- Email: [email protected]
Induction of suppressor cells in autologous mixed lymphocyte culture (AMLC) in humans
CELLULAR
IMMUNOLOGY
75, 122- 133 ( 1983)
Induction of Suppressor Cells in Autologous Mixed Lymphocyte Culture (AMLC) in Humans FELDZGERITTA PAZDERKA,’ ALMARIO ANGELES, THAVISAKDI KOVITHAVONGS, AND JOHN B. DOSSETOR Clinical
and Transplant Immunology Group, Departments of Medicine University of Alberta, Edmonton, Alberta T6G 2E1,
Received August
IO, 1982; accepted
October
and Laboratory Canada
Medicine,
29, I982
T cells stimulated for 6-7 days in autologous mixed lymphocyte culture (AMLC) showed suppressive effects when added to fresh mixed cultures where autologous lymphocytes (A) were stimulated by Mitomycin C-treated allogeneic lymphocytes (Xm), in a ratio of A:Xm:AMLCactivated cells of 1: 1 :OS. Both cytotoxic and proliferative activities in second cultures, as assayed after 6 days of incubation, were significantly inhibited (percentage suppression of cytotoxic activity observed in 17 experiments was 75.3 f 22.4; percentage suppression of proliferation was 60.6 * 18.2). Suppressor cells (SC) generated in AMLC were Mitomycin C sensitive and nonspecific in their action; not only A/Xm but also X/Am and X/Ym cultures were suppressed to the same extent. AMLC-Activated cells showed a considerable degree of proliferation in response to alloantigens but failed to express any cytotoxic activity against autologous or allogeneic phytohemagglutinin blasts. Thus, the inhibitory effect observed in this system is not due to cytotoxic elimination of responding or stimulating cells in the second culture but rather reflects a true regulatory (suppressive) mechanism. INTRODUCTION It has been demonstrated repeatedly (1, 2) that human T lymphocytes can be stimulated by autologous non-T lymphocytes to proliferate in culture (AMLC).* AMLC has been shown to possess two main attributes of immune responses: immunologic specificity and memory (3). Although the biologic significance of this phenomenon is still uncertain, it has been interpreted by some authors as reflecting a mechanism by which T lymphocytes might maintain self-tolerance in vivo (4) and as representing a measure of T-cell regulatory ability (3, 5). In addition to their proliferative response, a variety of effector and regulatory functions have been attributed to AMLC-activated T cells. Under certain circum’ To whom correspondence should be addressed. * Abbreviations used: AET, 2-aminoethylisothiouronium bromide hydrobromide; AMLC, autologous mixed lymphocyte culture; AMLR, autologous mixed lymphocyte reaction; CLL, chronic lymphatic leukemia; CML, cell-mediated lymphocytotoxicity; Con A, concanavalin A; E, effector; FCS, fetal calf serum; Ig, immunoglobulin; MHC, major histocompatibility complex; MLC, mixed lymphocyte culture; PBL, peripheral blood lymphocytes; PHA, phytohemagglutinin; PWM, pokeweed mitogen; SC, suppressor cell; SI, stimulation index; SIg, surface immunoglobulin; SLE, systemic lupus erythematosus; SRBC, sheep red blood cells; T, target. 122 OOOS-8749/83/010122-12$03.00/O Copyright 0 1983 by Academx Press, Inc. All rights of reproduction in any fom reserved
AMLC-INDUCED
SUPPRESSOR
CELLS
123
stances, these cells may exert cytotoxic effects (6,7), helper effects (8,9) or suppression (4, lo- 12). The deficiency of AMLC in a variety of autoimmune disorders (13-l 5) has been thought to be especially important in view of regulatory functions of AMLCactivated cells. These findings have been interpreted as a possible major pathway for the immunoregulatory imbalance generally observed in these diseases. We have chosen AMLC as a model of induction of suppressor cell activity in order to study the suppressor phenomenon in vitro, in a system which may be relevant to a regulatory mechanism in viva. The advantage of this system is that T cells which are activated in AMLC, as a rule, do not exert cytotoxic effect on autologous or allogeneic PHA-treated targets ( 16, 17). This means that inhibitory effects exerted by suppressor cells cannot be attributed simply to cell-mediated lysis of stimulating cells, as is often the case with allo-MLC-induced suppressor cells (18-20). In this communication, we present data on the generation of SC in AMLC and on proliferative and cytotoxic activities when these cells are added to fresh allogeneic MLC. The specificity of suppression in terms of HLA-A,B,C and DR antigens shared by stimulating and target cells was also investigated. This study was performed with norma individuals. Subsequently, we intend to use this model to analyze suppressor cell activity in patients with autoimmune diseases. MATERIALS
Donors and HLA typing. Donors panel. HLA-A,B,C (21, 22).
and DR
typing
AND
METHODS
were healthy volunteers from our local HLA was performed using standardized methods
Cell preparations. Mononuclear cells were separated from heparinized blood by Ficoll-Hypaque method (23). Separation of cells into T and non-T fractions was performed either by AET-treated SRBC rosetting (24) or by incubation on nylon wool columns (25). With the rosetting technique, the rosetted cell fraction gave 9095% T cells as detected by rerosetting with less than 5% SIg-positive B cells, as demonstrated by indirect fluorescence with polyvalent anti-human Ig, and less than 5% monocytes by a&dine orange staining. The nonrosetted fraction consisted of 4565% B cells and 25-45% monocytes. Contamination by T-cells was less than 2%. The nylon wool separation technique provided cell fractions of comparable purity with a somewhat lower percentage of monocytes in the non-T-cell fraction (18 to 30%). In several experiments, where the effect of adherent cells on SC inducibility was studied, adherent cells were first isolated by adherence to plastic petri dishes (26) before carrying out further cell separation procedures. Cell cultures. Suppressor cehs were induced by autologous MLC, where equal numbers of responding (T) and stimulating (non-T) cells were cocultured at a concentration 1 X lo6 cells/ml in RPM1 1640 (Gibco, Grand Island, N.Y.), and supplemented with lo- 15% heat-inactivated pooled human serum (Medical Specialties Lab, Boston, Mass.), 100 units/ml penicillin, and 100 pg/ml streptomycin (Gibco, Grand Island, N.Y.). Cultures were established in 17 X loo-mm Falcon plastic culture tubes (Becton-Dickinson, Oxnard, Calif.) and incubated for 6 to 7 days at 37°C in 5% COZ and 95% humidity. Inactivation of stimulating cells was achieved with Mitomycin C (Sigma, St. Louis, MO.) at a concentration 25 pg/ml.
124
PAZDERKA
ET
AL.
Second (allogeneic) MLC was set up under the same conditions except that both responding and stimulating cells used were unseparated PBL. Cytotoxic lymphocytes were generated in culture tubes ( 17 X 100 mm); whereas proliferative responses were simultaneously measured in microcultures in U-bottom microtiter trays (Nunc, Denmark), using lo* responders and 10’ stimulators in 0.2 ml of medium per well (in triplicates). Cultures were assayed on Day 6 using a Skatron harvester (Lierbyen, Norway), 16 to 18 hr after addition of tritiated thymidine (Amersham, England), at 1 &i per well. Thymidine incorporation was measured in a 12 16 Rackbeta II liquid scintillation counter (LKB Wallac). The degree of cell proliferation was estimated on the basis of stimulation index (SI), calculated as median cpm in allostimulated culture median background cpm ’ Background radioactivity was measured, as a rule, in a culture of responding cells in the presence of Mitomycin-treated responders. CML assay. Cytotoxic cells from 17 X loo-mm tubes were collected, washed, and suspended in medium supplemented with 10% FCS. The cell concentration was adjusted to obtain an effector:target ratio of 50: 1. Target cells were PBL incubated for 72 hr in the presence of 1% PHA (Gibco, Grand Island, N.Y.) before labeling with 51Cr (50-100 &i for 16 hr) and adjusting to 0.4 X lo6 cells/ml (after three washes). The assay was performed (in duplicate) in 12 X 75-mm glass test tubes (Fisher Scientific) with 500 ~1 of effector cell and 25 ~1 of target cell suspensions. Tubes were incubated at 37°C for 4 hr after which 2 ml of cold saline was added. They were then spun at 200g for 5 min in a refrigerated centrifuge. Supernatants were decanted into fresh tubes for radioactive counting of both supernatants and pellets in a 1270 Rackgamma II counter (LKB Wallac). Spontaneous “Cr release was measured in tubes where target cells were incubated without effector cells. Maximum release was determined by treating labeled target cells with hemolyte (Fisher Scientific). The percentage of “Cr release was determined as cpm supematant x 100. cpm supematant + cpm pellet The percentage of specific lysis was calculated as % experimental release - % spontaneous release x 100. % maximum release - % spontaneous release Experimental protocol. T cells were cultured for 6-7 days alone or with Mitomycin C-inactivated autologous non-T-cells. At the end of this culture period, viable cells from each culture were counted, washed, and added to a fresh allogeneic MLC, at a concentration of lo6 cells/ml with or without Mitomycin C treatment. To establish a dose-response curve, precultured cells were added at various proportions initially. In later experiments, the ratio of responder:stimulator:precultured cells was fixed at 1: 1:0.5. The proliferation and cytotoxicity in the second cultures were assayed after 6 days of incubation. Inhibition of proliferation was determined by comparing SI in the second cultures without added precultured cells (positive control) to those where precultured cells were added: SI (positive control) - SI (with precultured cell added) x looo/ 0. SI (positive control)
AMLC-INDUCED
Similarly,
inhibition
of cytotoxicity
SUPPRESSOR
125
CELLS
was estimated as
% specific lysis (pos. control) - % specific lysis (with precultured % specific lysis (positive control)
cells added)
x 100%. RESULTS In preliminary experiments, AMLC-activated T cells caused inhibition of proliferation and cytotoxic activity to a comparable extent in second cultures in which either separated T cells or PBL were used as responding cells (data not shown). Subsequent experiments were therefore performed using PBL as responders. In our first series of experiments, shown in Table 1, AMLC-activated cells caused profound inhibition of both cytotoxic and proliferative responses by autologous PBL to allogeneic stimulators. Inhibition was Mitomycin sensitive, since pretreatment of precultured cells with Mitomycin C reduced the degree of suppression in both assays but more markedly in proliferation. In many cases, Mitomycin C treatment of suppressor cells actually resulted in enhanced proliferation. In 17 experiments, inhibition TABLE
1
Effect of AMLC-Activated Cells on Cytotoxic of Autologous Responders
and Proliferative in MLC
Responses
CML
Expt I
2
Cells added
to MLC
5
% Inhibition
SI
% Inhibition
AMLC-activated AMLC-activated,
43.4 23.6 42.4
45.6 2.8
17.6 9.4 22.1
46.6 -25.6
AMLC-activated AMLC-activated,
22.5 -1.4 10.3
100.0 54.2
7.3 1.4 8.5
80.8 -16.4
AMLC-activated AMLC-activated,
42.0 0.7 14.7
98.3 65.0
4.5 1.9 4.6
57.8 -2.2
AMLC-activated AMLC-activated,
34. I 7.6 28.4
77.7 16.7
20.0 2.1 6.4
89.5 68.0
AMLC-activated AMLC-activated,
31.2 9.3 22.6
70.2 21.6
28.9 11.5 29.5
60.2 -2.1
-
3
4
% Cytotoxicity
MLC
Note. T cells were cultured for 7 days in the presence of autologous (non-T), cells and then added, after or without Mitomycin C treatment, to a second culture containing autologous responders and allogeneic stimulators. HLA profiles of cells used in AMLC (A) and of allogeneic stimulators (X) were as follows: Expt 1: A-A2, 11; BW35, W44; CW4, 5; DR4; X-A2, 26; B7, 27; CW2; DR2, 4. Expt 2: A-A2, W33; B8, W58; CW3; DR3; X-A2; BW35, W52; CW3; DR2, W9. Expt 3: A-A2, W33; B8, W58; CW3, DR3; X-Al 1, W32; BW35, W44, CW4, 5; DR3, 4. Expt 4: A-Al 1, 28; BW35, W42; CW4; DR3, 4; X-A2, W33; B8, W58; CW3; DR3. Expt 5: A-A2; B7, W57; CW6; DR4, 7; X-A2, W33, B8, W58; CW3; DR3.
126
PAZDERKA
ET
AL.
of 75.3 + 22.4 in CML and 60.6 + 18.2 in MLC became 32.8 + 35.7 (for CML) and -12.3 f 47.4 (for MLC) when AMLC-activated cells were treated with Mitomycin C before addition to second cultures. Since all responding and stimulating cells used in our experiments were HLAA,B,C and DR typed, it was possible to analyze whether AMLC-induced suppressor cells are restricted by MHC. No such restriction was found. Also, there was no quantitative difference either in CML or in MLC when correlation was sought between the amount of inhibition and the number of HLA-A,B,C and/or DR antigens shared by the original stimulator with the stimulator of the second culture (Table 1). Thus, the action of these suppressor cells appeared to be nonspecific (in terms of HLA) in this autologous responder system. When precultured cells (from AMLC) were added to a second culture containing allogeneic cells as responders, with either autologous to AMLC cells or third-party PBL as stimulators (Table 2), AMLC-activated cells were equally suppressive to all allogeneic responses irrespective of type of stimulator. Again, suppression was Mitomycin sensitive. Only CML activity was assayed in this series of experiments. We concluded that suppressor cells induced in AMLC are nonspecific, and that valid tests for suppressor cell activity may be performed in second cultures composed of entirely unrelated cells. This consideration is important from the point of view of reliably measuring AMLC suppressor cell activity in patients using cells of healthy individuals in the second culture (thereby also avoiding a second bleeding of the patient). The system also permits the assay of AMLC suppressor activity independently of possible impairment of patient’s own allogeneic responsiveness. TABLE Effect of AMLC-Activated
Cells on Cytotoxic
2 Response
of Allogeneic
Responders
CML Cells
added
%
CML
%
Expt
to X/Am
Cytotoxicity
Inhibition
1
AMLC-activated AMLC-activated,
19.8 6.3 21.5
68.2 -12.5
2 ND
3 4
in MLC
Cells added to X/Ym
% Cytotoxicity
% Inhibition
AMLC-activated AMLC-activated,,,
40.0 13.2 38.9
67.0 2.7
AMLC-activated AMLC-activated,
56.7 -4.4 68.5
100.0 -20.8
AMLC-activated
80.4 24.3
69.8
AMLC-activated
74.8 39.1
47.7
AMLC-activated
35.1 9.5
72.9
AMLC-activated
65.1 34.8
46.5
Note. Cells activated in AMLC were added to a second culture containing allogeneic responders (X) stimulated either by PBL from original stimulator (Am) or by third party allogeneic cells (Ym). HLA profiles of A, X, and Y cells were as follows: Experiment 1. A-Al, W24; B37, W39; CW6; DR4, 7; XA28, W32; B14, W55; DRW6, WlO; Y-A2, 29; BW44, W55; CW3; DRl, 3. Experiment 2. A-A2, 11; BW35; W44; CW4, 5; DR4; X-AI 1, W32, BW35, W44; CW4, 5; DR3, 4; Y-A2, W24; BW60, W62; CW3; DRI, 2. Experiment 3. A-A2, W33; B8, W58; CW3; DR3; X-AI 1, W32; BW35, W44; CW4, 5; DR3,4; Y-A2, W33; BW44, W63; DR2, 7. Experiment 4. A-A2, 11; BW35, W44; CW4, 5; DR4; XAl 1, W32; BW35, W44; CW4, 5; DR3, 4; Y-A2, W33; BW44; W63; DR2, 7. ND, not done.
AMLC-INDUCED
SUPPRESSOR
CELLS
127
Figure 1 depicts a dose-response study, in four experiments, of suppression and the number of added AMLC suppressor cells, from a ratio of 1: 1:O. 1 to I : 1:O.S. A 55.5% inhibition of CML activity was obtained with as few as 10% of added AMLCactivated cells. Kinetics studies of SC generation are summarized in Fig. 2. Suppression became apparent after 3 days of incubation in AMLC; peak activity was reached on Day 6. In these experiments, suppressor cells were added to a second culture at Day 0. We have not examined the kinetics of adding precultured cells at various time intervals after establishing the second culture. It is known, however, that addition of AMLCactivated cells to an MLC 24 hr prior to the termination of the culture does not result in inhibition of either proliferation or cytotoxicity (1 I). This indicates that suppression cannot be attributed to toxic effects exerted by activated cells on the second culture. To assess the proliferative and the cytotoxic potentials of AMLC-stimulated cells, corresponding parameters were assayed by (a) assaying AMLC-precultured cells as responders in second cultures and (b) measuring the cytotoxic activity of these cells against prospective responders and stimulators to be used in a second culture. Table 3 shows that AMLC precultured cells are fully capable of proliferation in the second culture when confronted by allogeneic stimulators. A certain amount of proliferation occurs also in response to autologous PBL. This is not surprising since PBL contain B cells and monocytes, both of which have been implicated as stimulators of autologous MLC activity. Therefore, cell proliferation (as measured by thymidine uptake) in cultures containing added precultured cells must be interpreted with caution, taking into account proliferative activities of both cell populations in the mixture against both autologous and allogeneic stimulators (see Discussion). To explore whether the inhibitory effect of AMLC-activated cells is due to the development of cytotoxic cells during AMLC, precultured cells were tested for cytotoxicity against PHA blasts prepared from PBL of prospective responders and stimulators. The amount of killing of autologous PHA blasts never exceeded 15% (range of variation from -5.1 to 13.6%). The percentage of killing of allogeneic blasts varied from -2.7 to 12.2%. Thus, it is unlikely that the suppression observed in second
100
1
0Proportion of AMLC activated Added to Second Culture
0.50 Cells
FIG. I. Suppressionof the cytotoxic response of autologous cells by AMLC-activated cells: dose-response curve. T cells were incubated in the presence of autologous (non-T)m cells for 7 days and then added to a second MLC (A/Xm) at various proportions. Numbers of responding and stimulating cells in all cultures were kept constant. Each point and bars represent mean + SE obtained in four independent experiments.
128
PAZDERKA
2 0
ET AL.
60_
’ 6 60.G ri 40i3 3 :
20-
Days
I I I I , 3 4 5 6 7 of Incubation in AMLC
FIG. 2. Kinetics of generation of SC in AMLC. AMLC-activated T cells were harvested at various time intervals after initiation of AMLC and added to a second culture (A/Xm) at a ratio 1:1:0.5 (see Materials and Methods). Each point and bars represent mean f SE obtained in three independent experiments.
cultures is due to cytotoxic elimination of either the responding or the stimulating cells, especially when one takes into account that in these cytotoxicity experiments the E:T ratio was 50: 1, whereas in the second culture that ratio would be 1:2. Actually, a significant amount of suppression was observed even at ratios as low as 1: 10 (Fig. 1). Figure 3 shows the lack of correlation between the amount of cytotoxic activity of the precultured cells against either prospective responders or stimulators and their suppressive effect in second cultures. Furthermore, cytotoxicity of the precultured cells was not affected by Mitomycin C treatment (data not shown), whereas their suppressive activity was Mitomycin sensitive. Our experiments are in agreement with those of Smith and Knowlton (11) in showing that stimulation of T-cells by autologous non-T cells is a prerequisite for the generation of suppressor cells in this system. Neither unseparated PBL nor purified T cells, precultured alone, showed any significant suppression (Table 4), as compared TABLE 3 Proliferative Response of AMLC-Precultured
Cells when Added to a Second Culture Stimulators
Expt
Line
Responders
Am
Xm
(AMLCprecultured cells),,,
I
1 2 3 4
A A + AMLC-precultured cells A + (AMLC-precultured cells),,, AMLC-precultured cells
1120” 2700 1060 2860
23,220 26,830 23,190 16,340
1670 3850 350 830
20.7 9.9 21.9 19.7
II
5 6 7 8
A A + AMLC-precultured cells A + (AMLC-precultured cells), AMLC-precultured cells
920 970 530 350
16,220 9,080 11,730 6,480
950 800 100 p80
17.6 9.4 22.1 13.5
Note. Underlined values are the background counts used for calculation of SI against Xm. ’ Median cpm.
SI
AMLC-INDUCED 100
SUPPRESSOR 0
1"
. 0.
l 0
.
2 800 -
.
' 60g _ Fi? h 404 f.n .xc m-
.0 0
1,
1
-50
0
% Killing
FIG. 3. Absence of correlation effect. T cells were activated in responders and stimulators was ratio in all experiments was 50:
129
CELLS
5.0
by AMLC
10.0 - activated
15.0 Cells
between cytotoxic activity of AMLC-activated cells and their suppressive AMLC for 7 days. Their cytotoxicity against PHA blasts of prospective assayed prior to addition of these cells to a second culture (A/Xm). E:T 1. 0, Allogeneic targets (X); 0, autologous targets (A).
to the suppressive effect of AMLC-activated cells. The role of adherent cells in the generation of suppressor activity in autologous MLC is not entirely clear. The amount of suppression exerted by AMLC-activated T cells was, in our experience, comparable irrespective of whether T cells were separated by SRBC-rosetting or by the nylon TABLE Comparison
of Suppressive
4
Effect of Various Cell Preparations Activity in the Second Culture
on the Cytotoxic
CML Expt
Precultured cells added to a second culture
1
PBL AMLC-activated
2
3
4
5
a Adherent
35.7 43.2 23.5
-21.0 34.2
7.9 76.1 55.3 19.7
T
45.6 42.0 10.9 20.4 36.6
AMLC-activated T alone
T
24.5 15.8 29.6
35.5 -20.8
AMLC-activated T alone
T
21.0 3.8 20.1
81.9 4.3
79.7 59.0 -2.2 75.7
26.0 100.0 5.0
37.2 11.2
69.9
AMLC-activated cells removed
T
% Inhibition
PBL AMLC-activated T alone B alone
PBL” AMLC-activated Adherent cells
6
% Cytotoxicity
before
T’
TO
separation
of T and non-T
cells.
130
PAZDERKA
ET AL.
wool technique (data not shown). However, neither of these techniques results in the complete removal of adherent cells from the stimulating cell population. Experiments in which adherent cells were removed prior to T and non-T-cell separation (experiments 5, 6, Table 4) suggest that the presence of adherent cells in autologous MLC is not crucial for the generation of suppressor activity. DISCUSSION Proliferation of T cells upon stimulation with autologous non-T cells in vitro is a well-established phenomenon (1, 2). AMLR is believed to be driven by Ia-like determinants on stimulating non-T cells (27, 28). Although the biologic significance of this phenomenon is still uncertain, the ability of T cells to proliferate in AMLC has been thought to represent a measure of T-cell regulatory ability (2, 3, 5). It has been demonstrated that the predominant functional consequence of activation of T cells in AMLC is the generation of suppressor T cells. These cells are able to suppress both proliferative response and immunoglobulin synthesis in PWM-driven cultures of autologous B cells (12) and inhibit proliferation and generation of cytotoxic cells upon stimulation of autologous T cells by allogeneic stimulators (10, 11). It was also shown that Con A-induced suppressor cells are drawn from the population of T lymphocytes activated in AMLC (4). Our studies confirm previous observations (10, 11) that AMLC-activated cells are able to suppress both proliferation and cytotoxic cell generation by autologous lymphocytes stimulated in MLC. Suppressor cells generated in this system are Mitomycin C sensitive, indicating that proliferation of suppressor cells is probably necessary in this system. This finding is in contrast to some studies demonstrating resistance of in vitro-induced suppressor cells to Mitomycin C and irradiation, using different SC induction techniques (29, 30) but consistent with the results of Smith and Knowlton (11) and Innes et al. (4). The action of suppressor cells induced in AMLC appears to be nonspecific in the sense that the response to any allogeneic stimulator is suppressed, irrespective of HLA-A,B,C and DR profile (Table 1). This finding is in agreement with several previous observations demonstrating the lack of specificity after in vitro induction of suppressor cells in humans (29, 30). Sasportes et al. (31) were able to demonstrate DR-specific suppression but only after repeated in vitro sensitization by allogeneic cells. AMLC-induced suppressor cells are able to inhibit not only the response of autologous lymphocytes, but are equally suppressive to allogeneic responders against any stimulator (Table 2). T cells from autologous MLC do not lose their capacity to proliferate when used as responders in a second culture (Table 3). It has been shown by Kozak et al. (32) that T cells from autologous MLC are composed of cells that were activated in AMLC (autoreactive) and cells that remained nonactivated during AMLC (alloreactive). In their experiments, it was possible to fractionate these cell populations using Percoll density gradient centrifugation and analyze them separately. Autoreactive cells, when used as responders in a second MLC, showed secondary proliferative kinetics against both autologous and allogeneic stimuli, while alloreactive cells responded only to allogeneic stimulators in a primary fashion. Since separation of AMLC-stimulated T cells was not attempted in our experiments, proliferation of these cells when used as responders in a second culture presumably reflects the mixed activities of secondary
AMLC-INDUCED
SUPPRESSOR
CELLS
131
response by autoreactive cells and primary response by alloreactive cells. Secondary response was not explored in our experiments, since our main purpose was to analyze the effect of AMLC-stimulated cells on allogeneic reactivity. In a 6-day culture, AMLCprecultured cells showed a primary proliferative response against allogeneic stimulators, Xm, in all experiments, giving SI comparable to that obtained in A/Xm group (Table 3). This response most likely reflects the activity of alloreactive cells and is not directly related to activity of SC. However, persistence of alloreactivity in AMLCstimulated cell population complicates the assessment of SC effect in MLC. It should be kept in mind that the response to Xm in the experimental group [A + AMLCprecultured cells] is, in fact, composed of two components: proliferation of A cells and proliferation of AMLC-precultured cells against the allogeneic stimulator Xm. If the primary response by precultured (alloreactive) cells is sufficiently high, the total cpm response in this group may be even higher than in A/Xm group (for example, line 2, Table 3) and thus, no suppressive effect is apparent judging by cpm. However, it should be kept in mind that over 50% of proliferative response in this group is, in fact, contributed by AMLC-precultured cells (lines 4 and 8, Table 3). Thus, proliferation of A cells against Xm in this group is considerably reduced as compared to A/Xm group (lines 1 and 5). In order to evaluate the effect of suppressor cells on second MLC, combination [AMLC-precultured cells/Xm] should always be included as an important control and, when estimating the percentage of suppression, the values should be adjusted for the amount of proliferation contributed by alloreactive cells. Another factor complicating the interpretation of MLC results is the continuous proliferation of AMLCactivated cells against autologous A cells. Under the conditions of our experiments (6-day incubation in a second culture), a secondary response against autologous stimulators was not observed, and the amount of proliferation in [AMLC-stimulated cells/ Am] varied from experiment to experiment. Where it was expressed, it resulted in the increase of background cpm and thus, affected the SI. Therefore, to evaluate the net amount of suppression in MLC, the response in experimental group should be adjusted for both proliferation of AMLC-stimulated cells against allogeneic stimulator and for increased background in the autologous control. The fact that AMLC-activated cells retain their ability to proliferate in response to allogeneic stimulators is in agreement with the conclusion of Sakane and Green (10) that autologous and allogeneic MLR are for the most part independently mediated by two different T-cell populations in human peripheral blood, although they may partially overlap. More recent studies, using monoclonal antibodies against specific T-cell subsets, show that autoactivated T cells are drawn predominantly from the OKT4-reactive subset, whereas alloactivated T cells are drawn from both OKT4 (helper/inducer) and OKTS (cytotoxic/suppressor) positive subsets (32). Similar conclusions were also reached by Engleman ez al. (33) using different sets of monoclonal antibodies. Suppression induced in allogeneic MLC is often explained as a result of cytotoxic elimination of the stimulating cells by effector cells generated in the inducing culture (18-20, 34). However, most investigators failed to demonstrate cytotoxic cells in AMLC (16, 17). In this study, we were likewise unable to demonstrate the cytotoxic effect of AMLC-activated cells against autologous PHA blasts. Furthermore, no cytotoxic activity against allogeneic PHA blasts could be demonstrated (Fig. 3). One
132
PAZDERKA
ET AL.
may speculate that the cytotoxic effect of these cells, while not demonstrable in a short-term (4 hr) assay, may be accumulating during incubation in. a second culture. To test this possibility, AMLC-activated cells were restimulated by autologous or allogeneic stimulators and tested in CML 3 and 6 days later. No significant killing was observed (data not shown). Thus, inhibitory action of AMLC-activated cells is not a consequence of simple cytotoxic elimination of either responding or stimulating cell populations in the second culture, but truly represents a regulatory suppressive mechanism. Participation by various cell populations in the induction and mediation of sup pressor effect, as observed in various in vitro models, is still a matter of controversy. In AMLC, in particular, the role of adherent cells is uncertain. On one hand, macrophages have been demonstrated to serve as potent stimulators of AMLR in humans (35-37). On the other hand, however, elimination of macrophages from stimulator cells results in higher proliferative responses (2, 38). This phenomenon has been interpreted by Smolen et al. (39) as evidence of suppressive effect of macrophages on AMLC. However, involvement of macrophages in the generation of cells which inhibit responses to allogeneic stimulators has not been studied in depth. We have approached this question by comparing suppression generated in T/non-Tm-stimulated cultures from which adherent cells were removed before further separation to that when adherent cells were present, No appreciable difference was found between these two types of cultures (Table 4, experiments 5 and 6). Thus, adherent cells in AMLC do not appear to be crucial for generation of suppressor cells as assayed in alloantigen-stimulated system. However, our technique for macrophage depletion may not have been sufficient to achieve complete removal of macrophages, and a small number of these cells may be enough for generation of suppressor effect as demonstrated in a Con A model by Innes et al. (4). Studies using specific monoclonal antibodies to macrophages are necessary in order to answer this question conclusively. The mediators of suppression induced in AMLC appear to be T lymphocytes (4, 10, 11). In our experiments, in agreement with observations of Smith and Knowlton (1 l), maximal suppressor activity was generated only in cultures where T cells were stimulated with autologous non-T cells (Table 4). Suppressor cells were not, as a rule, recovered from cultures of purified T or non-T cells cultured alone. Low suppressive activity occasionally observed in the cultures of T cells alone was most probably due to contaminating non-T cells in the T-cell preparation. Thus, activation of T cells by autologous non-T cells is a prerequisite to the generation of suppressor cells in this system. Our findings support the notion that the stimulation noted in AMLC reflects an important immunoregulatory mechanism. The absence of this reaction in patients with CLL (40) and SLE (13,4 1) is in agreement with the hypothesis that regulatory failure of the immune system could be important in the pathogenesis of certain lymphoproliferative and autoimmune diseases. The application of this system to the study of immune status of such patients is warranted. ACKNOWLEDGMENTS This work was supported by MRC of Canada. We wish to express our thanks to Ms. J. Clark for efficient secretarial help.
REFERENCES I. Opelz, G., Kiuchi, M., Takasugi, M., and Terasaki, P. I., J. Exp. Med. 142, 1327, 1976. 2. Kuntz, M. M., Innes, J. B., and Weksler, M. E., J. Exp. Med. 143, 1042, 1976.
AMLC-INDUCED 3. 4. 5. 6. 7. 8. 9. 10. Il. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22,
23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41.
SUPPRESSOR
CELLS
133
Weksler, M. E., and Kozak, R., J. Exp. Med. 146, 1833, 1977. Irmes, J. B., Kuntz, M. M., Kim, Y. T., and Weksler, M. E., J. Clin. Invest. 64, 1608, 1979. Smolen, J. S., Luger, T. A., Chused, T. M., and Steinberg, A. D., J. Clin. Invest. 68, 1601, 1981. Miller, R. A., and Kaplan, H. S., J. Immunol. 121, 2165, 1978. Tomonari, K., J. Immunol. 124, 1111, 1980. Chiorazzi, N., Fu, S. M., and Kunkel, H. G., J. Exp. Med. 149, 1543, 1979. Hausmann, P. B., and Stobo, J. D., J. Exp. Med. 149, 1537, 1979. Sakane, T., and Green, I., J. Immunol. 123, 584, 1979. Smith, J. B., and Knowlton, R. P., J. Immunol. 123, 419, 1979. James, S. P., Yenokida, G. G., Graeff, A. S., Elson, C. O., and Strobcr, W., J. Immunoi. 127, 2605, 1981. Sakane, T., Steinberg, A. D., and Green, I., Proc. Nat. Acad. Sci. USA 75, 3464, 1978. Miyasaka, N., Sauvezie, B., Pierce, D. A., Daniels, T. E., and Talal, N., J. Clin. Invest 66, 928, 1980. James, S. P., Elson, C. D., Waggoner, J. G., Jones, E. A., and Strober, W., J. C/in. Invest 66, 1305, 1980. Vande Stouwe, R. A., Kunkel, H. G., Halper, J. P., and Weksler, M. E., J. Exp. Med. 146, 1809, 1977. Katz, P., Haynes, B. F., and Fauci, A. S., J. Immunol. 121, 1998, 1978. Hodes, R. J., Nadler, L. M., and Hathcock, K. S., Immunology 119, 961, 1977. Fitch, F. W., Engers, H. D., Cerottini, J. C., and Brunner, K. T., J. Immunol. 116, 716. 1976. Sinclair, N. R., Lees, R. K., Wheeler, M. E., Vichos, E. E., and Fung, F. Y., Cell. Immunol. 27, 153, 1976. Mittal, K. K., Mickey, M. R., Singal, D. P., and Terasaki, P. I., Transplantation 6, 913, 1968. Kissmeyer-Nielsen, F., and Dick, H., In “Histocompatibility Techniques” (H. Dick and F. KissmeyerNielsen, Eds.), pp. 9-38, Elsevier/North-Holland Biomedical Press, Amsterdam/New York/Oxford, 1979. Boyum, A., Stand. J. Clin. Lab. Invest. Suppl. 97, 17, 1968. Pellegrino, M. A., Ferrone, S., Dierich, M. P., and Reisfeld, R. A., Clin. Immunol. Immunopathol. 3, 324, 1975. Lowry, R. P., Person, A. E., Goguen, J. E., Carpenter, C. B., and Garovoy, M. R., Transplant. Proc. 10, 833, 1978. Tomonari, K., Wakisaka, A., and Aizawa, M., J. Immunol. 125, 1596, 1980. Bergholtz, B., Albrechtsen, D., and Thorsby, E., Tissue Antigens 10, 27, 1977. Gottlieb, A. B., Fu, S. M., Yu, D. T., Wang, C. Y., Halpcr, J. P., and Kunkel, H. G., J. Immunol. 123, 1497, 1979. Hirschberg, H., and Thorsby, E., Stand. J. Immunol. 6, 809, 1977. Shou, L., Schwartz, A., and Good, R. A., J. Exp. Med. 143, 1100, 1976. Sasportes, M., Fradelizi, D., and Dausset, J., Nature (London) 276, 502, 1978. Kozak, R. W., Moody, C. E., Staiano-Coico, L., and Weksler, M. E., J. Immunol. 128, 1723, 1982. Engleman, E. G., Benike, C. J., Grumet, F. C., and Evans, R. L., J. Immunol. 127, 2 124, 198 I. Sheehy, M. J., Mawas, C., and Charmot, D., J. Immunol. 122, 2198, 1979. Beale, M. G., MacDermott, R. P., Stacey, M. C., Nash, J. S., Hahn, B. H., Seiden, M. V., BurkholderJacobs, S. L., and Pletscher-Lowenstein, L. S., J. Immunol. 124, 227, 1980. MacDermott, R. P., and Stacey, M. C., J. Immunol. 126, 729, 1981. Dock, N. L., and Davey, F. R., Clin. Immunol. Immunopathol. 17, 123, 1980. Katz, P., and Fauci, A. S., J. Immunol. 123, 2270, 1979. Smolen, J. S., Sharrow, S. O., Reeves, J. P., Boegel, W. A., and Steinberg, A. D., J. Immunol. 127, 1987, 1981. Smith, J. B., Knowlton, R. P., and Koons, L. S., J. Nat. Cancer Inst. 58, 579, 1977. Kuntz, M. M., Innes, J. B., and Weksler, M. E., Clin. Res. 24, 448A, 1976.
IMMUNOLOGY
75, 122- 133 ( 1983)
Induction of Suppressor Cells in Autologous Mixed Lymphocyte Culture (AMLC) in Humans FELDZGERITTA PAZDERKA,’ ALMARIO ANGELES, THAVISAKDI KOVITHAVONGS, AND JOHN B. DOSSETOR Clinical
and Transplant Immunology Group, Departments of Medicine University of Alberta, Edmonton, Alberta T6G 2E1,
Received August
IO, 1982; accepted
October
and Laboratory Canada
Medicine,
29, I982
T cells stimulated for 6-7 days in autologous mixed lymphocyte culture (AMLC) showed suppressive effects when added to fresh mixed cultures where autologous lymphocytes (A) were stimulated by Mitomycin C-treated allogeneic lymphocytes (Xm), in a ratio of A:Xm:AMLCactivated cells of 1: 1 :OS. Both cytotoxic and proliferative activities in second cultures, as assayed after 6 days of incubation, were significantly inhibited (percentage suppression of cytotoxic activity observed in 17 experiments was 75.3 f 22.4; percentage suppression of proliferation was 60.6 * 18.2). Suppressor cells (SC) generated in AMLC were Mitomycin C sensitive and nonspecific in their action; not only A/Xm but also X/Am and X/Ym cultures were suppressed to the same extent. AMLC-Activated cells showed a considerable degree of proliferation in response to alloantigens but failed to express any cytotoxic activity against autologous or allogeneic phytohemagglutinin blasts. Thus, the inhibitory effect observed in this system is not due to cytotoxic elimination of responding or stimulating cells in the second culture but rather reflects a true regulatory (suppressive) mechanism. INTRODUCTION It has been demonstrated repeatedly (1, 2) that human T lymphocytes can be stimulated by autologous non-T lymphocytes to proliferate in culture (AMLC).* AMLC has been shown to possess two main attributes of immune responses: immunologic specificity and memory (3). Although the biologic significance of this phenomenon is still uncertain, it has been interpreted by some authors as reflecting a mechanism by which T lymphocytes might maintain self-tolerance in vivo (4) and as representing a measure of T-cell regulatory ability (3, 5). In addition to their proliferative response, a variety of effector and regulatory functions have been attributed to AMLC-activated T cells. Under certain circum’ To whom correspondence should be addressed. * Abbreviations used: AET, 2-aminoethylisothiouronium bromide hydrobromide; AMLC, autologous mixed lymphocyte culture; AMLR, autologous mixed lymphocyte reaction; CLL, chronic lymphatic leukemia; CML, cell-mediated lymphocytotoxicity; Con A, concanavalin A; E, effector; FCS, fetal calf serum; Ig, immunoglobulin; MHC, major histocompatibility complex; MLC, mixed lymphocyte culture; PBL, peripheral blood lymphocytes; PHA, phytohemagglutinin; PWM, pokeweed mitogen; SC, suppressor cell; SI, stimulation index; SIg, surface immunoglobulin; SLE, systemic lupus erythematosus; SRBC, sheep red blood cells; T, target. 122 OOOS-8749/83/010122-12$03.00/O Copyright 0 1983 by Academx Press, Inc. All rights of reproduction in any fom reserved
AMLC-INDUCED
SUPPRESSOR
CELLS
123
stances, these cells may exert cytotoxic effects (6,7), helper effects (8,9) or suppression (4, lo- 12). The deficiency of AMLC in a variety of autoimmune disorders (13-l 5) has been thought to be especially important in view of regulatory functions of AMLCactivated cells. These findings have been interpreted as a possible major pathway for the immunoregulatory imbalance generally observed in these diseases. We have chosen AMLC as a model of induction of suppressor cell activity in order to study the suppressor phenomenon in vitro, in a system which may be relevant to a regulatory mechanism in viva. The advantage of this system is that T cells which are activated in AMLC, as a rule, do not exert cytotoxic effect on autologous or allogeneic PHA-treated targets ( 16, 17). This means that inhibitory effects exerted by suppressor cells cannot be attributed simply to cell-mediated lysis of stimulating cells, as is often the case with allo-MLC-induced suppressor cells (18-20). In this communication, we present data on the generation of SC in AMLC and on proliferative and cytotoxic activities when these cells are added to fresh allogeneic MLC. The specificity of suppression in terms of HLA-A,B,C and DR antigens shared by stimulating and target cells was also investigated. This study was performed with norma individuals. Subsequently, we intend to use this model to analyze suppressor cell activity in patients with autoimmune diseases. MATERIALS
Donors and HLA typing. Donors panel. HLA-A,B,C (21, 22).
and DR
typing
AND
METHODS
were healthy volunteers from our local HLA was performed using standardized methods
Cell preparations. Mononuclear cells were separated from heparinized blood by Ficoll-Hypaque method (23). Separation of cells into T and non-T fractions was performed either by AET-treated SRBC rosetting (24) or by incubation on nylon wool columns (25). With the rosetting technique, the rosetted cell fraction gave 9095% T cells as detected by rerosetting with less than 5% SIg-positive B cells, as demonstrated by indirect fluorescence with polyvalent anti-human Ig, and less than 5% monocytes by a&dine orange staining. The nonrosetted fraction consisted of 4565% B cells and 25-45% monocytes. Contamination by T-cells was less than 2%. The nylon wool separation technique provided cell fractions of comparable purity with a somewhat lower percentage of monocytes in the non-T-cell fraction (18 to 30%). In several experiments, where the effect of adherent cells on SC inducibility was studied, adherent cells were first isolated by adherence to plastic petri dishes (26) before carrying out further cell separation procedures. Cell cultures. Suppressor cehs were induced by autologous MLC, where equal numbers of responding (T) and stimulating (non-T) cells were cocultured at a concentration 1 X lo6 cells/ml in RPM1 1640 (Gibco, Grand Island, N.Y.), and supplemented with lo- 15% heat-inactivated pooled human serum (Medical Specialties Lab, Boston, Mass.), 100 units/ml penicillin, and 100 pg/ml streptomycin (Gibco, Grand Island, N.Y.). Cultures were established in 17 X loo-mm Falcon plastic culture tubes (Becton-Dickinson, Oxnard, Calif.) and incubated for 6 to 7 days at 37°C in 5% COZ and 95% humidity. Inactivation of stimulating cells was achieved with Mitomycin C (Sigma, St. Louis, MO.) at a concentration 25 pg/ml.
124
PAZDERKA
ET
AL.
Second (allogeneic) MLC was set up under the same conditions except that both responding and stimulating cells used were unseparated PBL. Cytotoxic lymphocytes were generated in culture tubes ( 17 X 100 mm); whereas proliferative responses were simultaneously measured in microcultures in U-bottom microtiter trays (Nunc, Denmark), using lo* responders and 10’ stimulators in 0.2 ml of medium per well (in triplicates). Cultures were assayed on Day 6 using a Skatron harvester (Lierbyen, Norway), 16 to 18 hr after addition of tritiated thymidine (Amersham, England), at 1 &i per well. Thymidine incorporation was measured in a 12 16 Rackbeta II liquid scintillation counter (LKB Wallac). The degree of cell proliferation was estimated on the basis of stimulation index (SI), calculated as median cpm in allostimulated culture median background cpm ’ Background radioactivity was measured, as a rule, in a culture of responding cells in the presence of Mitomycin-treated responders. CML assay. Cytotoxic cells from 17 X loo-mm tubes were collected, washed, and suspended in medium supplemented with 10% FCS. The cell concentration was adjusted to obtain an effector:target ratio of 50: 1. Target cells were PBL incubated for 72 hr in the presence of 1% PHA (Gibco, Grand Island, N.Y.) before labeling with 51Cr (50-100 &i for 16 hr) and adjusting to 0.4 X lo6 cells/ml (after three washes). The assay was performed (in duplicate) in 12 X 75-mm glass test tubes (Fisher Scientific) with 500 ~1 of effector cell and 25 ~1 of target cell suspensions. Tubes were incubated at 37°C for 4 hr after which 2 ml of cold saline was added. They were then spun at 200g for 5 min in a refrigerated centrifuge. Supernatants were decanted into fresh tubes for radioactive counting of both supernatants and pellets in a 1270 Rackgamma II counter (LKB Wallac). Spontaneous “Cr release was measured in tubes where target cells were incubated without effector cells. Maximum release was determined by treating labeled target cells with hemolyte (Fisher Scientific). The percentage of “Cr release was determined as cpm supematant x 100. cpm supematant + cpm pellet The percentage of specific lysis was calculated as % experimental release - % spontaneous release x 100. % maximum release - % spontaneous release Experimental protocol. T cells were cultured for 6-7 days alone or with Mitomycin C-inactivated autologous non-T-cells. At the end of this culture period, viable cells from each culture were counted, washed, and added to a fresh allogeneic MLC, at a concentration of lo6 cells/ml with or without Mitomycin C treatment. To establish a dose-response curve, precultured cells were added at various proportions initially. In later experiments, the ratio of responder:stimulator:precultured cells was fixed at 1: 1:0.5. The proliferation and cytotoxicity in the second cultures were assayed after 6 days of incubation. Inhibition of proliferation was determined by comparing SI in the second cultures without added precultured cells (positive control) to those where precultured cells were added: SI (positive control) - SI (with precultured cell added) x looo/ 0. SI (positive control)
AMLC-INDUCED
Similarly,
inhibition
of cytotoxicity
SUPPRESSOR
125
CELLS
was estimated as
% specific lysis (pos. control) - % specific lysis (with precultured % specific lysis (positive control)
cells added)
x 100%. RESULTS In preliminary experiments, AMLC-activated T cells caused inhibition of proliferation and cytotoxic activity to a comparable extent in second cultures in which either separated T cells or PBL were used as responding cells (data not shown). Subsequent experiments were therefore performed using PBL as responders. In our first series of experiments, shown in Table 1, AMLC-activated cells caused profound inhibition of both cytotoxic and proliferative responses by autologous PBL to allogeneic stimulators. Inhibition was Mitomycin sensitive, since pretreatment of precultured cells with Mitomycin C reduced the degree of suppression in both assays but more markedly in proliferation. In many cases, Mitomycin C treatment of suppressor cells actually resulted in enhanced proliferation. In 17 experiments, inhibition TABLE
1
Effect of AMLC-Activated Cells on Cytotoxic of Autologous Responders
and Proliferative in MLC
Responses
CML
Expt I
2
Cells added
to MLC
5
% Inhibition
SI
% Inhibition
AMLC-activated AMLC-activated,
43.4 23.6 42.4
45.6 2.8
17.6 9.4 22.1
46.6 -25.6
AMLC-activated AMLC-activated,
22.5 -1.4 10.3
100.0 54.2
7.3 1.4 8.5
80.8 -16.4
AMLC-activated AMLC-activated,
42.0 0.7 14.7
98.3 65.0
4.5 1.9 4.6
57.8 -2.2
AMLC-activated AMLC-activated,
34. I 7.6 28.4
77.7 16.7
20.0 2.1 6.4
89.5 68.0
AMLC-activated AMLC-activated,
31.2 9.3 22.6
70.2 21.6
28.9 11.5 29.5
60.2 -2.1
-
3
4
% Cytotoxicity
MLC
Note. T cells were cultured for 7 days in the presence of autologous (non-T), cells and then added, after or without Mitomycin C treatment, to a second culture containing autologous responders and allogeneic stimulators. HLA profiles of cells used in AMLC (A) and of allogeneic stimulators (X) were as follows: Expt 1: A-A2, 11; BW35, W44; CW4, 5; DR4; X-A2, 26; B7, 27; CW2; DR2, 4. Expt 2: A-A2, W33; B8, W58; CW3; DR3; X-A2; BW35, W52; CW3; DR2, W9. Expt 3: A-A2, W33; B8, W58; CW3, DR3; X-Al 1, W32; BW35, W44, CW4, 5; DR3, 4. Expt 4: A-Al 1, 28; BW35, W42; CW4; DR3, 4; X-A2, W33; B8, W58; CW3; DR3. Expt 5: A-A2; B7, W57; CW6; DR4, 7; X-A2, W33, B8, W58; CW3; DR3.
126
PAZDERKA
ET
AL.
of 75.3 + 22.4 in CML and 60.6 + 18.2 in MLC became 32.8 + 35.7 (for CML) and -12.3 f 47.4 (for MLC) when AMLC-activated cells were treated with Mitomycin C before addition to second cultures. Since all responding and stimulating cells used in our experiments were HLAA,B,C and DR typed, it was possible to analyze whether AMLC-induced suppressor cells are restricted by MHC. No such restriction was found. Also, there was no quantitative difference either in CML or in MLC when correlation was sought between the amount of inhibition and the number of HLA-A,B,C and/or DR antigens shared by the original stimulator with the stimulator of the second culture (Table 1). Thus, the action of these suppressor cells appeared to be nonspecific (in terms of HLA) in this autologous responder system. When precultured cells (from AMLC) were added to a second culture containing allogeneic cells as responders, with either autologous to AMLC cells or third-party PBL as stimulators (Table 2), AMLC-activated cells were equally suppressive to all allogeneic responses irrespective of type of stimulator. Again, suppression was Mitomycin sensitive. Only CML activity was assayed in this series of experiments. We concluded that suppressor cells induced in AMLC are nonspecific, and that valid tests for suppressor cell activity may be performed in second cultures composed of entirely unrelated cells. This consideration is important from the point of view of reliably measuring AMLC suppressor cell activity in patients using cells of healthy individuals in the second culture (thereby also avoiding a second bleeding of the patient). The system also permits the assay of AMLC suppressor activity independently of possible impairment of patient’s own allogeneic responsiveness. TABLE Effect of AMLC-Activated
Cells on Cytotoxic
2 Response
of Allogeneic
Responders
CML Cells
added
%
CML
%
Expt
to X/Am
Cytotoxicity
Inhibition
1
AMLC-activated AMLC-activated,
19.8 6.3 21.5
68.2 -12.5
2 ND
3 4
in MLC
Cells added to X/Ym
% Cytotoxicity
% Inhibition
AMLC-activated AMLC-activated,,,
40.0 13.2 38.9
67.0 2.7
AMLC-activated AMLC-activated,
56.7 -4.4 68.5
100.0 -20.8
AMLC-activated
80.4 24.3
69.8
AMLC-activated
74.8 39.1
47.7
AMLC-activated
35.1 9.5
72.9
AMLC-activated
65.1 34.8
46.5
Note. Cells activated in AMLC were added to a second culture containing allogeneic responders (X) stimulated either by PBL from original stimulator (Am) or by third party allogeneic cells (Ym). HLA profiles of A, X, and Y cells were as follows: Experiment 1. A-Al, W24; B37, W39; CW6; DR4, 7; XA28, W32; B14, W55; DRW6, WlO; Y-A2, 29; BW44, W55; CW3; DRl, 3. Experiment 2. A-A2, 11; BW35; W44; CW4, 5; DR4; X-AI 1, W32, BW35, W44; CW4, 5; DR3, 4; Y-A2, W24; BW60, W62; CW3; DRI, 2. Experiment 3. A-A2, W33; B8, W58; CW3; DR3; X-AI 1, W32; BW35, W44; CW4, 5; DR3,4; Y-A2, W33; BW44, W63; DR2, 7. Experiment 4. A-A2, 11; BW35, W44; CW4, 5; DR4; XAl 1, W32; BW35, W44; CW4, 5; DR3, 4; Y-A2, W33; BW44; W63; DR2, 7. ND, not done.
AMLC-INDUCED
SUPPRESSOR
CELLS
127
Figure 1 depicts a dose-response study, in four experiments, of suppression and the number of added AMLC suppressor cells, from a ratio of 1: 1:O. 1 to I : 1:O.S. A 55.5% inhibition of CML activity was obtained with as few as 10% of added AMLCactivated cells. Kinetics studies of SC generation are summarized in Fig. 2. Suppression became apparent after 3 days of incubation in AMLC; peak activity was reached on Day 6. In these experiments, suppressor cells were added to a second culture at Day 0. We have not examined the kinetics of adding precultured cells at various time intervals after establishing the second culture. It is known, however, that addition of AMLCactivated cells to an MLC 24 hr prior to the termination of the culture does not result in inhibition of either proliferation or cytotoxicity (1 I). This indicates that suppression cannot be attributed to toxic effects exerted by activated cells on the second culture. To assess the proliferative and the cytotoxic potentials of AMLC-stimulated cells, corresponding parameters were assayed by (a) assaying AMLC-precultured cells as responders in second cultures and (b) measuring the cytotoxic activity of these cells against prospective responders and stimulators to be used in a second culture. Table 3 shows that AMLC precultured cells are fully capable of proliferation in the second culture when confronted by allogeneic stimulators. A certain amount of proliferation occurs also in response to autologous PBL. This is not surprising since PBL contain B cells and monocytes, both of which have been implicated as stimulators of autologous MLC activity. Therefore, cell proliferation (as measured by thymidine uptake) in cultures containing added precultured cells must be interpreted with caution, taking into account proliferative activities of both cell populations in the mixture against both autologous and allogeneic stimulators (see Discussion). To explore whether the inhibitory effect of AMLC-activated cells is due to the development of cytotoxic cells during AMLC, precultured cells were tested for cytotoxicity against PHA blasts prepared from PBL of prospective responders and stimulators. The amount of killing of autologous PHA blasts never exceeded 15% (range of variation from -5.1 to 13.6%). The percentage of killing of allogeneic blasts varied from -2.7 to 12.2%. Thus, it is unlikely that the suppression observed in second
100
1
0Proportion of AMLC activated Added to Second Culture
0.50 Cells
FIG. I. Suppressionof the cytotoxic response of autologous cells by AMLC-activated cells: dose-response curve. T cells were incubated in the presence of autologous (non-T)m cells for 7 days and then added to a second MLC (A/Xm) at various proportions. Numbers of responding and stimulating cells in all cultures were kept constant. Each point and bars represent mean + SE obtained in four independent experiments.
128
PAZDERKA
2 0
ET AL.
60_
’ 6 60.G ri 40i3 3 :
20-
Days
I I I I , 3 4 5 6 7 of Incubation in AMLC
FIG. 2. Kinetics of generation of SC in AMLC. AMLC-activated T cells were harvested at various time intervals after initiation of AMLC and added to a second culture (A/Xm) at a ratio 1:1:0.5 (see Materials and Methods). Each point and bars represent mean f SE obtained in three independent experiments.
cultures is due to cytotoxic elimination of either the responding or the stimulating cells, especially when one takes into account that in these cytotoxicity experiments the E:T ratio was 50: 1, whereas in the second culture that ratio would be 1:2. Actually, a significant amount of suppression was observed even at ratios as low as 1: 10 (Fig. 1). Figure 3 shows the lack of correlation between the amount of cytotoxic activity of the precultured cells against either prospective responders or stimulators and their suppressive effect in second cultures. Furthermore, cytotoxicity of the precultured cells was not affected by Mitomycin C treatment (data not shown), whereas their suppressive activity was Mitomycin sensitive. Our experiments are in agreement with those of Smith and Knowlton (11) in showing that stimulation of T-cells by autologous non-T cells is a prerequisite for the generation of suppressor cells in this system. Neither unseparated PBL nor purified T cells, precultured alone, showed any significant suppression (Table 4), as compared TABLE 3 Proliferative Response of AMLC-Precultured
Cells when Added to a Second Culture Stimulators
Expt
Line
Responders
Am
Xm
(AMLCprecultured cells),,,
I
1 2 3 4
A A + AMLC-precultured cells A + (AMLC-precultured cells),,, AMLC-precultured cells
1120” 2700 1060 2860
23,220 26,830 23,190 16,340
1670 3850 350 830
20.7 9.9 21.9 19.7
II
5 6 7 8
A A + AMLC-precultured cells A + (AMLC-precultured cells), AMLC-precultured cells
920 970 530 350
16,220 9,080 11,730 6,480
950 800 100 p80
17.6 9.4 22.1 13.5
Note. Underlined values are the background counts used for calculation of SI against Xm. ’ Median cpm.
SI
AMLC-INDUCED 100
SUPPRESSOR 0
1"
. 0.
l 0
.
2 800 -
.
' 60g _ Fi? h 404 f.n .xc m-
.0 0
1,
1
-50
0
% Killing
FIG. 3. Absence of correlation effect. T cells were activated in responders and stimulators was ratio in all experiments was 50:
129
CELLS
5.0
by AMLC
10.0 - activated
15.0 Cells
between cytotoxic activity of AMLC-activated cells and their suppressive AMLC for 7 days. Their cytotoxicity against PHA blasts of prospective assayed prior to addition of these cells to a second culture (A/Xm). E:T 1. 0, Allogeneic targets (X); 0, autologous targets (A).
to the suppressive effect of AMLC-activated cells. The role of adherent cells in the generation of suppressor activity in autologous MLC is not entirely clear. The amount of suppression exerted by AMLC-activated T cells was, in our experience, comparable irrespective of whether T cells were separated by SRBC-rosetting or by the nylon TABLE Comparison
of Suppressive
4
Effect of Various Cell Preparations Activity in the Second Culture
on the Cytotoxic
CML Expt
Precultured cells added to a second culture
1
PBL AMLC-activated
2
3
4
5
a Adherent
35.7 43.2 23.5
-21.0 34.2
7.9 76.1 55.3 19.7
T
45.6 42.0 10.9 20.4 36.6
AMLC-activated T alone
T
24.5 15.8 29.6
35.5 -20.8
AMLC-activated T alone
T
21.0 3.8 20.1
81.9 4.3
79.7 59.0 -2.2 75.7
26.0 100.0 5.0
37.2 11.2
69.9
AMLC-activated cells removed
T
% Inhibition
PBL AMLC-activated T alone B alone
PBL” AMLC-activated Adherent cells
6
% Cytotoxicity
before
T’
TO
separation
of T and non-T
cells.
130
PAZDERKA
ET AL.
wool technique (data not shown). However, neither of these techniques results in the complete removal of adherent cells from the stimulating cell population. Experiments in which adherent cells were removed prior to T and non-T-cell separation (experiments 5, 6, Table 4) suggest that the presence of adherent cells in autologous MLC is not crucial for the generation of suppressor activity. DISCUSSION Proliferation of T cells upon stimulation with autologous non-T cells in vitro is a well-established phenomenon (1, 2). AMLR is believed to be driven by Ia-like determinants on stimulating non-T cells (27, 28). Although the biologic significance of this phenomenon is still uncertain, the ability of T cells to proliferate in AMLC has been thought to represent a measure of T-cell regulatory ability (2, 3, 5). It has been demonstrated that the predominant functional consequence of activation of T cells in AMLC is the generation of suppressor T cells. These cells are able to suppress both proliferative response and immunoglobulin synthesis in PWM-driven cultures of autologous B cells (12) and inhibit proliferation and generation of cytotoxic cells upon stimulation of autologous T cells by allogeneic stimulators (10, 11). It was also shown that Con A-induced suppressor cells are drawn from the population of T lymphocytes activated in AMLC (4). Our studies confirm previous observations (10, 11) that AMLC-activated cells are able to suppress both proliferation and cytotoxic cell generation by autologous lymphocytes stimulated in MLC. Suppressor cells generated in this system are Mitomycin C sensitive, indicating that proliferation of suppressor cells is probably necessary in this system. This finding is in contrast to some studies demonstrating resistance of in vitro-induced suppressor cells to Mitomycin C and irradiation, using different SC induction techniques (29, 30) but consistent with the results of Smith and Knowlton (11) and Innes et al. (4). The action of suppressor cells induced in AMLC appears to be nonspecific in the sense that the response to any allogeneic stimulator is suppressed, irrespective of HLA-A,B,C and DR profile (Table 1). This finding is in agreement with several previous observations demonstrating the lack of specificity after in vitro induction of suppressor cells in humans (29, 30). Sasportes et al. (31) were able to demonstrate DR-specific suppression but only after repeated in vitro sensitization by allogeneic cells. AMLC-induced suppressor cells are able to inhibit not only the response of autologous lymphocytes, but are equally suppressive to allogeneic responders against any stimulator (Table 2). T cells from autologous MLC do not lose their capacity to proliferate when used as responders in a second culture (Table 3). It has been shown by Kozak et al. (32) that T cells from autologous MLC are composed of cells that were activated in AMLC (autoreactive) and cells that remained nonactivated during AMLC (alloreactive). In their experiments, it was possible to fractionate these cell populations using Percoll density gradient centrifugation and analyze them separately. Autoreactive cells, when used as responders in a second MLC, showed secondary proliferative kinetics against both autologous and allogeneic stimuli, while alloreactive cells responded only to allogeneic stimulators in a primary fashion. Since separation of AMLC-stimulated T cells was not attempted in our experiments, proliferation of these cells when used as responders in a second culture presumably reflects the mixed activities of secondary
AMLC-INDUCED
SUPPRESSOR
CELLS
131
response by autoreactive cells and primary response by alloreactive cells. Secondary response was not explored in our experiments, since our main purpose was to analyze the effect of AMLC-stimulated cells on allogeneic reactivity. In a 6-day culture, AMLCprecultured cells showed a primary proliferative response against allogeneic stimulators, Xm, in all experiments, giving SI comparable to that obtained in A/Xm group (Table 3). This response most likely reflects the activity of alloreactive cells and is not directly related to activity of SC. However, persistence of alloreactivity in AMLCstimulated cell population complicates the assessment of SC effect in MLC. It should be kept in mind that the response to Xm in the experimental group [A + AMLCprecultured cells] is, in fact, composed of two components: proliferation of A cells and proliferation of AMLC-precultured cells against the allogeneic stimulator Xm. If the primary response by precultured (alloreactive) cells is sufficiently high, the total cpm response in this group may be even higher than in A/Xm group (for example, line 2, Table 3) and thus, no suppressive effect is apparent judging by cpm. However, it should be kept in mind that over 50% of proliferative response in this group is, in fact, contributed by AMLC-precultured cells (lines 4 and 8, Table 3). Thus, proliferation of A cells against Xm in this group is considerably reduced as compared to A/Xm group (lines 1 and 5). In order to evaluate the effect of suppressor cells on second MLC, combination [AMLC-precultured cells/Xm] should always be included as an important control and, when estimating the percentage of suppression, the values should be adjusted for the amount of proliferation contributed by alloreactive cells. Another factor complicating the interpretation of MLC results is the continuous proliferation of AMLCactivated cells against autologous A cells. Under the conditions of our experiments (6-day incubation in a second culture), a secondary response against autologous stimulators was not observed, and the amount of proliferation in [AMLC-stimulated cells/ Am] varied from experiment to experiment. Where it was expressed, it resulted in the increase of background cpm and thus, affected the SI. Therefore, to evaluate the net amount of suppression in MLC, the response in experimental group should be adjusted for both proliferation of AMLC-stimulated cells against allogeneic stimulator and for increased background in the autologous control. The fact that AMLC-activated cells retain their ability to proliferate in response to allogeneic stimulators is in agreement with the conclusion of Sakane and Green (10) that autologous and allogeneic MLR are for the most part independently mediated by two different T-cell populations in human peripheral blood, although they may partially overlap. More recent studies, using monoclonal antibodies against specific T-cell subsets, show that autoactivated T cells are drawn predominantly from the OKT4-reactive subset, whereas alloactivated T cells are drawn from both OKT4 (helper/inducer) and OKTS (cytotoxic/suppressor) positive subsets (32). Similar conclusions were also reached by Engleman ez al. (33) using different sets of monoclonal antibodies. Suppression induced in allogeneic MLC is often explained as a result of cytotoxic elimination of the stimulating cells by effector cells generated in the inducing culture (18-20, 34). However, most investigators failed to demonstrate cytotoxic cells in AMLC (16, 17). In this study, we were likewise unable to demonstrate the cytotoxic effect of AMLC-activated cells against autologous PHA blasts. Furthermore, no cytotoxic activity against allogeneic PHA blasts could be demonstrated (Fig. 3). One
132
PAZDERKA
ET AL.
may speculate that the cytotoxic effect of these cells, while not demonstrable in a short-term (4 hr) assay, may be accumulating during incubation in. a second culture. To test this possibility, AMLC-activated cells were restimulated by autologous or allogeneic stimulators and tested in CML 3 and 6 days later. No significant killing was observed (data not shown). Thus, inhibitory action of AMLC-activated cells is not a consequence of simple cytotoxic elimination of either responding or stimulating cell populations in the second culture, but truly represents a regulatory suppressive mechanism. Participation by various cell populations in the induction and mediation of sup pressor effect, as observed in various in vitro models, is still a matter of controversy. In AMLC, in particular, the role of adherent cells is uncertain. On one hand, macrophages have been demonstrated to serve as potent stimulators of AMLR in humans (35-37). On the other hand, however, elimination of macrophages from stimulator cells results in higher proliferative responses (2, 38). This phenomenon has been interpreted by Smolen et al. (39) as evidence of suppressive effect of macrophages on AMLC. However, involvement of macrophages in the generation of cells which inhibit responses to allogeneic stimulators has not been studied in depth. We have approached this question by comparing suppression generated in T/non-Tm-stimulated cultures from which adherent cells were removed before further separation to that when adherent cells were present, No appreciable difference was found between these two types of cultures (Table 4, experiments 5 and 6). Thus, adherent cells in AMLC do not appear to be crucial for generation of suppressor cells as assayed in alloantigen-stimulated system. However, our technique for macrophage depletion may not have been sufficient to achieve complete removal of macrophages, and a small number of these cells may be enough for generation of suppressor effect as demonstrated in a Con A model by Innes et al. (4). Studies using specific monoclonal antibodies to macrophages are necessary in order to answer this question conclusively. The mediators of suppression induced in AMLC appear to be T lymphocytes (4, 10, 11). In our experiments, in agreement with observations of Smith and Knowlton (1 l), maximal suppressor activity was generated only in cultures where T cells were stimulated with autologous non-T cells (Table 4). Suppressor cells were not, as a rule, recovered from cultures of purified T or non-T cells cultured alone. Low suppressive activity occasionally observed in the cultures of T cells alone was most probably due to contaminating non-T cells in the T-cell preparation. Thus, activation of T cells by autologous non-T cells is a prerequisite to the generation of suppressor cells in this system. Our findings support the notion that the stimulation noted in AMLC reflects an important immunoregulatory mechanism. The absence of this reaction in patients with CLL (40) and SLE (13,4 1) is in agreement with the hypothesis that regulatory failure of the immune system could be important in the pathogenesis of certain lymphoproliferative and autoimmune diseases. The application of this system to the study of immune status of such patients is warranted. ACKNOWLEDGMENTS This work was supported by MRC of Canada. We wish to express our thanks to Ms. J. Clark for efficient secretarial help.
REFERENCES I. Opelz, G., Kiuchi, M., Takasugi, M., and Terasaki, P. I., J. Exp. Med. 142, 1327, 1976. 2. Kuntz, M. M., Innes, J. B., and Weksler, M. E., J. Exp. Med. 143, 1042, 1976.
AMLC-INDUCED 3. 4. 5. 6. 7. 8. 9. 10. Il. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22,
23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41.
SUPPRESSOR
CELLS
133
Weksler, M. E., and Kozak, R., J. Exp. Med. 146, 1833, 1977. Irmes, J. B., Kuntz, M. M., Kim, Y. T., and Weksler, M. E., J. Clin. Invest. 64, 1608, 1979. Smolen, J. S., Luger, T. A., Chused, T. M., and Steinberg, A. D., J. Clin. Invest. 68, 1601, 1981. Miller, R. A., and Kaplan, H. S., J. Immunol. 121, 2165, 1978. Tomonari, K., J. Immunol. 124, 1111, 1980. Chiorazzi, N., Fu, S. M., and Kunkel, H. G., J. Exp. Med. 149, 1543, 1979. Hausmann, P. B., and Stobo, J. D., J. Exp. Med. 149, 1537, 1979. Sakane, T., and Green, I., J. Immunol. 123, 584, 1979. Smith, J. B., and Knowlton, R. P., J. Immunol. 123, 419, 1979. James, S. P., Yenokida, G. G., Graeff, A. S., Elson, C. O., and Strobcr, W., J. Immunoi. 127, 2605, 1981. Sakane, T., Steinberg, A. D., and Green, I., Proc. Nat. Acad. Sci. USA 75, 3464, 1978. Miyasaka, N., Sauvezie, B., Pierce, D. A., Daniels, T. E., and Talal, N., J. Clin. Invest 66, 928, 1980. James, S. P., Elson, C. D., Waggoner, J. G., Jones, E. A., and Strober, W., J. C/in. Invest 66, 1305, 1980. Vande Stouwe, R. A., Kunkel, H. G., Halper, J. P., and Weksler, M. E., J. Exp. Med. 146, 1809, 1977. Katz, P., Haynes, B. F., and Fauci, A. S., J. Immunol. 121, 1998, 1978. Hodes, R. J., Nadler, L. M., and Hathcock, K. S., Immunology 119, 961, 1977. Fitch, F. W., Engers, H. D., Cerottini, J. C., and Brunner, K. T., J. Immunol. 116, 716. 1976. Sinclair, N. R., Lees, R. K., Wheeler, M. E., Vichos, E. E., and Fung, F. Y., Cell. Immunol. 27, 153, 1976. Mittal, K. K., Mickey, M. R., Singal, D. P., and Terasaki, P. I., Transplantation 6, 913, 1968. Kissmeyer-Nielsen, F., and Dick, H., In “Histocompatibility Techniques” (H. Dick and F. KissmeyerNielsen, Eds.), pp. 9-38, Elsevier/North-Holland Biomedical Press, Amsterdam/New York/Oxford, 1979. Boyum, A., Stand. J. Clin. Lab. Invest. Suppl. 97, 17, 1968. Pellegrino, M. A., Ferrone, S., Dierich, M. P., and Reisfeld, R. A., Clin. Immunol. Immunopathol. 3, 324, 1975. Lowry, R. P., Person, A. E., Goguen, J. E., Carpenter, C. B., and Garovoy, M. R., Transplant. Proc. 10, 833, 1978. Tomonari, K., Wakisaka, A., and Aizawa, M., J. Immunol. 125, 1596, 1980. Bergholtz, B., Albrechtsen, D., and Thorsby, E., Tissue Antigens 10, 27, 1977. Gottlieb, A. B., Fu, S. M., Yu, D. T., Wang, C. Y., Halpcr, J. P., and Kunkel, H. G., J. Immunol. 123, 1497, 1979. Hirschberg, H., and Thorsby, E., Stand. J. Immunol. 6, 809, 1977. Shou, L., Schwartz, A., and Good, R. A., J. Exp. Med. 143, 1100, 1976. Sasportes, M., Fradelizi, D., and Dausset, J., Nature (London) 276, 502, 1978. Kozak, R. W., Moody, C. E., Staiano-Coico, L., and Weksler, M. E., J. Immunol. 128, 1723, 1982. Engleman, E. G., Benike, C. J., Grumet, F. C., and Evans, R. L., J. Immunol. 127, 2 124, 198 I. Sheehy, M. J., Mawas, C., and Charmot, D., J. Immunol. 122, 2198, 1979. Beale, M. G., MacDermott, R. P., Stacey, M. C., Nash, J. S., Hahn, B. H., Seiden, M. V., BurkholderJacobs, S. L., and Pletscher-Lowenstein, L. S., J. Immunol. 124, 227, 1980. MacDermott, R. P., and Stacey, M. C., J. Immunol. 126, 729, 1981. Dock, N. L., and Davey, F. R., Clin. Immunol. Immunopathol. 17, 123, 1980. Katz, P., and Fauci, A. S., J. Immunol. 123, 2270, 1979. Smolen, J. S., Sharrow, S. O., Reeves, J. P., Boegel, W. A., and Steinberg, A. D., J. Immunol. 127, 1987, 1981. Smith, J. B., Knowlton, R. P., and Koons, L. S., J. Nat. Cancer Inst. 58, 579, 1977. Kuntz, M. M., Innes, J. B., and Weksler, M. E., Clin. Res. 24, 448A, 1976.