Superantigen presentation by human T cell clones
Eur. J. Immunol. 1992. 22: 2033-2039
Roberto Nisini, Paolo M. Matricardi, Andrea Fattorossi, Roberto Biselli and Raffaele D’Amelio D.A.S.R.S. Rep. Medicina, Lab. Immunologia, Pomezia Roma
2033
Presentation of superantigen by human T cell clones: a model of T-T cell interaction Superantigens (SAg) interact withT lymphocytes bearing particular vj sequences as part of their T cell receptor (TcR). The interaction, however, requires the presence of major histocompatibility complex (MHC) class I1 molecules on antigen-presenting cell (APC). In peculiar circumstances, MHC class 11+T cell clones (TCC) have been shown to present peptides and selected antigens interacting with antigen-specific TCC in the absence of APC. In this report we studied the capacity of SAg to mediate a T-T cell interaction, investigating the TCC ability to present a panel of staphylococcal enteroxins (SE) independently of the presence of added APC. Upon exposure to a broad range of SE concentrations, MHC class IIf TCC showed an intense proliferative response even in the absence of professional APC. Diverse SE optimally stimulated responder TCC at different concentrations. The proliferation was inhibited by anti-DR monoclonal antibodies, both in the presence and in the absence of APC.The SE activation of TCC in the absence of APC induced the same series of phenotypic variations as that observed following the TCC stimulation with APC. Irradiated TCC efficiently presented membrane-bound SE to responder TCC as well as professional APC. These results show that a single cell of a given clone effectively presents the SE to other cells of the same clone, and provide evidence that SAg can efficiently mediate T-T cell interaction. In addition, the possibility also exists that one cell of the clone can actually undergo an auto-stimulation via SAg-mediated interactions between its own TcR and MHC class I1 molecule. It has recently been suggested that the Vb-selective depletion of T cells observed in acquired immunodeficiency syndrome (AIDS) patients might be a consequence of the interaction between a human immunodeficiency virus (H1V)-encoded SAg and T cells expressing a SAg complementary V,.We suggest that the hypothesized HIV-encoded SAg might mediate T-T cell interactions that could play a relevant role in the Vpelective depletion of T lymphocytes observed in HIV-infected patients.
1 Introduction
maternally transmitted mouse retroviral genes encode SAg responsible for an intrathymic deletion of discrete sets of Staphylococcal enterotoxins (SE) and mouse retroviral- T cells bearing certain Vp families of TcR [20-231. Conseencoded proteins such as minor lymphocyte stimulating quently, the activation/depletion of T cells expressing a (MIS) gene products, have been defined as superantigens particular Vp is diagnostic for the presence of SAg [24].The (SAg), because of their unusual MHC-TcR binding and recent observation that the CD4+ T cell depletion in AIDS consequent T cell activaton [ll.All SAg share the ability to patients is selective for certain Vp families [25] led to bind to human and mouse MHC class 11proteins [2-61 and speculations on the possible pathogenetic role of a hypothe SAg-MHC complex shows specificity for a peculiar part thetical HIV-encoded SAg. It has also been hypothesized of theTcR,Vp [3, 71. SAg activateT cells irrespective of their that upon activation, HIV-infected CD4+ T cells would antigen specificity, via a selective interaction with mono- express MHC class I1 molecules together with retroviral morphic regions of MHC class I1 molecules on APC and SAg and “interact” [24] with other CD4+ TcR+ Tcells with v j regions of TcR on T cells [8-141: it has been expressing particular y3, and that this interaction could be suggested that the mitogenic activity of SAg probably responsible for a selective T cell depletion. According to resides in their three-dimensional structure, able to stabil- this hypothesis, the HIV-encoded SAg should share most of ize the TcR-MHC class I1 binding [lo, 15-17]. It has also the properties of other known SAg.The presentation of the been shown in mice that in vivo administration of SE [18]or HIV-encoded SAg might have different functional conseMls-disparate cells [19] leads first to an activation and then quences depending on whether the presentation occurs in to a depletion of CD4+ cells expressing the SAg-comple- the thymus or in the periphery and/or on the type of mentary vj. Furthermore, rccent data have shown that SAg-presenting cells. Activated humanT lymphocytes and T cell clones (TCC) express MHC class I1 molecules [26] and elegant experiments have shown that TCC act as accessory cells in presenting antigens upon fulfillment of [I 103241 certain conditions. For example, TCC can process and Correspondence:Roberto Nisini. Lab. Immunologia, Rep. Medi- present those antigens that are targeted to their own surface cina, DASRS, Alto Pratica di Mare. 1-00040 Pomczia Roma. receptors, thus bypassing thcir lack of uptake capacity [27, 281. TCC can also effectively present antigens that, by Italy binding directly to MHC class I1 molecules, do not need to Abbreviations: SAg: Superantigen SE: Staphylococcal entero- be internalized and processed, as is the case for peptidetoxins TCC: T cell clones. sized antigens [29]. 0 VCH Vcrlagsgesellschaft mbH, D-6940 Weinheim, 1992
0014-2980/92/0808-2033$3.50+ .25/0
2034
R. Nisini, P. M . Matricardi, A. Fattorossi et al.
In this work, we used a panel of SE to study T-Tcell interaction. We observed that tetanus toxoid (TT)-specific MHC class 11+ CD4+ TCC are able to present SE and to proliferate even in the absence of additional APC, indicating that SAg can mediate an effective T-T cell interaction. The possibility is envisaged that a T-T cell interaction mediated by the hypothesized HIV-encoded SAg could take place and possibly play a relevant role in determining the selective depletion of V, families of TcR-bearing T lymphocyte observed in AIDS patients.
2 Materials and methods 2.1 Antigens Staphylococcal enterotoxin A (SEA), B (SEB), C1 (SECI), C2 (SEC2), D (SED) and E (SEE) were purchases from Serva Fine Biochemicals (New York, NY). TT was kindly provided by Dr. A. Podda, (Sclavo SPA, Siena, Italy)
2.2 Media RPMI 1640 (Flow Laboratories, Irvine, Scotland) supplemented with 100 U/ml penicillin, 100 mg/ml streptomycin, 2 mM L-glutamine, 1 mM sodium pyruvate, 1% nonessential amino acids 5 x lo-' M 2-ME and 10% FCS (Gibco Lab., Grand Island, NY) was used for culturing EBVtransformed lines and for proliferation assays (medium B), whilc 2-ME and FCS were substituted by 50U/ml IL-2 (kindly provided by Hoffmann-La Roche, Inc. (Basel, Switzerland) and 5 % autologous serum for culturing TCC (medium T). 2.3 Cloned human TT-reactive CD4+ T lymphocytes Twenty-five milliliters of heparinized blood was obtained from a normal donor (PM) recently vaccinated with TT (Anatetall, Sclavo SPA, Siena, Italy) and PBMC were purified on a density gradient (Lymphoprep, Nycomed Pharma AS, Oslo, Norway). Cells were resuspended at 1O6/rnlin medium T without IL-2, and 100 or 200 pl were dispensed in 96-well flat-bottom plastic plates (Costar, Cambridge, MA). TTwas added at 20 pg/ml (final concentration) to half of the cultures. After 6 days, 2 or 4 U/well TL-2 were added to the cultures. After 4 additional days, cultures showing a significant cell growth were considered positive. Cells were counted and cloned by limiting dilution at 3, 1 and 0.3 cells/well in medium T supplemented with 5 x lo5 irradiated PBMC and 1 pg/ml PHA (Wellcome Diagnostics, Dartford, GB). After 10-15 days, growing cultures wcre expanded in medium Tand finally tested for TT specificity in a proliferation assay using irradiated autologous PBMC pre-pulsed or not with TT at 20 pg/ml. TT-specific clones were expanded and maintained in culture with 25-35-day cycles of restimulation with PHA and irradiated PBMC. Clonality was assessed by a subcloning test that showed 100 % TTspecificity of growing subcloned cells. TT-rpecific CD4+ TCC (PM 1.1, PM 1.13, PM 1.16, PM 1.20, PM 3.11) screened for responsiveness to SE in the presence of PBMC as APC, were used for proliferation experiments.
Eur. J. lmmunol. 1992. 22: 2033-2039
2.4 CD8+ TCC A PHA-stimulated cell line was assessed from the same donor (PM) and cloned as previously described. Growing clones were screened for CD8 antigen by mAb staining and flow cytometric analysis. Positive TCC were expanded and maintained in the same condition as CD4+ cells.
2.5 EBV-transformed B lymphocyte cell line (LCL) A LCL was established from the same donor (PM) by infecting lo6 PBMC in medium B supplemented with 5 pg/ml cyclosporin A , with EBV-containing supernatant of the B.869 cell line culture.
2.6 Proliferation assay Proliferative responses were measured in triplicate cultures (220 pl) in 96-well flat-bottom plates that routinely contained 3 x lo4 responder TCC and either lo5 irradiated PBMC or 3 x lo4 LCL as APC. TT was added at 200.625 pg/well (twofold dilutions) SE at 1 pg-1 pg/well (tenfold dilutions) and PHA at 9-0.1 pglwell (threefold dilutions). In selected experiments, TCC were used without APC. In this case, TCC were used for proliferation assays not less than 20 days after the last restimulation cycle. In some experiments, TCC were layered onto a Lymphoprep gradient and centrifuged prior to their utilization in proliferation experiments in order to eliminate nonviable cells. For pre-pulsing experiments, 2 x 106TCC, PBMC or LCL were incubated on ice with SEA, SEB (100 ng/ml) or TT (20 pg/ml). After 1 h of incubation, TCC, PBMC and LCL were washed four times with RPMI supplemented with 1% FCS, irradiated with 1500, 3500 and 4500 rad, respectively, washed again and finally resuspended at 4 x 106 in medium B. The medium from the last washing was collected and tested for mitogenic activity due to residual unbound SE using nonirradiated PBMC or TCC. Twofold dilutions of pre-pulsed cells were used to measure the proliferative response of 3 x lo4 TCC. In some experiments, macrocultures of 2 x lo6 TCC in medium B were incubated with SEA (100 ng/ml) and SEB (1000 ng/ml) with or without 2 X 106APC. TCC were then washed five times and resuspended at 3 x 105/ml for proliferation assay or cultured for an additional 36 h in medium B or medium T before testing their proliferative response to SE, PHA and TT. In inhibition experiments, 20 pl/well of culture medium supernatant of the anti-DR mAb producing hybridomas L245, L227 [30], 2.06 [31] and 9.3F10 [32] (ATCC) or 10 pl/well of anti-DQ and anti-DP mAb (Becton Dickinson, Milano, Italy) were added. In these experiments SE or PHA were added either together with the mAb or after I-h incubation of cells with mAb. In all proliferation assays, tritiated thymidine (Amersham, Int., Amersham, GB) was added at 0.5 mCi/well after 48-h culture, and cells were harvested 18 h later. Results were expressed as mean cpm of triplicates.
Superantigen presentation by human T cell clones
Eur. J. lmmunol. 1992. 22: 2033-2039
Table 1. Proliferative response to different SE of 7 clones without APC. In parenthesis are reported the responses with APC
2.7 slCr-release assay The cytolytic activity of CD8+ and CD4+ TCC was determined by 51Cr release of target autologus LCL, as previously described [33]. Briefly, LCL was labeled with V r at 37°C for 1 h and incubated with SEE and SED (100 pg/ml) for 1 h. After four further washings, LCL were resuspended at S x 104cell/ml in medium B and combined with the CD4+ or CD8+ effector TCC at the effector: target ratios 20 :1 and 10 :1. Cultures of 200 pl were assessed in round-bottom 96-well plates for a period of S h at 37°C. After centrifugation at 400 X g for S min, 100 p1 of supernatant was removed for determination of 51Cr release. Spontaneous release was determined by incubating target cells without effectors, while maximal release was determined by treating the labeled target cells with 0.1 YONP40. The following formula was used for calculating the percentage of specific lysis: Yo Specific release
=
cpm (experimental) cpm (max release)
2035
-
cpm (spont release) % cpm (spont. release)
The spontaneous release was always less than 10 YOof the maximal release. 2.8 Phenotypic analysis After 1, 24 or 48 h of incubation, aliquots of cells were harvested from macrocultures which consisted of 2 x lo6 TCC in medium B containing SEA (100, 10 and 1ng/ml) and SEB (1000,100 and 10 ng/ml), with or without 2 X 106 APC. The aliquots were washed twice with cold PBS plus 0.1% BSA and 0.01% NaN3 and stained with FITCconjugated murine mAb anti-TcR cdp, anti-DR, anti IL-2 receptor (CD2S) and anti-CD4. FITC-conjugated mouse Ig control (Becton Dickinson) served for background determination. After washings, cells were resuspended in medium suitable for FACScan (Becton Dickinson) analysis supplemented with 5 pg/ml propidium iodide (PI) to visualize better dead cells. Fluorescence intensity was by evaluated by computerized analysis of histrograms generated by SO00 viable cells [34].
Enteroxtoxin SEA SEB SECl SEC2 SED
SEE
CD4+ TCC +(+> +(+I -(&I -4-) +(+) +(+I a(+) -(-I +(+) +(+I -(-) -(-I +(+) +(+) -(-) -(-) +(+) +(+) n.d.a) n.d.
PM 1.20 PM 1.16 PM 1.13 PM 3.11 PM 1.1 CDS+ TCC PM 30.1 PM 18.1
-(-I -(-I
-(+I -t+)
-(+) -(+)
+(+I
+(+I
+(+) +(+) +(+) n.d.
+(+) +(+) +(+) n.d.
--(+I *(+I -(+I -(+I
-(-) -(-)
The proliferative response of 3 X lo4 TCC was arbitrarily considered + when, at the optimal SE concentration. resulted in cpm > 10.000, or & when resulted in 3.000 < cpm < 10.000. The background was always less than 1.000 cpm when PBMC were used as APC and less then 200 when TCC were used without APC. a) n.d. = not done.
concentration of SE was different for the diverse clones (data not shown). We could rule out the possibility that contaminating irradiated PBMC, remnant of the last restimulation cycle, might be acting as APC for SAg presenta-
nglrnl
3 Results
Figure 1. Proliferative response of the TT-specific CD4+ TCC PM 31.1 to SEA and SEB in the absence of APC.
3.1 TCC self presentation of SE TT-specific human CD4+ TCC PM1.1 and 3.11 proliferated in the absence of APC when incubated with a wide concentration range of SEA and SEB (Fig. 1). The TCC proliferation induced by SEA or SEB without APC showed a dose-response curve comparable to that obtained in the presence of APC (Fig. 2). At the optimal concentration, the TCC self-presentation of SEA (1 ng/ml) and SEB (100 ng/ml), induced a proliferation of the same order of magnitude as that obtained with 20 pg/ml TTpresented by autologous APC (Fig. 2 ) . Similar results (Fig. 3) were obtained with the series of CD4+ antigen-specific clones tested with SEA, SEB, SED and SEE (Table 1). None of the TCC tested proliferated with SECl and SEC2. It was possible to observe that, if one SE induced the stimulation of oneTCC, the optimal concentration was the same with or without APC (Fig. 2 and 3). On the contrary, the optimal
200
P
-
8."
,,,,A
g+"~ ,... ......... ......., . . . ....., . . . . . . *,,.,N
07 0
-
01
1
...,
1
10
........, . . . . ...., 1000
100
......"l
10000
[SEE or TT]
Figure 2. Proliferative response of the TT-specific CD4+ TCC PM 1.1to SEB with or without APC (PBMC) and toTT in the presence of autologous PBMC. SEB was added at 1 wg-1 . - pg/well (tenfold dilutionsrand R a t 20-0.625 bg/well (twofold dilutions). I -
2036
Eur. J. Immunol. 1992. 22: 2033-2039
R. Nisini, P. M. Matricardi, A. Fattorossi et al.
Figure 3. Proliferation of four TT-specific CD4+ and two CD8+ TCC to SEE in the presence (A) or absence (B) of APC (LCL 3 x 104/well). The optimal concentration of SEE was the same for each clone either in the presence or absence of APC. CDS+ TCC did not show a SEE-induced proliferation. For each clone, the increase of the concentration over the optimal, resulted in a decrease of the proliferative response.
120 116
113 311 18 1 (CD81 30 1 (COB]
5ee
tion. First, the contaminating irradiated PBMC underwent a 30-60-fold dilution due to the TCC active growth. Secondly, TCC were used for proliferation assays not less than 20 days after the restimulation, i.e. after a lapse of time in which irradiated cells die. TCC did not proliferate with TT at 20 pglml in the absence of added APC 20 days after the last restimulation with irradiated autologous PBMC (data not shown). Moreover, the SE-mediated proliferation of TCC in the absence of APC was even enhanced when clones were separated from remnants of filler cells through a centrifugation step onto a Lymphoprep gradient before use (data not shown). In addition, the phenotypic analysis of TCC showed that virtually all the cells were co-expressing the surface markers CD4+ and TcR, and PI exclusion showed a more than 98% cell viability. Finally, when pre-pulsing experiments were performed (see below), the efficiency of presentation of TCC was about two orders of magnitude higher than that of PBMC, so that the presentation ability could not be ascribed to contaminating PBMC, but to SAg-mediated T-T cell interactions. Increasing the concentration of SE beyond the optimal doses resulted in a decrease of the proliferative response in the presence or absence of APC (Figs. 2 and 3). To test the possible causes of nonresponsiveness to high doses of SE,TCC were incubated o/n with SEA (100 ng/ml) or SEB (1000 ng/ml) without added APC. After overnight incubation and SE removal by extensive washing, TCC were capable to proliferate in response to IL-2, and if APC were added. TCC, on the other hand, showed a marked hyporesponsiveness to SE, TT and PHA in the presence or absence of APC. The responsiveness was partially restored after 36 h of culture in medium T, following removal of SE (manuscript in preparation) .These preliminary results indicate that at high concentrations of SE the decrease of proliferative response was not due to SE toxicity and that an SE-mediated self-cytotoxity can be excluded.
see
P 160 140
120
300
0,
250
E
200
B
150 40
100 50
0
0 1
, I
10
100
1
1000
!?
120:
6
100: ‘
80 -
n
\; i
5
U T C C P H A
........*,..
1
100
10
1000
[SEE]
PEE1
TCCSED
;
180
-
150-
90
-
60
-
30
-
120
.
0-
Figure 4. The anti-DR mAb L243, L227 or Y.3F10 inhibited the SE-induced proliferation either in the presence (A) or in the absence (B-D) of APC. All mAb were used at 20 pl/well of culture medium supernatant of producing hybridomas. Twofold dilutions of the mAb did not inhibit the proliferative response of TCC obtained with 0.3 pg/ml of PHA (C), but a dose-dependent mAb inhibition of proliferation was observed in cultures stimulated with 10 @well of SED.
The SE-induced proliferation was inhibited by the anti-DR mAb L243, L227 and 9.3F10, but not by mAb 2.06, anti-DP and anti-DQ (Figs. 4a and b). The inhibition was dose dependent and not due to mAb toxicity since the proliferative response to PHA was only slightly affected (Figs. 4c and d). To test the relative efficiency of SAg presentation of various APC, thc same concentrations of TCC, LCL or PBMC incubated with SE for 1 h on ice, irradiated and extensively washed, were used as APC to compare the relative cfficiency of SE presentation to responder TCC. As shown in Fig. 5, TCC were less efficient than LCL and more
# APC
Figure 5. Proliferative response of 3 x lo4TCC (PM1.l) was used to compare the presentation ability of irradiated TCC, LCL or PBL pre-pulsed with 100 ng/ml of SEB, as detailed in Sect. 2.6. Pre-pulsed APC were diluted twofold starting from 4 X lo5 cells well.
Superantigen presentation by human T cell clones
Eur. J. Immunol. 1992. 22: 2033-2039
efficient than PBMC in presenting SE to the same responder TCC, when the number of pre-pulsed APC was lower then 2 x 104/well. However, when the number of prepulsed APC was greater than 2 x 104/well, the highest proliferative response was obtained with PBMC and the relative presentation efficiency was PBMC > LCL > TCC. The medium collected after the last washing of pre-pulsed APC did not show any mitogenic activity due to unbound SE, when tested with TCC or nonirradiated PBMC (data not shown), indicating that the proliferation observed with SE pre-pulsed APC was due to cell-bound SE. Taken together, these results indicate that a single cell of the clone effectively presented the SE to cells of the same clone. 3.2 SE-mediated cytotoxicity CD4+ TCC able to actively proliferate when stimulated with SEE and SED and CD8+ TCC able to proliferate in the presence of APC with SED but not with SEE (Table l), were tested for SE-induced cytotoxicity in a WR-release assay, using SEE- and SED-pre-pulsed labeled autologous LCL as target. As shown in Fig. 6, CD4+ TCC did not show any significant SE-mediated cytotoxic activity. On the other hand, CD8+ TCC showed a marked SE-mediated cytotoxicity only when the LCL were pre-pulsed with SED. 80
1
-20 ,
CD8 20 1
CD8 10 1
CD4 20 1
CD4 10 1
Figure 6. CD4+ TT-specific TCC did not show any significant SE-mediated cytotoxicity, even when target LCL were pre-pulsed with the same SE as that which induced a proliferative response. On the other hand, CD8+ TCC, showed an SE-mediated cytotoxicity when target LCL were pre-pulsed with an SE (SED) that induced a CDW TCC proliferation in the presence of APC.
3.3 Phenotypic modulation induced by SE activation of TCC
To determine whether the self presentation by TCC of SE was followed by the same phenotypic variation accompanying TCC proliferation in the presence of APC, TCC incubated with SEA or SEB were stained and analyzed by flow cytometry after 24 and 48 h (Fig. 7). Preliminary experiments had shown that an incubation of TCC with SEA or SEB for 1 h on ice, that was sufficient to induce proliferation, did not interfere with the binding of mAb. Thus, any variation of the phenotypic profile could be regarded as representing active modifications and not a shielding of relevant epitopes by SE. As depicted in Fig. 7,
TcR
DR
2037
IL-2 R
Fluorescence intensity Figure 7. Modulation of TCC surface antigens after a 48-h culture with SEA without APC.TCC were culturedinmedium B with SEA (100,lO or 1ng/ml) or SEB (1000,100 or 10 ng/ml) with or without APC. After 1, 24 or 48 h of culture, aliquots were washed and stained with mAb anti-TcR, anti-DR and anti-IL-2R as detailed in Sect. 2.8. Similar phenotypic variations were observed with SEB and in the presence of APC (PBMC). (background: -----; control without SEA: "...; SEA 1 ng/ml: .....'.....: SEA 100 ng/ml:
-1. SE stimulation induced modifications in the phenotypic profile of TCC, irrespective of the presence of added APC with a dose- and time-dependent up-regulation of DR and CD25, and a parallel down-regulation of TcR.
4 Discussion We demonstrate that TT-specific MHC class II+ CD4+ human T clones proliferate in the absence of added APC when stimulated with suitable concentrations of SE. The TCC proliferation in response to SE, obtained in the absence of additional APC, exhibited a dose-response curve comparable to that obtained with the addition of APC. Furthermore, the TCC self-presentation of SE induced a proliferation of the same order of magnitude as that obtained with the autologous PBMC presentation of TT. Anti-DR mAb inhibited at different degrees of efficiency the SE-induced proliferation both in the presence and in the absence of APC. Conversely, anti-DP or anti-DQ mAb did not interfere with SE-induced proliferation. These results indicate that the main constraint for SE presentation in our experiments is the expression of the D R isotype of MHC class I1 molecules [2-61. Finally, the TCC self-presentation of SE induced a series of phenotypic modifications on TCC that overlapped those occurring after the SE stimulation in the presence of APC. These data indicate that TCC present SE and provide sufficient accessory signals, thus acting as SAg-presenting cells. The differences in the SE presentation efficiency among TCC, LCL and PBMC, observed in pre-pulsing experiments, could be explained by differences in the expression of MHC class I1 molecules, that actually function as cellular receptors for SE [ 2 ] . Indeed, the expression of MHC class II- molecules/cell is the highest on LCL [35, 361 and the lowest on PBMC, due to the great number of MHC class II-cells in this latter cell population. However, PBMC showed the highest SE presentation efficacy that might reflect their highest capacity to provide ligands for TCC adhesion molecules [37] and co-stimulatory signals [38], in addition to the expression of MHC-antigen or SAg complexes. All together our results strongly suggest that cells of a given clone effectively present the SE to other cells of the
2038
Eur. J. Immunol. 1992. 22: 2033-2039
R. Nisini. P. M. Matricardi, A. Fattorossi et al.
retroviral SAg and “interact” [24] with other TcR+ T cells, becoming responsible for the observed deletion of T cells. In such a situation, CD4+ T cells would be the prominent SAg-presenting cells, as these cells are the main target of HIV Our results indicate that human CD4+ TCC present SAg and demonstrate that SAg-mediated interactions can take place among T cells. The outcome of this interaction was, in our experiments, the proliferation of responder TCC. However, it is not possible to exclude that the in vivo presentation of an HIV-encoded SAg by infected CD4+ Tcell might lead to deletion of SAg complementary Vp Tcells. Recent findings [19] suggest that an immune response to a SAg in vivo can eventually culminate in marked tolerance and clonal elimination. With a similar mechanism, activated HIV-infected T cells might present the hypothesized SAg, inducing first a clonal expansion and then a deletion of SAg-complementary Vp cells. What the functional consequences of the in vivo presentation by Tcells are and whether HIV-infected cells can deliver adequate accessory signals remains unknown. Anergy is a consequence of mature T cells confronting antigen on Our results are in line with previously reported [27, 281 inappropriate or damaged APC [42,43]. The hypothesized studies that emphasized the APC function of TCC [29]. HIV-SAg presentation in vivo by activated infected “daOur results, however, are in contrast with those reported by maged” CD4+ T cells might not provide adequate accessoO’Hehir and Lamb [39] who failed to obtain significant ry signals and induce anergy. A long-lasting anergic state, TCC proliferation upon SE stimulation in the absence of due to SAg-prolonged presentation in chronic HIV-infected APC. The reasons for such discrepancies are not clear, but patients, might lead to the elimination of anergic cells [19]. some points, regarding the TCC self-presentation of SE, Our results raise the possibility that MHC class II+ cells deserve consideration. First, a cytotoxic activity of theTCC expressing a SAg-complementary Vp may stimulate themused as SE-presenting cells should be excluded. It has been selves. The auto-stimulation of an HIV-infected cell might reported that IL-2-maintained antigen-specific CD4+ TCC deliver inappropriate signals due to a SAg-mediated TcR may acquire cytotoxic properties [40] while retaining the and MHC class I1 cross-linking, leading to anergy or cell antigen-induced proliferation capacity. It is conceivable death by apoptosis [44] that, in mature mouse T cells with that these clones might exert cytolytic activity toward cross-linked CD4+ molecules, has been shown to follow SE-presenting cells. If cytotoxicT cells are at the same time TcR stimulation [45]. The hypothesized HIV SAg, possibly SE-presenting and responder cells, proliferation might not in association with other retroviral products such as the result upon SE stimulation as a consequence of SE- CD4 ligand gp120, might mediate the auto-stimulation of mediated self-cytotoxicity [14]. This was not the case of infected T cell expressing the complementary Vb and lead TCC used for our experiments. Second, TCC express to a selective CD4+ T cell depletion with the mechanisms different levels of MHC class I1 molecules at different times that have been suggested in the panergic imnesia hypotheof the restimulation cycle, and the MHC class 11 molecules sis [46] of AIDS development. expresvion is the main constraint of SAg presentation ability. Thus, at different times after the last restimulation, In summary, the main finding in this report is that SAg TCC might present SE with a different efficiency and, as a mediate T-T cell interaction andor T cell auto-interaction. consequence, the SE optimal concentration must be care- If the presence of an HIV-encoded SAg produced in CD4+ fully chosen to avoid nonresponsiveness.Third, the binding T cells is confirmed, in future studies it will be critical to affinity of the SE to the MHC class I1 molecules [4] and to examine the consequences of the SAg-mediated T-T cell the V, families of the responder clones may differ among interaction, in order to understand the pathogenesis of the laboratories and also contribute to the observed discrepan- Vp-selective helper T cells depletion observed in AIDS cies [39]. patients.
same clone. This view is substantiated by results of pre-pulsing experiments showing that irradiated, SAg pre-pulsed TCC were able to stimulate nonirradiated, non-SAg-pulsed TCC. However, the possibility exists that, in addition to the T-T cell interaction, cells expressing the SAg-complementary V, may undergo a SAg-mediated auto-stimulation process. This is of particular interest because an auto-stimulated cell would modulate its membrane receptors and possibly decrease its ability of interacting with other cells. In keeping with this view, we observed that, at high concentrations, SE did not induce TCC proliferation, although they did induce a phenotypic modulation, similar to that observed in optimally SE-stimulated, proliferating TCC. After overnight incubation with high SE concentration, moreover, TCC were still viable and able to proliferate upon addition of IL-2 or APC, provided that the unbound SE was removed by extensive washings (R. Nisini, manuscript in preparation). This suggests that at high SE concentrations, the TCC autostimulation might be favored and result in the loss of the presentation capacity.
Several reports have clearly demonstrated that the activatioddepletion of T cells expressing a particular Vp is diagnostic for the presence of a SAg [20-24, 41].The recent finding that the Tcell depletion in AIDS patients is selective for certainY3 families [25] led to speculate on the possible pathogenetic role of a hypothetical HIV-encoded SAg. The possibility that an SAg might be responsible for a selective depletion of T cells in an adult acquired retroviral infection, raises the question of what condition might be decisive in determining whether the final outcome of SAg-mediated T cell activation is proliferation or deletion of SAg-complementary Vp-bearing cells. It has been hypothesized, that, upon activation, HIV-infected CD4+ T cells would express MHC class I1 molecules together with the
The authors would like to thank, K Barnaba and M . Paroli for helpful criticism and suggestion, A. Vincenti for his advice and continuous encouragement and A . Di Pietro and R. Vitalone for excellent technical assistance.
Received February 5, 1992; in revised form March 25, 1992.
5 References 1 Choi,Y., Kotzin, B., Herron, L., Callahn, J. and Marrack, P., Proc. Natl. Acad. Sci. USA 1989. 86: 8941. 2 Mollick; J. A . , Cook, R. G. and Rich, R. R., Science 1989.244: 817.
Eur. J. Immunol. 1992. 22: 2033-2039
3 Marrack, l? and Kappler, J., Science 1990. 248: 707. 4 Herrman, T., Accolla, R. and MacDonald, R., Eur. J. Immunol. 1989. 19: 2171. 5 Frascr, J. D., Nature 1989. 339: 221. 6 Fisher, H., Dohlsten, M., Lindvall, M., Sjorgren, H. and Carlsson, R., J. Immunol. 1989. 142: 3151. 7 Choi,Y., Herman, A , , Di Giusto, D.,Wade,T., Marrack, l? and Kappler, J., Science 1900. 346: 471. 8 Mollick, J. A . , Chintagumpala. M., Cook, R. G. and Rich, R. R., J. Immunol. 1991. 146: 463. 9 Fleischer, B., Schrezenmeier, H. and Conradt, P., Cell. Immunol. 1989. 120: 92. 10 Fleischer, B. and Schrezenmeier, H., J. Exp. Med. 1988. 167: 1697. 11 Kappler, J., Kotzin, B.. Herron, L., Gelfand. E. W., Bigler, R. D., Boylston, A , , Carrel, S., Posnett, D. N., Choi,Y. and Marrack, F!, Science 1989. 244: 811. 12 Fleischer. B., Gerardy-Schahn, R., Metzroth, B., Carrel, S., Gerlach, D. and Kiihler, W., J. Immunol. 1990. 146: 11. 13 Imanishi, K., Igarashi, H. and Uchiyama,T., J. Immunol. 1990. 145: 3170. 14 Dohlsten, M., Hedlund, G. and Kalland, T., Immunol. Today 1991. 12: 147. 15 Carlsson, R., Ficher, H. and Sjegren, O., J. Immunol. 1988, 140: 2484. 16 Hufnagle. W. 0.. Tremaine, M. T. and Betley, M., Infect. Immun. 1991. 59: 2126. 17 Pontzer, C. H., Russel, J. H. and Johnson, H. M., J. Immunol. 1989. 143: 280. 18 Kavabe,Y., Ochi, A., Nature 1991. 349: 245. 19 Webb. S. R.. Morris, C., Sprent, J., Cell 1990. 63: 1249. 20 Marrack, l?. Kushir, E., Kappler, J., Nature 1991. 349: 524. 21 Francke1,W. N . , Rudy, C., Coffin, J. M., Huber, B. T., Nature 1991. 349: 526. 22 Woodland, D. L., Happ, M. l?, Gollob, K. J., Palmer, E., Nature 1991. 349: 529. 23 Dyson, l? J., Knight, A . M., Fairchild, S., Simpson, E., Tomonari, K., Nature 1991. 349: 531. 24 Imberti, L., Sottini, A., Bettinardi, A., Puoti, M. and Primi, D., Science 1991. 254: 860. 25 Janaway. C.. Nature 1991. 349: 459.
Superantigen presentation by human T cell clones
2039
26 Evans, R. L., Faldetta,T. J., Humphreys, R. E., Pratt, D. M., Yunis, E. J. and Schlossman, S. F., J. Exp. Med. 1979. 148: 1440. 27 Lanzavecchia, A., Abrignani, S., Scheidegger, D., Obrist, R., Dorken, B. and Moldenhauer, G., J. Exp. Med. 1988. 1647: 345. 28 Lanzavecchia, A , , Roosnek, E., Gregory,T., Berman, l? and Abrignani, S., Nature 1988. 334: 530. 29 Hewitt, C. R. A. and Feldman, M., J. Immunol. 1989. 142: 1429. 30 Lampson, L. A. and Levy, R., J. Immunol. 1980. 125: 293. 31 Charron, D. J. and McDevitt, H. O., Fed. Proc. 1979. 38: 1280. 32 Van Voorhis,W. C., Steinman, R. M., Hair, L. S., Luban, J., Witmer, M. D., Koide, S., Cohn, Z. A , , J. Exp. Med. 1983. 158: 126. 33 Barnaba,V., Franco, A., Alberti, A . , Balsano, C., Benvenuto, R., Balsano, F., J. Immunol. 1989. 143: 2650. 34 Fattorossi, A , , Nisini, R., Pizzolo, J. G., D’Amelio, R.. Cytometry 1989. 10: 320. 35 Pawelec, G. and Biihring, H. J., Cell. Immunol. 1990. 127: 520. 36 Robbins, l?, Maino,V. C., Warner, N. L. and Brodsky, F. M., J. Immunol. 1988. 141: 1281. 37 Springer,T. A., Nature 1991. 346: 425. 38 Weaver, C. T. and Unanue, E. R., Immunol. Today 1990. 11: 49. 39 O’Hehir, R. E. and Lamb, J. R., Proc. Natl. Acad. Sci. USA 1990. 87: 8884. 40 Fleisher, D., Nature 1984. 308: 365. 41 Janeway, C. A., Yagi, J., Conrad, F!, Katz, M. E., Jones, B., Vroegop, S. and Buxser, S., Immunol. Rev. 1989. 107: 61. 42 Mueller, D. L., Jenkins, M. K., Schwartz, R. H., Annu. Rev. Immunol. 1989. 7: 445. 43 Schwartz, R. H., Cell 1989. 57: 1073. 44 Terai, C., Kornbluth, R. S., Pauza, D., Richman, D. D., Carson, D. A., J. Clin. Invest. 1991. 87: 1710. 45 Newell, M. K., Haughn, L. J., Maroun, C. R., Julius, M. H., Nature 1990. 347: 286. 46 Ascher, M. S. and Sheppard, 11. W., J. AIDS 1990. 3: 177.
Eur. J. Immunol. 1992. 22: 2033-2039
Roberto Nisini, Paolo M. Matricardi, Andrea Fattorossi, Roberto Biselli and Raffaele D’Amelio D.A.S.R.S. Rep. Medicina, Lab. Immunologia, Pomezia Roma
2033
Presentation of superantigen by human T cell clones: a model of T-T cell interaction Superantigens (SAg) interact withT lymphocytes bearing particular vj sequences as part of their T cell receptor (TcR). The interaction, however, requires the presence of major histocompatibility complex (MHC) class I1 molecules on antigen-presenting cell (APC). In peculiar circumstances, MHC class 11+T cell clones (TCC) have been shown to present peptides and selected antigens interacting with antigen-specific TCC in the absence of APC. In this report we studied the capacity of SAg to mediate a T-T cell interaction, investigating the TCC ability to present a panel of staphylococcal enteroxins (SE) independently of the presence of added APC. Upon exposure to a broad range of SE concentrations, MHC class IIf TCC showed an intense proliferative response even in the absence of professional APC. Diverse SE optimally stimulated responder TCC at different concentrations. The proliferation was inhibited by anti-DR monoclonal antibodies, both in the presence and in the absence of APC.The SE activation of TCC in the absence of APC induced the same series of phenotypic variations as that observed following the TCC stimulation with APC. Irradiated TCC efficiently presented membrane-bound SE to responder TCC as well as professional APC. These results show that a single cell of a given clone effectively presents the SE to other cells of the same clone, and provide evidence that SAg can efficiently mediate T-T cell interaction. In addition, the possibility also exists that one cell of the clone can actually undergo an auto-stimulation via SAg-mediated interactions between its own TcR and MHC class I1 molecule. It has recently been suggested that the Vb-selective depletion of T cells observed in acquired immunodeficiency syndrome (AIDS) patients might be a consequence of the interaction between a human immunodeficiency virus (H1V)-encoded SAg and T cells expressing a SAg complementary V,.We suggest that the hypothesized HIV-encoded SAg might mediate T-T cell interactions that could play a relevant role in the Vpelective depletion of T lymphocytes observed in HIV-infected patients.
1 Introduction
maternally transmitted mouse retroviral genes encode SAg responsible for an intrathymic deletion of discrete sets of Staphylococcal enterotoxins (SE) and mouse retroviral- T cells bearing certain Vp families of TcR [20-231. Conseencoded proteins such as minor lymphocyte stimulating quently, the activation/depletion of T cells expressing a (MIS) gene products, have been defined as superantigens particular Vp is diagnostic for the presence of SAg [24].The (SAg), because of their unusual MHC-TcR binding and recent observation that the CD4+ T cell depletion in AIDS consequent T cell activaton [ll.All SAg share the ability to patients is selective for certain Vp families [25] led to bind to human and mouse MHC class 11proteins [2-61 and speculations on the possible pathogenetic role of a hypothe SAg-MHC complex shows specificity for a peculiar part thetical HIV-encoded SAg. It has also been hypothesized of theTcR,Vp [3, 71. SAg activateT cells irrespective of their that upon activation, HIV-infected CD4+ T cells would antigen specificity, via a selective interaction with mono- express MHC class I1 molecules together with retroviral morphic regions of MHC class I1 molecules on APC and SAg and “interact” [24] with other CD4+ TcR+ Tcells with v j regions of TcR on T cells [8-141: it has been expressing particular y3, and that this interaction could be suggested that the mitogenic activity of SAg probably responsible for a selective T cell depletion. According to resides in their three-dimensional structure, able to stabil- this hypothesis, the HIV-encoded SAg should share most of ize the TcR-MHC class I1 binding [lo, 15-17]. It has also the properties of other known SAg.The presentation of the been shown in mice that in vivo administration of SE [18]or HIV-encoded SAg might have different functional conseMls-disparate cells [19] leads first to an activation and then quences depending on whether the presentation occurs in to a depletion of CD4+ cells expressing the SAg-comple- the thymus or in the periphery and/or on the type of mentary vj. Furthermore, rccent data have shown that SAg-presenting cells. Activated humanT lymphocytes and T cell clones (TCC) express MHC class I1 molecules [26] and elegant experiments have shown that TCC act as accessory cells in presenting antigens upon fulfillment of [I 103241 certain conditions. For example, TCC can process and Correspondence:Roberto Nisini. Lab. Immunologia, Rep. Medi- present those antigens that are targeted to their own surface cina, DASRS, Alto Pratica di Mare. 1-00040 Pomczia Roma. receptors, thus bypassing thcir lack of uptake capacity [27, 281. TCC can also effectively present antigens that, by Italy binding directly to MHC class I1 molecules, do not need to Abbreviations: SAg: Superantigen SE: Staphylococcal entero- be internalized and processed, as is the case for peptidetoxins TCC: T cell clones. sized antigens [29]. 0 VCH Vcrlagsgesellschaft mbH, D-6940 Weinheim, 1992
0014-2980/92/0808-2033$3.50+ .25/0
2034
R. Nisini, P. M . Matricardi, A. Fattorossi et al.
In this work, we used a panel of SE to study T-Tcell interaction. We observed that tetanus toxoid (TT)-specific MHC class 11+ CD4+ TCC are able to present SE and to proliferate even in the absence of additional APC, indicating that SAg can mediate an effective T-T cell interaction. The possibility is envisaged that a T-T cell interaction mediated by the hypothesized HIV-encoded SAg could take place and possibly play a relevant role in determining the selective depletion of V, families of TcR-bearing T lymphocyte observed in AIDS patients.
2 Materials and methods 2.1 Antigens Staphylococcal enterotoxin A (SEA), B (SEB), C1 (SECI), C2 (SEC2), D (SED) and E (SEE) were purchases from Serva Fine Biochemicals (New York, NY). TT was kindly provided by Dr. A. Podda, (Sclavo SPA, Siena, Italy)
2.2 Media RPMI 1640 (Flow Laboratories, Irvine, Scotland) supplemented with 100 U/ml penicillin, 100 mg/ml streptomycin, 2 mM L-glutamine, 1 mM sodium pyruvate, 1% nonessential amino acids 5 x lo-' M 2-ME and 10% FCS (Gibco Lab., Grand Island, NY) was used for culturing EBVtransformed lines and for proliferation assays (medium B), whilc 2-ME and FCS were substituted by 50U/ml IL-2 (kindly provided by Hoffmann-La Roche, Inc. (Basel, Switzerland) and 5 % autologous serum for culturing TCC (medium T). 2.3 Cloned human TT-reactive CD4+ T lymphocytes Twenty-five milliliters of heparinized blood was obtained from a normal donor (PM) recently vaccinated with TT (Anatetall, Sclavo SPA, Siena, Italy) and PBMC were purified on a density gradient (Lymphoprep, Nycomed Pharma AS, Oslo, Norway). Cells were resuspended at 1O6/rnlin medium T without IL-2, and 100 or 200 pl were dispensed in 96-well flat-bottom plastic plates (Costar, Cambridge, MA). TTwas added at 20 pg/ml (final concentration) to half of the cultures. After 6 days, 2 or 4 U/well TL-2 were added to the cultures. After 4 additional days, cultures showing a significant cell growth were considered positive. Cells were counted and cloned by limiting dilution at 3, 1 and 0.3 cells/well in medium T supplemented with 5 x lo5 irradiated PBMC and 1 pg/ml PHA (Wellcome Diagnostics, Dartford, GB). After 10-15 days, growing cultures wcre expanded in medium Tand finally tested for TT specificity in a proliferation assay using irradiated autologous PBMC pre-pulsed or not with TT at 20 pg/ml. TT-specific clones were expanded and maintained in culture with 25-35-day cycles of restimulation with PHA and irradiated PBMC. Clonality was assessed by a subcloning test that showed 100 % TTspecificity of growing subcloned cells. TT-rpecific CD4+ TCC (PM 1.1, PM 1.13, PM 1.16, PM 1.20, PM 3.11) screened for responsiveness to SE in the presence of PBMC as APC, were used for proliferation experiments.
Eur. J. lmmunol. 1992. 22: 2033-2039
2.4 CD8+ TCC A PHA-stimulated cell line was assessed from the same donor (PM) and cloned as previously described. Growing clones were screened for CD8 antigen by mAb staining and flow cytometric analysis. Positive TCC were expanded and maintained in the same condition as CD4+ cells.
2.5 EBV-transformed B lymphocyte cell line (LCL) A LCL was established from the same donor (PM) by infecting lo6 PBMC in medium B supplemented with 5 pg/ml cyclosporin A , with EBV-containing supernatant of the B.869 cell line culture.
2.6 Proliferation assay Proliferative responses were measured in triplicate cultures (220 pl) in 96-well flat-bottom plates that routinely contained 3 x lo4 responder TCC and either lo5 irradiated PBMC or 3 x lo4 LCL as APC. TT was added at 200.625 pg/well (twofold dilutions) SE at 1 pg-1 pg/well (tenfold dilutions) and PHA at 9-0.1 pglwell (threefold dilutions). In selected experiments, TCC were used without APC. In this case, TCC were used for proliferation assays not less than 20 days after the last restimulation cycle. In some experiments, TCC were layered onto a Lymphoprep gradient and centrifuged prior to their utilization in proliferation experiments in order to eliminate nonviable cells. For pre-pulsing experiments, 2 x 106TCC, PBMC or LCL were incubated on ice with SEA, SEB (100 ng/ml) or TT (20 pg/ml). After 1 h of incubation, TCC, PBMC and LCL were washed four times with RPMI supplemented with 1% FCS, irradiated with 1500, 3500 and 4500 rad, respectively, washed again and finally resuspended at 4 x 106 in medium B. The medium from the last washing was collected and tested for mitogenic activity due to residual unbound SE using nonirradiated PBMC or TCC. Twofold dilutions of pre-pulsed cells were used to measure the proliferative response of 3 x lo4 TCC. In some experiments, macrocultures of 2 x lo6 TCC in medium B were incubated with SEA (100 ng/ml) and SEB (1000 ng/ml) with or without 2 X 106APC. TCC were then washed five times and resuspended at 3 x 105/ml for proliferation assay or cultured for an additional 36 h in medium B or medium T before testing their proliferative response to SE, PHA and TT. In inhibition experiments, 20 pl/well of culture medium supernatant of the anti-DR mAb producing hybridomas L245, L227 [30], 2.06 [31] and 9.3F10 [32] (ATCC) or 10 pl/well of anti-DQ and anti-DP mAb (Becton Dickinson, Milano, Italy) were added. In these experiments SE or PHA were added either together with the mAb or after I-h incubation of cells with mAb. In all proliferation assays, tritiated thymidine (Amersham, Int., Amersham, GB) was added at 0.5 mCi/well after 48-h culture, and cells were harvested 18 h later. Results were expressed as mean cpm of triplicates.
Superantigen presentation by human T cell clones
Eur. J. lmmunol. 1992. 22: 2033-2039
Table 1. Proliferative response to different SE of 7 clones without APC. In parenthesis are reported the responses with APC
2.7 slCr-release assay The cytolytic activity of CD8+ and CD4+ TCC was determined by 51Cr release of target autologus LCL, as previously described [33]. Briefly, LCL was labeled with V r at 37°C for 1 h and incubated with SEE and SED (100 pg/ml) for 1 h. After four further washings, LCL were resuspended at S x 104cell/ml in medium B and combined with the CD4+ or CD8+ effector TCC at the effector: target ratios 20 :1 and 10 :1. Cultures of 200 pl were assessed in round-bottom 96-well plates for a period of S h at 37°C. After centrifugation at 400 X g for S min, 100 p1 of supernatant was removed for determination of 51Cr release. Spontaneous release was determined by incubating target cells without effectors, while maximal release was determined by treating the labeled target cells with 0.1 YONP40. The following formula was used for calculating the percentage of specific lysis: Yo Specific release
=
cpm (experimental) cpm (max release)
2035
-
cpm (spont release) % cpm (spont. release)
The spontaneous release was always less than 10 YOof the maximal release. 2.8 Phenotypic analysis After 1, 24 or 48 h of incubation, aliquots of cells were harvested from macrocultures which consisted of 2 x lo6 TCC in medium B containing SEA (100, 10 and 1ng/ml) and SEB (1000,100 and 10 ng/ml), with or without 2 X 106 APC. The aliquots were washed twice with cold PBS plus 0.1% BSA and 0.01% NaN3 and stained with FITCconjugated murine mAb anti-TcR cdp, anti-DR, anti IL-2 receptor (CD2S) and anti-CD4. FITC-conjugated mouse Ig control (Becton Dickinson) served for background determination. After washings, cells were resuspended in medium suitable for FACScan (Becton Dickinson) analysis supplemented with 5 pg/ml propidium iodide (PI) to visualize better dead cells. Fluorescence intensity was by evaluated by computerized analysis of histrograms generated by SO00 viable cells [34].
Enteroxtoxin SEA SEB SECl SEC2 SED
SEE
CD4+ TCC +(+> +(+I -(&I -4-) +(+) +(+I a(+) -(-I +(+) +(+I -(-) -(-I +(+) +(+) -(-) -(-) +(+) +(+) n.d.a) n.d.
PM 1.20 PM 1.16 PM 1.13 PM 3.11 PM 1.1 CDS+ TCC PM 30.1 PM 18.1
-(-I -(-I
-(+I -t+)
-(+) -(+)
+(+I
+(+I
+(+) +(+) +(+) n.d.
+(+) +(+) +(+) n.d.
--(+I *(+I -(+I -(+I
-(-) -(-)
The proliferative response of 3 X lo4 TCC was arbitrarily considered + when, at the optimal SE concentration. resulted in cpm > 10.000, or & when resulted in 3.000 < cpm < 10.000. The background was always less than 1.000 cpm when PBMC were used as APC and less then 200 when TCC were used without APC. a) n.d. = not done.
concentration of SE was different for the diverse clones (data not shown). We could rule out the possibility that contaminating irradiated PBMC, remnant of the last restimulation cycle, might be acting as APC for SAg presenta-
nglrnl
3 Results
Figure 1. Proliferative response of the TT-specific CD4+ TCC PM 31.1 to SEA and SEB in the absence of APC.
3.1 TCC self presentation of SE TT-specific human CD4+ TCC PM1.1 and 3.11 proliferated in the absence of APC when incubated with a wide concentration range of SEA and SEB (Fig. 1). The TCC proliferation induced by SEA or SEB without APC showed a dose-response curve comparable to that obtained in the presence of APC (Fig. 2). At the optimal concentration, the TCC self-presentation of SEA (1 ng/ml) and SEB (100 ng/ml), induced a proliferation of the same order of magnitude as that obtained with 20 pg/ml TTpresented by autologous APC (Fig. 2 ) . Similar results (Fig. 3) were obtained with the series of CD4+ antigen-specific clones tested with SEA, SEB, SED and SEE (Table 1). None of the TCC tested proliferated with SECl and SEC2. It was possible to observe that, if one SE induced the stimulation of oneTCC, the optimal concentration was the same with or without APC (Fig. 2 and 3). On the contrary, the optimal
200
P
-
8."
,,,,A
g+"~ ,... ......... ......., . . . ....., . . . . . . *,,.,N
07 0
-
01
1
...,
1
10
........, . . . . ...., 1000
100
......"l
10000
[SEE or TT]
Figure 2. Proliferative response of the TT-specific CD4+ TCC PM 1.1to SEB with or without APC (PBMC) and toTT in the presence of autologous PBMC. SEB was added at 1 wg-1 . - pg/well (tenfold dilutionsrand R a t 20-0.625 bg/well (twofold dilutions). I -
2036
Eur. J. Immunol. 1992. 22: 2033-2039
R. Nisini, P. M. Matricardi, A. Fattorossi et al.
Figure 3. Proliferation of four TT-specific CD4+ and two CD8+ TCC to SEE in the presence (A) or absence (B) of APC (LCL 3 x 104/well). The optimal concentration of SEE was the same for each clone either in the presence or absence of APC. CDS+ TCC did not show a SEE-induced proliferation. For each clone, the increase of the concentration over the optimal, resulted in a decrease of the proliferative response.
120 116
113 311 18 1 (CD81 30 1 (COB]
5ee
tion. First, the contaminating irradiated PBMC underwent a 30-60-fold dilution due to the TCC active growth. Secondly, TCC were used for proliferation assays not less than 20 days after the restimulation, i.e. after a lapse of time in which irradiated cells die. TCC did not proliferate with TT at 20 pglml in the absence of added APC 20 days after the last restimulation with irradiated autologous PBMC (data not shown). Moreover, the SE-mediated proliferation of TCC in the absence of APC was even enhanced when clones were separated from remnants of filler cells through a centrifugation step onto a Lymphoprep gradient before use (data not shown). In addition, the phenotypic analysis of TCC showed that virtually all the cells were co-expressing the surface markers CD4+ and TcR, and PI exclusion showed a more than 98% cell viability. Finally, when pre-pulsing experiments were performed (see below), the efficiency of presentation of TCC was about two orders of magnitude higher than that of PBMC, so that the presentation ability could not be ascribed to contaminating PBMC, but to SAg-mediated T-T cell interactions. Increasing the concentration of SE beyond the optimal doses resulted in a decrease of the proliferative response in the presence or absence of APC (Figs. 2 and 3). To test the possible causes of nonresponsiveness to high doses of SE,TCC were incubated o/n with SEA (100 ng/ml) or SEB (1000 ng/ml) without added APC. After overnight incubation and SE removal by extensive washing, TCC were capable to proliferate in response to IL-2, and if APC were added. TCC, on the other hand, showed a marked hyporesponsiveness to SE, TT and PHA in the presence or absence of APC. The responsiveness was partially restored after 36 h of culture in medium T, following removal of SE (manuscript in preparation) .These preliminary results indicate that at high concentrations of SE the decrease of proliferative response was not due to SE toxicity and that an SE-mediated self-cytotoxity can be excluded.
see
P 160 140
120
300
0,
250
E
200
B
150 40
100 50
0
0 1
, I
10
100
1
1000
!?
120:
6
100: ‘
80 -
n
\; i
5
U T C C P H A
........*,..
1
100
10
1000
[SEE]
PEE1
TCCSED
;
180
-
150-
90
-
60
-
30
-
120
.
0-
Figure 4. The anti-DR mAb L243, L227 or Y.3F10 inhibited the SE-induced proliferation either in the presence (A) or in the absence (B-D) of APC. All mAb were used at 20 pl/well of culture medium supernatant of producing hybridomas. Twofold dilutions of the mAb did not inhibit the proliferative response of TCC obtained with 0.3 pg/ml of PHA (C), but a dose-dependent mAb inhibition of proliferation was observed in cultures stimulated with 10 @well of SED.
The SE-induced proliferation was inhibited by the anti-DR mAb L243, L227 and 9.3F10, but not by mAb 2.06, anti-DP and anti-DQ (Figs. 4a and b). The inhibition was dose dependent and not due to mAb toxicity since the proliferative response to PHA was only slightly affected (Figs. 4c and d). To test the relative efficiency of SAg presentation of various APC, thc same concentrations of TCC, LCL or PBMC incubated with SE for 1 h on ice, irradiated and extensively washed, were used as APC to compare the relative cfficiency of SE presentation to responder TCC. As shown in Fig. 5, TCC were less efficient than LCL and more
# APC
Figure 5. Proliferative response of 3 x lo4TCC (PM1.l) was used to compare the presentation ability of irradiated TCC, LCL or PBL pre-pulsed with 100 ng/ml of SEB, as detailed in Sect. 2.6. Pre-pulsed APC were diluted twofold starting from 4 X lo5 cells well.
Superantigen presentation by human T cell clones
Eur. J. Immunol. 1992. 22: 2033-2039
efficient than PBMC in presenting SE to the same responder TCC, when the number of pre-pulsed APC was lower then 2 x 104/well. However, when the number of prepulsed APC was greater than 2 x 104/well, the highest proliferative response was obtained with PBMC and the relative presentation efficiency was PBMC > LCL > TCC. The medium collected after the last washing of pre-pulsed APC did not show any mitogenic activity due to unbound SE, when tested with TCC or nonirradiated PBMC (data not shown), indicating that the proliferation observed with SE pre-pulsed APC was due to cell-bound SE. Taken together, these results indicate that a single cell of the clone effectively presented the SE to cells of the same clone. 3.2 SE-mediated cytotoxicity CD4+ TCC able to actively proliferate when stimulated with SEE and SED and CD8+ TCC able to proliferate in the presence of APC with SED but not with SEE (Table l), were tested for SE-induced cytotoxicity in a WR-release assay, using SEE- and SED-pre-pulsed labeled autologous LCL as target. As shown in Fig. 6, CD4+ TCC did not show any significant SE-mediated cytotoxic activity. On the other hand, CD8+ TCC showed a marked SE-mediated cytotoxicity only when the LCL were pre-pulsed with SED. 80
1
-20 ,
CD8 20 1
CD8 10 1
CD4 20 1
CD4 10 1
Figure 6. CD4+ TT-specific TCC did not show any significant SE-mediated cytotoxicity, even when target LCL were pre-pulsed with the same SE as that which induced a proliferative response. On the other hand, CD8+ TCC, showed an SE-mediated cytotoxicity when target LCL were pre-pulsed with an SE (SED) that induced a CDW TCC proliferation in the presence of APC.
3.3 Phenotypic modulation induced by SE activation of TCC
To determine whether the self presentation by TCC of SE was followed by the same phenotypic variation accompanying TCC proliferation in the presence of APC, TCC incubated with SEA or SEB were stained and analyzed by flow cytometry after 24 and 48 h (Fig. 7). Preliminary experiments had shown that an incubation of TCC with SEA or SEB for 1 h on ice, that was sufficient to induce proliferation, did not interfere with the binding of mAb. Thus, any variation of the phenotypic profile could be regarded as representing active modifications and not a shielding of relevant epitopes by SE. As depicted in Fig. 7,
TcR
DR
2037
IL-2 R
Fluorescence intensity Figure 7. Modulation of TCC surface antigens after a 48-h culture with SEA without APC.TCC were culturedinmedium B with SEA (100,lO or 1ng/ml) or SEB (1000,100 or 10 ng/ml) with or without APC. After 1, 24 or 48 h of culture, aliquots were washed and stained with mAb anti-TcR, anti-DR and anti-IL-2R as detailed in Sect. 2.8. Similar phenotypic variations were observed with SEB and in the presence of APC (PBMC). (background: -----; control without SEA: "...; SEA 1 ng/ml: .....'.....: SEA 100 ng/ml:
-1. SE stimulation induced modifications in the phenotypic profile of TCC, irrespective of the presence of added APC with a dose- and time-dependent up-regulation of DR and CD25, and a parallel down-regulation of TcR.
4 Discussion We demonstrate that TT-specific MHC class II+ CD4+ human T clones proliferate in the absence of added APC when stimulated with suitable concentrations of SE. The TCC proliferation in response to SE, obtained in the absence of additional APC, exhibited a dose-response curve comparable to that obtained with the addition of APC. Furthermore, the TCC self-presentation of SE induced a proliferation of the same order of magnitude as that obtained with the autologous PBMC presentation of TT. Anti-DR mAb inhibited at different degrees of efficiency the SE-induced proliferation both in the presence and in the absence of APC. Conversely, anti-DP or anti-DQ mAb did not interfere with SE-induced proliferation. These results indicate that the main constraint for SE presentation in our experiments is the expression of the D R isotype of MHC class I1 molecules [2-61. Finally, the TCC self-presentation of SE induced a series of phenotypic modifications on TCC that overlapped those occurring after the SE stimulation in the presence of APC. These data indicate that TCC present SE and provide sufficient accessory signals, thus acting as SAg-presenting cells. The differences in the SE presentation efficiency among TCC, LCL and PBMC, observed in pre-pulsing experiments, could be explained by differences in the expression of MHC class I1 molecules, that actually function as cellular receptors for SE [ 2 ] . Indeed, the expression of MHC class II- molecules/cell is the highest on LCL [35, 361 and the lowest on PBMC, due to the great number of MHC class II-cells in this latter cell population. However, PBMC showed the highest SE presentation efficacy that might reflect their highest capacity to provide ligands for TCC adhesion molecules [37] and co-stimulatory signals [38], in addition to the expression of MHC-antigen or SAg complexes. All together our results strongly suggest that cells of a given clone effectively present the SE to other cells of the
2038
Eur. J. Immunol. 1992. 22: 2033-2039
R. Nisini. P. M. Matricardi, A. Fattorossi et al.
retroviral SAg and “interact” [24] with other TcR+ T cells, becoming responsible for the observed deletion of T cells. In such a situation, CD4+ T cells would be the prominent SAg-presenting cells, as these cells are the main target of HIV Our results indicate that human CD4+ TCC present SAg and demonstrate that SAg-mediated interactions can take place among T cells. The outcome of this interaction was, in our experiments, the proliferation of responder TCC. However, it is not possible to exclude that the in vivo presentation of an HIV-encoded SAg by infected CD4+ Tcell might lead to deletion of SAg complementary Vp Tcells. Recent findings [19] suggest that an immune response to a SAg in vivo can eventually culminate in marked tolerance and clonal elimination. With a similar mechanism, activated HIV-infected T cells might present the hypothesized SAg, inducing first a clonal expansion and then a deletion of SAg-complementary Vp cells. What the functional consequences of the in vivo presentation by Tcells are and whether HIV-infected cells can deliver adequate accessory signals remains unknown. Anergy is a consequence of mature T cells confronting antigen on Our results are in line with previously reported [27, 281 inappropriate or damaged APC [42,43]. The hypothesized studies that emphasized the APC function of TCC [29]. HIV-SAg presentation in vivo by activated infected “daOur results, however, are in contrast with those reported by maged” CD4+ T cells might not provide adequate accessoO’Hehir and Lamb [39] who failed to obtain significant ry signals and induce anergy. A long-lasting anergic state, TCC proliferation upon SE stimulation in the absence of due to SAg-prolonged presentation in chronic HIV-infected APC. The reasons for such discrepancies are not clear, but patients, might lead to the elimination of anergic cells [19]. some points, regarding the TCC self-presentation of SE, Our results raise the possibility that MHC class II+ cells deserve consideration. First, a cytotoxic activity of theTCC expressing a SAg-complementary Vp may stimulate themused as SE-presenting cells should be excluded. It has been selves. The auto-stimulation of an HIV-infected cell might reported that IL-2-maintained antigen-specific CD4+ TCC deliver inappropriate signals due to a SAg-mediated TcR may acquire cytotoxic properties [40] while retaining the and MHC class I1 cross-linking, leading to anergy or cell antigen-induced proliferation capacity. It is conceivable death by apoptosis [44] that, in mature mouse T cells with that these clones might exert cytolytic activity toward cross-linked CD4+ molecules, has been shown to follow SE-presenting cells. If cytotoxicT cells are at the same time TcR stimulation [45]. The hypothesized HIV SAg, possibly SE-presenting and responder cells, proliferation might not in association with other retroviral products such as the result upon SE stimulation as a consequence of SE- CD4 ligand gp120, might mediate the auto-stimulation of mediated self-cytotoxicity [14]. This was not the case of infected T cell expressing the complementary Vb and lead TCC used for our experiments. Second, TCC express to a selective CD4+ T cell depletion with the mechanisms different levels of MHC class I1 molecules at different times that have been suggested in the panergic imnesia hypotheof the restimulation cycle, and the MHC class 11 molecules sis [46] of AIDS development. expresvion is the main constraint of SAg presentation ability. Thus, at different times after the last restimulation, In summary, the main finding in this report is that SAg TCC might present SE with a different efficiency and, as a mediate T-T cell interaction andor T cell auto-interaction. consequence, the SE optimal concentration must be care- If the presence of an HIV-encoded SAg produced in CD4+ fully chosen to avoid nonresponsiveness.Third, the binding T cells is confirmed, in future studies it will be critical to affinity of the SE to the MHC class I1 molecules [4] and to examine the consequences of the SAg-mediated T-T cell the V, families of the responder clones may differ among interaction, in order to understand the pathogenesis of the laboratories and also contribute to the observed discrepan- Vp-selective helper T cells depletion observed in AIDS cies [39]. patients.
same clone. This view is substantiated by results of pre-pulsing experiments showing that irradiated, SAg pre-pulsed TCC were able to stimulate nonirradiated, non-SAg-pulsed TCC. However, the possibility exists that, in addition to the T-T cell interaction, cells expressing the SAg-complementary V, may undergo a SAg-mediated auto-stimulation process. This is of particular interest because an auto-stimulated cell would modulate its membrane receptors and possibly decrease its ability of interacting with other cells. In keeping with this view, we observed that, at high concentrations, SE did not induce TCC proliferation, although they did induce a phenotypic modulation, similar to that observed in optimally SE-stimulated, proliferating TCC. After overnight incubation with high SE concentration, moreover, TCC were still viable and able to proliferate upon addition of IL-2 or APC, provided that the unbound SE was removed by extensive washings (R. Nisini, manuscript in preparation). This suggests that at high SE concentrations, the TCC autostimulation might be favored and result in the loss of the presentation capacity.
Several reports have clearly demonstrated that the activatioddepletion of T cells expressing a particular Vp is diagnostic for the presence of a SAg [20-24, 41].The recent finding that the Tcell depletion in AIDS patients is selective for certainY3 families [25] led to speculate on the possible pathogenetic role of a hypothetical HIV-encoded SAg. The possibility that an SAg might be responsible for a selective depletion of T cells in an adult acquired retroviral infection, raises the question of what condition might be decisive in determining whether the final outcome of SAg-mediated T cell activation is proliferation or deletion of SAg-complementary Vp-bearing cells. It has been hypothesized, that, upon activation, HIV-infected CD4+ T cells would express MHC class I1 molecules together with the
The authors would like to thank, K Barnaba and M . Paroli for helpful criticism and suggestion, A. Vincenti for his advice and continuous encouragement and A . Di Pietro and R. Vitalone for excellent technical assistance.
Received February 5, 1992; in revised form March 25, 1992.
5 References 1 Choi,Y., Kotzin, B., Herron, L., Callahn, J. and Marrack, P., Proc. Natl. Acad. Sci. USA 1989. 86: 8941. 2 Mollick; J. A . , Cook, R. G. and Rich, R. R., Science 1989.244: 817.
Eur. J. Immunol. 1992. 22: 2033-2039
3 Marrack, l? and Kappler, J., Science 1990. 248: 707. 4 Herrman, T., Accolla, R. and MacDonald, R., Eur. J. Immunol. 1989. 19: 2171. 5 Frascr, J. D., Nature 1989. 339: 221. 6 Fisher, H., Dohlsten, M., Lindvall, M., Sjorgren, H. and Carlsson, R., J. Immunol. 1989. 142: 3151. 7 Choi,Y., Herman, A , , Di Giusto, D.,Wade,T., Marrack, l? and Kappler, J., Science 1900. 346: 471. 8 Mollick, J. A . , Chintagumpala. M., Cook, R. G. and Rich, R. R., J. Immunol. 1991. 146: 463. 9 Fleischer, B., Schrezenmeier, H. and Conradt, P., Cell. Immunol. 1989. 120: 92. 10 Fleischer, B. and Schrezenmeier, H., J. Exp. Med. 1988. 167: 1697. 11 Kappler, J., Kotzin, B.. Herron, L., Gelfand. E. W., Bigler, R. D., Boylston, A , , Carrel, S., Posnett, D. N., Choi,Y. and Marrack, F!, Science 1989. 244: 811. 12 Fleischer. B., Gerardy-Schahn, R., Metzroth, B., Carrel, S., Gerlach, D. and Kiihler, W., J. Immunol. 1990. 146: 11. 13 Imanishi, K., Igarashi, H. and Uchiyama,T., J. Immunol. 1990. 145: 3170. 14 Dohlsten, M., Hedlund, G. and Kalland, T., Immunol. Today 1991. 12: 147. 15 Carlsson, R., Ficher, H. and Sjegren, O., J. Immunol. 1988, 140: 2484. 16 Hufnagle. W. 0.. Tremaine, M. T. and Betley, M., Infect. Immun. 1991. 59: 2126. 17 Pontzer, C. H., Russel, J. H. and Johnson, H. M., J. Immunol. 1989. 143: 280. 18 Kavabe,Y., Ochi, A., Nature 1991. 349: 245. 19 Webb. S. R.. Morris, C., Sprent, J., Cell 1990. 63: 1249. 20 Marrack, l?. Kushir, E., Kappler, J., Nature 1991. 349: 524. 21 Francke1,W. N . , Rudy, C., Coffin, J. M., Huber, B. T., Nature 1991. 349: 526. 22 Woodland, D. L., Happ, M. l?, Gollob, K. J., Palmer, E., Nature 1991. 349: 529. 23 Dyson, l? J., Knight, A . M., Fairchild, S., Simpson, E., Tomonari, K., Nature 1991. 349: 531. 24 Imberti, L., Sottini, A., Bettinardi, A., Puoti, M. and Primi, D., Science 1991. 254: 860. 25 Janaway. C.. Nature 1991. 349: 459.
Superantigen presentation by human T cell clones
2039
26 Evans, R. L., Faldetta,T. J., Humphreys, R. E., Pratt, D. M., Yunis, E. J. and Schlossman, S. F., J. Exp. Med. 1979. 148: 1440. 27 Lanzavecchia, A., Abrignani, S., Scheidegger, D., Obrist, R., Dorken, B. and Moldenhauer, G., J. Exp. Med. 1988. 1647: 345. 28 Lanzavecchia, A , , Roosnek, E., Gregory,T., Berman, l? and Abrignani, S., Nature 1988. 334: 530. 29 Hewitt, C. R. A. and Feldman, M., J. Immunol. 1989. 142: 1429. 30 Lampson, L. A. and Levy, R., J. Immunol. 1980. 125: 293. 31 Charron, D. J. and McDevitt, H. O., Fed. Proc. 1979. 38: 1280. 32 Van Voorhis,W. C., Steinman, R. M., Hair, L. S., Luban, J., Witmer, M. D., Koide, S., Cohn, Z. A , , J. Exp. Med. 1983. 158: 126. 33 Barnaba,V., Franco, A., Alberti, A . , Balsano, C., Benvenuto, R., Balsano, F., J. Immunol. 1989. 143: 2650. 34 Fattorossi, A , , Nisini, R., Pizzolo, J. G., D’Amelio, R.. Cytometry 1989. 10: 320. 35 Pawelec, G. and Biihring, H. J., Cell. Immunol. 1990. 127: 520. 36 Robbins, l?, Maino,V. C., Warner, N. L. and Brodsky, F. M., J. Immunol. 1988. 141: 1281. 37 Springer,T. A., Nature 1991. 346: 425. 38 Weaver, C. T. and Unanue, E. R., Immunol. Today 1990. 11: 49. 39 O’Hehir, R. E. and Lamb, J. R., Proc. Natl. Acad. Sci. USA 1990. 87: 8884. 40 Fleisher, D., Nature 1984. 308: 365. 41 Janeway, C. A., Yagi, J., Conrad, F!, Katz, M. E., Jones, B., Vroegop, S. and Buxser, S., Immunol. Rev. 1989. 107: 61. 42 Mueller, D. L., Jenkins, M. K., Schwartz, R. H., Annu. Rev. Immunol. 1989. 7: 445. 43 Schwartz, R. H., Cell 1989. 57: 1073. 44 Terai, C., Kornbluth, R. S., Pauza, D., Richman, D. D., Carson, D. A., J. Clin. Invest. 1991. 87: 1710. 45 Newell, M. K., Haughn, L. J., Maroun, C. R., Julius, M. H., Nature 1990. 347: 286. 46 Ascher, M. S. and Sheppard, 11. W., J. AIDS 1990. 3: 177.
