibodies, and ultraviolet irradiation have met with varying success (4, 5). In addition, several investigators have suggested that any changes in immunogenicity as a result of organ pretreatment may alter antigen expression on the graft parenchymal cells [6]. To study antigen recognition we have used two models--in uitro, the mixed lymphocyte hepatocyte culture (MLHC) [7], and in. uivo, the sponge matrix allograft model [8,9]. When used as a stimulator, the hepatocyte has several advantages over solid organs and other single-cell suspensions. First, Percoll-purified murine hepatocytes are free of contaminating class II+ passenger leukocytes [ 71. Therefore, T cell responses cannot be secondary to direct recognition of allogeneic MHC class II molecules on donor APCs. In addition, by using purified hepatocytes from untreated donors, we have avoided the potential adverse effects of graft pretreatment on hepatocyte antigens. Finally, murine hepatocytes are MHC class I+, class III, so even if they were capable of supplying a second signal, they have no way of supplying the first (i.e., they are MHC class III). Thus, our system
The immunogenicity of major histocompatihility complex (MHC) class I+ class II- hepatocytes is controversial. We studied the effect of pretransplant donorspecific sensitization with either purified hepatocytes (HC) or splenocytes (Spl) on subsequent skin allograft survival. Five million Percoll-purified DBA HC or 10 X lo6 DBA Spl were injected into C57BL/6 recipients either intraperitoneally (ip) or into a sponge matrix allograft. Twelve days later, sensitized mice received a DBA skin graft. On the same day, allogeneic (DBA) and syngeneic (BL/6) skin grafts were placed on naive BL/6 mice. In naive BL/6 mice, allogeneic skin graft survival was 7.8 + 0.5 days (n = 4), and syngeneic survival was indefinite (n = 5). Skin graft survival (mean f SD in days) in recipients sensitized with hepatocytes ip was 6.0 ? 1.2 days (n = 5) compared with 5.6 + 0.5 days in recipients sensitized with splenocytes ip. Similarly, graft survival in recipients that received hepatocytes into a sponge matrix allograft was 5.67 ?Z1 days (n = 6) compared with 5.2 + 1.1 days (n = 8) in those that received splenocytes into the sponge. There was no difference in graft survival between mice sensitized with HC vs Spl, nor between mice injected ip vs with the sponge. All sensitized mice experienced accelerated graft rejection compared with naive controls (P < 0.000). These results demonstrate that purified MHC class I+, class II- murine HCs are immunogenic in vivo. Sensitization with donor-specific HCs led to accelerated rejection of subsequent skin grafts, similar to the accelerated rejection seen after sensitization with MHC ClaSS I+ and ChSS II+ Sph?nOCyteS. 0 1992 Academic Press, Inc.
INTRODUCTION T cell activation by alloantigen is a two-signal process [l]. The first signal is delivered through the T cell recep-
1To whom reprint requests should be addressed at University of Minnesota, 420 Delaware Street SE, Box 328 Mayo, Minneapolis, MN 55455. 0022-4804/92 $4.00 Copyright 0 1992 by Academic Press, All rights of reproduction in any form
182 Inc. reserved.
ALMOND
ET AL.:
TABLE Injection 1. 2. 3. 4. 5. 6.
site
DBA HC SMA DBA SP SMA DBAHCip DBASPip None None
PRETRANSPLANT
SENSITIZATION
1
Skin
No. of mice
Days to rejection
DBA DBA DBA DBA B6 DBA
6 5 5 8 5 4
5.67 + l.O* 5.2 + l.l* 6.0 f 1.2* 5.6 k 0.5* Indefinite 7.8 + 0.5
* P < 0.05 vs line 6.
would allow us to determine if (a) passenger leukocytes are required for allograft rejection and (b) there is evidence for indirect antigen presentation in the MLHC. Previous reports from our laboratory have demonstrated that Percoll-purified MHC class I+ class III murine hepatocytes are immunogenic both in vitro [7] and in uiuo [9]. In primary culture, allogeneic hepatocytes stimulate the proliferation of both CD4+ and CDB+ T cells and the development of CDBf cytotoxic T lymphocytes (CTLs) [ 71. These CTLs are specific for the class I molecule on hepatocytes: they lyse allogeneic, but not syngeneic or third party targets [7, 91. Finally, T cells harvested from primary MLHCs undergo a second set response when restimulated with allogeneic hepatocytes in vitro [9]. To determine if this same phenomenon occurred in uiuo, donor-specific hepatocytes or splenocytes were injected pretransplant. Twelve days later, sensitized mice received an allogeneic skin transplant. All sensitized mice rejected their skin grafts in an accelerated fashion, compared with controls, suggesting that hepatocytes are immunogenic in uivo. MATERIAL
AND
METHODS
Animals. Mice were purchased from the National Institutes of Health (Bethesda, MD) and maintained in our animal facility in accordance with the National Research Council’s guide for animal care. The K, I, D alleles of the mouse strains used were as follows: C57BL/6 (b, b, b) and DBA/S (d, d, d). Hepatocyte harvest. Six-week-old DBA/2 mice were anesthetized with ip Nembutal. The abdomen was prepped with a 70% alcohol solution, shaved, and opened widely. Hepatocytes were harvested using a modification of Seglen’s perfusion technique [lo]. Briefly, a 22-gauge Jelco was inserted into the inferior vena cava, the portal vein was divided, and the suprahepatic vena cava was ligated. The liver was flushed with an ethyleneglycol-bis-N,N,N’,N’-tetraacetic acid (EGTA) calciumfree salt solution warmed to 37°C at a rate of 4 cc/min for 5 min, followed by a 15min perfusion with a 0.05% collagenase (Sigma type I, 300 units/mg) solution. The liver was then removed, minced, and passed over a 40-
AND
SKIN
GRAFT
183
SURVIVAL
pm filter. The resulting suspension was washed with RPM1 and centrifuged at 35g three times. Hepatocyte purification. The hepatocyte suspension was layered onto a 60% Percoll gradient and centrifuged at 140g for 5 min. The supernatant was discarded and the hepatocytes were harvested, counted, and viability was determined by trypan blue exclusion. Preparations with viability > 90% were used for immunizations. Splenocyte preparation. DBA/2 mice were killed by CO, overdose. The spleen was removed aseptically and placed in a petri dish. Then it was crushed with a glass stopper, filtered (40 pm) to remove parts of the capsule, and washed three times with DMEM and centrifuged between washes. Sponge matrix allografts. Our technique has previously been described [B, 111. Briefly, polyurethane sponges (approximately 1 cm3) were implanted subcutaneously on the backs of C57BL/6 mice. Five days later, hepatocytes or splenocytes were injected into the sponge. Immunizations. C57BL/6 mice were injected with either 5 x lo6 DBA hepatocytes or 10 X lo6 DBA splenocytes. These cell numbers were selected based on previous in uivo studies [ 111. Two routes of administration were used: ip and into a sponge matrix allograft. Skin grafts were performed 12 days later. Skin transplants. Donors were killed by CO, overdose. Allo- and syngeneic skin grafts were harvested from DBA and C57BL/6 mice, respectively, and cut into approximately 1 X l-cm sections. Recipients were anesthetized with a 10% Nembutal solution. The flank was prepped and shaved. A 1 X l-cm area of skin was removed and the defect grafted. Grafts were not sewn in place. Sensitized mice received allogeneic skin grafts, and naive animals either allogeneic or syngeneic skin grafts. Grafts were covered with Vaseline gauze and inspected every day beginning on postoperative Day 3. Graft inspections were conducted under anesthesia. Rejection was defined as loss of > 50% of the graft or a “white” graft. At the initiation of the study, there were eight animals in each of the four pretreated groups. In the naive recipients, there were initially six animals in each of the two groups. Animals were removed from the study if anesthetic death occurred (n = 3), if the graft was chewed off (n = 4), or if the graft had “slipped” off (n = 4). Statistics. The Gehan test was used to analyze graft survival data. RESULTS
Pretransplant, Donor-Specific Accelerated Rejection
Sensitization
Leads to
Allogeneic grafts on naive animals survived for 7.8 + 0.5 days (range 7 to 8 days, n = 4) (P = 0.0018 vs spleno-
184
FIG. 1. splenocytes
JOURNAL
Skin grafts on post-transplant (middle right), naive recipient
OF SURGICAL
RESEARCH:
53, NO. 2, AUGUST
1992
Day 5: ip hepatocytes (upper left), ip splenocytes (upper right), SMA hepatocytes (lower left), and syngeneic control (lower right).
cytes and 0.0057 vs hepatocytes). Skin allografts were rejected in an accelerated fashion by all sensitized recipients (Table 1, Fig. l-4). Graft survival in recipients sensitized (either ip or with SMA) with splenocytes was 5.5 +- 0.7 days (range 4 to 6 days, n = 13); with hepatocytes, 5.8 f 1 days (range 4 to 7 days, n = 11). There was no difference in graft survival between mice sensitized with hepatocytes vs splenocytes. Sensitization Occurs after Injection ip or into a SMA
VOL.
of Donor Cells either
Skin allografts were rejected in an equal and accelerated fashion by mice injected (either ip or SMA) with donor cells (hepatocytes or splenocytes) (Table 1, Fig. l-4). Graft survival in mice sensitized ip with spleno-
(middle left), SMA
cytes was 5.6 + 0.5 days (range 5 to 6 days, n = 8); with hepatocytes, 6 -t 1.2 days (range 4 to 7 days, n = 5) (P = 0.26). Graft survival in mice sensitized with SMA with splenocytes was 5.2 + 1.1 days (range 4 to 6 days, n = 5); with hepatocytes, 5.7 ? 1 days (range 4 to 7 days, n = 6) (P = 0.48). DISCUSSION
The diagnosis of acute liver allograft rejection is based on three pathologic findings: endothelialitis, a portal infiltrate, and injury to > 50% of the bile ducts [ 121. Hepatocytes, which form the bulk of the liver parenchyma, are relatively unaffected by the rejection process. Several reasons have been put forth to explain this apparent difference in immunogenicity. The most popular expla-
ALMOND
ET AL.:
PRETRANSPLANT
SENSITIZATION
AND
SKIN
GRAFT
SURVIVAL
185
nge Math
FIG. 2. splenocytes
Skin grafts on post-transplant (middle right), naive recipient
Day 8: ip hepatocytes (upper left), ip splenocytes (upper right), SMA hepatocytes (lower left), and syngeneic control (lower right).
nation is that there are differences in MHC antigen expression between hepatocytes and other cells within the liver and this difference is related to the observed difference in immunogenicity. Endothelial and bile duct epithelial cells, for example, express class II MHC on their cell surface. In contrast, hepatocytes are MHC class I+ and express little, if any, class II antigen [ 131. It has also been suggested that a pure class I stimulus by itself is not immunogenic and may, in fact, be immunosuppressive [ 14,151. Finally, there is evidence of an immunosuppressive factor in hepatocyte cytosol, which may be released at the time of cell death [ 161. The immunogenicity of hepatocytes is a controversial subject, the resolution of which will have a direct impact on two aspects of transplantation: hepatocyte transplantation and possibly the induction of donor-specific hyporesponsiveness. In an effort to bring hepatocyte
(middle left), SMA
transplantation to clinical trials, we have been studying the immunogenicity of Percoll-purified murine hepatocytes. In previous experiments, we [7, 9, 111 and others [ 171, however, have demonstrated that Percoll-purified, allogeneic, MHC class I+, class III murine hepatocytes are capable of eliciting T cell responses in vitro. Murine hepatocytes stimulate the proliferation of allogeneic T cells in a primary MLHC. In addition, responder cells harvested from primary MLHC demonstrate a second set response when restimulated with allogeneic, but not third party hepatocytes, in a standard primed lymphocyte assay. Hepatocytes can also stimulate the development of CDB+ allospecific cytotoxic T lymphocytes (CTLs) in uitro. This process requires the presence of CD4+ T cells and responder APCs; removal of either from culture abrogates the development of cytotoxicity. In fact, at opti-
186
JOURNAL
FIG.
3.
Syngeneic controls,
OF SURGICAL
post-transplant
RESEARCH:
Day 14.
ma1 cell numbers, MHC class If, class II+ hepatocytes appear to be more immunogenic than MHC class I+, class II+ splenocytes [ll]. Interestingly, this difference in immunogenicity correlates well with the increased class I antigen expression on hepatocytes compared with splenocytes. Similar results have been obtained in viva with the sponge matrix allograft model. The injection of allogeneic Percoll-purified murine hepatocytes into subcutaneous sponge matrix allografts (implanted 5 days prior) leads to the accumulation of alloreactive cells within the sponge. Effector cells harvested from the sponge on Day 12 are CDB+ and allospecific; they kill allogeneic but not third party targets. Treating the recipients with either silica (which is taken up by macrophages and interferes with their phagocytic function) or CD4 monoclonal antibody (which depletes CD4+ T cells) significantly decreases cytotoxicity. The value of the sponge matrix allograft model is that graft-infiltrating cells can be easily obtained, analyzed, and tested in functional assays. The trade-off is that the model does not provide a method for measuring graft function or graft survival. That is to say, although there are alloreactive cells within the sponge which are specific for the injected allogeneic cells, it is not known if these cells could mediate the rejection of a functioning allograft. To address this concern, we set up an in vivo correlate of our primed lymphocyte assay. Recipients were sensitized with either donor hepatocytes or splenocytes; subsequent skin allograft survival was compared with allograft survival in unprimed recipients. If our previous in vitro data were correct, skin grafts on mice primed with hepatocytes should undergo accelerated rejection when compared with unprimed recipients.
VOL.
53, NO. 2, AUGUST
1992
The results of this experiment indicate that, in a murine skin allograft model, donor-specific pretreatment with 5 X lo6 hepatocytes or 10 X lo6 splenocytes 12 days prior to transplantation leads to accelerated skin graft rejection, compared with unsensitized allograft controls. In addition, there was no difference in the time to rejection between animals pretreated with hepatocytes vs splenocytes. These data extend our previous observations in two important ways. First, the data suggest that allogeneic MHC class I+, class III hepatocytes can activate responding T cells and that once activated, these cells can and do mediate allograft rejection. Second, the data suggest that indirect antigen presentation is as effective as direct antigen presentation in activating responding T cells. A single-cell suspension of Percoll-purified MHC class I+, class III murine hepatocytes is devoid of contaminating class II+ “passenger leukocytes” and therefore cannot activate alloreactive CD4+ T cells directly. Instead, hepatocyte antigens must be processed and presented by responder APC and then presented to responding CD4+ T cells in the context of self-class II MHC (indirect recognition). In contrast, a bulk splenocyte preparation contains a mixture of cells including class II+ passenger leukocytes. Therefore, this cell preparation should be capable of activating alloreactive T cells (i.e., CD4+ T cells which can recognize allogeneic class II antigens directly without the need for antigen processing). The data from this experiment suggest that either method is equally effective in activating responding T cells.
FIG.
4.
Allogeneic
controls,
post-transplant
Day 14.
ALMOND
ET AL.: PRETRANSPLANT
SENSITIZATION
Other investigators have studied the effect of donorspecific hepatocyte transfusions on allograft survival. In a rat heart allograft model, priming with donor hepatocytes 72 hr pretransplant had no effect on subsequent heart allograft survival [14]. In contrast, rat renal allograft survival was prolonged in rats receiving lo7 donorspecific hepatocytes 7 days pretransplant [El. Similar results were obtained by using purified membranebound class I molecules (isolated from liver cells) in the form of protein micelles [14]. This donor-specific prolongation of allograft survival was also demonstrated in a mouse heart allograft model by pretreating the recipient with 5 X lo6 syngeneic cells transfected with the donor MHC class I or class II gene [14]. Pretreatment with higher or lower doses had less effect or no effect on graft survival. Our findings are not inconsistent with these previously published reports [ 14, 151. Skin grafts, in contrast to heart allografts, are not immediately vascularized and instead require neovascularization for survival. In addition, skin grafts are more immunogenic than vascularized grafts. Allograft survival in unsensitized murine recipients supports this hypothesis, with skin allografts being rejected faster (7.8 +- 0.5 days) than heart (9.1 + 0.7 days) [14] or renal allografts (10 days) [15]. Finally, the prolongation seen in graft survival in these models is the result of long-term survival in a few animals. Taken together with our data, this would suggest that the difference between inducing hyporesponsiveness and sensitization is dependent on the antigenic load, timing of antigen dosing, route of antigen administration, and immunogenicity of the organ transplanted. REFERENCES 1. Goodman, J. W., and Sercarz, E. E. The complex of structures involved in T cell activation. Annu. Reu. Immunol. 1: 475, 1983. 2.
3.
Lafferty, K. F., Prowse, S. J., Simeonovic, C. J., and Warren, H. S. Immunobiology of tissue transplantation: A return to the passenger leukocyte theory. Annu. Reu. Zmmunol. 1: 143, 1983. Gill, R. G. Role of passenger leukocytes in islet allograft immunity. Clin. Transplant. 4: 176, 1990.
AND
SKIN
GRAFT
SURVIVAL
187
4. Gores, P. F., Sutherland,
D. E. R., Platt, J. L., and Bach, F. H. Depletion of donor Ia+ cells before transplantation does not prolong islet allograft survival. J. Immunol. 137: 1482, 1986.
5. Faustman, D. L., Hauptfeld,
V., Lacy, P., and Davie, J. Prolongation of murine islet allograft survival by pretreatment of islets with antibody directed to Ia determinants. Proc. Nutl. Acud. Sci. USA 78: 5156,1981. 6. Hullett, D. A., and Sollinger, H. W. Modification of allograft antigenicity. Transplant. Proc. 22(4): 1926, 1990. 7. Bumgardner, G. L., Chen, S., Hoffman, R., Cahill, D. C., So, S. K., Platt, J., Bach, F. H., and Ascher, N. L. Afferent and efferent pathways in T cell responses to MHC Class I+, II-hepatocytes. Transplantation 47: 163, 1989. 8. Roberts, P. J., and Hayry, P. Effector mechanisms in allograft rejection. Assembly of sponge matrix allografts. Cell. Immunol.
26:160,1976. 9. Bumgardner, G. L., Ascher, N. L., Cahill, D., Chen, S., Payne, W. D., and Matas, A. J. Analysis of the in vitro and in vivo immune response to purified hepatocytes. Transplantation 49(z): 429, 1990. 10. Seglen, P. 0. Preparation of isolated rat liver cells. Methods Cell.
Biol. 13: 29, 1976. 11. Almond, P. Stephen, Bumgardner,
G., Cunningham, T., Chen, S., Payne, W., and Matas, A. Correlation of class I antigen expression with immunogenicity. Curr. Surg. 47: 262, 1990. 12. Snover, D. C. Liver transplantation. In George E. Sale (Ed.), The Pathology of Organ Transplantation. MA: Butterworth, 1990. Pp.
103-132. 13. Lautenschlager,
I., Nyman, N., Vaananen, H., Lehto, V. P., and Hayry, P. Antigenic and immunogenic components in rat liver. Scund. J. Immunol. 17: 61, 1983. 14. Foster, S., Wood, K. J., and Morris, P. J. Comparison of the effect of protein micelles containing purified class I MHC antigen and a cytosolic preparation containing water soluble class I molecules on rat renal allograft survival. Trun.splunt. Proc. 21:
375,1989. 15. Madsen, J. C., Wood, K. J., Superina,
R. A., and Morris, P. J. Induction of immunological unresponsiveness using recipient cells transfected with donor class I or class II MHC genes. Transplant. Proc. 21:477, 1989. 16. Bumgardner, G. L., Dunn, G., Chen, S., Cahill, D., and Ascher, N. L. Immunosuppressive effects of murine liver cytosol in vivo. Surg. Forum 39: 370, 1988. 17. So, S. K., Wilkes, L. M., Platt, J. L., Ascher, N. L., and Simmons, R. L. Purified hepatocytes can stimulate allospecific cytotoxic T lymphocytes in a mixed lymphocyte-hepatocyte culture. Trunsplant. Proc. 19: 251, 1987.
