- Email: [email protected]
Regulatory responses in contact sensitivity: Afferent suppressor T cells inhibit the activation of efferent suppressor T cells
CELLULAR
IMMUNOLOGY
132,400-4 10 ( 199 1)
Regulatory Responses in Contact Sensitivity: Afferent Suppressor Cells inhibit the Activation of Efferent Suppressor T Cells WLODZIMIERZ
PTAK,*
T
CHARLES A. JANEWAY, JR.,? JANUSZ MARCINKIEWICZ,* AND PATRICK M. FLOOD-~
*Department of Immunology, Copernicus Medical School, Cracow, Poland, and tDepartment oj Pathology and Howard Hughes Medical Institute, Yale University School of Medicine, New Haven, Connecticut 06520 Received June 18, 1990; accepted September 14, 1990 Two types of suppressor cells regulate the contact sensitivity (CS) response to picryl chloride (PCL). Afferent suppressor T cells (Ts-aff) inhibit the generation of CS responses to PCL, while efferent suppressor T cells (Ts-eff) inhibit the activity of Th 1 cells that mediate CS reaction. Intravenous injection of mice with TNP-substituted peritoneal exudate cells (TNP-PEC) induces Ts-elf cells that block the adoptive transfer of contact sensitivity. The induction of Ts-eff cells is prevented by the presence of Ts-aff cells, which in turn are induced by the injection of TNP-PEC coupled with antibodies of the IgG2a and IgG2b isotype (TNP-PEC-Ab). If an animal is injected with TNP-PEC prior to or simultaneously with TNP-PEC-Ab, it generates only Ts-aff cells, while if it is injected with TNP-PEC alone or TNP-PEC prior to TNP-PEC-Ab, it generates Ts-effcells. Ts-aff cells effect only the generation of Ts-eff cells, as the addition of Ts-aff cells to assays for Tseff cells has no inhibitory effect on the suppressive effects of Ts-eff cells in adoptive transfer. Our experiments show that Ts-aff cells induced by TNP-PEC-Ab are phenotypically either Lyt l+2or Lyt l-2+, but only the latter inhibit the generation of Ts-eff cells in vivo. The Ts-aff cells that inhibit Ts-eff activity adhere to the lectin Vicia villosa (VV), while Ts-eff cells are VV nonadherent. In addition, Ts-aff cells can prevent the generation of Ts-eff to linked haptens presented on the same PEC. it appears that a cascade of Ts cell interactions are involved in the regulation of CS responses. 0 1991 Academic Press, Inc.
INTRODUCTION
The quantity and quality of the immune response are determined by the activity of two functionally distinct T cell populations: suppressor T cells that eliminate responses and contrasuppressor T cells that protect responding cells from suppressor cell influences (1). The interaction of these regulatory cells forms a very complex circuit in which the activities of one subpopulation of cells may influence the activity of other subpopulations of cells. In assessingthe nature of a contact sensitivity (CS)’ response,it is crucial to determine how a particular subpopulation of cells can regulate the activity of other cells. The ’ Abbreviations used: APC, antigen-presenting cell; C, rabbit complement; CS, contact sensitivity; DNP, dinitrophenyl; iv, intravenous; Lytl cells, cells treated with anti-Lyt2 mAb + C; Lyt2 cells, cells treated with anti-Lytl mAb + C, mAb, monoclonal antibody; OX, 4-ethoxymethylene-2-phenyloxazolone; PCL, picryl chloride; PEC, peritoneal exudate cells; TNBSA, trinitrobenzene sulfonic acid; Th, T helper cells; TNP, trinitrophenyl; TNP-PEC, TNP-haptenated peritoneal exudate cells; TNP-PEC-Ab, TNP haptenated PEC treated with anti-TNP IgG2a monoclonal antibody; Ts, T suppressor cell; Ts-aff, afferent T suppressor cells; Ts-eff, efferent T suppressor cells; VV, Vicia villosa; VV+, V. villosa adherent; VV-, V. villosa nonadherent.
0008-8749/9 I $3.00 Copyright 0 1991 by Academic Press, Inc. All rights of reproduction in any form reaved
REGULATION
OF SUPPRESSION
IN CS
401
regulation of the CS response to haptens such as TNP or DNP is characterized by the activity of two distinct suppressor T cell populations: (i) afferent suppressor cells (Tsaff), which prevent CS responses by blocking the priming of cells to antigen in naive animals (3); and (ii) efferent suppressor T cells (Ts-eff) which block CS by inactivating cells previously primed to antigen (4-6). Ts-eff cells, are generated by intravenous injection of water-soluble hapten derivatives or by iv injection of haptenated spleen or peritoneal exudate (PEC) cells (4-6). Ts-aff cells, on the other hand, are generated by injection of TNP-haptenated PEC treated with anti-TNP monoclonal antibodies of the IgG2a or IgG2b isotype, which form cell surface bound immune complexes
(3, 7). We noticed that while injection of TNP-PEC-IgG2a resulted in Ts-aff cells, no Tseff cells were generated, despite the fact that large numbers of uncomplexed haptenic groups were still present on the antibody-treated TNP-PEC (7). This suggested to us that the generation of Ts-aff cells might somehow inhibit the generation of Ts-eff cells in viva. We report here that Ts-aff cells, in addition to their previously established inhibition of effector lymphocytes that mediate CS reaction (Th l), do in fact inhibit the generation of Ts-eff cells. Ts-aff cells have no effect on the functional activity of either Th 1 or Ts-eff cells once these cells are induced. MATERIALS
AND
METHODS
Mice. CBA/J male mice were purchased from Jackson Laboratories (Bar Harbor, ME) or from the breeding unit, Department of Immunology, Cracow, Poland, and were used between 6 and 8 weeks of age. Each experimental group consisted of four or five mice. Reagents. The following reagents were used: twice-recrystallized picryl chloride (trinitrophenyl (TNP) chloride; PCL) (Chemtronics, Swannanoa, NC); trinitrobenzene sulfonic acid (TNBSA) (Eastman Organic Chemicals, Rochester, NY); Sephadex G200 and Sepharose 4B (Pharmacia, Fine Chemicals, Piscataway, NJ); and 4-ethoxymethylene-2-phenyloxazolone (OX) (British Drug House, Gallard-Schlesinger Chemical Mfg. Corp., Carle Place, NY). Peritoneal exudate cells (PEC). PEC were collected 4 days after intraperitoneal injection of Mark01 52 oil. Red blood cells were lysed with NH4Cl buffer prior to haptenation. PEC were labeled with TNP and oxazolone (OX) as described previously (5). When PEC were double labeled (i.e., with TNP and OX), TNP substitution was followed by OX substitution. TNP-PEC were incubated for 30 min at room temperature with an affinity-purified anti-TNP IgG2a antibody (3 fig per 1 X lo7 cells unless otherwise stated). Cells were washed in phosphate-buffered saline and 1 X lo7 cells were injected iv into recipients. Antibodies. Monoclonal anti-Lyt 1.1 and anti-Lyt2. I antibodies were kindly provided by Dr. F. W. Shen, Sloan-Kettering Institute, New York. Monoclonal anti-TNP IgG2a (32.1) hybridoma cells were obtained from Dr. Julian Fleischman, Washington University School of Medicine, St. Louis, Missouri. Antibodies were affinity purified by absorbtion onto a TNP-KLH-Sepharose 4B affinity column and were eluted with 3 A4 guanidine. Treutment of regulatory cefls. Spleen cells from mice immunized with TNP-PEC or TNP-PEC-IgG2a (TNP-PEC-Ab) were treated with the anti-Lyt 1.1 or the antiLyt2.1 antibody for 30 min at room temperature, washed, and then incubated for 30
402
PTAK
ET AL.
min at 37°C in rabbit complement (C) (prescreened for low natural cytotoxicity to mouse spleen cells). Animals were injected with the equivalent of 7 X 10’ untreated spleen cells. Active sensitization and adoptive transfer of contact sensitivity. Animals were skin sensitized with PCL or OX by painting the shaved abdomen as well as four feet with 0.15 ml of 5% PCL or 3% OX (in 3:l ethanoLacetone) and were tested for contact sensitivity by application of 0.8% (w/v) PCL or OX in olive oil to both ears 4 days later (8). Ear swelling was assessedby measuring the earsimmediately before application and 24 hr later using an engineer’s micrometer. Results are expressedin units of 10e3 cm + SD. In each experiment, ears of separategroups of nonimmunized controls were similarly challenged and measured. The background swelling response in the control group averaged about 1 to 2 units and was subtracted from the swelling response of the experimental animal group so the results are presented as a net increase in ear swelling. In adoptive transfer, 7 X 10’ spleen and lymph node cells of animals skin painted 4 days previously with PCL or OX were injected iv into recipients that were ear challenged immediately, and ear swelling was measured 24 hr later. In results, group A always represents the positive untreated control group. In the afferent suppressor cell assay,this group representsthe group that was only contact sensitized with PCL or OX and received no regulatory cells. In the efferent suppressor cell assay,this group received only cells from animals skin painted 4 days previously with PCL or OX. In all figures, results are also expressed as percentage responses,which are calculated according to the formula: (experimental - negative control)/(positive control - negative control) X 100. Testfor efirent suppression (Ts-ef). Spleen and lymph node cells (7 X IO’) from animals immunized by skin painting were mixed with 5 X 10’ spleen cells from mice injected 7 days previously with hapten-substituted PEC. This cell mixture was incubated for 30 min at 37°C and was then injected into naive recipient mice. Animals were challenged immediately on the ears and were assayed for contact sensitivity 24 hr later (5). Testfor a&rent suppression (Ts-af). Spleen cells (5 X 10’) from mice being assayed for Ts-aff activity were injected iv into naive recipient mice, which were immediately sensitized by skin painting with PCL or OX. Animals were ear challenged 4 days later and contact sensitivity was assessedat 24 hr (3, 7). Separation of cells on the Vicia villosa (VV) lectin. VV was prepared as described previously (9). Separation of cells into V. villosa adherent (VV’) and nonadherent (VV) populations was done as described in detail previously (3, 7). Generally, the lectin adherent cells represent 8-20% of the immune spleen cell population. Statistics. Evaluation of the statistical significance of the results was done using the double-tailed Student’s t test. RESULTS Ts-eflcells are generated by TNP-PEC only under conditions in which Ts-afcells are not generated. Earlier reports had shown that iv injection of TNP-PEC cells led to the production of Ts-eff cells, while pretreatment of these TNP-PEC with anti-TNP mAb of the IgG2a or IgG2b isotype leads to only Ts-aff cells (3, 7). We examined the relationship between Ts-aff generation and Ts-eff inhibition by titrating the amount of IgG2a antibody used to coat the TNP-PEC (Fig. 1). TNP-labeled PEC (10’) were
REGULATION
TNPPEC
An,,bady
Afferent
Suppressor
Cell
OF SUPPRESSION
IN CS
Assay
Efferent
403 Suppressor
Cell
Assay
-I (100) ( 91) (
22)
( 25) ( 39) ( ‘51) ( 84)
(Units
of swelling
x IO
-3
cm)
FIG,. 1. Titration of IgG2a monoclonal anti-TNP antibody needed to induce Ts-aff cell activity when incubated with TNP-PEC. TNP-PEC (I X IO’) from CBA mice were incubated with the indicated amount of affinity-purified anti-TNP-IgG2a mAb before iv injection into syngeneic mice. Six days later the spleen cells were assayed for activity. The assay for Ts-aff consisted of iv injection of 5 X IO’ spleen cells from TNP-PEC+-Ab recipient animals into naive syngeneic mice and skin painting these mice immediately with PCL. Contact sensitivity to PCL was measured 4 days later. The assay for Ts-eff ceils consisted of combining 5 X 10’ spleen cells from TNP-PEC+-Ab recipient animals with 7 X IO’ spleen and lymph node cells from mice skin sensitized with PCL 4 days earlier. These cell mixtures were incubated for 30 min at 37°C then injected iv into naive recipients that were then immediately challenged with PCL, and the ear swelling (DTH) reaction was measured 24 hr later. In this and in the following figures, nonspecific ear swelling of negative control (nontreated) mice that were similarly challenged with antigen was subtreacted from experimental values (see Materials and Methods). Thus, results are presented as the net increase in ear swelling in units X lo-’ cm f SD. In no case did the injection of TNP-PEC or TNP-PEC-Ab induce a significant contact sensitivity reaction by itself. In Student’s t test. in the afferent suppressor cell assay, group B was significantly different than groups C, D, and E (P < 0.00 1) and F (P < 0.0 I); while in the efferent suppressor cell assay. group A was significantly different than groups B and G (P < 0.00 I) and group F (P < 0.01).
incubated with various doses of affinity-purified anti-TNP IgG2a mAb, starting at 20 pg to as low as 0.03 pg, and then injecting these TNP-PEC-Ab complexes into naive mice. Ts-aff cell activity or Ts-eff activity was assayed as described under Materials and Methods. The results in Fig. 1 show that IO7 TNP-PEC treated with anti-TNP IgG2a mAb with doses as low as 0.8 PLgwere effective in generating Ts-aff but not Tseff cells, while a dose of 0.03 pg did not generate Ts-aff but did generate Ts-eff cells. A dose of 0.8 PLgof IgG2 mAb represents approximately 3 X 10” antibody molecules and, as shown previously (3) lo7 PEC are substituted with approximately 3 X lOI TNP haptenic groups. Therefore, assuming that all anti-TNP mAb molecules bind to a TNP haptenic group on the surface of the APC, our results show that saturation of approximately 1% of the TNP molecules on the surface of the TNP-haptenated PEC with antibody completely prevents the generation of Ts-eff cells. Interestingly, TNPPEC treated with anti-TNP IgG2a mAb at a dose of 0.15 PLggenerated intermediate activities of both Ts-aff and Ts-eff cells. These results suggest a direct correlation between the activation of Ts-aff activity and the inhibition of Ts-eff cell activity. Ts-afcells inhibit the generation of Ts-efcells. In order to test whether the lack of Ts-eff activity in TNP-PEC-Ab treated mice was simply due to TNP-PEC-Ab not activating Ts-eff or whether this treatment actually blocked Ts-eff activation, we injected TNP-PEC, TNP-PEC-Ab, or equal numbers of both iv into naive animals and assayed their spleen cells for both Ts-aff and Ts-eff. The results in Fig. 2 show that while TNP-
404
PTAK Induction Regulatory
of Cells Afferent
GrOUD
TNP-PEC
ET AL.
Suppressor
Cell Assay
Efferent Suppressor
Cell Assay
TNP-PEC-Ab
2
0
4
10
8
6
(Units
0
of Swelling
4
2
6
8
x id 3 cm)
FIG. 2. Injection of TNP-PEC incubated with IgG2a anti-TNP monoclonal antibody prevents the induction of Ts-eff cells. CBA mice were injected iv with 1 X 10’ TNP-PEC incubated in BSS, I X 10’ TNP-PEC incubated before injection with 3 pg of IgG2a anti-TNP monoclonal antibody, or a mixture of both (IO’ each). Seven days later, cells of these mice were tested for afferent and efferent suppressor cell activity as described in Fig. 1. In Student’s t test in the afferent suppressor cell assay, group B was significantly different than groups C and D (P < 0.001); while in the efferent suppressor cell assay, group A was significantly different than group B (P < 0.001).
PEC induced only Ts-eff cells, TNP-PEC-Ab induced only Ts-aff cells. Figure 2 also shows that an equal number of TNP-PEC and TNP-PEC-Ab injected iv simultaneously resulted in the exclusive generation of Ts-aff cells with no Ts-eff cells. In the next experiment (Fig. 3) we examined whether the injection of TNP-PECIgG2a prior to or subsequent to TNP-PEC inhibited Ts-eff cell activity. Naive mice were injected on Day 0 with either TNP-PEC-Ab or TNP-PEC, and again on Day 6.
Treatment of donors of regulatory cells
Afferent
Suppressor
Efferent Suppressor
Cell Assay
Cell Assay
Group A e c 0 E F G H 0
2
4
6
(Units
8
10
of swelling
2
0
x IO
4
6
8
-3 cm)
FIG. 3. Ts-eff cell activity generated prior to Ts-aff cell induction is not suppressed by Ts-aff cells. CBA mice were injected with 1 X 10’ TNP-PEC or 1 X 10’ TNP-PEC incubated in 3 ng IgG2a anti-TNP monoclonal antibody on Day 0 as indicated. On Day 6, recipient mice received an injection with the identical dose of either TNP-PEC or TNP-PEC incubated with 3 pg of IgG2a anti-TNP monoclonal antibody as indicated. The group marked with an asterisk was injected on Day 2 rather than Day 6. On Day 12, spleen cells of these mice were tested for the presence of afferent and efferent suppressor cells as described in Fig. 1. In Student’s t test, in the afferent suppressor cell assay, group A was significantly different than groups B, C, E, F, and H (P < 0.001); while in the efferent suppressor cell assay, group A was significantly different than groups D, E and G (P < 0.00 I ) and group F (P < 0.05).
REGULATION
OF SUPPRESSION
405
IN CS
On Day 12 the mice were tested for Ts-aff and Ts-eff activity. Our results indicate that injection of TNP-PEC-Ab induced good Ts-aff activity (group C) but not Ts-eff activity, even when animals were subsequently injected with TNP-PEC (group B). In contrast, animals injected with TNP-PEC prior to injection with TNP-PEC-Ab (group E) showed both Ts-aff and Ts-eff activity when assayed. Interestingly, injection of TNP-PEC-Ab only 2 days after TNP-PEC significantly inhibited the generation of Tseff cells (group F). These results suggest that inhibition of Ts-eff cell activity was not due to a lack of APC for Ts-eff activation. They demonstrate rather that procedures that activate Ts-aff cells inhibit the generation of functional Ts-eff cells when performed before, at the same time, or shortly after procedures that activate Ts-eff cells. Ts-affcells do not act by masking Ts-efactivity. However, to rule out the possibility that Ts-aff cells that inhibit Ts-eff generation were functioning by masking or “contrasuppressing” the activity of Ts-eff cells, we took advantage of the fact that contrasuppressor T cells and Ts-aff cells adhere to the lectin V. villosa while Ts-eff cells do not (7, 10). We injected mice iv with TNP-PEC, TNP-PEC-Ab, or a mixture of TNPPEC and TNP-PEC-Ab, separated spleen cells of each group on the basis of their adherence to V. villosa, and then assayed cell fractions for Ts-aff and Ts-eff cell activity. The results in Fig. 4 show that TNP-PEC-injected animals produced only Ts-eff cells and that these cells were V. villosa nonadherent (group A). Animals injected with TNP-PEC-Ab produced only Ts-aff cells (group B) and these cells, in contrast to Tseff cells, were V. villosa adherent. Animals injected with equal numbers of TNP-PEC and TNP-PEC-Ab produced only Ts-aff cell activity, even when separated into V. villosa adherent and nonadherent fractions (group C). These results indicate that the generation of Ts-aff cells does not mask Ts-eff cell activity, but rather inhibits the generation of Ts-eff cells. Ts-a$‘cells do not suppress Ts-efcell activity. We then tested whether Ts-aff cells inhibited the activity of previously activated Ts-eff cells (Fig. 5). Ts-eff cells significantly suppressed the adoptive transfer of contact sensitivity by PCL-immune cells, while Ts-aff cells had no effect (groups B and C). When Ts-aff and Ts-eff cells were mixed
B
TNP~PEC-Ab
C
TNP-PEC + TNP-PEC-Ab
( 92) (100) ( 98) ( 401 ( 39) ( 891 ( 421 ( 341 ( 92) 0
2
4
6
El
(Units
of
IO swelling
0 x IO
2
4
6
8
-3 cm)
FIG. 4. Inhibition of Ts-etf cell activity by induction of Ts-aff cell is not due to the “masking” of Ts-eff cell activity by Ts-aff cells. CBA mice were injected iv with I X IO’ TNP-PEC, I X IO’ TNP-PEC incubated in 3 pg of IgG2a anti-TNP monoclonal antibody, or a mixture of 5 X 10” TNP-PEC from both the treated and the untreated groups as in Fig. 2. Seven days later, spleen cells of these mice were separated on V. vi//osa (VV) lectin plates into VV+ and VV populations, and each fraction (equivalent of 7 X 10’ cells) was tested for afferent and efferent suppressor cell activity as described in Fig. 1. In Student’s t test, group A was significantly different than groups B and C (P < 0.001).
406
PTAK
ET AL.
PCI immune cells Incubated with: FUSl
Se3md
2
0 (Units
4
6
of swelling
x 10
8
-3
cm)
FIG. 5. Ts-aff cells do not inhibit the functional activity of Ts-eff cells. CBA mice were injected with 1 X 10’ TNP-PEC to generate Ts-eff cells or with 1 X IO’ TNP-PEC incubated with 3 fig of IgG2a anti-TNP monoclonal antibody to generate Ts-aff cells. Six days later, 5 X 10’ spleen cells from these mice were incubated for 60 min at 37°C with 7 X IO’ spleen and lymph node cells from animals skin painted 4 days earlier with PCL (groups B and C). Cell mixtures were washed and incubated for a second 60-min period with 5 X IO7 spleen cells exhibiting Ts-aff or Ts-eff activity (groups E and F). Group D: PCL-immune cells were incubated with 5 X IO’ spleen cells from TNP-PEC-immunized animals simultaneously with 5 X IO’ spleen cells from TNP-PEC-Ab-immunized animals. In Student’s t test, group A was significantly different than groups B and D-F (P < 0.00 I).
together prior to addition to PCL-immune cells, the suppressive activity of Ts-eff cells was retained (group D). Preincubation of PCL-immune cells with Ts-aff cells likewise did not prevent the suppression of PCL immune cells by Ts-eff cells (group E). The subsequent addition of Ts-aff cells did not reverse the suppressive effects of Ts-eff cells on PCL-immune cells (group F). Therefore, it appears that Ts-aff cells have no effect on the activity of mature Ts-eff cells. Lytl-2+ but not the LytS2- Ts-afcells inhibit Ts-eflcell generation. We phenotyped the cells from the Ts-aff population that inhibited the generation of functional Ts-eff
Group A s c 0
2
0 (Units
4 of swelling
6
8
x 16 3 cm)
FIG. 6. Only Lytl-,2+ Ts-aff cells inhibit the generation of Ts-eff cells. CBA mice were injected with 1 X IO’ TNP-PEC treated with 3 pg of IgG2a anti-TNP monoclonal antibody and their spleen cells were removed 7 days later. Spleen cells were treated with anti-Lytl mAb (Lyt2 cells) or anti-Lyt2 mAb (Lytl cells) mAb + C, and 7 X 10’ splenic cell equivalents were then injected iv into syngeneic recipient mice. Recipients simultaneously received iv injections of I X 10’ TNP-PEC, and the activity of Ts-eff cells was tested 6 days later as detailed in Fig. 1. In Student’s t test, group A was significantly different than groups B and G (P < 0.001).
REGULATION
OF SUPPRESSION IN CS
407
cells (Fig. 6). Ts-aff cells were generated in naive CBA mice by iv injection of TNPPEC-Ab, and their spleen cells were transferred into naive recipients. These recipients were then injected with TNP-PEC in order to generate Ts-eff cells. While naive mice first injected with Ts-aff cells did not develop Ts-eff cells when injected with TNPPEC (group C), and treatment of the same Ts-aff cells with anti-Lytl mAb + C before transfer (Lyt 2+ cells) still blocked Ts-eff cell production (groups D and E), treatment of Ts-aff cells with anti-Lyt2 mAb + C before transfer (Lyt l+ cells) allowed the generation of Ts-eff cells (groups F and G). Therefore, even though the Ts-aff cells which inhibit CS reactions have been phenotyped as both Lytl+,2- and Lytl-,2+ (7) only Lyt l-,2+ Ts-aff cells inhibit the generation of Ts-eff cells.
TNP-specijc Ts-ajfcelIs prevent the getieration of oxazolone-spec$c Ts-eHee1f.s on/~. if‘ TNP and OX are linked on the same carrier. We tested the antigen specificity of inhibition of Ts-eff cell generation by measuring the inhibition of OX-specific Ts-eff cell generation by TNP-specific Ts-aff cells. Mice were injected with a mixture of TNPPEC and OX-PEC or with PEC that were colabeled with both haptens (TNP-OXPEC). The results in Fig. 7 show that mice injected with both TNP-PEC and OXPEC (group B) or mice injected with TNP-OX-PEC (group C) generated Ts-eff cells specific for both TNP and OX. When TNP-PEC were treated with anti-TNP IgG2a mAb and coinjected with OX-PEC (group B), only OX-specific Ts-eff cells could be detected. These results demonstrate that the activity of Ts-eff cell generation is antigen specific since TNP-PEC-Ab injection inhibits only TNP-Ts-eff cell generation and not OX-Ts-eff cell generation. Interestingly, injection of TNP-OX-PEC treated with antiTNP IgG2a mAb (group E) inhibits the generation of Ts-eff cells specific for either TNP or OX. These results parallel those obtained earlier for Ts-aff cells which inhibit CS reactions (see Ref. (7)) and suggest that the immune complexes bound to the surface of an APC also induce Ts-aff cells to other haptenic determinants, provided these determinants are present on the same PEC as the antigen-antibody complex. DISCUSSION The suppression of the CS response to PCL in normal adult mice is mediated by at least two distinct populations of suppressor T cells. These two cell types can be distinguished both on the basis of functional activity and by their adherence to the lectin I’. viffosa (7). The first population consisted of both Lyt 1+,2- and Lyt-2+ T cells that adhere to VV and that suppress the generation of, but not the effector phase of, a CS response (7). These cells, termed Ts-aff cells, have also been described in a different antigenic system (2, 11). The second population of Ts cells does not adhere to VV (7). These Ts-eff cells are Lyt l-2+ (12) and suppress only the effector phase and not the induction phase of a CS response (3, 5, 6). In addition to the functional difference between these two cell populations, the conditions that activate Ts-aff and Ts-eff cells are distinct. It has previously been found that the type of Ts cells induced by antigenic challenge can depend on the chemical nature of the tolerogen used (e.g., DNBSA vs TNBSA (2-6) the epitope density of the hapten-modified cell (3-5, 13), the dose of tolerogen used to induce suppressor cells (2, 4) the time after antigenic challenge (2, 3, 7), and whether the hapten-modified cells are syngeneic or allogeneic (3-5, 13). In the TNP system, however, Ts-eff cells are generated by intravenous injection of tolerizing doses of TNBSA or TNP-conjugated splenic or peritoneal exudate APC (3, 5-7) while Ts-aff cells are generated by TNP-conjugated IgG (3) or TNPPEC treated with anti-TNP mAb of the IgG2a or IgG2b isotype (7).
408
PTAK
Ofl”d”C,lO” GP3”PMD& OfTs-effCells
ET AL.
ox
PCI
A
NOcellsadded
B
TNP-PEC +OX-PEC
( 18)
c
TNP-OX-PEC
( 39)
(100)
D TM-PEC-Ab +OX-PEC
E
(
TNP-OX-PEC-Ab 2
4
6
8
10
iL 0
(100)
( 0) (25) ( 10) (88)
82)
( 90)
cc 0
--I
1
2
4
G
8
(Units of swelling x 10. 3 cm) RG. 7. Anti-TNP IgG2a monoclonal antibodies on the surface of TNP-PEC prevent the development of Ts-eff to the OX hapten if TNP and OX are on the same PEC. CBA mice were injected iv with 1 x IO7 TNP-PEC, OX-PEC, or with PEC that were double labeled with TNP and OX (TNP-OX-PEC) incubated prior to injection with 3 pg of IgG2a anti-TNP monoclonal antibody. Seven days later, Ts-eff cell activity was examined in injected animals by mixing 5 X 10’ of their spleen cells with 7 X 10’ spleen and lymph node cells from PCL- or OX-immune animals and injecting these cell mixtures into naive recipients. Recipient mice were then immediately challenged with the immunizing hapten, and the adoptive response was measured 24 hr later. In Student’s t test, OX group B was significantly different than group E (P < 0.001).
Miller et al. (2) have shown that Ts-eff and then Ts-aff cells are induced by an intravenous injection of syngeneic DNP-coupled lymphoid cells, with Ts-eff cells appearing around Day 7 followed by Ts-aff cells around Day 14. Ts-aff cells can in fact be induced by infusion of Ts-eff cells, suggesting a cellular circuit in which Ts-eff cells are first induced by antigenic challenge, followed by a wave of Ts-aff cells. Our results, both those presented here and those reported earlier (3, 7), suggest two more factors that influence both Ts-aff and Ts-eff cell generation: (i) the functional cell type that is initially activated, since while Ts-eff cell activation may favor the subsequent activation of Ts-aff cells (2), Ts-aff cell activity does not allow the generation of Ts-eff cells; and (ii) the nature of the humoral response to the hapten, in that hapten presented to the immune system in the absence of antibody responses favors Ts-eff cell generation (3, 7), while hapten complexed to IgG immunoglobulins (11, 14) or macrophage-bound immune complexes of the IgG2a or IgG2b isotype favor Ts-aff cell generation (3, 7). The mechanism by which Ts-aff cells block activation of Ts-eff cells is not yet clear, and because our experiments demonstrate only that cell populations containing Tsaff cells prevent the generation of functional Ts-eff cells, the cell interactions leading to this effect may be highly complex. A number of systems that exhibit feedback inhibition of suppressor T cell activity have been described ( 1, 15, 16). Initially, it was found by Gershon and co-workers that T suppressor-effector cells active in regulating SRBC-specific antibody responses in vitro suppress Ts-inducer cells ( 17). Ts-aff cells likewise seem to block the induction of Ts-eff activity since mature Ts-eff cells fail to be suppressed by Ts-aff cells. This model suggests that Ts-aff cells prevent both the generation of CS responses and the generation of regulatory cells which would suppress it, thereby conserving immunologic energy by preventing an immunoregulatory response that is not needed. We have never found in the TNP system that Ts-eff cell generation preceeds and is required for Ts-aff cell activity (W. Ptak, unpublished results), although it is clear from our results and those of others (2) that a mutually antagonistic relationship exists between Ts-aff and Ts-eff cells. Such interregulation of T suppressor cell subsets has also been found in CS responses to DNP (2) and to
REGULATION
OF SUPPRESSION
IN CS
409
antibody responses to SRBC ( 1) and phosphorylcholine (18, 19). Our present results suggest that this antagonistic relationship between Ts-aff and Ts-eff cells is not due to simply the “masking” or the contrasuppressing of Ts-eff cell activity by Ts-aff cells since separation of Ts-aff from Ts-eff containing fractions on I’. villosa lectin does not expose latent suppressive activity in the Ts-eff cell population. Therefore, it appears that Ts-aff cells prevent the generation of functional Ts-eff cells in viva Interestingly, previously published data, as well as our own studies, have shown that humoral and cell-mediated activities are regulated independently and reciprocally (7, 10, 21-23) and that immune complexes of different immunoglobulin isotypes play a mutually antagonistic role in dictating these activities by determining which regulatory T cell response is generated. Thus while IgG2a and IgG2b augment humoral response and suppress DTH, IgG 1 augments DTH and suppresses antibody production. These complex interactions may represent an important key to the signals that regulate immunity, allowing the system to “fine tune” the response. Our present results show that when a second hapten (OX) was present on molecules bearing the TNP antigenIgG2a antibody complex, Ts-aff cells specific for OX, as well as to TNP, were generated. These cells suppressed both the OX-specific CS responses and the OX-specific Ts-eff cells. We have previously shown that when TNP and OX are present on the same molecule, complexes of anti-TNP antibodies of the IgM or IgGl isotype will induce contrasuppressor T cells specific for both TNP and OX (10). We interpret these findings as supporting the notion that antibody isotypes present on the surface of haptenated APC serve to focus immunoglobulin-recognizing T cells to the site of activation. T cells that recognize immunoglobulin have been previously described (23-29) although the specificity and functional activity of these cells are still a mystery. Our results suggest that independent recognition of antibody isotypes and antigen on the surface of the APC plays a role in the generation of Tcs and Ts-aff cells (9, 7, 10) and perhaps other Ig-recognizing cells as well. Clearly, the interrelationship of Ts subsets in CS responses is extremely complex. Not only are there regulatory T cells that suppress Th 1 cells, but they also suppress other Ts cell subsets as well. While it is not clear whether the same Ts-aff cell can suppress the generation of both the Th 1 and Ts-eff cells, these Ts-aff cells appear to be identical by phenotype (Lyt I-,2’, VV adherent), parameters of induction (TNPPEC-IgG2a/b), and kinetics of appearance (they appear by Day 7 after immunization). It is not clear why the immune system requires two functionally distinct Ts cell subsets, Ts-aff and Ts-eff cells, and how and why these subsets interplay in controlling a contact sensitivity response. Studies aimed at answering these questions are in progress. ACKNOWLEDGMENTS This work was supported by The Howard Hughes Medical Institute, by Grants CA29606 and CA 16359 from the National Institutes of Health, Bethesda, Maryland, and by grants from the Polish Academy of Sciences (6.1). the Polish Anti-Cancer Program (I I .5-96). and a NIH grant (Polish-American Agreement, O5- I 15-N).
REFERENCES I. 2. 3. 4.
Green, D. R., Flood. P. M.. and Gershon. R. K., Annu. Rev. Immunol. I, 439. 1983. Miller. S. D.. Butler, L. D., and Claman, H. N., J. ItnmztnoC 129, 461. 1982. Ptak, W., Bereta, M., Ptak, M.. Iverson. G. M., and Green, D. R., J. Immund 135, 23 12, 1986. Claman, H. N., J. Immnnol. 116, 704. 1976.
410
PTAK ET AL.
5. Ptak, W., Rewicka, M., and Roiycka, D., Immunobiology 156,400, 1979. 6. Zembala, M., and Asherson, G. L., Eur. J. Immunol. 4, 799, 1974. 7. Ptak, W., Janeway, C. A., and Flood, P. M., J Immunol. 141, 756, 1988. 8. Asherson, G. L., and Ptak, W., Immunology 15, 405, 1968. 9. Kimura, A. K., Wigzell, H., Holmquest, G., Ersson, B., and Carlsson, P., J. Exp. Med. 149,473, 1979. 10. Ptak, W., Flood, P. M., Janeway, C. A., Marcinkiewicz, J., and Green, D. R., J. Zmmunol. 141, 756, 1988. 11. Marcinkiewicz, J., Bereta, M., Malinowski, J., and Ptak, W., Eur. J. Immunol. 14, 759, 1984. 12. Ptak, W., Ptak, M., Rosenstein, R. W., and Gershon, R. K., Proc. Natl. Acad. Sci. USA 79,2375, 1982. 13. Miller, S. D., Sy, M. S., and Claman, H. N., J. Immunol. 121, 274, 1978. 14. Ptak, W., and Rozycka, D., Eur. J. lmmunol. 7, 855, 1977. 15. Dorf, M. E., and Benacerraf, B., Annu. Rev. Immunol. 2, 127, 1984. 16. Germain, R. N., and Benacerraf, B., Stand. J. Immunol. 13, 1, 1981. 17. Green, D. R., Gershon, R. K., and Eardley, D. D., Proc. Natl. Acad. Sci. USA 78, 3819, 1981. 18. Bottomly, H. K., Mathieson, B. J., and Mosier, D. E., J. Exp. Med. 148, 1216, 1978. 19. Kim, B. S., and Greenberg, J. A., J. Exp. Med. 154, 809, 1981. 20. Murgita, R. A., and Vas, S. L., Immunology 22, 319, 1972. 21. Coulie, P. G., and VanSnick, J., Eur. J. Immunol. 15, 793, 1985. 22. Sinclair, N. R. S., Pharmacol. Ther. 4, 355, 1979. 23. Gershon, R. K., In “Immunoglobulin Idiotypes and Their Expression” (C. A. Janeway, Jr., E. E. Sercarz, H. Wigzell, and C. F. Fox, Eds.), pp. 467-477. Academic Press,New York, 1981. 24. Sy, M. S., Lowy, A., Hayglass, K. T., Janeway, C. A., Greene, M. I., and Benacerraf, B., Proc. Natl. Acad. Sci. USA 81,3846, 1984. 25. Flood, P. M., Janeway, C. A., and Gershon, R. K., J. Mol. Cell. Immunol. 1, 167, 1984. 26. Janeway, C. A., Murgita, R. A., Weinbaum, F. I., Asofsky, R., and Wigzell, H., Proc. Nat/. Acad. Sci. USA 74,4582, 1977. 27. Bottomly, H. K., Janeway, C. A., Mathieson, B. J., and Mosier, D. E., Eur. J. Immunol. 10, 159, 1980. 28. Rosenberg, Y. J., and Asofsky, R., Eur. J. Immunol. 11, 705, 1981. 29. Ptak, W., Flood, P. M., and Janeway, C. A., J. Mol. Cell. Immunol. 4, 281, 1990.
IMMUNOLOGY
132,400-4 10 ( 199 1)
Regulatory Responses in Contact Sensitivity: Afferent Suppressor Cells inhibit the Activation of Efferent Suppressor T Cells WLODZIMIERZ
PTAK,*
T
CHARLES A. JANEWAY, JR.,? JANUSZ MARCINKIEWICZ,* AND PATRICK M. FLOOD-~
*Department of Immunology, Copernicus Medical School, Cracow, Poland, and tDepartment oj Pathology and Howard Hughes Medical Institute, Yale University School of Medicine, New Haven, Connecticut 06520 Received June 18, 1990; accepted September 14, 1990 Two types of suppressor cells regulate the contact sensitivity (CS) response to picryl chloride (PCL). Afferent suppressor T cells (Ts-aff) inhibit the generation of CS responses to PCL, while efferent suppressor T cells (Ts-eff) inhibit the activity of Th 1 cells that mediate CS reaction. Intravenous injection of mice with TNP-substituted peritoneal exudate cells (TNP-PEC) induces Ts-elf cells that block the adoptive transfer of contact sensitivity. The induction of Ts-eff cells is prevented by the presence of Ts-aff cells, which in turn are induced by the injection of TNP-PEC coupled with antibodies of the IgG2a and IgG2b isotype (TNP-PEC-Ab). If an animal is injected with TNP-PEC prior to or simultaneously with TNP-PEC-Ab, it generates only Ts-aff cells, while if it is injected with TNP-PEC alone or TNP-PEC prior to TNP-PEC-Ab, it generates Ts-effcells. Ts-aff cells effect only the generation of Ts-eff cells, as the addition of Ts-aff cells to assays for Tseff cells has no inhibitory effect on the suppressive effects of Ts-eff cells in adoptive transfer. Our experiments show that Ts-aff cells induced by TNP-PEC-Ab are phenotypically either Lyt l+2or Lyt l-2+, but only the latter inhibit the generation of Ts-eff cells in vivo. The Ts-aff cells that inhibit Ts-eff activity adhere to the lectin Vicia villosa (VV), while Ts-eff cells are VV nonadherent. In addition, Ts-aff cells can prevent the generation of Ts-eff to linked haptens presented on the same PEC. it appears that a cascade of Ts cell interactions are involved in the regulation of CS responses. 0 1991 Academic Press, Inc.
INTRODUCTION
The quantity and quality of the immune response are determined by the activity of two functionally distinct T cell populations: suppressor T cells that eliminate responses and contrasuppressor T cells that protect responding cells from suppressor cell influences (1). The interaction of these regulatory cells forms a very complex circuit in which the activities of one subpopulation of cells may influence the activity of other subpopulations of cells. In assessingthe nature of a contact sensitivity (CS)’ response,it is crucial to determine how a particular subpopulation of cells can regulate the activity of other cells. The ’ Abbreviations used: APC, antigen-presenting cell; C, rabbit complement; CS, contact sensitivity; DNP, dinitrophenyl; iv, intravenous; Lytl cells, cells treated with anti-Lyt2 mAb + C; Lyt2 cells, cells treated with anti-Lytl mAb + C, mAb, monoclonal antibody; OX, 4-ethoxymethylene-2-phenyloxazolone; PCL, picryl chloride; PEC, peritoneal exudate cells; TNBSA, trinitrobenzene sulfonic acid; Th, T helper cells; TNP, trinitrophenyl; TNP-PEC, TNP-haptenated peritoneal exudate cells; TNP-PEC-Ab, TNP haptenated PEC treated with anti-TNP IgG2a monoclonal antibody; Ts, T suppressor cell; Ts-aff, afferent T suppressor cells; Ts-eff, efferent T suppressor cells; VV, Vicia villosa; VV+, V. villosa adherent; VV-, V. villosa nonadherent.
0008-8749/9 I $3.00 Copyright 0 1991 by Academic Press, Inc. All rights of reproduction in any form reaved
REGULATION
OF SUPPRESSION
IN CS
401
regulation of the CS response to haptens such as TNP or DNP is characterized by the activity of two distinct suppressor T cell populations: (i) afferent suppressor cells (Tsaff), which prevent CS responses by blocking the priming of cells to antigen in naive animals (3); and (ii) efferent suppressor T cells (Ts-eff) which block CS by inactivating cells previously primed to antigen (4-6). Ts-eff cells, are generated by intravenous injection of water-soluble hapten derivatives or by iv injection of haptenated spleen or peritoneal exudate (PEC) cells (4-6). Ts-aff cells, on the other hand, are generated by injection of TNP-haptenated PEC treated with anti-TNP monoclonal antibodies of the IgG2a or IgG2b isotype, which form cell surface bound immune complexes
(3, 7). We noticed that while injection of TNP-PEC-IgG2a resulted in Ts-aff cells, no Tseff cells were generated, despite the fact that large numbers of uncomplexed haptenic groups were still present on the antibody-treated TNP-PEC (7). This suggested to us that the generation of Ts-aff cells might somehow inhibit the generation of Ts-eff cells in viva. We report here that Ts-aff cells, in addition to their previously established inhibition of effector lymphocytes that mediate CS reaction (Th l), do in fact inhibit the generation of Ts-eff cells. Ts-aff cells have no effect on the functional activity of either Th 1 or Ts-eff cells once these cells are induced. MATERIALS
AND
METHODS
Mice. CBA/J male mice were purchased from Jackson Laboratories (Bar Harbor, ME) or from the breeding unit, Department of Immunology, Cracow, Poland, and were used between 6 and 8 weeks of age. Each experimental group consisted of four or five mice. Reagents. The following reagents were used: twice-recrystallized picryl chloride (trinitrophenyl (TNP) chloride; PCL) (Chemtronics, Swannanoa, NC); trinitrobenzene sulfonic acid (TNBSA) (Eastman Organic Chemicals, Rochester, NY); Sephadex G200 and Sepharose 4B (Pharmacia, Fine Chemicals, Piscataway, NJ); and 4-ethoxymethylene-2-phenyloxazolone (OX) (British Drug House, Gallard-Schlesinger Chemical Mfg. Corp., Carle Place, NY). Peritoneal exudate cells (PEC). PEC were collected 4 days after intraperitoneal injection of Mark01 52 oil. Red blood cells were lysed with NH4Cl buffer prior to haptenation. PEC were labeled with TNP and oxazolone (OX) as described previously (5). When PEC were double labeled (i.e., with TNP and OX), TNP substitution was followed by OX substitution. TNP-PEC were incubated for 30 min at room temperature with an affinity-purified anti-TNP IgG2a antibody (3 fig per 1 X lo7 cells unless otherwise stated). Cells were washed in phosphate-buffered saline and 1 X lo7 cells were injected iv into recipients. Antibodies. Monoclonal anti-Lyt 1.1 and anti-Lyt2. I antibodies were kindly provided by Dr. F. W. Shen, Sloan-Kettering Institute, New York. Monoclonal anti-TNP IgG2a (32.1) hybridoma cells were obtained from Dr. Julian Fleischman, Washington University School of Medicine, St. Louis, Missouri. Antibodies were affinity purified by absorbtion onto a TNP-KLH-Sepharose 4B affinity column and were eluted with 3 A4 guanidine. Treutment of regulatory cefls. Spleen cells from mice immunized with TNP-PEC or TNP-PEC-IgG2a (TNP-PEC-Ab) were treated with the anti-Lyt 1.1 or the antiLyt2.1 antibody for 30 min at room temperature, washed, and then incubated for 30
402
PTAK
ET AL.
min at 37°C in rabbit complement (C) (prescreened for low natural cytotoxicity to mouse spleen cells). Animals were injected with the equivalent of 7 X 10’ untreated spleen cells. Active sensitization and adoptive transfer of contact sensitivity. Animals were skin sensitized with PCL or OX by painting the shaved abdomen as well as four feet with 0.15 ml of 5% PCL or 3% OX (in 3:l ethanoLacetone) and were tested for contact sensitivity by application of 0.8% (w/v) PCL or OX in olive oil to both ears 4 days later (8). Ear swelling was assessedby measuring the earsimmediately before application and 24 hr later using an engineer’s micrometer. Results are expressedin units of 10e3 cm + SD. In each experiment, ears of separategroups of nonimmunized controls were similarly challenged and measured. The background swelling response in the control group averaged about 1 to 2 units and was subtracted from the swelling response of the experimental animal group so the results are presented as a net increase in ear swelling. In adoptive transfer, 7 X 10’ spleen and lymph node cells of animals skin painted 4 days previously with PCL or OX were injected iv into recipients that were ear challenged immediately, and ear swelling was measured 24 hr later. In results, group A always represents the positive untreated control group. In the afferent suppressor cell assay,this group representsthe group that was only contact sensitized with PCL or OX and received no regulatory cells. In the efferent suppressor cell assay,this group received only cells from animals skin painted 4 days previously with PCL or OX. In all figures, results are also expressed as percentage responses,which are calculated according to the formula: (experimental - negative control)/(positive control - negative control) X 100. Testfor efirent suppression (Ts-ef). Spleen and lymph node cells (7 X IO’) from animals immunized by skin painting were mixed with 5 X 10’ spleen cells from mice injected 7 days previously with hapten-substituted PEC. This cell mixture was incubated for 30 min at 37°C and was then injected into naive recipient mice. Animals were challenged immediately on the ears and were assayed for contact sensitivity 24 hr later (5). Testfor a&rent suppression (Ts-af). Spleen cells (5 X 10’) from mice being assayed for Ts-aff activity were injected iv into naive recipient mice, which were immediately sensitized by skin painting with PCL or OX. Animals were ear challenged 4 days later and contact sensitivity was assessedat 24 hr (3, 7). Separation of cells on the Vicia villosa (VV) lectin. VV was prepared as described previously (9). Separation of cells into V. villosa adherent (VV’) and nonadherent (VV) populations was done as described in detail previously (3, 7). Generally, the lectin adherent cells represent 8-20% of the immune spleen cell population. Statistics. Evaluation of the statistical significance of the results was done using the double-tailed Student’s t test. RESULTS Ts-eflcells are generated by TNP-PEC only under conditions in which Ts-afcells are not generated. Earlier reports had shown that iv injection of TNP-PEC cells led to the production of Ts-eff cells, while pretreatment of these TNP-PEC with anti-TNP mAb of the IgG2a or IgG2b isotype leads to only Ts-aff cells (3, 7). We examined the relationship between Ts-aff generation and Ts-eff inhibition by titrating the amount of IgG2a antibody used to coat the TNP-PEC (Fig. 1). TNP-labeled PEC (10’) were
REGULATION
TNPPEC
An,,bady
Afferent
Suppressor
Cell
OF SUPPRESSION
IN CS
Assay
Efferent
403 Suppressor
Cell
Assay
-I (100) ( 91) (
22)
( 25) ( 39) ( ‘51) ( 84)
(Units
of swelling
x IO
-3
cm)
FIG,. 1. Titration of IgG2a monoclonal anti-TNP antibody needed to induce Ts-aff cell activity when incubated with TNP-PEC. TNP-PEC (I X IO’) from CBA mice were incubated with the indicated amount of affinity-purified anti-TNP-IgG2a mAb before iv injection into syngeneic mice. Six days later the spleen cells were assayed for activity. The assay for Ts-aff consisted of iv injection of 5 X IO’ spleen cells from TNP-PEC+-Ab recipient animals into naive syngeneic mice and skin painting these mice immediately with PCL. Contact sensitivity to PCL was measured 4 days later. The assay for Ts-eff ceils consisted of combining 5 X 10’ spleen cells from TNP-PEC+-Ab recipient animals with 7 X IO’ spleen and lymph node cells from mice skin sensitized with PCL 4 days earlier. These cell mixtures were incubated for 30 min at 37°C then injected iv into naive recipients that were then immediately challenged with PCL, and the ear swelling (DTH) reaction was measured 24 hr later. In this and in the following figures, nonspecific ear swelling of negative control (nontreated) mice that were similarly challenged with antigen was subtreacted from experimental values (see Materials and Methods). Thus, results are presented as the net increase in ear swelling in units X lo-’ cm f SD. In no case did the injection of TNP-PEC or TNP-PEC-Ab induce a significant contact sensitivity reaction by itself. In Student’s t test. in the afferent suppressor cell assay, group B was significantly different than groups C, D, and E (P < 0.00 1) and F (P < 0.0 I); while in the efferent suppressor cell assay. group A was significantly different than groups B and G (P < 0.00 I) and group F (P < 0.01).
incubated with various doses of affinity-purified anti-TNP IgG2a mAb, starting at 20 pg to as low as 0.03 pg, and then injecting these TNP-PEC-Ab complexes into naive mice. Ts-aff cell activity or Ts-eff activity was assayed as described under Materials and Methods. The results in Fig. 1 show that IO7 TNP-PEC treated with anti-TNP IgG2a mAb with doses as low as 0.8 PLgwere effective in generating Ts-aff but not Tseff cells, while a dose of 0.03 pg did not generate Ts-aff but did generate Ts-eff cells. A dose of 0.8 PLgof IgG2 mAb represents approximately 3 X 10” antibody molecules and, as shown previously (3) lo7 PEC are substituted with approximately 3 X lOI TNP haptenic groups. Therefore, assuming that all anti-TNP mAb molecules bind to a TNP haptenic group on the surface of the APC, our results show that saturation of approximately 1% of the TNP molecules on the surface of the TNP-haptenated PEC with antibody completely prevents the generation of Ts-eff cells. Interestingly, TNPPEC treated with anti-TNP IgG2a mAb at a dose of 0.15 PLggenerated intermediate activities of both Ts-aff and Ts-eff cells. These results suggest a direct correlation between the activation of Ts-aff activity and the inhibition of Ts-eff cell activity. Ts-afcells inhibit the generation of Ts-efcells. In order to test whether the lack of Ts-eff activity in TNP-PEC-Ab treated mice was simply due to TNP-PEC-Ab not activating Ts-eff or whether this treatment actually blocked Ts-eff activation, we injected TNP-PEC, TNP-PEC-Ab, or equal numbers of both iv into naive animals and assayed their spleen cells for both Ts-aff and Ts-eff. The results in Fig. 2 show that while TNP-
404
PTAK Induction Regulatory
of Cells Afferent
GrOUD
TNP-PEC
ET AL.
Suppressor
Cell Assay
Efferent Suppressor
Cell Assay
TNP-PEC-Ab
2
0
4
10
8
6
(Units
0
of Swelling
4
2
6
8
x id 3 cm)
FIG. 2. Injection of TNP-PEC incubated with IgG2a anti-TNP monoclonal antibody prevents the induction of Ts-eff cells. CBA mice were injected iv with 1 X 10’ TNP-PEC incubated in BSS, I X 10’ TNP-PEC incubated before injection with 3 pg of IgG2a anti-TNP monoclonal antibody, or a mixture of both (IO’ each). Seven days later, cells of these mice were tested for afferent and efferent suppressor cell activity as described in Fig. 1. In Student’s t test in the afferent suppressor cell assay, group B was significantly different than groups C and D (P < 0.001); while in the efferent suppressor cell assay, group A was significantly different than group B (P < 0.001).
PEC induced only Ts-eff cells, TNP-PEC-Ab induced only Ts-aff cells. Figure 2 also shows that an equal number of TNP-PEC and TNP-PEC-Ab injected iv simultaneously resulted in the exclusive generation of Ts-aff cells with no Ts-eff cells. In the next experiment (Fig. 3) we examined whether the injection of TNP-PECIgG2a prior to or subsequent to TNP-PEC inhibited Ts-eff cell activity. Naive mice were injected on Day 0 with either TNP-PEC-Ab or TNP-PEC, and again on Day 6.
Treatment of donors of regulatory cells
Afferent
Suppressor
Efferent Suppressor
Cell Assay
Cell Assay
Group A e c 0 E F G H 0
2
4
6
(Units
8
10
of swelling
2
0
x IO
4
6
8
-3 cm)
FIG. 3. Ts-eff cell activity generated prior to Ts-aff cell induction is not suppressed by Ts-aff cells. CBA mice were injected with 1 X 10’ TNP-PEC or 1 X 10’ TNP-PEC incubated in 3 ng IgG2a anti-TNP monoclonal antibody on Day 0 as indicated. On Day 6, recipient mice received an injection with the identical dose of either TNP-PEC or TNP-PEC incubated with 3 pg of IgG2a anti-TNP monoclonal antibody as indicated. The group marked with an asterisk was injected on Day 2 rather than Day 6. On Day 12, spleen cells of these mice were tested for the presence of afferent and efferent suppressor cells as described in Fig. 1. In Student’s t test, in the afferent suppressor cell assay, group A was significantly different than groups B, C, E, F, and H (P < 0.001); while in the efferent suppressor cell assay, group A was significantly different than groups D, E and G (P < 0.00 I ) and group F (P < 0.05).
REGULATION
OF SUPPRESSION
405
IN CS
On Day 12 the mice were tested for Ts-aff and Ts-eff activity. Our results indicate that injection of TNP-PEC-Ab induced good Ts-aff activity (group C) but not Ts-eff activity, even when animals were subsequently injected with TNP-PEC (group B). In contrast, animals injected with TNP-PEC prior to injection with TNP-PEC-Ab (group E) showed both Ts-aff and Ts-eff activity when assayed. Interestingly, injection of TNP-PEC-Ab only 2 days after TNP-PEC significantly inhibited the generation of Tseff cells (group F). These results suggest that inhibition of Ts-eff cell activity was not due to a lack of APC for Ts-eff activation. They demonstrate rather that procedures that activate Ts-aff cells inhibit the generation of functional Ts-eff cells when performed before, at the same time, or shortly after procedures that activate Ts-eff cells. Ts-affcells do not act by masking Ts-efactivity. However, to rule out the possibility that Ts-aff cells that inhibit Ts-eff generation were functioning by masking or “contrasuppressing” the activity of Ts-eff cells, we took advantage of the fact that contrasuppressor T cells and Ts-aff cells adhere to the lectin V. villosa while Ts-eff cells do not (7, 10). We injected mice iv with TNP-PEC, TNP-PEC-Ab, or a mixture of TNPPEC and TNP-PEC-Ab, separated spleen cells of each group on the basis of their adherence to V. villosa, and then assayed cell fractions for Ts-aff and Ts-eff cell activity. The results in Fig. 4 show that TNP-PEC-injected animals produced only Ts-eff cells and that these cells were V. villosa nonadherent (group A). Animals injected with TNP-PEC-Ab produced only Ts-aff cells (group B) and these cells, in contrast to Tseff cells, were V. villosa adherent. Animals injected with equal numbers of TNP-PEC and TNP-PEC-Ab produced only Ts-aff cell activity, even when separated into V. villosa adherent and nonadherent fractions (group C). These results indicate that the generation of Ts-aff cells does not mask Ts-eff cell activity, but rather inhibits the generation of Ts-eff cells. Ts-a$‘cells do not suppress Ts-efcell activity. We then tested whether Ts-aff cells inhibited the activity of previously activated Ts-eff cells (Fig. 5). Ts-eff cells significantly suppressed the adoptive transfer of contact sensitivity by PCL-immune cells, while Ts-aff cells had no effect (groups B and C). When Ts-aff and Ts-eff cells were mixed
B
TNP~PEC-Ab
C
TNP-PEC + TNP-PEC-Ab
( 92) (100) ( 98) ( 401 ( 39) ( 891 ( 421 ( 341 ( 92) 0
2
4
6
El
(Units
of
IO swelling
0 x IO
2
4
6
8
-3 cm)
FIG. 4. Inhibition of Ts-etf cell activity by induction of Ts-aff cell is not due to the “masking” of Ts-eff cell activity by Ts-aff cells. CBA mice were injected iv with I X IO’ TNP-PEC, I X IO’ TNP-PEC incubated in 3 pg of IgG2a anti-TNP monoclonal antibody, or a mixture of 5 X 10” TNP-PEC from both the treated and the untreated groups as in Fig. 2. Seven days later, spleen cells of these mice were separated on V. vi//osa (VV) lectin plates into VV+ and VV populations, and each fraction (equivalent of 7 X 10’ cells) was tested for afferent and efferent suppressor cell activity as described in Fig. 1. In Student’s t test, group A was significantly different than groups B and C (P < 0.001).
406
PTAK
ET AL.
PCI immune cells Incubated with: FUSl
Se3md
2
0 (Units
4
6
of swelling
x 10
8
-3
cm)
FIG. 5. Ts-aff cells do not inhibit the functional activity of Ts-eff cells. CBA mice were injected with 1 X 10’ TNP-PEC to generate Ts-eff cells or with 1 X IO’ TNP-PEC incubated with 3 fig of IgG2a anti-TNP monoclonal antibody to generate Ts-aff cells. Six days later, 5 X 10’ spleen cells from these mice were incubated for 60 min at 37°C with 7 X IO’ spleen and lymph node cells from animals skin painted 4 days earlier with PCL (groups B and C). Cell mixtures were washed and incubated for a second 60-min period with 5 X IO7 spleen cells exhibiting Ts-aff or Ts-eff activity (groups E and F). Group D: PCL-immune cells were incubated with 5 X IO’ spleen cells from TNP-PEC-immunized animals simultaneously with 5 X IO’ spleen cells from TNP-PEC-Ab-immunized animals. In Student’s t test, group A was significantly different than groups B and D-F (P < 0.00 I).
together prior to addition to PCL-immune cells, the suppressive activity of Ts-eff cells was retained (group D). Preincubation of PCL-immune cells with Ts-aff cells likewise did not prevent the suppression of PCL immune cells by Ts-eff cells (group E). The subsequent addition of Ts-aff cells did not reverse the suppressive effects of Ts-eff cells on PCL-immune cells (group F). Therefore, it appears that Ts-aff cells have no effect on the activity of mature Ts-eff cells. Lytl-2+ but not the LytS2- Ts-afcells inhibit Ts-eflcell generation. We phenotyped the cells from the Ts-aff population that inhibited the generation of functional Ts-eff
Group A s c 0
2
0 (Units
4 of swelling
6
8
x 16 3 cm)
FIG. 6. Only Lytl-,2+ Ts-aff cells inhibit the generation of Ts-eff cells. CBA mice were injected with 1 X IO’ TNP-PEC treated with 3 pg of IgG2a anti-TNP monoclonal antibody and their spleen cells were removed 7 days later. Spleen cells were treated with anti-Lytl mAb (Lyt2 cells) or anti-Lyt2 mAb (Lytl cells) mAb + C, and 7 X 10’ splenic cell equivalents were then injected iv into syngeneic recipient mice. Recipients simultaneously received iv injections of I X 10’ TNP-PEC, and the activity of Ts-eff cells was tested 6 days later as detailed in Fig. 1. In Student’s t test, group A was significantly different than groups B and G (P < 0.001).
REGULATION
OF SUPPRESSION IN CS
407
cells (Fig. 6). Ts-aff cells were generated in naive CBA mice by iv injection of TNPPEC-Ab, and their spleen cells were transferred into naive recipients. These recipients were then injected with TNP-PEC in order to generate Ts-eff cells. While naive mice first injected with Ts-aff cells did not develop Ts-eff cells when injected with TNPPEC (group C), and treatment of the same Ts-aff cells with anti-Lytl mAb + C before transfer (Lyt 2+ cells) still blocked Ts-eff cell production (groups D and E), treatment of Ts-aff cells with anti-Lyt2 mAb + C before transfer (Lyt l+ cells) allowed the generation of Ts-eff cells (groups F and G). Therefore, even though the Ts-aff cells which inhibit CS reactions have been phenotyped as both Lytl+,2- and Lytl-,2+ (7) only Lyt l-,2+ Ts-aff cells inhibit the generation of Ts-eff cells.
TNP-specijc Ts-ajfcelIs prevent the getieration of oxazolone-spec$c Ts-eHee1f.s on/~. if‘ TNP and OX are linked on the same carrier. We tested the antigen specificity of inhibition of Ts-eff cell generation by measuring the inhibition of OX-specific Ts-eff cell generation by TNP-specific Ts-aff cells. Mice were injected with a mixture of TNPPEC and OX-PEC or with PEC that were colabeled with both haptens (TNP-OXPEC). The results in Fig. 7 show that mice injected with both TNP-PEC and OXPEC (group B) or mice injected with TNP-OX-PEC (group C) generated Ts-eff cells specific for both TNP and OX. When TNP-PEC were treated with anti-TNP IgG2a mAb and coinjected with OX-PEC (group B), only OX-specific Ts-eff cells could be detected. These results demonstrate that the activity of Ts-eff cell generation is antigen specific since TNP-PEC-Ab injection inhibits only TNP-Ts-eff cell generation and not OX-Ts-eff cell generation. Interestingly, injection of TNP-OX-PEC treated with antiTNP IgG2a mAb (group E) inhibits the generation of Ts-eff cells specific for either TNP or OX. These results parallel those obtained earlier for Ts-aff cells which inhibit CS reactions (see Ref. (7)) and suggest that the immune complexes bound to the surface of an APC also induce Ts-aff cells to other haptenic determinants, provided these determinants are present on the same PEC as the antigen-antibody complex. DISCUSSION The suppression of the CS response to PCL in normal adult mice is mediated by at least two distinct populations of suppressor T cells. These two cell types can be distinguished both on the basis of functional activity and by their adherence to the lectin I’. viffosa (7). The first population consisted of both Lyt 1+,2- and Lyt-2+ T cells that adhere to VV and that suppress the generation of, but not the effector phase of, a CS response (7). These cells, termed Ts-aff cells, have also been described in a different antigenic system (2, 11). The second population of Ts cells does not adhere to VV (7). These Ts-eff cells are Lyt l-2+ (12) and suppress only the effector phase and not the induction phase of a CS response (3, 5, 6). In addition to the functional difference between these two cell populations, the conditions that activate Ts-aff and Ts-eff cells are distinct. It has previously been found that the type of Ts cells induced by antigenic challenge can depend on the chemical nature of the tolerogen used (e.g., DNBSA vs TNBSA (2-6) the epitope density of the hapten-modified cell (3-5, 13), the dose of tolerogen used to induce suppressor cells (2, 4) the time after antigenic challenge (2, 3, 7), and whether the hapten-modified cells are syngeneic or allogeneic (3-5, 13). In the TNP system, however, Ts-eff cells are generated by intravenous injection of tolerizing doses of TNBSA or TNP-conjugated splenic or peritoneal exudate APC (3, 5-7) while Ts-aff cells are generated by TNP-conjugated IgG (3) or TNPPEC treated with anti-TNP mAb of the IgG2a or IgG2b isotype (7).
408
PTAK
Ofl”d”C,lO” GP3”PMD& OfTs-effCells
ET AL.
ox
PCI
A
NOcellsadded
B
TNP-PEC +OX-PEC
( 18)
c
TNP-OX-PEC
( 39)
(100)
D TM-PEC-Ab +OX-PEC
E
(
TNP-OX-PEC-Ab 2
4
6
8
10
iL 0
(100)
( 0) (25) ( 10) (88)
82)
( 90)
cc 0
--I
1
2
4
G
8
(Units of swelling x 10. 3 cm) RG. 7. Anti-TNP IgG2a monoclonal antibodies on the surface of TNP-PEC prevent the development of Ts-eff to the OX hapten if TNP and OX are on the same PEC. CBA mice were injected iv with 1 x IO7 TNP-PEC, OX-PEC, or with PEC that were double labeled with TNP and OX (TNP-OX-PEC) incubated prior to injection with 3 pg of IgG2a anti-TNP monoclonal antibody. Seven days later, Ts-eff cell activity was examined in injected animals by mixing 5 X 10’ of their spleen cells with 7 X 10’ spleen and lymph node cells from PCL- or OX-immune animals and injecting these cell mixtures into naive recipients. Recipient mice were then immediately challenged with the immunizing hapten, and the adoptive response was measured 24 hr later. In Student’s t test, OX group B was significantly different than group E (P < 0.001).
Miller et al. (2) have shown that Ts-eff and then Ts-aff cells are induced by an intravenous injection of syngeneic DNP-coupled lymphoid cells, with Ts-eff cells appearing around Day 7 followed by Ts-aff cells around Day 14. Ts-aff cells can in fact be induced by infusion of Ts-eff cells, suggesting a cellular circuit in which Ts-eff cells are first induced by antigenic challenge, followed by a wave of Ts-aff cells. Our results, both those presented here and those reported earlier (3, 7), suggest two more factors that influence both Ts-aff and Ts-eff cell generation: (i) the functional cell type that is initially activated, since while Ts-eff cell activation may favor the subsequent activation of Ts-aff cells (2), Ts-aff cell activity does not allow the generation of Ts-eff cells; and (ii) the nature of the humoral response to the hapten, in that hapten presented to the immune system in the absence of antibody responses favors Ts-eff cell generation (3, 7), while hapten complexed to IgG immunoglobulins (11, 14) or macrophage-bound immune complexes of the IgG2a or IgG2b isotype favor Ts-aff cell generation (3, 7). The mechanism by which Ts-aff cells block activation of Ts-eff cells is not yet clear, and because our experiments demonstrate only that cell populations containing Tsaff cells prevent the generation of functional Ts-eff cells, the cell interactions leading to this effect may be highly complex. A number of systems that exhibit feedback inhibition of suppressor T cell activity have been described ( 1, 15, 16). Initially, it was found by Gershon and co-workers that T suppressor-effector cells active in regulating SRBC-specific antibody responses in vitro suppress Ts-inducer cells ( 17). Ts-aff cells likewise seem to block the induction of Ts-eff activity since mature Ts-eff cells fail to be suppressed by Ts-aff cells. This model suggests that Ts-aff cells prevent both the generation of CS responses and the generation of regulatory cells which would suppress it, thereby conserving immunologic energy by preventing an immunoregulatory response that is not needed. We have never found in the TNP system that Ts-eff cell generation preceeds and is required for Ts-aff cell activity (W. Ptak, unpublished results), although it is clear from our results and those of others (2) that a mutually antagonistic relationship exists between Ts-aff and Ts-eff cells. Such interregulation of T suppressor cell subsets has also been found in CS responses to DNP (2) and to
REGULATION
OF SUPPRESSION
IN CS
409
antibody responses to SRBC ( 1) and phosphorylcholine (18, 19). Our present results suggest that this antagonistic relationship between Ts-aff and Ts-eff cells is not due to simply the “masking” or the contrasuppressing of Ts-eff cell activity by Ts-aff cells since separation of Ts-aff from Ts-eff containing fractions on I’. villosa lectin does not expose latent suppressive activity in the Ts-eff cell population. Therefore, it appears that Ts-aff cells prevent the generation of functional Ts-eff cells in viva Interestingly, previously published data, as well as our own studies, have shown that humoral and cell-mediated activities are regulated independently and reciprocally (7, 10, 21-23) and that immune complexes of different immunoglobulin isotypes play a mutually antagonistic role in dictating these activities by determining which regulatory T cell response is generated. Thus while IgG2a and IgG2b augment humoral response and suppress DTH, IgG 1 augments DTH and suppresses antibody production. These complex interactions may represent an important key to the signals that regulate immunity, allowing the system to “fine tune” the response. Our present results show that when a second hapten (OX) was present on molecules bearing the TNP antigenIgG2a antibody complex, Ts-aff cells specific for OX, as well as to TNP, were generated. These cells suppressed both the OX-specific CS responses and the OX-specific Ts-eff cells. We have previously shown that when TNP and OX are present on the same molecule, complexes of anti-TNP antibodies of the IgM or IgGl isotype will induce contrasuppressor T cells specific for both TNP and OX (10). We interpret these findings as supporting the notion that antibody isotypes present on the surface of haptenated APC serve to focus immunoglobulin-recognizing T cells to the site of activation. T cells that recognize immunoglobulin have been previously described (23-29) although the specificity and functional activity of these cells are still a mystery. Our results suggest that independent recognition of antibody isotypes and antigen on the surface of the APC plays a role in the generation of Tcs and Ts-aff cells (9, 7, 10) and perhaps other Ig-recognizing cells as well. Clearly, the interrelationship of Ts subsets in CS responses is extremely complex. Not only are there regulatory T cells that suppress Th 1 cells, but they also suppress other Ts cell subsets as well. While it is not clear whether the same Ts-aff cell can suppress the generation of both the Th 1 and Ts-eff cells, these Ts-aff cells appear to be identical by phenotype (Lyt I-,2’, VV adherent), parameters of induction (TNPPEC-IgG2a/b), and kinetics of appearance (they appear by Day 7 after immunization). It is not clear why the immune system requires two functionally distinct Ts cell subsets, Ts-aff and Ts-eff cells, and how and why these subsets interplay in controlling a contact sensitivity response. Studies aimed at answering these questions are in progress. ACKNOWLEDGMENTS This work was supported by The Howard Hughes Medical Institute, by Grants CA29606 and CA 16359 from the National Institutes of Health, Bethesda, Maryland, and by grants from the Polish Academy of Sciences (6.1). the Polish Anti-Cancer Program (I I .5-96). and a NIH grant (Polish-American Agreement, O5- I 15-N).
REFERENCES I. 2. 3. 4.
Green, D. R., Flood. P. M.. and Gershon. R. K., Annu. Rev. Immunol. I, 439. 1983. Miller. S. D.. Butler, L. D., and Claman, H. N., J. ItnmztnoC 129, 461. 1982. Ptak, W., Bereta, M., Ptak, M.. Iverson. G. M., and Green, D. R., J. Immund 135, 23 12, 1986. Claman, H. N., J. Immnnol. 116, 704. 1976.
410
PTAK ET AL.
5. Ptak, W., Rewicka, M., and Roiycka, D., Immunobiology 156,400, 1979. 6. Zembala, M., and Asherson, G. L., Eur. J. Immunol. 4, 799, 1974. 7. Ptak, W., Janeway, C. A., and Flood, P. M., J Immunol. 141, 756, 1988. 8. Asherson, G. L., and Ptak, W., Immunology 15, 405, 1968. 9. Kimura, A. K., Wigzell, H., Holmquest, G., Ersson, B., and Carlsson, P., J. Exp. Med. 149,473, 1979. 10. Ptak, W., Flood, P. M., Janeway, C. A., Marcinkiewicz, J., and Green, D. R., J. Zmmunol. 141, 756, 1988. 11. Marcinkiewicz, J., Bereta, M., Malinowski, J., and Ptak, W., Eur. J. Immunol. 14, 759, 1984. 12. Ptak, W., Ptak, M., Rosenstein, R. W., and Gershon, R. K., Proc. Natl. Acad. Sci. USA 79,2375, 1982. 13. Miller, S. D., Sy, M. S., and Claman, H. N., J. Immunol. 121, 274, 1978. 14. Ptak, W., and Rozycka, D., Eur. J. lmmunol. 7, 855, 1977. 15. Dorf, M. E., and Benacerraf, B., Annu. Rev. Immunol. 2, 127, 1984. 16. Germain, R. N., and Benacerraf, B., Stand. J. Immunol. 13, 1, 1981. 17. Green, D. R., Gershon, R. K., and Eardley, D. D., Proc. Natl. Acad. Sci. USA 78, 3819, 1981. 18. Bottomly, H. K., Mathieson, B. J., and Mosier, D. E., J. Exp. Med. 148, 1216, 1978. 19. Kim, B. S., and Greenberg, J. A., J. Exp. Med. 154, 809, 1981. 20. Murgita, R. A., and Vas, S. L., Immunology 22, 319, 1972. 21. Coulie, P. G., and VanSnick, J., Eur. J. Immunol. 15, 793, 1985. 22. Sinclair, N. R. S., Pharmacol. Ther. 4, 355, 1979. 23. Gershon, R. K., In “Immunoglobulin Idiotypes and Their Expression” (C. A. Janeway, Jr., E. E. Sercarz, H. Wigzell, and C. F. Fox, Eds.), pp. 467-477. Academic Press,New York, 1981. 24. Sy, M. S., Lowy, A., Hayglass, K. T., Janeway, C. A., Greene, M. I., and Benacerraf, B., Proc. Natl. Acad. Sci. USA 81,3846, 1984. 25. Flood, P. M., Janeway, C. A., and Gershon, R. K., J. Mol. Cell. Immunol. 1, 167, 1984. 26. Janeway, C. A., Murgita, R. A., Weinbaum, F. I., Asofsky, R., and Wigzell, H., Proc. Nat/. Acad. Sci. USA 74,4582, 1977. 27. Bottomly, H. K., Janeway, C. A., Mathieson, B. J., and Mosier, D. E., Eur. J. Immunol. 10, 159, 1980. 28. Rosenberg, Y. J., and Asofsky, R., Eur. J. Immunol. 11, 705, 1981. 29. Ptak, W., Flood, P. M., and Janeway, C. A., J. Mol. Cell. Immunol. 4, 281, 1990.