- Email: [email protected]
Transplantation in miniature swine: analysis of graft-infiltrating lymphocytes provides evidence for local suppression
Transplantation in Miniature Swine: Analysis of Graft-Infiltrating Lymphocytes Provides Evidence for Local Suppression Bruce R. Rosengard, Eric O. Kortz, Philip C. Guzzetta, Thoralf M. Sundt III, Christina A. Ojikutu, Richard B. Alexander, and David H. Sachs
ABSTRACT: Previous studies from this laboratory have demonstrated that swine tolerant of class I disparate renal allografts show peripheral antidonor cellular reactivity which can be augmented by skin grafting. To assess the possibility of local suppression, cell-mediated lymphocytotoxicity of graft-infiltrating lymphocytes was compared to that of peripheral blood lymphocytes from three tolerant and four acutely rejecting recipients of class 1-disparate renal allografts. Mixed lymphocyte cultures using peripheral blood lymphocytes or graft-infiltrating lymphocytes and an equal number of irradiated peripheral blood lymphocyte stimulators were incubated for 6 days and tested in a 6-hr 51Cr release assay. Graft-infiltrating lymphocytes from rejecting animals had potent antidonor cell-mediated lymphocytotoxic activity with or without in vitro stimulation. Anti-thirdparty reactivity was seen with appropriate stimulation, suggesting heterogeneity of graft-infiltrating lymphocyte cultures. Peripheral blood lymphocytes from rejectors generated donor-specific cellmediated lymphocytotoxicity. Graft-infiltrating lymphocytes from tolerant animals generated no antidonor cell-mediated lymphocytotoxicity with or without in vitro stimulation, but generated an anti-third-party response. Peripheral blood lymphocytes from tolerant animals displayed both antidonor and anti-third-party reactivity with appropriate in vitro stimulation. These data support the hypothesis that local suppression may contribute significantly to maintenance of tolerance to class I disparate renal allografts in miniature swine. ABBREVIATIONS BUN blood urea nitrogen CML cell-mediated lymphocytotoxicity
GILS PBL
graft-infiltrating lymphocytes peripheral blood lymphocytes
INTRODUCTION Spontaneous tolerance develops in approximately 3 5 % of miniature swine receiving o n e haplotype class I plus minor antigen disparate renal allografts [1-3]. The clinical course of the tolerant animals is usually characterized by a transient elevation of blood urea nitrogen (BUN) and serum creatinine, at which time
From the Transplantation Biology Section, Immunology Branch and Surgery Branch, National Cancer Institute, National Institutes of Health, Bethesda, Maryland. Address reprint requests to Dr. David H. Sachs, Immunology Branch, National Cancer Institute, NIH, 9000 Rockville Pike, Building 10, Room 4B13, Bethesda, MD 20892. Received December1, 1989; acceptedJanuary 3, 1990.
HumanImmunology28, 153-158 (1990) Published by ElsevierScience PublishingCo, Inc.,
153 0198-8859/90/$0.00
154
B.R. Rosengard et al. antidonor cytotoxic antibody of the IgM subclass is generated [1,2]. Renal function returns and the IgM antibody disappears in tolerant animals, as compared to acute rejectors, in which case there is an antibody class switch to IgG and progressive, lethal renal failure [1,2]. Analysis of in vitro cellular responses of peripheral blood lymphocytes (PBL) from tolerant animals suggested that absence of antidonor reactivity in graft acceptors was related to a deficiency in class I-restricted helper-T-cell activity [4]. This finding was consistent with the failure of antibody class switch and suggested that a variable helper precursor frequency might be the mechanism underlying spontaneous tolerance. Alternatively, a variable suppressor precursor frequency could explain differences in helper activity. Analysis of long-term tolerant recipients provided evidence for nonspecific suppressor cells in the PBL of such animals [5]. In subsequent studies, long-term tolerant recipients were grafted with skin that was donor class I and third-party class II in an attempt to break tolerance. The skin was promptly rejected, PBL showed marked antidonor cell-mediated lymphocytotoxic (CML), and a transient elevation of BUN and serum creatinine ensued [6]. The presence of peripheral antidonor activity without subsequent graft rejection suggested local suppression or graft adaptation as possible explanations. In this study, we have compared the in vitro cellular reactivity of graft-infiltrating lymphocytes (GILS) from longterm tolerant recipients and from acute rejectors to investigate the possibility of local suppression.
MATERIALS A N D METHODS
Animals. Transplant donors and recipients were chosen from the National Institutes of Health herd of partially inbred miniature swine at 5-7 months of age. Previous reports from this laboratory have characterized the immunogenetics of this herd [7] and have described intra-major histocompatibility complex recombinant haplotypes [8]. The disparity between donors and recipients was class I plus minor antigens in all cases, with most transplants being SLA~ --> SLA~d. One tolerant animal, an SLAdd recipient of an SLAgggraft, received a 12-day course of cyclosporine in the immediate postoperative period as a means of inducing tolerance. Recipients underwent bilateral nephrectomy and received an orthotopic left renal transplant as previously described [9]. Graft function was assessed postoperatively by frequent determinations of BUN and serum creatinine. Grafts were sterilely harvested from both tolerant and rejecting animals at a second laparotomy, at which time animals were either retransplanted or euthanized.
Media. Tissue culture medium used for bulk CML assays as well as for propagation of GILS consisted of RPMI-1640 (Gibco, Grand Island, NY) supplemented with 100 U/ml penicillin, 135 Izg/ml streptomycin, 50 tzg/ml gentamicin, 10 mM HEPES, 14.4 mM 2-mercaptoethanol, 2 mM l-glutamine, 1 mM sodium pyruvate, and 6% fetal pig serum. Recombinant interleukin 2 (rIL-2) (generously provided by Cetus, Stanford, CA), at concentrations of 10, 100, and 1000 U/ml, was added to the complete medium for culturing GILS. CML test medium consisted of basal medium Eagle (Gibco) supplemented with 100 U/ml penicillin, 135/.¢g/ml streptomycin, 10 mM HEPES, and 6% newborn cal£ serum.
Preparation and propagation of GILS. Sterilely explanted kidneys were minced and were then digested with a mixture of 1 mg/ml collagenase type IV (Sigma, St. Louis, MO), 0.1 mg/ml hyaluronidase type V (Sigma), and 0.02 mg/ml deoxyribonuclease I type V (Sigma) in RPMI-1640 (Gibco) for 4 hr at 20°C. The suspension was filtered, and mononuclear cells were plated at a concentration of 2.5 ×
Suppression of CML by Graft-Infiltrating Cells
155
105 cells/ml in 24-well plates. Cultures were incubated at 37°C in 8% CO2 and 100% humidity. Cultures were examined daily and were split when they became confluent. Half of the media in each well was changed every 2-3 days, and most of the cultures were stimulated at weekly intervals with 1 × 106 irradiated (2500 rad) donor-type PBL. GILS were also obtained from needle biopsy specimens by mincing the biopsy cores and placing a piece of tissue in each well of a 24well plate with medium. When cellular outgrowth was observed, the cells were harvested and then resuspended at 2.5 x 105 cells/ml. From this point on, cultures from biopsy specimens were treated identically to cultures derived from enzymatic digestion.
Cellular responses. CML assays were performed as previously described [4,10]. Briefly, mixed lymphocyte cultures were prepared using 24-well plates with responder and stimulator concentrations of 4 x 106 cells per well. Responders were either GILS or PBL, and stimulators were irradiated (2500 rad) PBL. Cultures were incubated for 6 days at 37°C in 8% CO 2 and 100% humidity. Bulk cultures were harvested, and effector cells were tested on 24-hr phytohemagglutinin blasts from PBL that were labeled with 5*Cr. The tests were run in triplicate at four different effector-to-target ratios (100: 1, 50: 1, 25 : 1, 12 : 1). The results were expressed as percent specific lysis: % specific lysis _- experimental release (cpm) - spontaneous release (cpm) x 100 maximum release (cpm) - spontaneous release (cpm) RESULTS CML A c t i v i t y of Rejecting Recipient PBL and GILS Four experiments were performed, each utilizing PBL from different rejecting animals as responders. PBL were sampled as early as postoperative day 14 and as late as day 37. In each case three different stimulators were used with the following disparities: donor class I, donor class I plus third-party class II, and third-party class I plus third-party class II. All effectors were tested on targets bearing stimulator class I as well as on self targets, which served as negative controls for lymphokine-activated killer (LAK) activity. In all four cases, significant specific lysis was seen with stimulation by donor class I plus third-party class II and with stimulation by full third party. Stimulation by donor class I failed to elicit a CML response in all cases. A representative experiment is depicted in Figure I(A). This pattern of reactivity in the PBL has been previously described [4] and is characteristic of acutely rejecting animals. GILS from the same four animals were tested for CML activity with and without in vitro stimulation. In all four experiments unstimulated GILS had potent CML activity against donor class I targets and demonstrated no anti-third party reactivity or LAK activity (data not shown). GILS stimulated by donor class I showed markedly augmented lytic activity, and third-party stimulation also elicited specific anti-third party activity. Stimulated GILS showed no LAK activity (data not shown). A representative experiment is shown in Figure I(B).
CML Activity of Tolerant Recipient PBL and GILS PBL from three different tolerant animals were tested in a series of experiments. Two of the animals, SLA ~d recipients of SLAP kidneys, were greater than 2 years posttransplant at the time of testing. The third animal (an SLAdd recipient of
156
B.R. Rosengard et al.
A.
40
' '"' '-......
A. " "-...... ""diL
f~
>-
"
30
~'"..
•
.--I
I¢
o (J IM o. t~
" 'A.
20
10
i
i
L
L
100:1
50:.1
25:1
12:1
EFFECTOR TO TARGET RATIO
80 70
m
60
>•J
50
o_ u.
40
i.iJ o.
30
FIGURE 1 CML of GILS versus PBL isolated from a rejector of a class I-disparate renal allograft. (A) CML activity of recipient PBL against targets class I matched to stimulators with the following disparities: donor class I (~), donor class I plus third-party class I1 (__m__), and third-party class I plus thirdparty class II (.../k...). (B) CML activity of unstimulated recipient GILS against targets bearing donor class I (------0---)and against targets class I matched to stimulators with the following disparities: donor class I plus third-party class II (__L_) and third-party class I plus third party class II (.--A"-).
2O 10
~
. . . . . -.., 'AA ..... ,..., .... J
i
i
100:1
50:1
25:1
L 12:1
EFFECTORTO TARGET RATIO
an SLA m kidney) was tested throughout his postoperative course until graft explantation on postoperative day 113. The same series of stimulators and targets that were used to study rejector PBL were employed. In four of four experiments, specific lysis was generated following stimulation by either donor class I plus third-party class II or third-party class I plus third-party class II. Stimulation by donor class I alone failed to elicit a CML response in all cases. A representative experiment is depicted in Figure 2(A). GILS from the same three animals were tested for CML activity with and without in vitro stimulation. In three experiments, unstimulated GILS demonstrated no CML activity against donor class I or third-party class I. In contrast to PBL, GILS stimulated by either donor class I or donor class I plus third-party class II also showed no specific lytic activity. Thirdparty stimulation elicited specific anti-third-party activity. LAK activity was not seen in any assay (data not shown). A representative experiment is shown in Figure 2(B).
Suppression of CML by Graft-Infiltrating Cells
40
A. A
m
30
157
~"-....
..I ~ o
20
ILl
~11 ~
1.................. -*.I
50:1
25:1
~ 10
00:1
12:1
EFFECTOR T O T A R G E T RATIO
40
>.
B.
30
..J
o
_g
o
20
Ill a.
A....,... "........
t~
'"".-.. .*...
FIGURE 2 CML of GILS versus PBL isolated from a long-term tolerant recipient of a class I-disparate renal allograft. (A) CML activity of recipient PBL against targets class I matched to stimulators with the following disparities: donor class I ( ~ ) , donor class I plus third-party class II (--II--), and third party class I plus third-party class II (" "A"-). (B) CML activity of unstimulated recipient GILS against targets bearing donor class I (---0--) and against targets class I matched to stimulators with the following disparities: donor class I plus third-party class II (--II--) and third-party class I plus third-party class II (...A..-).
10
.........
I,
.........- 4
i
i
J
i
100:1
50:. 1
25:1
12:1
EFFECTOR T O T A R G E T RATIO
DISCUSSION Previous studies from this laboratory suggested that variable levels of class I-specific helper activity determine the fate of class I-disparate renal allografts [4]. Differences in helper function could result from a deficiency of helper cell precursors, the presence of active suppression, or both. Studies of long-term tolerant animals showed the presence of antidonor cellular reactivity in the absence of renal dysfunction [4]. Furthermore, reactivity could be augmented with skin grafts bearing donor class I and third-party class II, leading to transient renal dysfunction without inducing graft rejection [6]. These findings were suggestive of local suppression. Nonspecific suppressor activity was found in the PBL of long-term tolerant animals [5], and subsequent experiments have found apparent specific suppressor activity in the PBL of tolerant animals in the early postoperative period [6].'The present study has examined the in vitro cellular reactivity of GILS to assess the possibility of local suppression as the basis of these results. The absence ofantidonor CML activity in GILS extracted from tolerated organs
158
B.R. Rosengard et al. could be explained by either clonal deletion or suppression. The presence of antidonor activity in the PBL of tolerant animals argues against clonal deletion as the explanation for the absence of activity in GILS. Preliminary coculture experiments (data not shown) have demonstrated specific suppression of both CML and MLR by GILS extracted from tolerant animals. The finding that GILS from tolerant animals develop anti-third-party cytotoxicity following appropriate stimulation confirms both the functional viability and the heterogeneity of these cultures. Heterogeneity of GILS has been documented previously in several limiting dilution studies, and specific antidonor cells were shown to represent less than 1% of total graft infiltrate [11,12]. The potent antidonor CML activity demonstrated in all unstimulated GILS derived from rejecting organs suggests that cytotoxic T cells play a major role in the pathogenesis of acute graft rejection in miniature swine. Graft resident suppressor T cells have been previously described in other tolerance systems [13]. The precise mechanism of suppression is unclear, but could involve anti-idiotypic T cells, which have been reported in other systems [14], or could be antigen dependent. In our system, donor-specific suppression was present in PBL early after grafting and then disappeared [6]. Preliminary findings suggest that donor-specific suppression remains in GILS after disappearance from PBL. Further studies will be directed toward determining the mechanism of this suppression and toward isolation and characterization of the cells involved. If such cells can be cloned, they could provide a tool for adoptive immunotherapy as a means of preventing graft rejection.
REFERENCES 1. Pescovitz MD, AuchinclossH Jr, ThistlethwaiteJR Jr, et al: Transplant Proc 15:1124, 1983. 2. Pescovitz MD, Thistlethwaite JR Jr, Auchincloss H Jr et al: J Exp Med 160:1495, 1984. 3. Rosengard BR, Ojikutu CA, Kortz EO, et al: Manuscript in preparation. 4. Kortz EO, Sakamoto K, Suzuki T, et al: Transplantation, in press. 5. Kortz EO, Sachs DH: Curt Surg 46:107, 1989. 6. Rosengard BR, Kortz EO, Guzzetta PC, et al: Manuscript in preparation. 7. Sachs DH, Leight GS, Cone J, et al: Transplantation 22:559, 1976. 8. Pennington LR, Lunney JK, Sachs DH: Transplantation 31:66, 1981. 9. Kirkman RL, Colvin R.B, Flye MW, et al: Transplantation 28:18, 1979. 10. Kirkman RL, Colvin RB, Flye MW, et al: Transplantation 28:24, 1979. 11. Orosz CG, Zinn NE, Sirinek L, et al: Transplantation 41:75, 1986. 12. Manca F, Ferry B, Jaakkola M, et al: ScandJ Immunol 25:255, 1987. 13. Yoshimura N, Kahan BD: Transplantation 40:708, 1985. 14. Lancaster F, Chiu YL, Batchelor JR: Nature 315:326, 1986.
ABSTRACT: Previous studies from this laboratory have demonstrated that swine tolerant of class I disparate renal allografts show peripheral antidonor cellular reactivity which can be augmented by skin grafting. To assess the possibility of local suppression, cell-mediated lymphocytotoxicity of graft-infiltrating lymphocytes was compared to that of peripheral blood lymphocytes from three tolerant and four acutely rejecting recipients of class 1-disparate renal allografts. Mixed lymphocyte cultures using peripheral blood lymphocytes or graft-infiltrating lymphocytes and an equal number of irradiated peripheral blood lymphocyte stimulators were incubated for 6 days and tested in a 6-hr 51Cr release assay. Graft-infiltrating lymphocytes from rejecting animals had potent antidonor cell-mediated lymphocytotoxic activity with or without in vitro stimulation. Anti-thirdparty reactivity was seen with appropriate stimulation, suggesting heterogeneity of graft-infiltrating lymphocyte cultures. Peripheral blood lymphocytes from rejectors generated donor-specific cellmediated lymphocytotoxicity. Graft-infiltrating lymphocytes from tolerant animals generated no antidonor cell-mediated lymphocytotoxicity with or without in vitro stimulation, but generated an anti-third-party response. Peripheral blood lymphocytes from tolerant animals displayed both antidonor and anti-third-party reactivity with appropriate in vitro stimulation. These data support the hypothesis that local suppression may contribute significantly to maintenance of tolerance to class I disparate renal allografts in miniature swine. ABBREVIATIONS BUN blood urea nitrogen CML cell-mediated lymphocytotoxicity
GILS PBL
graft-infiltrating lymphocytes peripheral blood lymphocytes
INTRODUCTION Spontaneous tolerance develops in approximately 3 5 % of miniature swine receiving o n e haplotype class I plus minor antigen disparate renal allografts [1-3]. The clinical course of the tolerant animals is usually characterized by a transient elevation of blood urea nitrogen (BUN) and serum creatinine, at which time
From the Transplantation Biology Section, Immunology Branch and Surgery Branch, National Cancer Institute, National Institutes of Health, Bethesda, Maryland. Address reprint requests to Dr. David H. Sachs, Immunology Branch, National Cancer Institute, NIH, 9000 Rockville Pike, Building 10, Room 4B13, Bethesda, MD 20892. Received December1, 1989; acceptedJanuary 3, 1990.
HumanImmunology28, 153-158 (1990) Published by ElsevierScience PublishingCo, Inc.,
153 0198-8859/90/$0.00
154
B.R. Rosengard et al. antidonor cytotoxic antibody of the IgM subclass is generated [1,2]. Renal function returns and the IgM antibody disappears in tolerant animals, as compared to acute rejectors, in which case there is an antibody class switch to IgG and progressive, lethal renal failure [1,2]. Analysis of in vitro cellular responses of peripheral blood lymphocytes (PBL) from tolerant animals suggested that absence of antidonor reactivity in graft acceptors was related to a deficiency in class I-restricted helper-T-cell activity [4]. This finding was consistent with the failure of antibody class switch and suggested that a variable helper precursor frequency might be the mechanism underlying spontaneous tolerance. Alternatively, a variable suppressor precursor frequency could explain differences in helper activity. Analysis of long-term tolerant recipients provided evidence for nonspecific suppressor cells in the PBL of such animals [5]. In subsequent studies, long-term tolerant recipients were grafted with skin that was donor class I and third-party class II in an attempt to break tolerance. The skin was promptly rejected, PBL showed marked antidonor cell-mediated lymphocytotoxic (CML), and a transient elevation of BUN and serum creatinine ensued [6]. The presence of peripheral antidonor activity without subsequent graft rejection suggested local suppression or graft adaptation as possible explanations. In this study, we have compared the in vitro cellular reactivity of graft-infiltrating lymphocytes (GILS) from longterm tolerant recipients and from acute rejectors to investigate the possibility of local suppression.
MATERIALS A N D METHODS
Animals. Transplant donors and recipients were chosen from the National Institutes of Health herd of partially inbred miniature swine at 5-7 months of age. Previous reports from this laboratory have characterized the immunogenetics of this herd [7] and have described intra-major histocompatibility complex recombinant haplotypes [8]. The disparity between donors and recipients was class I plus minor antigens in all cases, with most transplants being SLA~ --> SLA~d. One tolerant animal, an SLAdd recipient of an SLAgggraft, received a 12-day course of cyclosporine in the immediate postoperative period as a means of inducing tolerance. Recipients underwent bilateral nephrectomy and received an orthotopic left renal transplant as previously described [9]. Graft function was assessed postoperatively by frequent determinations of BUN and serum creatinine. Grafts were sterilely harvested from both tolerant and rejecting animals at a second laparotomy, at which time animals were either retransplanted or euthanized.
Media. Tissue culture medium used for bulk CML assays as well as for propagation of GILS consisted of RPMI-1640 (Gibco, Grand Island, NY) supplemented with 100 U/ml penicillin, 135 Izg/ml streptomycin, 50 tzg/ml gentamicin, 10 mM HEPES, 14.4 mM 2-mercaptoethanol, 2 mM l-glutamine, 1 mM sodium pyruvate, and 6% fetal pig serum. Recombinant interleukin 2 (rIL-2) (generously provided by Cetus, Stanford, CA), at concentrations of 10, 100, and 1000 U/ml, was added to the complete medium for culturing GILS. CML test medium consisted of basal medium Eagle (Gibco) supplemented with 100 U/ml penicillin, 135/.¢g/ml streptomycin, 10 mM HEPES, and 6% newborn cal£ serum.
Preparation and propagation of GILS. Sterilely explanted kidneys were minced and were then digested with a mixture of 1 mg/ml collagenase type IV (Sigma, St. Louis, MO), 0.1 mg/ml hyaluronidase type V (Sigma), and 0.02 mg/ml deoxyribonuclease I type V (Sigma) in RPMI-1640 (Gibco) for 4 hr at 20°C. The suspension was filtered, and mononuclear cells were plated at a concentration of 2.5 ×
Suppression of CML by Graft-Infiltrating Cells
155
105 cells/ml in 24-well plates. Cultures were incubated at 37°C in 8% CO2 and 100% humidity. Cultures were examined daily and were split when they became confluent. Half of the media in each well was changed every 2-3 days, and most of the cultures were stimulated at weekly intervals with 1 × 106 irradiated (2500 rad) donor-type PBL. GILS were also obtained from needle biopsy specimens by mincing the biopsy cores and placing a piece of tissue in each well of a 24well plate with medium. When cellular outgrowth was observed, the cells were harvested and then resuspended at 2.5 x 105 cells/ml. From this point on, cultures from biopsy specimens were treated identically to cultures derived from enzymatic digestion.
Cellular responses. CML assays were performed as previously described [4,10]. Briefly, mixed lymphocyte cultures were prepared using 24-well plates with responder and stimulator concentrations of 4 x 106 cells per well. Responders were either GILS or PBL, and stimulators were irradiated (2500 rad) PBL. Cultures were incubated for 6 days at 37°C in 8% CO 2 and 100% humidity. Bulk cultures were harvested, and effector cells were tested on 24-hr phytohemagglutinin blasts from PBL that were labeled with 5*Cr. The tests were run in triplicate at four different effector-to-target ratios (100: 1, 50: 1, 25 : 1, 12 : 1). The results were expressed as percent specific lysis: % specific lysis _- experimental release (cpm) - spontaneous release (cpm) x 100 maximum release (cpm) - spontaneous release (cpm) RESULTS CML A c t i v i t y of Rejecting Recipient PBL and GILS Four experiments were performed, each utilizing PBL from different rejecting animals as responders. PBL were sampled as early as postoperative day 14 and as late as day 37. In each case three different stimulators were used with the following disparities: donor class I, donor class I plus third-party class II, and third-party class I plus third-party class II. All effectors were tested on targets bearing stimulator class I as well as on self targets, which served as negative controls for lymphokine-activated killer (LAK) activity. In all four cases, significant specific lysis was seen with stimulation by donor class I plus third-party class II and with stimulation by full third party. Stimulation by donor class I failed to elicit a CML response in all cases. A representative experiment is depicted in Figure I(A). This pattern of reactivity in the PBL has been previously described [4] and is characteristic of acutely rejecting animals. GILS from the same four animals were tested for CML activity with and without in vitro stimulation. In all four experiments unstimulated GILS had potent CML activity against donor class I targets and demonstrated no anti-third party reactivity or LAK activity (data not shown). GILS stimulated by donor class I showed markedly augmented lytic activity, and third-party stimulation also elicited specific anti-third party activity. Stimulated GILS showed no LAK activity (data not shown). A representative experiment is shown in Figure I(B).
CML Activity of Tolerant Recipient PBL and GILS PBL from three different tolerant animals were tested in a series of experiments. Two of the animals, SLA ~d recipients of SLAP kidneys, were greater than 2 years posttransplant at the time of testing. The third animal (an SLAdd recipient of
156
B.R. Rosengard et al.
A.
40
' '"' '-......
A. " "-...... ""diL
f~
>-
"
30
~'"..
•
.--I
I¢
o (J IM o. t~
" 'A.
20
10
i
i
L
L
100:1
50:.1
25:1
12:1
EFFECTOR TO TARGET RATIO
80 70
m
60
>•J
50
o_ u.
40
i.iJ o.
30
FIGURE 1 CML of GILS versus PBL isolated from a rejector of a class I-disparate renal allograft. (A) CML activity of recipient PBL against targets class I matched to stimulators with the following disparities: donor class I (~), donor class I plus third-party class I1 (__m__), and third-party class I plus thirdparty class II (.../k...). (B) CML activity of unstimulated recipient GILS against targets bearing donor class I (------0---)and against targets class I matched to stimulators with the following disparities: donor class I plus third-party class II (__L_) and third-party class I plus third party class II (.--A"-).
2O 10
~
. . . . . -.., 'AA ..... ,..., .... J
i
i
100:1
50:1
25:1
L 12:1
EFFECTORTO TARGET RATIO
an SLA m kidney) was tested throughout his postoperative course until graft explantation on postoperative day 113. The same series of stimulators and targets that were used to study rejector PBL were employed. In four of four experiments, specific lysis was generated following stimulation by either donor class I plus third-party class II or third-party class I plus third-party class II. Stimulation by donor class I alone failed to elicit a CML response in all cases. A representative experiment is depicted in Figure 2(A). GILS from the same three animals were tested for CML activity with and without in vitro stimulation. In three experiments, unstimulated GILS demonstrated no CML activity against donor class I or third-party class I. In contrast to PBL, GILS stimulated by either donor class I or donor class I plus third-party class II also showed no specific lytic activity. Thirdparty stimulation elicited specific anti-third-party activity. LAK activity was not seen in any assay (data not shown). A representative experiment is shown in Figure 2(B).
Suppression of CML by Graft-Infiltrating Cells
40
A. A
m
30
157
~"-....
..I ~ o
20
ILl
~11 ~
1.................. -*.I
50:1
25:1
~ 10
00:1
12:1
EFFECTOR T O T A R G E T RATIO
40
>.
B.
30
..J
o
_g
o
20
Ill a.
A....,... "........
t~
'"".-.. .*...
FIGURE 2 CML of GILS versus PBL isolated from a long-term tolerant recipient of a class I-disparate renal allograft. (A) CML activity of recipient PBL against targets class I matched to stimulators with the following disparities: donor class I ( ~ ) , donor class I plus third-party class II (--II--), and third party class I plus third-party class II (" "A"-). (B) CML activity of unstimulated recipient GILS against targets bearing donor class I (---0--) and against targets class I matched to stimulators with the following disparities: donor class I plus third-party class II (--II--) and third-party class I plus third-party class II (...A..-).
10
.........
I,
.........- 4
i
i
J
i
100:1
50:. 1
25:1
12:1
EFFECTOR T O T A R G E T RATIO
DISCUSSION Previous studies from this laboratory suggested that variable levels of class I-specific helper activity determine the fate of class I-disparate renal allografts [4]. Differences in helper function could result from a deficiency of helper cell precursors, the presence of active suppression, or both. Studies of long-term tolerant animals showed the presence of antidonor cellular reactivity in the absence of renal dysfunction [4]. Furthermore, reactivity could be augmented with skin grafts bearing donor class I and third-party class II, leading to transient renal dysfunction without inducing graft rejection [6]. These findings were suggestive of local suppression. Nonspecific suppressor activity was found in the PBL of long-term tolerant animals [5], and subsequent experiments have found apparent specific suppressor activity in the PBL of tolerant animals in the early postoperative period [6].'The present study has examined the in vitro cellular reactivity of GILS to assess the possibility of local suppression as the basis of these results. The absence ofantidonor CML activity in GILS extracted from tolerated organs
158
B.R. Rosengard et al. could be explained by either clonal deletion or suppression. The presence of antidonor activity in the PBL of tolerant animals argues against clonal deletion as the explanation for the absence of activity in GILS. Preliminary coculture experiments (data not shown) have demonstrated specific suppression of both CML and MLR by GILS extracted from tolerant animals. The finding that GILS from tolerant animals develop anti-third-party cytotoxicity following appropriate stimulation confirms both the functional viability and the heterogeneity of these cultures. Heterogeneity of GILS has been documented previously in several limiting dilution studies, and specific antidonor cells were shown to represent less than 1% of total graft infiltrate [11,12]. The potent antidonor CML activity demonstrated in all unstimulated GILS derived from rejecting organs suggests that cytotoxic T cells play a major role in the pathogenesis of acute graft rejection in miniature swine. Graft resident suppressor T cells have been previously described in other tolerance systems [13]. The precise mechanism of suppression is unclear, but could involve anti-idiotypic T cells, which have been reported in other systems [14], or could be antigen dependent. In our system, donor-specific suppression was present in PBL early after grafting and then disappeared [6]. Preliminary findings suggest that donor-specific suppression remains in GILS after disappearance from PBL. Further studies will be directed toward determining the mechanism of this suppression and toward isolation and characterization of the cells involved. If such cells can be cloned, they could provide a tool for adoptive immunotherapy as a means of preventing graft rejection.
REFERENCES 1. Pescovitz MD, AuchinclossH Jr, ThistlethwaiteJR Jr, et al: Transplant Proc 15:1124, 1983. 2. Pescovitz MD, Thistlethwaite JR Jr, Auchincloss H Jr et al: J Exp Med 160:1495, 1984. 3. Rosengard BR, Ojikutu CA, Kortz EO, et al: Manuscript in preparation. 4. Kortz EO, Sakamoto K, Suzuki T, et al: Transplantation, in press. 5. Kortz EO, Sachs DH: Curt Surg 46:107, 1989. 6. Rosengard BR, Kortz EO, Guzzetta PC, et al: Manuscript in preparation. 7. Sachs DH, Leight GS, Cone J, et al: Transplantation 22:559, 1976. 8. Pennington LR, Lunney JK, Sachs DH: Transplantation 31:66, 1981. 9. Kirkman RL, Colvin R.B, Flye MW, et al: Transplantation 28:18, 1979. 10. Kirkman RL, Colvin RB, Flye MW, et al: Transplantation 28:24, 1979. 11. Orosz CG, Zinn NE, Sirinek L, et al: Transplantation 41:75, 1986. 12. Manca F, Ferry B, Jaakkola M, et al: ScandJ Immunol 25:255, 1987. 13. Yoshimura N, Kahan BD: Transplantation 40:708, 1985. 14. Lancaster F, Chiu YL, Batchelor JR: Nature 315:326, 1986.