SUPPRESSION OF LYMPHOKINE PRODUCTION IN ANTI-MINOR HISTOCOMPATIBILITY ANTIGEN RESPONSES L. Yin and B.M. Chain AIlogeneic immunizations between mice strains that differ at multiple loci coding for minor histocompatibility antigens (mha) result in a variety of immune responses, including the induction of cytotoxic T cells, helper T cells, and delayed hypersensitivity reactions. However, high-dose intravenous immunization induces a population of cells that can suppress these responses. In this paper, we show that the transfer of this suppressor population in vivo is accompanied by a reduced ability of immune cells to produce the two lymphokines interleukin-2 (K-2) and interleukin3 (IL-3). In a different assay of suppressor function, suppressor cells were co-cultured with responder cells in vitro. Under these conditions, the presence of suppressor cells resulted in a lowered net production of IL-2, but not IL-3. Possible mechanisms for this phenomenon are discussed. Copyright o 1991 by W.B. Saunders Company
The mechanisms of T-cell-mediated immunosuppression remain one of the most problematic areas of cellular immunology. Thus, although the number of systems in which such suppression has been demonstrated continues to increase (reviewed in reference 1) the existence of a specific subpopulation of T cells with suppressor activity has not been convincingly demonstrated, and there has been little progress on molecular characterization of soluble suppressor “factors.“2’3 One area which has been relatively little explored has been to examine the changes in lymphokine biosynthesis that must underlie the overall changes in immune system function during suppression. A study of these changes may offer an approach to a molecular interpretation of suppressor phenomenology. In the present study we have examined the regulation of the production of two lymphokines, interleukin-2 (IL-2) and interleukin-3 (IL-3), in a system whose function has been extensively characterized in this laboratory. This model studies the allogeneic reaction that can be induced between two major histocompatibility complex (MHC)-compatible mice strains that differ by multiple minor histocompatibility antigens (mha). This response, which has been assayed as T helper function4 and T cytotoxic function5.6 and by T
ICRF Tumour Immunology Unit, Department of Biology, Medawar Building, University College London, Gower Street, London WClE 6BT, UK. Copyright 0 1991 by W.B. Saunders Company 1043-4666/91/0301-0007$05.00/O KEY WORDS: lymphokineianti-mha/suppression/IL-2/IL-3 CYTOKINE,
Vol. 3, No. 1 (January), 1991: pp 5-11
proliferatioq7 can be suppressed by spleen T cells from mice injected intravenously with high doses of alloantigen.5 Previous work has shown that the suppressor activity resides in the Thy 1’ CD8’ CD4- population, and is not IgH restricted.6 Suppression of the mha response can be demonstrated both in vivo and in vitro. In the absence of any previously published data on lymphokine production in this system, we have first established the parameters of IL-2 and IL-3 responses and the phenotype of the responding cell following both primary and secondary stimulation in vivo. We then demonstrate that protocols previously shown to induce suppression of the mha response both in vitro and in vivo result in profound suppression of net lymphokine production by responder cells. The pattern of suppression seen following in vivo suppression is rather different from that seen in vitro. Finally, we present some experiments that begin to address the mechanisms that may underlie the suppressor phenomena. RESULTS Lymphokine
Synthesis in Anti-mha
Responses
Cells from CBA mice immunized to BlO.BR minor alloantigen respond in vitro to BlO.BR spleen cells by the production of IL-2 and IL-3 (Fig. 1). This response occurs using either cells that had received a single primary immunization in vivo or cells that had been immunized 2 to 6 weeks previously and, therefore, were in the secondary phase of an immune response. The response is antigen specific and does not occur in 5
CYTOKINE,
6 1 Yin and Chain
A
cules, suggesting that the responding population contained cells restricted to both molecules. These results establish that the major lymphokineproducing cell under the conditions of these assayswas a CD4’, class II MHC-restricted cell. Similar results were obtained using lymph node cells taken 4 days after priming (results not shown).
IL-2 production
0 1 6 or
Vol. 3, No. 1 (January 1991: 5-11)
Suppressionof Lymphokine Responses 25 Number
B
5 of responder
10 cells
The lymphokine responses can be demonstrated either by in vivo transfer of suppressor cells to recipients at the time of immunization (Fig. 3A) or by co-culture of responder cells with suppressor cells (Fig. 3B). However, while IL-2 production is downregulated
20 (x 104)
IL-3 production 0 50 1
A
IL-2 productlo”
0 2OT
0 16 t x 0 10+
oooc
/ 25 Number
5 of responder
20
10 cells
(x
vI 5 D
012
; x 0
0.08
40
T \\
004.-
104) 0007
Figure 1. The production of lymphokines by lymph node and spleen cells from immune mice.
unimmunized animals or animals immunized with an irrelevant antigen (data not shown). In contrast, similar levels of spontaneous (“background”) production of lymphokines (particularly IL-3), were observed using either immune or naive mice. These background levels, which probably represent a mitogenic response to substances in the fetal calf serum, have not been subtracted from any data shown, but are given in the appropriate legends. Both IL-2 (Fig. 2A) and IL-3 production (not shown) by secondary responder cells was markedly inhibited by depletion of CD4’ T cells prior to culture. CD8’ depletion had a much smaller inhibitory effect. In agreement with these depletion experiments, the lymphokine production could be blocked by specific anti-class II MHC monoclonal antibodies added to the cell cultures (Fig. 2B). The response was partially inhibited by antibodies to both I-Ak and I-Ek mole-
Ii 5
c7
Varying numbers of lymph node cells, taken 4 days post priming (primary), or spleen cells taken 4 weeks post priming (secondary), were incubated with or without 4 X 10’ irradiated allogeneic (BlOBr) spleen cells (as a source of antigen) and assayed for (A) IL-2 and (B) IL-3 production as described in Materials and Methods. -A-, primary response with antigen; -A-, primary response without antigen; -O-, secondary response with antigen; -0-, secondary response without antigen.
n
6
complement
0 16-
Number only
10
20
L10
of responder Cells (X 1 04) I - CD4 depleted W
CDS depleted
IL-2 production
0 12.x P 8 0 2 0”
0 08
004..
ooo-
Figure 2. Phenotype and MHC restriction of responder cells. (A) The phenotype of responder cells. Spleen cells from mice immunized for a secondary response were depleted of CD4’ or CD8+ cells by treatment with antibody and complement. Equal numbers of viable cells from depleted or control (complement only) groups were co-cultured with antigen and assayed for IL-2 production. (B) MHC restriction of the allogeneic secondary lymphokine response. Spleen cells (4 x 10s) from immunized mice were cocultured with antigen in the presence of varying concentrations of anti-Class II MHC antibodies as shown. IL-2 response in the absence of antigen gave an optical density of 0.08 U. The antibodies tested were: 0, anti-I-A”, HB3 (from ATCC); n , anti-I-A’, H40481.3 (a gift from Dr. M. Pierres, Marseille). q , anti-I-Ek, HB32 (from ATCC). All antibodies were IgG fractions, purified from ascites on protein A.
Suppression of lymphokines in anti-mha responses / 7
IL-2 production
IL-3 production
. /
25
5
10
20
B
0.45
0.35
O
0 30 ‘0
0 0.08
20
40
.T.
o-
o\ \
I
0
O--O i
0 20
‘\
0’
0 0.15
0 00’ 2.5
5
10
20
Number
3.
10
0 25
0.00-e
Figure
5 (x104)
IL-3 production
040T
012 x .ez
cells
I-
.‘.\
z 0 0 z
2.5
of responder
IL-2 production
0.16
/
40
Number
0
050
The suppression
of lymphokine
responses
40 of
suppressor
25
5
cells
( ~10~)
10
20
40
in vivo and in vitro.
(A) Primary response, in vivo suppression. Mice received 5 x 10’ spleen cells from donors that had received a high dose of antigen intravenously four days previously or from naive donors. The recipients were then immediately immunized in the foot pads for generation of a primary immune response (see Materials and Methods). Variable number of lymph node cells from these two groups of animals (“suppressed” and “normals”) were co-cultured with antigen, and assayed for IL-2 and IL-3 production. -O-, “suppressed” mice; -O-, “normal” mice. (B) Secondary response, in vitro suppression. Variable numbers of spleen cells from mice receiving high doses of intravenous antigen four days previously (“suppressed,” see Materials and Methods) or from naive mice were co-cultured with 4 x lo4 immune spleen cells (secondary response) and antigen and assayed for IL-2 and IL-3 production. Lymphokine production in the absence of antigen (“background”) gave optical densities of 0.02 for IL-2 and 0.19 for IL-3. -0-, co-culture with cells from suppressed mice; -O-, co-culture with cells from naive mice.
in both assays of suppressor activity, IL-3 activity is downregulated only by in vivo protocols. We wished to ascertain whether the lack of IL-3 suppression observed in the in vitro mixing experiments was indeed due to the use of an in vitro assay or to the fact that suppression was being assayed during the secondary, rather than the primary, phase of the immune response. In vitro primaq responses against these antigens cannot easily be obtained. The secondary response was therefore suppressed in vivo as described above, by transferring suppressor cells immediately prior to boosting previously primed mice. Both IL-2 and IL-3 responses could be suppressed using this immunization protocol (not shown). The lack of IL-3
suppression observed during in vitro co-culture experiments was therefore unlikely to be due only to the differentiation stage of the responder T cell (i.e., 2 or more weeks post priming in our standard in vitro assay of suppression, as opposed to being the target of suppression actually at the time of priming, as is the case in our primary in vivo assay of suppression). As shown in Fig. 4, suppressor activity depended on the dose of immunizing antigen. The suppression was rather easily induced (sometimes with as little as 10’ BlO.BR spleen cells), and was always found after injecting 5 x 10’ cells or more. The induction of suppression in this system therefore seems to depend more on the route of injection than on antigen dose,
CYTOKINE,
8 I Yin and Chain
suppressor activity as measured in both in vivo and in vitro assays (data not shown). However, differences were observed after depleting CD4’ or CD8’ cells in the two types of assay. Depletion of suppressor cells with either anti-CD4 antibody, or anti-CD8 antibody prior to in vivo transfer into responder mice partially abrogated the suppression (Fig. 5A), suggesting that both cell types play a part in the generation of a full suppressive response in vivo. Depletion of CD8’ cells from lymph node populations of mice that had been previously suppressed in vivo did not, however, reverse the suppression (not shown). The suppressive effect of the transferred CD8’ population is therefore no longer reversible at the time the cells are tested for lymphokine production. In contrast, all in vitro suppressor activity could be removed by treatment with anti-CD8 antibody and complement, while CD4’ depletion had no effect (Fig. 5B).
IL-3 production
\
O--O
0.00 4
/ 0
105
107
106
Number
5x107
of unmunlsing
108
Vol. 3, No. 1 (January 1991: 5-11)
5x108
cells
Figure 4. Antigen dose-response for the induction of responder and suppressor activity. -0-, responder: Mice received varying doses of antigen (BlOBr spleen cells) intraperitoneally 2 weeks prior to assay. Spleen cells (4 x l@) from immunized mice were incubated with antigen, and assayed for IL-3 production. -O-, suppressor: Mice received varying doses of antigen intravenously. Four days later, 5 x 10’ spleen cells were transferred intravenously to recipients immediately prior to foot-pad immunization. After a further 4 days, lymph node cells (4 x 104) from recipients were co-cultured with antigen, and IL-3 production was measured. As shown, increased intravenous doses of antigen resulted in greater suppression of the primary response.
The Role of IL-2 Competition
in Suppression
One possible mechanism by which suppressor cells may function is by lymphokine competition.8P9 This mechanism is particularly significant since, in these experiments, we measure only net lymphokine production, which is a balance of uptake and secretion. While the importance of this mechanism cannot be readily assessed in vivo, it can be tested in vitro. As shown in Fig. 6, the addition of anti-IL-2 receptor antibody (TIB 222) enhanced IL-2 levels in the in vitro coculture suppression system and, in fact, completely abrogated the differences between cultures with control and suppressor cells. The addition of the same antibody to cells suppressed in vivo also raised IL-2
since certain doses of antigen administered intravenously or intraperitoneally can induce suppressor cells or responder cells, respectively. The Phenotype of Suppressor Cells Treatment of suppressor cells with anti-Thy-l antibody and complement completely removed all
B
A
IL-2 production
0.08
0 16
--
I
x .e : 0.06-m : 0 .i-’ 0.04.. 2
0.08
0
0 04
0.00 6
/
0.12 0 00 L 2.5
Number
of responder
cells
(x 104)
Number
5
10
of suppressor
20
40
cells
(~10~)
Figure 5. Phenotype of suppressor cell population. (A) In vivo suppression. (B) In vitro suppression. The activity of suppressor cells was measured as described in the legend to Fig. 3A and 3B, but suppressor cells were depleted of CD4’ and CD8’ cells by treatment with antibody and complement. -0-, normal spleen cells; -O-, suppressor population; -A-, CD4+-depleted suppressor population; -A-, CD8+-depleted suppressor population.
Suppression of lymphokines in anti-mha responses / 9 IL-2 production 0 16
,x
0 12
T iI
A/ n--n A------./
0 OOl’--25 Number
Figure
6.
The role of IL-2
5
10
of suppressor
adsorbtion
cells
in in vitro
20 (xl 04)
I 40
suppression.
In vitro suppression was assayed as described in the legend to Fig. 3B. Anti-IL-2 receptor antibody (TIB 222, from ATCC), in the form hybridoma supernatant, was added to some wells. -0-, co-culture with normal spleen cells; -A-, co-culture with normal spleen cells plus TIB 222; -O-, co-culture with suppressor spleen cells; -A-, co-culture with suppressor spleen cells plus TIB 222.
levels but did not abrogate suppression (not shown). This result argues that the effects of transferring suppressor cells in vivo are no longer reversible by the time lymphokine production is stimulated in vitro.
DISCUSSION This study represents a re-examination of a wellstudied T-cell suppression system, using as readout lymphokine production by responder cells. This approach has broadly confirmed previous in vivo studies, insofar as T cells immunized according to “suppressor” protocols downregulate lymphokine biosynthesis in responder T cells. Furthermore, they have conclusively established that the target of suppression is primarily the CD4’ “ helper” T cell, a fact which has been questioned in some recent discussions of suppression,1o and which is not consistent with models involving lymphokine antagonism at the effector stage of immune responses. l1 The lymphokines tested in this study are both likely to be of importance in the anti-mha functional assays previously used. Thus IL-2 plays a major role in the induction and amplification of both cytotoxic and delayed-type hypersensitivity reactions, while IL-3 is involved in the recruitment, proliferation, and differentiation of the nonhemopoeitic lineage cells involved in inflammatory responses. It is therefore likely that the greatly reduced ability of suppressed T cells to produce these lymphokines is critical in explaining their other depressed functional capabilities. A comparison of the in vitro and in vivo models of suppression shows a number of interesting differences. In vitro suppression results in decrease only of IL-3 levels, is a function predominantly of CD8+ T cells, and
can be reversed by blocking IL-2 consumption in the cultures. The simplest interpretation of these results is that the suppressor population contains CD8+ cells, which show a net IL-2 consumption. These cells thus compete for IL-2 and lower the free levels in the culture supernatants. We do not know why differences in immunization routes (intravenous versus intraperitoneal) should activate populations of cells that show, respectively, a net IL-2 consumption or a net IL-2 production. However, data from other systems indicate that it is possible to induce T-cell tolerance at a clonal level, and that such tolerant cells can selectively lose the ability to make IL-2 but not the ability to respond to (and hence utilize) this lymphokine.” Further studies are required to demonstrate whether such partially tolerant cells are present within the suppressor population used in these experiments. In contrast to this in vitro suppression, in vivo suppression requires the action of both CD4’ and CDS’ T cells; this interaction between these two cell types has been observed frequently in other suppressor cell systems. In vivo suppression also leads to impaired IL-3 production. In addition, in vivo suppression can no longer be reversed by removing suppressor cells or by blocking IL-2 consumption in culture. If in vivo suppression is to be interpreted in terms of the in vitro model, that is to say in terms of IL-2 competition, one must hypothesize that a lack of IL-2 at a critical stage in T-cell activation (taking place in vivo), results in a more basic failure in T-cell activation, which also affects IL-3 production and is not readily reversible. Alternatively, in vivo suppression may result from completely different mechanisms, for example from competition between suppressor cells and responder cells for limited amounts of antigen. Our future studies will therefore concentrate on trying to develop more realistic in vitro models of the immune microenvironments found in vivo and on assaying cellular interactions and lymphokine responses during suppression in vivo.
MATERIALS AND METHODS Mice CBA/H and BlO.BR mice were supplied by the Imperial Cancer Research Fund breeding colony, and were used when
between 6 and 12 weeks old. Donors and recipients were sex-matched within each transfer experiment. Anti-Minor
Responses-Lymphokine
Production
The generation of proliferative, cytotoxic, helper, and suppressor responses in CBA mice to the minor alloantigens of the BlO.BR strain (both strains have the H-2’ haplotype) has been described in detail elsewhere.“s6 The immunization protocols used here are based on these previous studies. Two
10 / yin and Chain types of immunization were used: (i) Primary responses. CBA mice were injected in the two hind footpads with 50 l.rl of BlO.BR spleen cell suspension containing 2.5 x lo6 cells that had been irradiated (2000 R) under a “Co source. Four days later, popliteal and inguinal lymph nodes (LN) were removed from 2 to 5 animals and pooled. (ii) Secondary responses. CBA mice were injected intraperitoneally with 10’ irradiated BlO.BR spleen cells. Two to six weeks later, the spleens were removed and the cells were cultured as above. In some experiments mice primed in this way received a further boost of 2 x lo6 spleen cells in the footpads four days prior to sacrifice. In this case lymph node cells (as for primary responses) instead of spleen cells were harvested and cultured.
Suppressionof Anti-mha Responses A population of cells with suppressor activity for CBA anti-BlO.BR responses was generated as described previously.(’Viable nucleated BlO.BR mice spleen cells (lo8 cells) were injected into the lateral tail vein of CBA mice. Four days later spleen cells from these mice were used as a source of suppressor cells for the response to B10 minor mha’s. These cells are referred to simply as “suppressor cells” in this paper. Cells from uninjected mice were used as controls.
Assaysof SuppressorActivity (i) Primary response. Suppressor cells (5 X 10’) or normal spleen cells were transferred intravenously into responder mice immediately prior to footpad immunization as described above. Lymphokine production of lymph node cells from mice receiving suppressor cells or normal spleen cellswas compared. (ii) Secondary response. In vitro: Spleen cells (4 x 10’) from mice immunized as in Anti-Minor Responses-Lymphokine Production, section ii (above) were co-cultured with various numbers of suppressor cells or normal cells, and lymphokine production in the two groups was compared. In vivo: suppressor cells or normal cells (5 x 10’) were transferred intravenously into mice immediately before boosting with antigen as described in Anti-Minor Responses-Lymphokine Production, section ii (above). In all experiments, single cell suspensions were prepared from spleen or lymph nodes and incubated at various cell densities in 200 ul of Iscove’smodification of Dulbecco’s medium supplemented with 5% fetal calf serum (Gibco), 2 x 10m5M 2-mercapthoethanol, 100 U/ml of penicillin, 100 p&ml of streptomycin. Antigen, in the form of 4 x 10’ irradiated (2,000 R) BlO.BR spleen cells, was added to some wells. Triplicate cultures were set up for each experimental point. Supernatants were collected after 48 h of incubation (in 5% CO,, humidified incubators at 37°C) and stored at -20°C for assayof IL-2 and/or IL-3. Preliminary experiments indicated that lymphokine production was optimal at 48 h and with this dose of antigen.
Anti-Thy 1, AntMD4, and Anti-CD8 Antibody Depletion of SuppressorCellsor ResponderCells Spleen cells were incubated at 2 x lO’/ml with anti-Thy1 (YBM29-2-1) anti-CD4 (RLi72.4), or anti-CD8 (3.169) anti-
CYTOKINE, Vol. 3, No. 1 (January1991:5-11) body for 20 min in ice. The cells were pelleted, resuspended in l/10 dilution of rabbit complement (Buxted), and incubated at 37°C for 45 min. Controls were treated with complement alone. The cells were washed twice and then used for further investigation. Preliminary studies showed that this treatment resulted in depletion of greater than 90% of the specific population targeted.
Assayof IL-2 and IL-3 Activities IL-2 and IL-3 activities were assayed by testing the ability of the culture supernatants collected from different experimental groups to support the growth of the IL-2dependent CTLL line or the IL-3-dependent FDCP-2 line. Cell viability of the indicator cell lines following culture in the test supernatants was assessedby reduction of the dye MIT by viable cells. The assays have been described in detail elsewhere.13Results have not been converted into lymphokine units, but are expressed simply as the mean optical density values (absorbance at 570 nm - absorbance at 630 nm) obtained from triplicate cultures of the stained indicator cell-line cultures. Results should not, therefore, be interpreted in a strictly quantitative way. Standard errors are not shown but were generally in the region of 10%. All figures show data from a single experiment, but each set of experiments was repeated a minimum of three times. Optical densities are approximately proportional to lymphokine titer up to values of 0.3 for IL-2 and 0.4 for IL-3. Five units of recombinant IL-2 and 10 U of IL-3 gave an optical density of approximately 0.1 optical density units in these assays.The CTLL line responds both to IL-2 and IL-4, but is very insensitive to IL-4. Furthermore, the IL-2 responses measured in this study could not be inhibited by a blocking antibody specific for IL-4. The FDCP-2 line did not respond to IL-2 or IL-4. It did, however, respond to both IL-3 and granulocyte-macrophage colony-stimulating factor (GMCSF). Since we did not have a specific antibody capable of blocking either of these activities, this study does not distinguish between the production of these two lymphokines. Acknowledgments L. Yin was supported by a Royal Society visiting Fellowship (China programme). B.M. Chain is supported by a Royal Society University Research Fellowship.
REFERENCES genes.
1. Ohveira DBG, Mitchison Clin Exp Immunol75:167-177.
NA
(1989)
Immune
suppression
2. Mitchison NA (1989) Is genes in the mouse. In Melchers et al. (eds) Progress in Immunology, Vol. VII, pp 845-857. 3. Moller G (1988) Immunol27:247-250.
Do
4. Crispe IN, Gascoigne murine T helper clone which vitro and in vivo. Immunology 5. Gascoigne
NRJ,
suppressor
T cells
exist?
F,
Stand
J
NRJ, Owens T (1984) An allospecihc can help both T and B cell responses in 5255-65.
Crispe
IN (1984)
Suppression
of the cyto-
Suppression of lymphokines in anti-mha responses / 11 toxic T cell response to minor alloantigens in vivo. Linked recognition by suppressor T cells. Eur J Immunol14:21O-215. 6. Gascoigne NRJ (1984) Suppression of the cytotoxic T cell response to minor alloantigens in vivo. II. Fine specificity of suppressor T cells and lack of restriction by immunoglobulin heavy chain-linked gene products. Em J Immunol 14:677-680. 7. Nanda NK (1989) Preferential restriction of minor alloantigen-specific suppressor T cells to I-E rather than I-A molecules. Immunology 68:163-168. 8. Gunter J, Haas W, von Boehmer H (1982) Suppression of T cell responses through competition for T cell growth factor (interleukin 2). Eur J Immunol12:247-249. 9. Palacios R, Moller G (1981) T cell growth factor abrogates
concanavalin A-induced suppressor cell function. J Exp Med 153: 1360-1365. 10. Lanzavecchia A (1989) Is suppression a function of class II restricted cytotoxic T cells. Immunology Today 10:157-159. 11. Rabin E, Mond J, Ohara J, Paul W (1986) Interferon-y inhibits the action of B cell stimulatory factor (BSF)-1 on resting B cells. J Immunol 137:1573-1576. 12. Essery G, Feldmann M, Lamb J (1988) Interleukin 2 can prevent and reverse antigen induced unresponsiveness in cloned human T cell clones. Immunology 64:413-417. 13. Marcinkiewicz J, Chain BM (1989) Antigen specific inhibition of IL-2 and IL-3 production in contact sensitivity to TNP. Immunology 68:185-189.
The mechanisms of T-cell-mediated immunosuppression remain one of the most problematic areas of cellular immunology. Thus, although the number of systems in which such suppression has been demonstrated continues to increase (reviewed in reference 1) the existence of a specific subpopulation of T cells with suppressor activity has not been convincingly demonstrated, and there has been little progress on molecular characterization of soluble suppressor “factors.“2’3 One area which has been relatively little explored has been to examine the changes in lymphokine biosynthesis that must underlie the overall changes in immune system function during suppression. A study of these changes may offer an approach to a molecular interpretation of suppressor phenomenology. In the present study we have examined the regulation of the production of two lymphokines, interleukin-2 (IL-2) and interleukin-3 (IL-3), in a system whose function has been extensively characterized in this laboratory. This model studies the allogeneic reaction that can be induced between two major histocompatibility complex (MHC)-compatible mice strains that differ by multiple minor histocompatibility antigens (mha). This response, which has been assayed as T helper function4 and T cytotoxic function5.6 and by T
ICRF Tumour Immunology Unit, Department of Biology, Medawar Building, University College London, Gower Street, London WClE 6BT, UK. Copyright 0 1991 by W.B. Saunders Company 1043-4666/91/0301-0007$05.00/O KEY WORDS: lymphokineianti-mha/suppression/IL-2/IL-3 CYTOKINE,
Vol. 3, No. 1 (January), 1991: pp 5-11
proliferatioq7 can be suppressed by spleen T cells from mice injected intravenously with high doses of alloantigen.5 Previous work has shown that the suppressor activity resides in the Thy 1’ CD8’ CD4- population, and is not IgH restricted.6 Suppression of the mha response can be demonstrated both in vivo and in vitro. In the absence of any previously published data on lymphokine production in this system, we have first established the parameters of IL-2 and IL-3 responses and the phenotype of the responding cell following both primary and secondary stimulation in vivo. We then demonstrate that protocols previously shown to induce suppression of the mha response both in vitro and in vivo result in profound suppression of net lymphokine production by responder cells. The pattern of suppression seen following in vivo suppression is rather different from that seen in vitro. Finally, we present some experiments that begin to address the mechanisms that may underlie the suppressor phenomena. RESULTS Lymphokine
Synthesis in Anti-mha
Responses
Cells from CBA mice immunized to BlO.BR minor alloantigen respond in vitro to BlO.BR spleen cells by the production of IL-2 and IL-3 (Fig. 1). This response occurs using either cells that had received a single primary immunization in vivo or cells that had been immunized 2 to 6 weeks previously and, therefore, were in the secondary phase of an immune response. The response is antigen specific and does not occur in 5
CYTOKINE,
6 1 Yin and Chain
A
cules, suggesting that the responding population contained cells restricted to both molecules. These results establish that the major lymphokineproducing cell under the conditions of these assayswas a CD4’, class II MHC-restricted cell. Similar results were obtained using lymph node cells taken 4 days after priming (results not shown).
IL-2 production
0 1 6 or
Vol. 3, No. 1 (January 1991: 5-11)
Suppressionof Lymphokine Responses 25 Number
B
5 of responder
10 cells
The lymphokine responses can be demonstrated either by in vivo transfer of suppressor cells to recipients at the time of immunization (Fig. 3A) or by co-culture of responder cells with suppressor cells (Fig. 3B). However, while IL-2 production is downregulated
20 (x 104)
IL-3 production 0 50 1
A
IL-2 productlo”
0 2OT
0 16 t x 0 10+
oooc
/ 25 Number
5 of responder
20
10 cells
(x
vI 5 D
012
; x 0
0.08
40
T \\
004.-
104) 0007
Figure 1. The production of lymphokines by lymph node and spleen cells from immune mice.
unimmunized animals or animals immunized with an irrelevant antigen (data not shown). In contrast, similar levels of spontaneous (“background”) production of lymphokines (particularly IL-3), were observed using either immune or naive mice. These background levels, which probably represent a mitogenic response to substances in the fetal calf serum, have not been subtracted from any data shown, but are given in the appropriate legends. Both IL-2 (Fig. 2A) and IL-3 production (not shown) by secondary responder cells was markedly inhibited by depletion of CD4’ T cells prior to culture. CD8’ depletion had a much smaller inhibitory effect. In agreement with these depletion experiments, the lymphokine production could be blocked by specific anti-class II MHC monoclonal antibodies added to the cell cultures (Fig. 2B). The response was partially inhibited by antibodies to both I-Ak and I-Ek mole-
Ii 5
c7
Varying numbers of lymph node cells, taken 4 days post priming (primary), or spleen cells taken 4 weeks post priming (secondary), were incubated with or without 4 X 10’ irradiated allogeneic (BlOBr) spleen cells (as a source of antigen) and assayed for (A) IL-2 and (B) IL-3 production as described in Materials and Methods. -A-, primary response with antigen; -A-, primary response without antigen; -O-, secondary response with antigen; -0-, secondary response without antigen.
n
6
complement
0 16-
Number only
10
20
L10
of responder Cells (X 1 04) I - CD4 depleted W
CDS depleted
IL-2 production
0 12.x P 8 0 2 0”
0 08
004..
ooo-
Figure 2. Phenotype and MHC restriction of responder cells. (A) The phenotype of responder cells. Spleen cells from mice immunized for a secondary response were depleted of CD4’ or CD8+ cells by treatment with antibody and complement. Equal numbers of viable cells from depleted or control (complement only) groups were co-cultured with antigen and assayed for IL-2 production. (B) MHC restriction of the allogeneic secondary lymphokine response. Spleen cells (4 x 10s) from immunized mice were cocultured with antigen in the presence of varying concentrations of anti-Class II MHC antibodies as shown. IL-2 response in the absence of antigen gave an optical density of 0.08 U. The antibodies tested were: 0, anti-I-A”, HB3 (from ATCC); n , anti-I-A’, H40481.3 (a gift from Dr. M. Pierres, Marseille). q , anti-I-Ek, HB32 (from ATCC). All antibodies were IgG fractions, purified from ascites on protein A.
Suppression of lymphokines in anti-mha responses / 7
IL-2 production
IL-3 production
. /
25
5
10
20
B
0.45
0.35
O
0 30 ‘0
0 0.08
20
40
.T.
o-
o\ \
I
0
O--O i
0 20
‘\
0’
0 0.15
0 00’ 2.5
5
10
20
Number
3.
10
0 25
0.00-e
Figure
5 (x104)
IL-3 production
040T
012 x .ez
cells
I-
.‘.\
z 0 0 z
2.5
of responder
IL-2 production
0.16
/
40
Number
0
050
The suppression
of lymphokine
responses
40 of
suppressor
25
5
cells
( ~10~)
10
20
40
in vivo and in vitro.
(A) Primary response, in vivo suppression. Mice received 5 x 10’ spleen cells from donors that had received a high dose of antigen intravenously four days previously or from naive donors. The recipients were then immediately immunized in the foot pads for generation of a primary immune response (see Materials and Methods). Variable number of lymph node cells from these two groups of animals (“suppressed” and “normals”) were co-cultured with antigen, and assayed for IL-2 and IL-3 production. -O-, “suppressed” mice; -O-, “normal” mice. (B) Secondary response, in vitro suppression. Variable numbers of spleen cells from mice receiving high doses of intravenous antigen four days previously (“suppressed,” see Materials and Methods) or from naive mice were co-cultured with 4 x lo4 immune spleen cells (secondary response) and antigen and assayed for IL-2 and IL-3 production. Lymphokine production in the absence of antigen (“background”) gave optical densities of 0.02 for IL-2 and 0.19 for IL-3. -0-, co-culture with cells from suppressed mice; -O-, co-culture with cells from naive mice.
in both assays of suppressor activity, IL-3 activity is downregulated only by in vivo protocols. We wished to ascertain whether the lack of IL-3 suppression observed in the in vitro mixing experiments was indeed due to the use of an in vitro assay or to the fact that suppression was being assayed during the secondary, rather than the primary, phase of the immune response. In vitro primaq responses against these antigens cannot easily be obtained. The secondary response was therefore suppressed in vivo as described above, by transferring suppressor cells immediately prior to boosting previously primed mice. Both IL-2 and IL-3 responses could be suppressed using this immunization protocol (not shown). The lack of IL-3
suppression observed during in vitro co-culture experiments was therefore unlikely to be due only to the differentiation stage of the responder T cell (i.e., 2 or more weeks post priming in our standard in vitro assay of suppression, as opposed to being the target of suppression actually at the time of priming, as is the case in our primary in vivo assay of suppression). As shown in Fig. 4, suppressor activity depended on the dose of immunizing antigen. The suppression was rather easily induced (sometimes with as little as 10’ BlO.BR spleen cells), and was always found after injecting 5 x 10’ cells or more. The induction of suppression in this system therefore seems to depend more on the route of injection than on antigen dose,
CYTOKINE,
8 I Yin and Chain
suppressor activity as measured in both in vivo and in vitro assays (data not shown). However, differences were observed after depleting CD4’ or CD8’ cells in the two types of assay. Depletion of suppressor cells with either anti-CD4 antibody, or anti-CD8 antibody prior to in vivo transfer into responder mice partially abrogated the suppression (Fig. 5A), suggesting that both cell types play a part in the generation of a full suppressive response in vivo. Depletion of CD8’ cells from lymph node populations of mice that had been previously suppressed in vivo did not, however, reverse the suppression (not shown). The suppressive effect of the transferred CD8’ population is therefore no longer reversible at the time the cells are tested for lymphokine production. In contrast, all in vitro suppressor activity could be removed by treatment with anti-CD8 antibody and complement, while CD4’ depletion had no effect (Fig. 5B).
IL-3 production
\
O--O
0.00 4
/ 0
105
107
106
Number
5x107
of unmunlsing
108
Vol. 3, No. 1 (January 1991: 5-11)
5x108
cells
Figure 4. Antigen dose-response for the induction of responder and suppressor activity. -0-, responder: Mice received varying doses of antigen (BlOBr spleen cells) intraperitoneally 2 weeks prior to assay. Spleen cells (4 x l@) from immunized mice were incubated with antigen, and assayed for IL-3 production. -O-, suppressor: Mice received varying doses of antigen intravenously. Four days later, 5 x 10’ spleen cells were transferred intravenously to recipients immediately prior to foot-pad immunization. After a further 4 days, lymph node cells (4 x 104) from recipients were co-cultured with antigen, and IL-3 production was measured. As shown, increased intravenous doses of antigen resulted in greater suppression of the primary response.
The Role of IL-2 Competition
in Suppression
One possible mechanism by which suppressor cells may function is by lymphokine competition.8P9 This mechanism is particularly significant since, in these experiments, we measure only net lymphokine production, which is a balance of uptake and secretion. While the importance of this mechanism cannot be readily assessed in vivo, it can be tested in vitro. As shown in Fig. 6, the addition of anti-IL-2 receptor antibody (TIB 222) enhanced IL-2 levels in the in vitro coculture suppression system and, in fact, completely abrogated the differences between cultures with control and suppressor cells. The addition of the same antibody to cells suppressed in vivo also raised IL-2
since certain doses of antigen administered intravenously or intraperitoneally can induce suppressor cells or responder cells, respectively. The Phenotype of Suppressor Cells Treatment of suppressor cells with anti-Thy-l antibody and complement completely removed all
B
A
IL-2 production
0.08
0 16
--
I
x .e : 0.06-m : 0 .i-’ 0.04.. 2
0.08
0
0 04
0.00 6
/
0.12 0 00 L 2.5
Number
of responder
cells
(x 104)
Number
5
10
of suppressor
20
40
cells
(~10~)
Figure 5. Phenotype of suppressor cell population. (A) In vivo suppression. (B) In vitro suppression. The activity of suppressor cells was measured as described in the legend to Fig. 3A and 3B, but suppressor cells were depleted of CD4’ and CD8’ cells by treatment with antibody and complement. -0-, normal spleen cells; -O-, suppressor population; -A-, CD4+-depleted suppressor population; -A-, CD8+-depleted suppressor population.
Suppression of lymphokines in anti-mha responses / 9 IL-2 production 0 16
,x
0 12
T iI
A/ n--n A------./
0 OOl’--25 Number
Figure
6.
The role of IL-2
5
10
of suppressor
adsorbtion
cells
in in vitro
20 (xl 04)
I 40
suppression.
In vitro suppression was assayed as described in the legend to Fig. 3B. Anti-IL-2 receptor antibody (TIB 222, from ATCC), in the form hybridoma supernatant, was added to some wells. -0-, co-culture with normal spleen cells; -A-, co-culture with normal spleen cells plus TIB 222; -O-, co-culture with suppressor spleen cells; -A-, co-culture with suppressor spleen cells plus TIB 222.
levels but did not abrogate suppression (not shown). This result argues that the effects of transferring suppressor cells in vivo are no longer reversible by the time lymphokine production is stimulated in vitro.
DISCUSSION This study represents a re-examination of a wellstudied T-cell suppression system, using as readout lymphokine production by responder cells. This approach has broadly confirmed previous in vivo studies, insofar as T cells immunized according to “suppressor” protocols downregulate lymphokine biosynthesis in responder T cells. Furthermore, they have conclusively established that the target of suppression is primarily the CD4’ “ helper” T cell, a fact which has been questioned in some recent discussions of suppression,1o and which is not consistent with models involving lymphokine antagonism at the effector stage of immune responses. l1 The lymphokines tested in this study are both likely to be of importance in the anti-mha functional assays previously used. Thus IL-2 plays a major role in the induction and amplification of both cytotoxic and delayed-type hypersensitivity reactions, while IL-3 is involved in the recruitment, proliferation, and differentiation of the nonhemopoeitic lineage cells involved in inflammatory responses. It is therefore likely that the greatly reduced ability of suppressed T cells to produce these lymphokines is critical in explaining their other depressed functional capabilities. A comparison of the in vitro and in vivo models of suppression shows a number of interesting differences. In vitro suppression results in decrease only of IL-3 levels, is a function predominantly of CD8+ T cells, and
can be reversed by blocking IL-2 consumption in the cultures. The simplest interpretation of these results is that the suppressor population contains CD8+ cells, which show a net IL-2 consumption. These cells thus compete for IL-2 and lower the free levels in the culture supernatants. We do not know why differences in immunization routes (intravenous versus intraperitoneal) should activate populations of cells that show, respectively, a net IL-2 consumption or a net IL-2 production. However, data from other systems indicate that it is possible to induce T-cell tolerance at a clonal level, and that such tolerant cells can selectively lose the ability to make IL-2 but not the ability to respond to (and hence utilize) this lymphokine.” Further studies are required to demonstrate whether such partially tolerant cells are present within the suppressor population used in these experiments. In contrast to this in vitro suppression, in vivo suppression requires the action of both CD4’ and CDS’ T cells; this interaction between these two cell types has been observed frequently in other suppressor cell systems. In vivo suppression also leads to impaired IL-3 production. In addition, in vivo suppression can no longer be reversed by removing suppressor cells or by blocking IL-2 consumption in culture. If in vivo suppression is to be interpreted in terms of the in vitro model, that is to say in terms of IL-2 competition, one must hypothesize that a lack of IL-2 at a critical stage in T-cell activation (taking place in vivo), results in a more basic failure in T-cell activation, which also affects IL-3 production and is not readily reversible. Alternatively, in vivo suppression may result from completely different mechanisms, for example from competition between suppressor cells and responder cells for limited amounts of antigen. Our future studies will therefore concentrate on trying to develop more realistic in vitro models of the immune microenvironments found in vivo and on assaying cellular interactions and lymphokine responses during suppression in vivo.
MATERIALS AND METHODS Mice CBA/H and BlO.BR mice were supplied by the Imperial Cancer Research Fund breeding colony, and were used when
between 6 and 12 weeks old. Donors and recipients were sex-matched within each transfer experiment. Anti-Minor
Responses-Lymphokine
Production
The generation of proliferative, cytotoxic, helper, and suppressor responses in CBA mice to the minor alloantigens of the BlO.BR strain (both strains have the H-2’ haplotype) has been described in detail elsewhere.“s6 The immunization protocols used here are based on these previous studies. Two
10 / yin and Chain types of immunization were used: (i) Primary responses. CBA mice were injected in the two hind footpads with 50 l.rl of BlO.BR spleen cell suspension containing 2.5 x lo6 cells that had been irradiated (2000 R) under a “Co source. Four days later, popliteal and inguinal lymph nodes (LN) were removed from 2 to 5 animals and pooled. (ii) Secondary responses. CBA mice were injected intraperitoneally with 10’ irradiated BlO.BR spleen cells. Two to six weeks later, the spleens were removed and the cells were cultured as above. In some experiments mice primed in this way received a further boost of 2 x lo6 spleen cells in the footpads four days prior to sacrifice. In this case lymph node cells (as for primary responses) instead of spleen cells were harvested and cultured.
Suppressionof Anti-mha Responses A population of cells with suppressor activity for CBA anti-BlO.BR responses was generated as described previously.(’Viable nucleated BlO.BR mice spleen cells (lo8 cells) were injected into the lateral tail vein of CBA mice. Four days later spleen cells from these mice were used as a source of suppressor cells for the response to B10 minor mha’s. These cells are referred to simply as “suppressor cells” in this paper. Cells from uninjected mice were used as controls.
Assaysof SuppressorActivity (i) Primary response. Suppressor cells (5 X 10’) or normal spleen cells were transferred intravenously into responder mice immediately prior to footpad immunization as described above. Lymphokine production of lymph node cells from mice receiving suppressor cells or normal spleen cellswas compared. (ii) Secondary response. In vitro: Spleen cells (4 x 10’) from mice immunized as in Anti-Minor Responses-Lymphokine Production, section ii (above) were co-cultured with various numbers of suppressor cells or normal cells, and lymphokine production in the two groups was compared. In vivo: suppressor cells or normal cells (5 x 10’) were transferred intravenously into mice immediately before boosting with antigen as described in Anti-Minor Responses-Lymphokine Production, section ii (above). In all experiments, single cell suspensions were prepared from spleen or lymph nodes and incubated at various cell densities in 200 ul of Iscove’smodification of Dulbecco’s medium supplemented with 5% fetal calf serum (Gibco), 2 x 10m5M 2-mercapthoethanol, 100 U/ml of penicillin, 100 p&ml of streptomycin. Antigen, in the form of 4 x 10’ irradiated (2,000 R) BlO.BR spleen cells, was added to some wells. Triplicate cultures were set up for each experimental point. Supernatants were collected after 48 h of incubation (in 5% CO,, humidified incubators at 37°C) and stored at -20°C for assayof IL-2 and/or IL-3. Preliminary experiments indicated that lymphokine production was optimal at 48 h and with this dose of antigen.
Anti-Thy 1, AntMD4, and Anti-CD8 Antibody Depletion of SuppressorCellsor ResponderCells Spleen cells were incubated at 2 x lO’/ml with anti-Thy1 (YBM29-2-1) anti-CD4 (RLi72.4), or anti-CD8 (3.169) anti-
CYTOKINE, Vol. 3, No. 1 (January1991:5-11) body for 20 min in ice. The cells were pelleted, resuspended in l/10 dilution of rabbit complement (Buxted), and incubated at 37°C for 45 min. Controls were treated with complement alone. The cells were washed twice and then used for further investigation. Preliminary studies showed that this treatment resulted in depletion of greater than 90% of the specific population targeted.
Assayof IL-2 and IL-3 Activities IL-2 and IL-3 activities were assayed by testing the ability of the culture supernatants collected from different experimental groups to support the growth of the IL-2dependent CTLL line or the IL-3-dependent FDCP-2 line. Cell viability of the indicator cell lines following culture in the test supernatants was assessedby reduction of the dye MIT by viable cells. The assays have been described in detail elsewhere.13Results have not been converted into lymphokine units, but are expressed simply as the mean optical density values (absorbance at 570 nm - absorbance at 630 nm) obtained from triplicate cultures of the stained indicator cell-line cultures. Results should not, therefore, be interpreted in a strictly quantitative way. Standard errors are not shown but were generally in the region of 10%. All figures show data from a single experiment, but each set of experiments was repeated a minimum of three times. Optical densities are approximately proportional to lymphokine titer up to values of 0.3 for IL-2 and 0.4 for IL-3. Five units of recombinant IL-2 and 10 U of IL-3 gave an optical density of approximately 0.1 optical density units in these assays.The CTLL line responds both to IL-2 and IL-4, but is very insensitive to IL-4. Furthermore, the IL-2 responses measured in this study could not be inhibited by a blocking antibody specific for IL-4. The FDCP-2 line did not respond to IL-2 or IL-4. It did, however, respond to both IL-3 and granulocyte-macrophage colony-stimulating factor (GMCSF). Since we did not have a specific antibody capable of blocking either of these activities, this study does not distinguish between the production of these two lymphokines. Acknowledgments L. Yin was supported by a Royal Society visiting Fellowship (China programme). B.M. Chain is supported by a Royal Society University Research Fellowship.
REFERENCES genes.
1. Ohveira DBG, Mitchison Clin Exp Immunol75:167-177.
NA
(1989)
Immune
suppression
2. Mitchison NA (1989) Is genes in the mouse. In Melchers et al. (eds) Progress in Immunology, Vol. VII, pp 845-857. 3. Moller G (1988) Immunol27:247-250.
Do
4. Crispe IN, Gascoigne murine T helper clone which vitro and in vivo. Immunology 5. Gascoigne
NRJ,
suppressor
T cells
exist?
F,
Stand
J
NRJ, Owens T (1984) An allospecihc can help both T and B cell responses in 5255-65.
Crispe
IN (1984)
Suppression
of the cyto-
Suppression of lymphokines in anti-mha responses / 11 toxic T cell response to minor alloantigens in vivo. Linked recognition by suppressor T cells. Eur J Immunol14:21O-215. 6. Gascoigne NRJ (1984) Suppression of the cytotoxic T cell response to minor alloantigens in vivo. II. Fine specificity of suppressor T cells and lack of restriction by immunoglobulin heavy chain-linked gene products. Em J Immunol 14:677-680. 7. Nanda NK (1989) Preferential restriction of minor alloantigen-specific suppressor T cells to I-E rather than I-A molecules. Immunology 68:163-168. 8. Gunter J, Haas W, von Boehmer H (1982) Suppression of T cell responses through competition for T cell growth factor (interleukin 2). Eur J Immunol12:247-249. 9. Palacios R, Moller G (1981) T cell growth factor abrogates
concanavalin A-induced suppressor cell function. J Exp Med 153: 1360-1365. 10. Lanzavecchia A (1989) Is suppression a function of class II restricted cytotoxic T cells. Immunology Today 10:157-159. 11. Rabin E, Mond J, Ohara J, Paul W (1986) Interferon-y inhibits the action of B cell stimulatory factor (BSF)-1 on resting B cells. J Immunol 137:1573-1576. 12. Essery G, Feldmann M, Lamb J (1988) Interleukin 2 can prevent and reverse antigen induced unresponsiveness in cloned human T cell clones. Immunology 64:413-417. 13. Marcinkiewicz J, Chain BM (1989) Antigen specific inhibition of IL-2 and IL-3 production in contact sensitivity to TNP. Immunology 68:185-189.
