Eur. J. Immunol. 1992. 22: 2295-2302
Human CD4 Tcell clones discriminated by different APC
2295
Tony Wyss-Coray, Christian Brander. Karin Frutig and Werner J. Pichler
Discrimination of human CD4 Tcell clones based on their reactivity with antigen-presenting T cells*
Institute for Clinical Immunology, Inselspital, Bern
In this report,we describe the discrimination of humanT cell clones based on their reactivity with activated Tcells as antigen-presenting cells (APC). CD4+ T cell clones specific for peptide P30 of tetanus toxin (amino acids 947-967) and restricted to the DP4 molecule were established and tested for proliferation to peptide presented either by peripheral blood mononuclear cells (PBMC), Epstein-Barr virus (EBV)-transformed B cells or major histocompatibility complex (MHC) class 11-expressingTcells. We found two sets of T cell clones: one set proliferated to peptide presentation by PBMC, EBV-transformed B cell lines (EBV-B cells) and MHC class 11+T cells (termed T-responder clones), while the other set of clones was only stimulated to proliferate, if the peptide was presented by PBMC or EBV-B cells, but not by T cells (T-nonresponder clones). Nevertheless, these T-nonresponder clones recognized P30 also on T cells, as revealed by Ca2+influx.The discrimination of the clones was not due to different avidities of the T cell receptors (TcR) of individual clones for the MHC-peptide complex as T-responder and T-nonresponder clones had similar dose-response curves to P30 presented by fixed EBV-B cell lines. Addition of cytokines [interleukin (1L)-1, IL-2, IL-4 and interferony] did not change the proliferative response of the clones, which was consistent throughout an observation period of > 4 months. T-nonresponder clones, exposed to P30 on MHC class 11-expressing T cells, became not anergic, as they could be restimulated by P30 presented on EBV-B cells. The measurement of a panel of T cell activation markers and adhesion molecules on T-responder and T-nonresponder clones revealed a higher expression of the CD28 molecule on theT-nonresponder clones.The data suggest that freshly cloned T cells can be differentiated by peptide presentation on classical (PBMC, EBV-B cells) or non-classical APC (class TI+ Tcells), and that this discrimination is further underlined by different levels of adhesion molecules.
1 Introduction MHC class I1 expression occurs constitutively on classical antigen-presenting cells (APC) like Langerhans cells of the skin, dendritic cells, macrophages and B cells [l-31. Other human cell types like, e.g. pancreatic islet cells, keratinocytes or endothelial cells can be induced to transiently express class TI molecules [4-51. The function of these so-called non-classical APC is not yet clear. It has been shown that under certain conditions, these cells induce nonresponsiveness or anergy in specific Tcells, which is explained by the lack of certain co-stimulatory signals ~ 671. , Activated T lymphocytes also express MHC class I1 structures and are thus potential APC [8]. Indeed, humanTcell clones have been shown to present synthetic peptides that fit into their class I1 molecules to peptide-specific T cells [9]. Intact protein antigens have to bind to theTcell surface to be internalized, processed and presented afterwards, as has been shown for gp120 of human immunodeficiency virus and recently for hepatitis B virus [lo, 111. Also peptides or whole antigen, covalently coupled to anti-CD4 [I 104801
*
This work was supported by Swiss National Science Foundation grant No. 31-28814-90 and Swiss Foundation for AIDS research.
Correspondence: Werner J. Pichler, Institute for Clinical Immunology, Inscfspilal, CH-3010 Bern, Switzerland
0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1992
antibodies, are taken up, processed and presented byT cells [12]. The bulk of peptides presented by activated human T cells are probably derived fromTcell proteins. Data from S.Weiss and Bogen [I31 and F! Panina-Bordignon et al. [14] suggested that class 11+ cells present self peptides and various self peptides could recently be eluted from murine class 11 molecules 1151. Since the processing machinery of Tcells seems to be intact, it is likely that class 11+Tcells also present self peptides in vivo. The response of a specific T cell to peptide presentation on activated T cells may, however, be different from a response induced by classical APC. Indeed, Lamb et al. [16] have shown that high antigen doses might induce a state of prolonged hyporesponsiveness in cloned T cells although this was clearly not the case in the above-mentioned reports [9-121. It is controversial whether predefined subsets of CD4+ Tcells exist in vivo or whether different functions of CD4+ T cells are stimulated depending on the mode of activation. On one hand, T cells carrying the same TcR can be induced to secrete different lymphokines depending either on the APC [ 171, on different antigens like superantigens or peptides [IS], or on anti-CD3 mAb [19] or single amino acid changes in a relevant peptide [20]. On the other hand, CD4+ Tcell clones can be divided into Thl and Th2subsets based on the different cytokine pattern produced. T h l cells secrete IL-2 and IFN-y while Th2 cells produce IL-4, IL-5 and IL-6 upon stimulation 1211. This distinction is widely accepted among murine CD4+ T cells and recent data showed that already single positive thymocytes produce different patterns of cytokines [22]. The 0014-2980/92/0909-2295$3.50+ .25/0
T. Wyss-Coray, C. Brander, K. Frutig and W. J. Pichler
Eur. J. Immunol. 1992. 22: 2295-2302
existence of human CD4+ Tcell subsets is less clear, although human Thl and Th2-like Tcell clones could be isolated from a variety of diseases. For instance, allergenspecific T cells that differ not only in lymphokine pattern but also in expression of CD28 have recently been described [23]. Whether a subset of Tcells, the mode of activation or a combination of both leads to the observed cytokine patterns inT cell clones has to be further analyzed. I n this context, it is of interest that adhesion molecules and their counter-receptors may be decisive in the activation of different signaling pathways in a given Tcell [24, 251.
and 10 pg/ml streptomycin and 100 U/ml penicillin (Amimed, Basel, Switzerland).This medium was enriched with 10 U/ml nIL-2 and 10 U/ml rIL-2 to culture Tcell clones (CM+).TheFCS medium used to culture EBV-transformed Bcell lines was essentially the same as the CM without L-glutamine and transferrin, but with 10% FCS (Gibco, Paisley, Scotland) instead of human serum.
22%
In this report, we describe two types of T cell clones which are specific for the same peptide and recognize the same class 11 molecule but differ in their proliferative response depending on the peptide-presenting cell. Thereby the use of Tcells as APC helps to discriminate between these two types of T cell clones: one set of clones reacts with peptide presented by PBMC or EBV-B cells but not by T cells as APC (called T-nonresponders), while the other clones react to the peptide presented by PBMC, EBV-B cells and T cells (T-responders). These subsets express different levels of CD28, which may be important in signal transduction [23, 261.
2 Materials and methods 2.1 Materials The tetanus toxin-derived peptides P2 (aa 830-843) and P30 (aa 947-967) of tetanus toxin were obtained from Multiple Peptide Systems (San Diego, CA). Both peptides have been described by S. Demotz et al. [28, 281. Peptides tt 947-960 and tt 953-967 were kindly provided by G. P. Corradin (Lausanne. Switzerland). Most mAb were obtained from Becton Dickinson (Rutherford, NY), either unlabeled or labeled with FITC or PE except anti-CD28 (clone CLB402, Janssen, Belgium), anti-DP2DP4 (Readysysteme, Zurzach, Switzerland), antiICAM-1, anti-LFA-1 (Inotech, Marseille, France) and anti-LFA-3 (Serotec, Oxford, GB). The anti-B7 mAb was kindly provided by Dr. A. S. Freedman (Dana-Farber Cancer Institute, Boston, MA) [29]. The anti-TcR ap mAb BMA031 was a generous gift of Dr. R. Kurrle (BehringWerke, Marburg, FRG). Polyclonal FITC-labeled goat anti-mouse antibody was obtained from Coulter (Hialeah, FL) . Cytokines were used as recombinant proteins except nIL-2 which was a kind gift of Dr. U. Schwulera (Biotest, Frankfurt/Main, FRG). rIL-1 was obtained from Biotest, rIL-2 and rIL-4 were a generous gift of Dr. D.Wran (Sandoz Research Institute, Vienna, Austria), IFN-y was a kind donation from Dr. R. Adolf (Bohringer Research Institute, Vienna, Austria).
2.3 Isolation of Tcell clones and EBV-B cell lines PBMC were freshly isolated by density gradient centrifugation on Lymphoprep (Nycomed, Oslo, Norway) from donor CB. Bulk cultures were set up with 1 X 106irradiated PBMC (3000 rad) and 0.5 x 106fresh PBMC together with 5 p~ peptide P2 or P30 in CM in 24-well plates (Falcon No 3047, Becton Dickinson). After 6 days, bulk cultures were restimulated with the same peptide at 5 pM with 1 X lo6 irradiated autologous PBMC (normally after thawing frozen cells previously stored in liquid nitrogen) and cloned by limiting dilution after another 3 days. Blast cells were seeded with 25 000 irradiated autologous PBMC and 2 p~ peptide at 0.3 cells/well into 96-well round-bottom plates (Falcon No. 3077) in CM+. Growing wells were expanded in CM+, transferred to 24-well plates and specificity was determined using irradiated PBMC as APC together with the relevant peptide. Specific clones were maintained in culture by biweekly restimulation with autologous EBVB cells (irradiated with 6000 rad), irradiated autologous or allogeneic PMNC and 5 p~ relevant peptide in CM+ in 24-well plates and culture flasks (50 ml and 250 ml, Falcon). EBV-B cell lines of various HLA-typed donors were obtained by transformation with supernatant of the EBVproducing cell line B95-8 (obtained from Dr. D. NeumannHaefelin, Freiburg, FRG). Cyclosporin A (1 pglml; Sandoz, Basel, Switzerland) was added to the infected PMNC to abolish T cell reaction to EBV, and outgrowing B cells were cultured in FCS medium.
2.4 Immunofluorescence analysis For direct immunofluorescence analysis, cells at 1 x 106/ml were washed once with PBS, 1% FCS and 0.02% NaN3 at 4°C and reconstituted with 1 ml of the same buffer. Cells (100 pl) were then incubated with 0.1 pg mAb for 25 min at 0 "C. After washing twice with the above buffer, cells were either fixed (staining with directly labeled antibodies) or incubated again with an FITC-labeled polyclonal goat anti-mouse antibody (indirect labeling) for another 25 min, washed again and fixed subsequently with the above buffer supplemented with 2% paraformaldehyde. The staining was performed in 96-well round-bottom microtiter plates. Cells were analyzed on an EPICS Profile I1 flow cytometer (Coulter).
2.5 [Ca2+]i measurement 2.2 Culture media
The complete culture medium (CM) was RPMI 1640 supplemented with 10% pooled human AB serum (Swiss Red Cross, Bern, Switzerland), 25 mM Hepes buffer, 2 mM L-glutamine, 25 &ml transferrin (Biotest, Dreieich, FRG)
Cells were loaded with fura-2 (Molecular Probes, Eugene, OR) by incubating 0.3 nmol f~ra-2110~ cells for 30 min at 37 "C. After centrifugation, cells were resuspended in Hanks' buffered salt solution (HBSS) containing 1 mM CaC12. Changes in the cytosolic free calcium concentration
Eur. J. Immunol. 1992. 22: 2295-2302
2297
Human CD4 Tcell clones discriminated by different APC
([Ca2+Ii)were calculated as described byV vonTscharner et al. [30]and expressed as % fura-2 saturation. 2.6 MHC restriction of T cell clones and proliferation assays The restriction element of T cell clones was elucidated using the following homozygous EBV-B cell lines as APC to present the relevant peptide: HOM2 (DR1, DP4) and W151 (DP2) kindly provided by Dr. E. Roosneck (Geneva, Switzerland) and our own established EBV-B cell lines that were oligotyped by Dr. J.-M. Thiercy, Geneva, Switzerland. T cell clones were stimulated by either directly adding the soluble peptide at the indicated concentrations; or PBMC, EBV-B (B-APC) cells or cloned Tcells (T-APC) (10 days after restimulation in culture) were incubated with or without 10 WM peptide at a cell concentration of 2 X lo6 cells/ml in 0.5 ml CM for 3 h at 37°C. The cells were then washed twice with HBSS. They were either fixed for 15 s with 0.05% glutaraldehyde in HBSS followed by 0.2 M L-glycine to stop the reaction or irradiated with 3000 rad (PBMC and T cells) or 6000 rad (EBV-B cells, = pulsing). In certain experiments the APC were first fixed and then incubated with the peptide. Fifty thousand cells used as APC were incubated with 50 000 specific cloned T cells (washed twice with HBSS) in 96-well plates and pulsed for 6 h with [‘HI thymidine after 2 to 4 days. The cultures were then harvested on glass fiber disks and counted in a microplate p-counter (lnotech Filter Counting Systsem INB-384, Inotech, Dottikon, Switzerland). Data are shown from duplicate or triplicate cultures from one experiment which was considered representative if results were reproduced in two additional experiments. Where indicated, the average of different experiments is shown.The data are expressed as cpm or stimulation index (SI) which was calculated as experimental cpm divided by unstimulated background cpm.
3 Results 3.1 Stimulation of Tcell clones with different APC During our previous work with activated CD4+ Tcells as APC, we made the observation that specificT cell clones can respond directly to its specific peptide, implying that the Tcell would present the peptide to a neighboring cell with the same specificityand vice versa [9,12].There is, however, controversy as to whether all CD4+ Tcells can be stimulated in this way [31].We tested a panel of 22 independent CD4+ BMA031+ (TcR afi) Tcell clones specific for the tctanus toxin-derived peptide P30 (aa 947-967). Fig. 1 shows that all these clones proliferated specifically to the P30 peptide presented by irradiated autologous PBMC (or irradiated autologous EBV-B cells; data not shown), but only 14 clones proliferated if the peptide was added without additional APC. The same clones proliferated also if the CD4+ T cell clone CB06 (which recognizes another epitope of tetanus toxin) was pulsed with P30 (see Table 1). Since
these clones can proliferate to class 11-expressing T cells presenting the P30 peptide, they are termed “T-responder” clones. In contrast, although specific for P30, eight clones did not proliferate to addition of P30 alone or to P30presenting T cells (“T-nonresponder”) . HLA DP4 could be revealed as the restriction element for all clones, since P30 was recognized with the DP4+ HLA homozygous EBV-Bcell line HOM2, but not the DP2+ EBV-B cell line W151 and as solely antibodies to DP or DP2,DP4 but not antibodies to DR blocked the proliferation (data not shown). In addition, we found that the clones responded to the peptide tt 947-960 but not to tt 953-967 which are fragments of P30 ([28] and data not shown).
T cell (CBOS)
PBMC
2
g.
L“ +
nil
c4 c9 c11 C15 C18 c19 C25 C26 C29 C31 c3a c39 C40 C41
I
I
c24 C28 c37 + c43
f
$
0
10
20
30
40
50 0
10
20
30
50 0
40
I+
10
I 20
30
40
50
StimulationIndex
Figure 1. Proliferation of Tcell clones is dependent on different APC. Tcell clones (50000 cells/well) were stimulated in the presence or absence of 2 KM peptide P30 together with the same number of irradiated autologous PBMC (PBMC; left panel) or with the irradiated CD4+, MHC class 11+ LTcell clone CB06 (50000 cells/well) (Tcell (CB06); middle panel) or without additional APC (nil; right panel). Proliferation was assessed as [3H]thymidine incorporation after 48 h. Bars represent stimulation index as a ratio of stimulation in the presence or absence of peptide. Background proliferation of control cultures in the absence of P30 was < 200 cpm for stimulations with T-APCS or no APC and < 500 cpm for PBMC.
Table 1. C37 and C38 as APC“)
APC
Stimulation index c37 C38
ERV CB CB06 c37 C38
13 0.8 I .2 1
30.8 17 15.1 23
a) APC were incubated with or without 5 KM P30 for 3 h, irradiated and washed extensively, 50000 APC were cocultivated with 50000 cells of C37 (T-nonresponder) or C38 (T-responder).
T. Wyss-Coray, C. Brander, K. Frutig and W. J. Pichler
2208
3.2 Dose response of T-responder and T-nonresponder clones
To test whether different TcR avidities for the MHC-P30 complex account for the phenomenon of T-responder and T-nonresponder clones, we performed dose-response curves with a selected number of clones from the two groups. Therefore, autologous EBV-B cells were fixed with glutaraldehyde, incubated for 3 h with different amounts of P30 and then washed extensively.The proliferation of the clones added to these peptide pulsed Bcells is shown in Fig. 2. The clones differ in the peptide concentration required for maximal proliferation implying different TcR avidities, but there is no correlation with T-responder or T-nonresponder clones. Specifically, some T-nonresponder clones reacted already to minute amounts of the peptide (e.g. clone C24), suggesting a very high avidity of these clones, which nevertheless did not react to peptide presentation on T cells. 12 5
7
Eur. J. Immunol. 1992. 22: 2295-2302 100
I
i
--1.
0
'
01
I
,
I
.1
1
10
Cytokines (Ulml)
Figure 3. Effect of cytokines on proliferation of C37 and C38.The T-nonresponder clone C37 (open symbols) and the T-responder clone C38 (closed symbols) were stimulated with 2 PM P30 in the presence of the indicated concentrations of the cytokines IL-1 (&A), IL-2 (O,.), IL-4 or IFN-y ( 0 . W ) . Proliferation after 48 h is shown as stimulation index. Proliferation of C37 without P30 was 78-132 cpm and 580 cpm with the highest IL-2 concentration. Proliferation of C38 without the peptide ranged from 77 to 180 cpm and reached 1650 cpm with the highest IL-2 concentration.
(0,e)
IL-2 lead to a spontaneous, peptide-independent proliferation of the clones, which leads to a lower stimulation index. The peptide-specific proliferation of clones C37 and C38 was also not affected by exogenously added cytokines if the peptide was presented by PBMC or EBV-B cells (not shown). 0
-" .-
t C15
.-
3.4 T-nonresponder clones can present P30
E
0
0.16
0.8
4
20
100
Peptide concentration (pM)
Figure 2. Dose-response curves of T-responder and T-nonresponder clones to peptide P30. Two million EBV-B cells/ml were fixed with glutaraldehyde and incubated with the indicated concentrations of the peptide P30 for 3 h and subsequently washed. Fifty thousand of these peptide-pulsed EBV-B cells were incubated with 50000 T-responder clones (open symbols; upper graph) or Tnonresponder clones (closed symbols; lower graph). Proliferation was assessed as ['HI thymidine incorporation after 48 h.
3.3 Effect of cytokines on proliferation
To test whether the failure of T-nonresponder clones to proliferate to presentation by MHC class IIf Tcells could be overcome by the addition of cytokines, we added the Tcell growth factors IL-2 or IL-4, or IL-1 which have been shown to possess co-stimulatory activity for T cell activation, or IFN-y that would eventually enhance peptide presentation.We selected two representative clones, one of each group: C37 as a T-nonresponder clone and C38 as a T-responder clone. As depicted in Fig. 3, none of the cytokines added had a stirnulatory or an inhibitory effect on peptide-specific proliferation. Only high concentrations of
To study whether the T-nonresponder clone C37 itself was not, or only in a very limited manner, capable of presenting P30,we evaluated the APC function of clone C37 for other clones (Table 1). Both, C37 and C38 showed a similar intensity and percentage of DP molecules suggesting a similar number of DP molecules per cell (see Table 2). Moreover, clone C37 was able to function as APC, as P30-pulsed C37 cells could stimulate C38 cells. A similar level of proliferation of clone C38 also resulted if the CD4+ Table 2. Phenotype of C37 and C38a) c37
C38
97 65
56
94 63 65 92 94 86 98 82
CD28
46
7
B7 CDS
1 83
30
CD3 CD25 CD44 DP CD2 LFA-3 LFA- 1 ICAM-1
Y1
90 96 70 97
1
a) Expression of activation markers and adhesion molecules was measured over a period of 12 days after restimulation with EBV-Bcells and P30 on a cytofluorometer as described in Sect. 2.4. The maximal expression observed during this period (day 9) as % positive cells is shown.
Eur. J. Immunol. 1992. 22: 2295-2302
Human CD4 Tcell clones discriminated by different APC
2299
T cell clone CB06 presented P30. In contrast, clone C37 did not proliferate to P30 presented by C38 or CB06. 3.5 [Ca2+Iimobilization in clone C37 with free peptide To see whether C37 recognizes P30 when presented on T cells, we measured the mobilization of intracellular Ca2+ ions. Indeed, P30 added in soluble form to fura-2-loaded C37 cells lead to a dose-dependent increase in Ca2+influx (Fig. 4) which implicates recognition of P30. Since the peptide P2 (aa 830-843), which binds to DR4 on these clones, did not raise Ca2+influx, we suggest that a specific recognition of P30 via TcR interaction takes place. C38 showed a similar raise in intracellular [Ca2+]iafter administration of soluble P30 (data not shown).
.IL
4
3
-
-
-
9
6
12
5
Time (days)
Figure 5. Influence of time interval after last restimulation on the ability to respond to the peptide. The proliferation of Tnonresponder clone C37 (open symbols) and T-responder clone C38 (closed symbols) was assessed after different time intervals since the last restimulation.Twenty-five thousand T cells/well were stimulated with 5 p~ peptide P30 alone (&A) or together with irradiated autologous EBV-B cells (50 000 cells/well) and P30 (0,O). Proliferation was assessed as [3H]thymidine incorporation after 48 h and expressed as stimulation index. Figure 4. [Ca2 mobilization in clone C37. Clone C37 was loaded with fura-2 as described in Sect. 2.5 and 5 pM or 25 KM of the E 0 c peptide P30 was added to the cells at the indicated points. Addition .of P2 induced no [Ca2+],(data not shown). Mobilization of [Ca2+I1 5 is calculated as described in Sect. 2.5 and shown as % of fura-2 v) c saturation. r
A (T-APC)
z
nil
3.6 Effect of time after restimulation
I
C (B-APC)
The discrimination between T-responder and T-nonresponder clones was consistently reproducible in repetitive experiments throughout a time period of > 4 months. It was not dependent on the time after the last restimulation. Tcell clones C37 and C38 were tested at different time points after the last restimulation to proliferate either to P30 directly or to P30-pulsed EBV-Bcells. Both clones showed the highest stimulation index with B cell presentation after 12 days (Fig. 5 ) . The optimum proliferation of C38 induced by P30 alone was after about 9 days, while C37 responded marginally (maximal SI 3.5) at day 3, which might be due to some remaining B cells from the restimulation procedure (not shown). 3.7 T-nonresponder clones become not anergic by T cell presentation The data of Fig. 6 show that theT-nonresponder clone C28, first exposed to P30 on T-APC, was still able to react to P30 presented on EBV-Bcells. The clone was exposed to unpulsed or P30-pulsed activated Tcells (CB06) or EBVB cells. Only B cell presentation induced a vigorous proliferation. After 48 h the clone was restimulated with P30pulsed EBV-B cells. Clone C28 could still proliferate after previous stimulation with peptide-pulsed T-APC, while the cells were refractory if they were first stimulated with peptide pulsed B-APC. This was most likely due to the
nil
P30
P30
1
D(BAPC)
I
.5 c
-E =
0
'El N
BG
m
134 P30
117 P30
P30
P30
Figure 6. T-nonresponder clones do not become anergic by peptide presentation on Tcells. In a first stimulation, the T-nonresponder clone C28 was exposed either to medium or P30 (10 pM) presented by the CD4+ Tcell clone CB06 (T-APC, graph A) or by autologous EBV-B cells (B-APC, graph B). Proliferation is shown thymidine incorporation. After 48 h all cultures as cpm from H]"[ were harvested, washed and restimulated (graph C and D) with P30 pulsed EBV-B cells. Corresponding cultures are one column upon another. The proliferation is expressed as stimulation index. Background proliferation of control cultures in the absence of P30 is indicated at the bottom of the columns in graphs C and D.
activation and transient down-regulation of the TcR. This clearly shows that no anergy is induced in the nonproliferating T cell clone.
3.8 Phenotype of C37 and C38 To define the surface characteristics of the two different types of Tcell clones represented by C37 and C38, we
2300
T. Wyss-Coray. C. Brander, K. Frutig and W. J. Pichler
Eur. J. Immunol. 1992. 22: 2295-2302
followcd some activation markers and adhesion molecules expressed on the cell surface from day 3 to day 12 after restimulation with PBMC plus P30. Table 2 shows surface expression measured by cytofluorometry at day 9 after rcstimulation, when cells were mainly used for experiments.
individual. Whether the observed higher frequency of T-responder to T-nonresponder clones was due to our culture conditions and specifically the choice of a TT peptide instead of the whole protein as stimulus is presently under investigation.
Among the activation markers, the 55-kDa component of the IL-2 receptor (CD25) disappears in both clones similarly,while DP is constantly expressed. Down-regulation of the TcR/CD3 complcx, which had been recently ascribed to be responsible for antigen nonresponsiveness in some cells [32].is no more relevant > 5 days after restimulation, since CD3 was constantly expressed at a high level in both clones. Of the adhesion molecules tested, CD28 and CD5 were found to be most different between clone C37 and C38. CD5 was detected on 83Y0 of C37 cells but only on 30% of C38 cells. Similarly, CD28 was expressed on 46% of C37 cells but only on 7% of C38 cells at day 9. ICAM-1 also showed a different expression on the two clones, whereas LFA-1, LFA-2 and LFA-3 as well as the B7 marker were more or less similarly expressed on the two clones. In general, all T-nonresponder clones tested had a high expression of CD28 (range between 21% and 73%) at day 9, whereas most T-responder clones had a low CD28 expression with < 7% positive cells. Four T-responder clones, however, stained as strong as the T-nonresponder clones for the CD28 molecule. The expression of CD5 seem? to be less restricted to one subset (data not shown).
4 Discussion T cells can be distinguished on the basis of different T cell rcccptors for antigen (af~or y6), distinct effector functions, diffcrent adhesion molecules (CD4 or CD8) or different patterns of cytokine secretion [1-31. Here we present data showing that CD4+ T cell clones can be differentiated by their reactivity to distinct APC and that the pattern of adhesion molecules may differ between these clones.
The clones were found to be specific for aa 947-967 of tetanus toxin presented on the DP4 molecule and recognize the recently described DP-restricted fragment aa 947-960 of P30.The fine specificityof the clones might in part differ, since not all clones respond to native tetanus toxin processed by autologous B cells or PBMC (data not shown). But these differences do not correlate with the reactivity to Tcell presentation. Moreover, fixation of APC prior to addition of the peptide (Fig. 2) did not changeT-responder or T-nonresponder discrimination. Therefore, one can rule out that the discrimination between T-responder and T-nonresponder clones is due to an additional T o r B cellspecific processing of the peptide. Different avidities of theTcR for the P30/HLA-DP complex are also not responsible for the distinction of Tcell clones. At present, theTcR avidity of a given clone can be measured only by dose-response curves for the specific peptide. With this assay, no correlation between TcR avidity and the capacity to proliferate to T-APC was observed. T-responder clones proliferate not only to MHC class IIexpressing Tcells pulsed with P30 but also to P30 added directly to the clone without additional APC. This stimulation is not due to classical APC remaining from the restimulation of the clone during the regular cell culture (Fig. 4), but to the fact that the clones themselves express class I1 in very high densities and can thus present P30. Indeed, the use of PBMC as APC, which contain normally only about 20% class II-expressing cells, yielded often lower proliferative responses than T-APC with > 80% class II-expressing cells.
T-nonresponder clones never proliferated to P30 without the addition of classical APC. Nevertheless, they recognized the peptide on T-APC as shown by Ca2+ influx induced by P30 added to the T-nonresponder clone C37. This Ca2+ influx is not due to peptide binding to MHC Based on our data we discriminate between two types of class I1 molecules, as this would require cross-linking of the T cell clones: “T-nonresponder” clones proliferate to pep- MHC class I1 molecules [33], which does not occur with tide presentation by EBV-B cells and PBMC (classical soluble peptide. In addition, peptide P2, which is not APC) but not to presentation by MHC class 11-expressing recognized by our clones but binds to their DR4 molecules, T cells, while “T-responder” clones proliferate to classical does not induce Ca2+ influx. Moreover, T-nonresponder APC but also toT cells as APC.This distinction between the clones can function as APC, as the T-nonresponder clone cloncs was stable throughout an observation period of > 4 C37 (as well as others, not shown) could present P30 to the months, as a T-responder clone never becamc a T-nonres- T-responder clone C38 (Table 1).Thus, the discrimination ponder or vice versu. The description of these clones may of peptide-specific CD4+ T cell clones intoT-responder and explain the contradictory data reported earlier with human T-nonresponder clones is a feature of the clones themselves ‘Tccll clones, some of which were reported to proliferate to and not of the APC. It is thus feasible that also other exogenously added free antigen or peptide [9, 121, while non-classical APC besides activated, MHC class IIother cloncs failed to react to soluble peptide [31]. expressing T cells are suitable to discriminate between the two types of T cell clones. This study is based on 22 Tcell clones, which were generated by in vitro stimulation of PBMC from one The finding that T-nonresponder clones can recognize and individual with the synthetic tetanus toxin peptide P30. present the peptide, but are nevertheless not stimulated to T-responder and T-nonresponder clones could also be proliferate upon T cell presentation suggests that the Tobserved in earlier studies with another individual, using responder and T-nonresponder clones differ in their the tetanus toxin epitope P2 (aa 830-843),which shows that requirement to be fully activated. T-responder clones may this subdivision is not restricted to the peptide or a certain require fewer signals for activation, as they already prol-
Eur. J. Immunol. 1YY2. 22: 2205-2302
Human CD4 Tccll clones discriminatcd by diffcrent APC
ifcratc to pcptidc presentation by Tcells. In contrast, T-nonresponder cells. although some of them appear to have a very high avidity for the peptide/MHC complex. require more signals to bc activated. which are obviously not provided if peptide presentation occurs o n Tcells. As the addition of IL-1. 1L-2. 1L-4 or IFN-y could not overcome the unresponsiveness to T cell presentation (Fig. 3). interactions with adhcsion molecules (or eventually other cytokines) might be neccsary to inducc proliferation in T-nonresponder cells. Since T-nonresponder clones do not react to peptide presentation by T-responder clones. the expression of adhesion molcculcs on thc rcactive but not on the presenting Tcells seems to be decisive. In this context. it is interesting that we found a different pattcrn of adhesion molecules on C37 and C38 Tcell clones. which are representative for T-nonrespondcr or T-rcsponder clones. Of special interest may be the higher lcvcl of CD28 o n T-nonresponder cells. as this surface molcculc is known to transducc a co-stimulatory signal to theTcell that is distinct from the TcRKD.3 signal [26]. The counterpart of CD28. t h e B7 or BBl antigen [20]. is not present onTcells which indicates that this signal might not be transmitted inT-Tcell interactions or that another ligand for CD28 exists. Eventually. only T-nonresponder clones require the signal through the CD28 molecules. whilcT-rcspondcr cloncs may not nccd this interaction to be activated. The finding of these Tcells in humans is rcminiscent of recently dcscribcd murine Tcell hybrids. which differ in thcir rcquirements for activation [Wl. Onc T hybrid was activated by rccognizing the MHC-peptide complex only. while thc activation of the other Tcell hybrid was dependent on an additional LFA-1-ICAM-1 intcraction 1341. ICAM- 1 was also cnhanced in the T-nonresponder clone C.37 in comparison with the T-responder C.38. It has to be tested whether this molcculc is involved in the observed phenomenon. Therc is much controversy about the function of class I1 molecules on Tcells since they appear to he present only in very low numbers on murine Tcells 13.5. 361 and almost nothing is known about their function in vivo. In contrast. human Tcells can express high levels of MHC class 11 molecules. Somc MHC class I I + Tcells are normally found o n freshly isolated human cells cithcr from blood. skin o r lung lavagc and their number might be dramatically augmcnted in diseases.
I n our expcriments the use of Tcells as APC allows for the discrimination betwccn the two types ofclones.This implies that Tcells are not classical APC. Nevertheless. they are functionally distinct from other cells with aberrant exprcssion of MHC class I1 molcculcs 14-61. which were shown to inducc ancrgy. Pcptidc presentation by Tcells did not induce anergy as restimulation with peptide on B cells was possiblc (Fig. 6) and addition of IL-2. which was shown to prevent anergy induction 161. had no effect. Our data emphasize that Tcclls can function as APC and can induce a prolifcrativc immune response in certain (T-respondcr) Tccll clones. Pulsing of Tcells with peptidc also stimulatcs rcsting. freshly isolated PBMC (own obscr-
2301
vation). At first sight this responder Tcell reaction. elicited by Tcells as APC. cannot be distinguished from the Tcell reaction to classical APC. However. it is impossible to maintain T-responder clones without regular addition of feeder cells. containing classical APC (own obscrvation). Furthermore. stimulation of T-responder or T-nonresponder clones on P30 pulsed T-APC does not induce apoptosis as judged by DNA-fragmentation analysis (F. Bettens. personal communication). Thus, also in responder Tcell clones the Tcell APC induce discrete hut decisively different responses. which nccd further clarification.
In conclusion. we describe two types of CD4+ Tcell clones. which can be distinguished by their reactivity with different APC. Our data suggest that T-rcsponder and T-nonresponder cells differ in their ability to react to antigen, which is presented by either B cells or Tcells. The description of these two types of Tcell cloncs may be helpful in analyzing the optimum rcquircments for Tcell stimulation and activation. as well as for the understanding of the regulation of T cell immunity. specifically if thcsc differently reacting Tcells can also he dcmonstrated in t-ivo.We are currently following these questions.
5 References I Brodsky. F. M . and Guagliardi. L. E.. .4nnu. Krr: ltnrnuriol. I Y Y I . 9: 707. 2 Weaver.T.W. and Unnnue. E. R.. Imrnunol. 1990. 11: 4Y. 3 Yewdell. J. W. and Bcnnink. J. R.. c‘c~ll1990. 62: 203. 4 Ba1.V.. Mclndoe. A . . Denton. G.. Hudson. D.. Lomhardi. G.. Lamh. J. and Lcchlcr. K . . Eur. J. Irnmunol. IYYO. 20: 18Y3.
5 Markmann. J.. Lo. D.. Naji. D.. Palmiter. R. D.. Brinster. R. L. and tiebcr-Katz. E.. Nrrrz4r.c. 1988. .?.?6: 476. 6 Jcnkins. M. K . and Schwartz. R. H.. .I. Exp. Mrci. 1987. 16-5:
302.
D. R . . Uhrdahl. K. B. and Jenkins. M. K . . J. lmrnunol. 1YY 1. 147. 3261. 8 Evans. R. L. Fa1dctta.T. J.. Humphreys. R. E.. Pratt. D. M.. Yunis. E. J. and Schlossman. S. F..J. Exp. M i d . 1978.148: 1440. Y Hcwitt. C . K. A. and Fcldmann. M.. J. Imrnunol. IY89. 14?:
7 DeSilva.
762. 10 Frdnco. A . . Paroli. M..Tcsta. U.. Benvenuto. R.. Peschle. C.. Balsoni. F. and Barnaha.V.. J. E.vp Mrd. 1092. 175: 1195. 1 1 Lanzavecchia. A . . Roosncck. E.. (3regory.T.. Berman. P and Abrignani. S.. Nururc. IYXX. M4: 530. 12 Wyss-Cor;iy.T.. Brander. C.. Bettcns. F.. Mijic. D. and Pichler. W. J.. Ci4. lmmunol. I Y Y I . 139: 268. 13 Weiss. S. nnd Bogen. B.. (PI1 I90 1. 64: 767. 14 Panina-Bordignon. P. Corradin. G.. Roosnek. E.. Sette. A . and Lanzavecchin. A , . Science I Y Y 1. 252: 1548. 15 Rudensky. A.Y.. Prcston-Hurlhurt. P. Hong. S. C.. Barlow. A . and Jancway Jr. C . A . . Nururc. 1YYl. .?53: 622. 16 Lamb. J. R.. Skidmorc. B. J.. Green N.. Chiller. J. M. and Feldmann. M.. J. Exp. Mcd. 1983. 167: 1434. 17 Gajewski. T. F.. Pinnas. M.. Wing. T. and Fitch. F. W.. J. lmmunol. 1901. 146: 1750. 1x Patarca. R..Wei. F.Y.. Iregui. M.V. and Cantor. H . . Proc. Nut/. Acacl. Sci. USA I Y Y I . 88: 2736.
T. Wyss-Coray, C. Brander, K. Frutig and W. J. Pichler
Eur. J. Immunol. 1992. 22: 2295-2302
19 Wyss, T., Bettens, F., Walker, Ch. and Pichler, W. J., Cell. lmmunol. 1990. 126: 91. 20 Evavold, B. D. and Allen, F! M., Science 1991. 252: 1308. 21 Mosmann,T. R., Cherwinski, H., Bond, M.W., Giedlin, M. A. and Coffman, R. L., J. Immunol. 1986. 136: 2348. 22 Bendelac, A. and Schwartz, R. H., Nature 1991. 353: 68. 23 Kapsenberg, M. L. ,Wierenga, E. A., Bos, J. D. and Jansen, H. M., Immunol. Today. 1991. 11: 392. 24 DeFranco, A. L., Nature 1991. 351: 603. 25 De Soma, M., Tilney, N. L. and Kupiec-Weglinski, J. W., Immunol. Today 1991. 12: 262. 26 Ledbetter, J. A., Parsons, M., Martin, P. J., Hansen, J. A., Rabinovitch, P. S. and June, C. H., J. Immunol. 1986.137: 3299. 27 Demotz, S., Lanzavecchia, A., Eisel, U., Niemann, H . , Widmann, C. and Corradin, G., J. Imrnunol. 1989. 142: 394. 28 Panina-Bordignon, F!, Tan, A., Termijtelen, A., Demotz, S., Corradin, G. and Lanzavecchia, A., Eur. J. Immunol. 1989.19: 2237.
29 Linsley, P. S., Clark, E. A. and Ledbetter, J. A., Proc. Natl. Acad. Sci. U S A 1990. 87: 5031. 30 VonTscharner,V., Deranleau, D. A. and Baggiolini, M., J. Biol. Chem. 1986. 261: 10163. 31 Ho, F! C., Mutch, D. A.,Winkel, K. D., Saul, A. J., Jones, G. L., Doran,T. L. and Rzepczyk, C. M., Eur. J. Immunol. 1990. 20: 477. 32 Schonrich, G., Kalinke, U., Momburg, F., Malissen, M., Schmitt-Verhulst, A. M., Malissen, B., Hammerling, G. J. and Arnold, B., Cell 1991. 65: 293. 33 Bdum, N., Martin, P. J., Schieven, G. J., Hansen, J. A. and Ledbetter, J. A., Eur. J. Immunol. 1991. 21: 123. 34 St.-Pierre,Y and Watts,T. H., J. Immunol. 1991. 147: 2875. 35 Reske-Kunz, A. B., Reske, K. and Rude, E., J. Immunol. 1986. 136: 2033. 36 Heuer, J. and Kolsch, E., J. Immunol. 1985. 134: 4031.
2302
Human CD4 Tcell clones discriminated by different APC
2295
Tony Wyss-Coray, Christian Brander. Karin Frutig and Werner J. Pichler
Discrimination of human CD4 Tcell clones based on their reactivity with antigen-presenting T cells*
Institute for Clinical Immunology, Inselspital, Bern
In this report,we describe the discrimination of humanT cell clones based on their reactivity with activated Tcells as antigen-presenting cells (APC). CD4+ T cell clones specific for peptide P30 of tetanus toxin (amino acids 947-967) and restricted to the DP4 molecule were established and tested for proliferation to peptide presented either by peripheral blood mononuclear cells (PBMC), Epstein-Barr virus (EBV)-transformed B cells or major histocompatibility complex (MHC) class 11-expressingTcells. We found two sets of T cell clones: one set proliferated to peptide presentation by PBMC, EBV-transformed B cell lines (EBV-B cells) and MHC class 11+T cells (termed T-responder clones), while the other set of clones was only stimulated to proliferate, if the peptide was presented by PBMC or EBV-B cells, but not by T cells (T-nonresponder clones). Nevertheless, these T-nonresponder clones recognized P30 also on T cells, as revealed by Ca2+influx.The discrimination of the clones was not due to different avidities of the T cell receptors (TcR) of individual clones for the MHC-peptide complex as T-responder and T-nonresponder clones had similar dose-response curves to P30 presented by fixed EBV-B cell lines. Addition of cytokines [interleukin (1L)-1, IL-2, IL-4 and interferony] did not change the proliferative response of the clones, which was consistent throughout an observation period of > 4 months. T-nonresponder clones, exposed to P30 on MHC class 11-expressing T cells, became not anergic, as they could be restimulated by P30 presented on EBV-B cells. The measurement of a panel of T cell activation markers and adhesion molecules on T-responder and T-nonresponder clones revealed a higher expression of the CD28 molecule on theT-nonresponder clones.The data suggest that freshly cloned T cells can be differentiated by peptide presentation on classical (PBMC, EBV-B cells) or non-classical APC (class TI+ Tcells), and that this discrimination is further underlined by different levels of adhesion molecules.
1 Introduction MHC class I1 expression occurs constitutively on classical antigen-presenting cells (APC) like Langerhans cells of the skin, dendritic cells, macrophages and B cells [l-31. Other human cell types like, e.g. pancreatic islet cells, keratinocytes or endothelial cells can be induced to transiently express class TI molecules [4-51. The function of these so-called non-classical APC is not yet clear. It has been shown that under certain conditions, these cells induce nonresponsiveness or anergy in specific Tcells, which is explained by the lack of certain co-stimulatory signals ~ 671. , Activated T lymphocytes also express MHC class I1 structures and are thus potential APC [8]. Indeed, humanTcell clones have been shown to present synthetic peptides that fit into their class I1 molecules to peptide-specific T cells [9]. Intact protein antigens have to bind to theTcell surface to be internalized, processed and presented afterwards, as has been shown for gp120 of human immunodeficiency virus and recently for hepatitis B virus [lo, 111. Also peptides or whole antigen, covalently coupled to anti-CD4 [I 104801
*
This work was supported by Swiss National Science Foundation grant No. 31-28814-90 and Swiss Foundation for AIDS research.
Correspondence: Werner J. Pichler, Institute for Clinical Immunology, Inscfspilal, CH-3010 Bern, Switzerland
0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1992
antibodies, are taken up, processed and presented byT cells [12]. The bulk of peptides presented by activated human T cells are probably derived fromTcell proteins. Data from S.Weiss and Bogen [I31 and F! Panina-Bordignon et al. [14] suggested that class 11+ cells present self peptides and various self peptides could recently be eluted from murine class 11 molecules 1151. Since the processing machinery of Tcells seems to be intact, it is likely that class 11+Tcells also present self peptides in vivo. The response of a specific T cell to peptide presentation on activated T cells may, however, be different from a response induced by classical APC. Indeed, Lamb et al. [16] have shown that high antigen doses might induce a state of prolonged hyporesponsiveness in cloned T cells although this was clearly not the case in the above-mentioned reports [9-121. It is controversial whether predefined subsets of CD4+ Tcells exist in vivo or whether different functions of CD4+ T cells are stimulated depending on the mode of activation. On one hand, T cells carrying the same TcR can be induced to secrete different lymphokines depending either on the APC [ 171, on different antigens like superantigens or peptides [IS], or on anti-CD3 mAb [19] or single amino acid changes in a relevant peptide [20]. On the other hand, CD4+ Tcell clones can be divided into Thl and Th2subsets based on the different cytokine pattern produced. T h l cells secrete IL-2 and IFN-y while Th2 cells produce IL-4, IL-5 and IL-6 upon stimulation 1211. This distinction is widely accepted among murine CD4+ T cells and recent data showed that already single positive thymocytes produce different patterns of cytokines [22]. The 0014-2980/92/0909-2295$3.50+ .25/0
T. Wyss-Coray, C. Brander, K. Frutig and W. J. Pichler
Eur. J. Immunol. 1992. 22: 2295-2302
existence of human CD4+ Tcell subsets is less clear, although human Thl and Th2-like Tcell clones could be isolated from a variety of diseases. For instance, allergenspecific T cells that differ not only in lymphokine pattern but also in expression of CD28 have recently been described [23]. Whether a subset of Tcells, the mode of activation or a combination of both leads to the observed cytokine patterns inT cell clones has to be further analyzed. I n this context, it is of interest that adhesion molecules and their counter-receptors may be decisive in the activation of different signaling pathways in a given Tcell [24, 251.
and 10 pg/ml streptomycin and 100 U/ml penicillin (Amimed, Basel, Switzerland).This medium was enriched with 10 U/ml nIL-2 and 10 U/ml rIL-2 to culture Tcell clones (CM+).TheFCS medium used to culture EBV-transformed Bcell lines was essentially the same as the CM without L-glutamine and transferrin, but with 10% FCS (Gibco, Paisley, Scotland) instead of human serum.
22%
In this report, we describe two types of T cell clones which are specific for the same peptide and recognize the same class 11 molecule but differ in their proliferative response depending on the peptide-presenting cell. Thereby the use of Tcells as APC helps to discriminate between these two types of T cell clones: one set of clones reacts with peptide presented by PBMC or EBV-B cells but not by T cells as APC (called T-nonresponders), while the other clones react to the peptide presented by PBMC, EBV-B cells and T cells (T-responders). These subsets express different levels of CD28, which may be important in signal transduction [23, 261.
2 Materials and methods 2.1 Materials The tetanus toxin-derived peptides P2 (aa 830-843) and P30 (aa 947-967) of tetanus toxin were obtained from Multiple Peptide Systems (San Diego, CA). Both peptides have been described by S. Demotz et al. [28, 281. Peptides tt 947-960 and tt 953-967 were kindly provided by G. P. Corradin (Lausanne. Switzerland). Most mAb were obtained from Becton Dickinson (Rutherford, NY), either unlabeled or labeled with FITC or PE except anti-CD28 (clone CLB402, Janssen, Belgium), anti-DP2DP4 (Readysysteme, Zurzach, Switzerland), antiICAM-1, anti-LFA-1 (Inotech, Marseille, France) and anti-LFA-3 (Serotec, Oxford, GB). The anti-B7 mAb was kindly provided by Dr. A. S. Freedman (Dana-Farber Cancer Institute, Boston, MA) [29]. The anti-TcR ap mAb BMA031 was a generous gift of Dr. R. Kurrle (BehringWerke, Marburg, FRG). Polyclonal FITC-labeled goat anti-mouse antibody was obtained from Coulter (Hialeah, FL) . Cytokines were used as recombinant proteins except nIL-2 which was a kind gift of Dr. U. Schwulera (Biotest, Frankfurt/Main, FRG). rIL-1 was obtained from Biotest, rIL-2 and rIL-4 were a generous gift of Dr. D.Wran (Sandoz Research Institute, Vienna, Austria), IFN-y was a kind donation from Dr. R. Adolf (Bohringer Research Institute, Vienna, Austria).
2.3 Isolation of Tcell clones and EBV-B cell lines PBMC were freshly isolated by density gradient centrifugation on Lymphoprep (Nycomed, Oslo, Norway) from donor CB. Bulk cultures were set up with 1 X 106irradiated PBMC (3000 rad) and 0.5 x 106fresh PBMC together with 5 p~ peptide P2 or P30 in CM in 24-well plates (Falcon No 3047, Becton Dickinson). After 6 days, bulk cultures were restimulated with the same peptide at 5 pM with 1 X lo6 irradiated autologous PBMC (normally after thawing frozen cells previously stored in liquid nitrogen) and cloned by limiting dilution after another 3 days. Blast cells were seeded with 25 000 irradiated autologous PBMC and 2 p~ peptide at 0.3 cells/well into 96-well round-bottom plates (Falcon No. 3077) in CM+. Growing wells were expanded in CM+, transferred to 24-well plates and specificity was determined using irradiated PBMC as APC together with the relevant peptide. Specific clones were maintained in culture by biweekly restimulation with autologous EBVB cells (irradiated with 6000 rad), irradiated autologous or allogeneic PMNC and 5 p~ relevant peptide in CM+ in 24-well plates and culture flasks (50 ml and 250 ml, Falcon). EBV-B cell lines of various HLA-typed donors were obtained by transformation with supernatant of the EBVproducing cell line B95-8 (obtained from Dr. D. NeumannHaefelin, Freiburg, FRG). Cyclosporin A (1 pglml; Sandoz, Basel, Switzerland) was added to the infected PMNC to abolish T cell reaction to EBV, and outgrowing B cells were cultured in FCS medium.
2.4 Immunofluorescence analysis For direct immunofluorescence analysis, cells at 1 x 106/ml were washed once with PBS, 1% FCS and 0.02% NaN3 at 4°C and reconstituted with 1 ml of the same buffer. Cells (100 pl) were then incubated with 0.1 pg mAb for 25 min at 0 "C. After washing twice with the above buffer, cells were either fixed (staining with directly labeled antibodies) or incubated again with an FITC-labeled polyclonal goat anti-mouse antibody (indirect labeling) for another 25 min, washed again and fixed subsequently with the above buffer supplemented with 2% paraformaldehyde. The staining was performed in 96-well round-bottom microtiter plates. Cells were analyzed on an EPICS Profile I1 flow cytometer (Coulter).
2.5 [Ca2+]i measurement 2.2 Culture media
The complete culture medium (CM) was RPMI 1640 supplemented with 10% pooled human AB serum (Swiss Red Cross, Bern, Switzerland), 25 mM Hepes buffer, 2 mM L-glutamine, 25 &ml transferrin (Biotest, Dreieich, FRG)
Cells were loaded with fura-2 (Molecular Probes, Eugene, OR) by incubating 0.3 nmol f~ra-2110~ cells for 30 min at 37 "C. After centrifugation, cells were resuspended in Hanks' buffered salt solution (HBSS) containing 1 mM CaC12. Changes in the cytosolic free calcium concentration
Eur. J. Immunol. 1992. 22: 2295-2302
2297
Human CD4 Tcell clones discriminated by different APC
([Ca2+Ii)were calculated as described byV vonTscharner et al. [30]and expressed as % fura-2 saturation. 2.6 MHC restriction of T cell clones and proliferation assays The restriction element of T cell clones was elucidated using the following homozygous EBV-B cell lines as APC to present the relevant peptide: HOM2 (DR1, DP4) and W151 (DP2) kindly provided by Dr. E. Roosneck (Geneva, Switzerland) and our own established EBV-B cell lines that were oligotyped by Dr. J.-M. Thiercy, Geneva, Switzerland. T cell clones were stimulated by either directly adding the soluble peptide at the indicated concentrations; or PBMC, EBV-B (B-APC) cells or cloned Tcells (T-APC) (10 days after restimulation in culture) were incubated with or without 10 WM peptide at a cell concentration of 2 X lo6 cells/ml in 0.5 ml CM for 3 h at 37°C. The cells were then washed twice with HBSS. They were either fixed for 15 s with 0.05% glutaraldehyde in HBSS followed by 0.2 M L-glycine to stop the reaction or irradiated with 3000 rad (PBMC and T cells) or 6000 rad (EBV-B cells, = pulsing). In certain experiments the APC were first fixed and then incubated with the peptide. Fifty thousand cells used as APC were incubated with 50 000 specific cloned T cells (washed twice with HBSS) in 96-well plates and pulsed for 6 h with [‘HI thymidine after 2 to 4 days. The cultures were then harvested on glass fiber disks and counted in a microplate p-counter (lnotech Filter Counting Systsem INB-384, Inotech, Dottikon, Switzerland). Data are shown from duplicate or triplicate cultures from one experiment which was considered representative if results were reproduced in two additional experiments. Where indicated, the average of different experiments is shown.The data are expressed as cpm or stimulation index (SI) which was calculated as experimental cpm divided by unstimulated background cpm.
3 Results 3.1 Stimulation of Tcell clones with different APC During our previous work with activated CD4+ Tcells as APC, we made the observation that specificT cell clones can respond directly to its specific peptide, implying that the Tcell would present the peptide to a neighboring cell with the same specificityand vice versa [9,12].There is, however, controversy as to whether all CD4+ Tcells can be stimulated in this way [31].We tested a panel of 22 independent CD4+ BMA031+ (TcR afi) Tcell clones specific for the tctanus toxin-derived peptide P30 (aa 947-967). Fig. 1 shows that all these clones proliferated specifically to the P30 peptide presented by irradiated autologous PBMC (or irradiated autologous EBV-B cells; data not shown), but only 14 clones proliferated if the peptide was added without additional APC. The same clones proliferated also if the CD4+ T cell clone CB06 (which recognizes another epitope of tetanus toxin) was pulsed with P30 (see Table 1). Since
these clones can proliferate to class 11-expressing T cells presenting the P30 peptide, they are termed “T-responder” clones. In contrast, although specific for P30, eight clones did not proliferate to addition of P30 alone or to P30presenting T cells (“T-nonresponder”) . HLA DP4 could be revealed as the restriction element for all clones, since P30 was recognized with the DP4+ HLA homozygous EBV-Bcell line HOM2, but not the DP2+ EBV-B cell line W151 and as solely antibodies to DP or DP2,DP4 but not antibodies to DR blocked the proliferation (data not shown). In addition, we found that the clones responded to the peptide tt 947-960 but not to tt 953-967 which are fragments of P30 ([28] and data not shown).
T cell (CBOS)
PBMC
2
g.
L“ +
nil
c4 c9 c11 C15 C18 c19 C25 C26 C29 C31 c3a c39 C40 C41
I
I
c24 C28 c37 + c43
f
$
0
10
20
30
40
50 0
10
20
30
50 0
40
I+
10
I 20
30
40
50
StimulationIndex
Figure 1. Proliferation of Tcell clones is dependent on different APC. Tcell clones (50000 cells/well) were stimulated in the presence or absence of 2 KM peptide P30 together with the same number of irradiated autologous PBMC (PBMC; left panel) or with the irradiated CD4+, MHC class 11+ LTcell clone CB06 (50000 cells/well) (Tcell (CB06); middle panel) or without additional APC (nil; right panel). Proliferation was assessed as [3H]thymidine incorporation after 48 h. Bars represent stimulation index as a ratio of stimulation in the presence or absence of peptide. Background proliferation of control cultures in the absence of P30 was < 200 cpm for stimulations with T-APCS or no APC and < 500 cpm for PBMC.
Table 1. C37 and C38 as APC“)
APC
Stimulation index c37 C38
ERV CB CB06 c37 C38
13 0.8 I .2 1
30.8 17 15.1 23
a) APC were incubated with or without 5 KM P30 for 3 h, irradiated and washed extensively, 50000 APC were cocultivated with 50000 cells of C37 (T-nonresponder) or C38 (T-responder).
T. Wyss-Coray, C. Brander, K. Frutig and W. J. Pichler
2208
3.2 Dose response of T-responder and T-nonresponder clones
To test whether different TcR avidities for the MHC-P30 complex account for the phenomenon of T-responder and T-nonresponder clones, we performed dose-response curves with a selected number of clones from the two groups. Therefore, autologous EBV-B cells were fixed with glutaraldehyde, incubated for 3 h with different amounts of P30 and then washed extensively.The proliferation of the clones added to these peptide pulsed Bcells is shown in Fig. 2. The clones differ in the peptide concentration required for maximal proliferation implying different TcR avidities, but there is no correlation with T-responder or T-nonresponder clones. Specifically, some T-nonresponder clones reacted already to minute amounts of the peptide (e.g. clone C24), suggesting a very high avidity of these clones, which nevertheless did not react to peptide presentation on T cells. 12 5
7
Eur. J. Immunol. 1992. 22: 2295-2302 100
I
i
--1.
0
'
01
I
,
I
.1
1
10
Cytokines (Ulml)
Figure 3. Effect of cytokines on proliferation of C37 and C38.The T-nonresponder clone C37 (open symbols) and the T-responder clone C38 (closed symbols) were stimulated with 2 PM P30 in the presence of the indicated concentrations of the cytokines IL-1 (&A), IL-2 (O,.), IL-4 or IFN-y ( 0 . W ) . Proliferation after 48 h is shown as stimulation index. Proliferation of C37 without P30 was 78-132 cpm and 580 cpm with the highest IL-2 concentration. Proliferation of C38 without the peptide ranged from 77 to 180 cpm and reached 1650 cpm with the highest IL-2 concentration.
(0,e)
IL-2 lead to a spontaneous, peptide-independent proliferation of the clones, which leads to a lower stimulation index. The peptide-specific proliferation of clones C37 and C38 was also not affected by exogenously added cytokines if the peptide was presented by PBMC or EBV-B cells (not shown). 0
-" .-
t C15
.-
3.4 T-nonresponder clones can present P30
E
0
0.16
0.8
4
20
100
Peptide concentration (pM)
Figure 2. Dose-response curves of T-responder and T-nonresponder clones to peptide P30. Two million EBV-B cells/ml were fixed with glutaraldehyde and incubated with the indicated concentrations of the peptide P30 for 3 h and subsequently washed. Fifty thousand of these peptide-pulsed EBV-B cells were incubated with 50000 T-responder clones (open symbols; upper graph) or Tnonresponder clones (closed symbols; lower graph). Proliferation was assessed as ['HI thymidine incorporation after 48 h.
3.3 Effect of cytokines on proliferation
To test whether the failure of T-nonresponder clones to proliferate to presentation by MHC class IIf Tcells could be overcome by the addition of cytokines, we added the Tcell growth factors IL-2 or IL-4, or IL-1 which have been shown to possess co-stimulatory activity for T cell activation, or IFN-y that would eventually enhance peptide presentation.We selected two representative clones, one of each group: C37 as a T-nonresponder clone and C38 as a T-responder clone. As depicted in Fig. 3, none of the cytokines added had a stirnulatory or an inhibitory effect on peptide-specific proliferation. Only high concentrations of
To study whether the T-nonresponder clone C37 itself was not, or only in a very limited manner, capable of presenting P30,we evaluated the APC function of clone C37 for other clones (Table 1). Both, C37 and C38 showed a similar intensity and percentage of DP molecules suggesting a similar number of DP molecules per cell (see Table 2). Moreover, clone C37 was able to function as APC, as P30-pulsed C37 cells could stimulate C38 cells. A similar level of proliferation of clone C38 also resulted if the CD4+ Table 2. Phenotype of C37 and C38a) c37
C38
97 65
56
94 63 65 92 94 86 98 82
CD28
46
7
B7 CDS
1 83
30
CD3 CD25 CD44 DP CD2 LFA-3 LFA- 1 ICAM-1
Y1
90 96 70 97
1
a) Expression of activation markers and adhesion molecules was measured over a period of 12 days after restimulation with EBV-Bcells and P30 on a cytofluorometer as described in Sect. 2.4. The maximal expression observed during this period (day 9) as % positive cells is shown.
Eur. J. Immunol. 1992. 22: 2295-2302
Human CD4 Tcell clones discriminated by different APC
2299
T cell clone CB06 presented P30. In contrast, clone C37 did not proliferate to P30 presented by C38 or CB06. 3.5 [Ca2+Iimobilization in clone C37 with free peptide To see whether C37 recognizes P30 when presented on T cells, we measured the mobilization of intracellular Ca2+ ions. Indeed, P30 added in soluble form to fura-2-loaded C37 cells lead to a dose-dependent increase in Ca2+influx (Fig. 4) which implicates recognition of P30. Since the peptide P2 (aa 830-843), which binds to DR4 on these clones, did not raise Ca2+influx, we suggest that a specific recognition of P30 via TcR interaction takes place. C38 showed a similar raise in intracellular [Ca2+]iafter administration of soluble P30 (data not shown).
.IL
4
3
-
-
-
9
6
12
5
Time (days)
Figure 5. Influence of time interval after last restimulation on the ability to respond to the peptide. The proliferation of Tnonresponder clone C37 (open symbols) and T-responder clone C38 (closed symbols) was assessed after different time intervals since the last restimulation.Twenty-five thousand T cells/well were stimulated with 5 p~ peptide P30 alone (&A) or together with irradiated autologous EBV-B cells (50 000 cells/well) and P30 (0,O). Proliferation was assessed as [3H]thymidine incorporation after 48 h and expressed as stimulation index. Figure 4. [Ca2 mobilization in clone C37. Clone C37 was loaded with fura-2 as described in Sect. 2.5 and 5 pM or 25 KM of the E 0 c peptide P30 was added to the cells at the indicated points. Addition .of P2 induced no [Ca2+],(data not shown). Mobilization of [Ca2+I1 5 is calculated as described in Sect. 2.5 and shown as % of fura-2 v) c saturation. r
A (T-APC)
z
nil
3.6 Effect of time after restimulation
I
C (B-APC)
The discrimination between T-responder and T-nonresponder clones was consistently reproducible in repetitive experiments throughout a time period of > 4 months. It was not dependent on the time after the last restimulation. Tcell clones C37 and C38 were tested at different time points after the last restimulation to proliferate either to P30 directly or to P30-pulsed EBV-Bcells. Both clones showed the highest stimulation index with B cell presentation after 12 days (Fig. 5 ) . The optimum proliferation of C38 induced by P30 alone was after about 9 days, while C37 responded marginally (maximal SI 3.5) at day 3, which might be due to some remaining B cells from the restimulation procedure (not shown). 3.7 T-nonresponder clones become not anergic by T cell presentation The data of Fig. 6 show that theT-nonresponder clone C28, first exposed to P30 on T-APC, was still able to react to P30 presented on EBV-Bcells. The clone was exposed to unpulsed or P30-pulsed activated Tcells (CB06) or EBVB cells. Only B cell presentation induced a vigorous proliferation. After 48 h the clone was restimulated with P30pulsed EBV-B cells. Clone C28 could still proliferate after previous stimulation with peptide-pulsed T-APC, while the cells were refractory if they were first stimulated with peptide pulsed B-APC. This was most likely due to the
nil
P30
P30
1
D(BAPC)
I
.5 c
-E =
0
'El N
BG
m
134 P30
117 P30
P30
P30
Figure 6. T-nonresponder clones do not become anergic by peptide presentation on Tcells. In a first stimulation, the T-nonresponder clone C28 was exposed either to medium or P30 (10 pM) presented by the CD4+ Tcell clone CB06 (T-APC, graph A) or by autologous EBV-B cells (B-APC, graph B). Proliferation is shown thymidine incorporation. After 48 h all cultures as cpm from H]"[ were harvested, washed and restimulated (graph C and D) with P30 pulsed EBV-B cells. Corresponding cultures are one column upon another. The proliferation is expressed as stimulation index. Background proliferation of control cultures in the absence of P30 is indicated at the bottom of the columns in graphs C and D.
activation and transient down-regulation of the TcR. This clearly shows that no anergy is induced in the nonproliferating T cell clone.
3.8 Phenotype of C37 and C38 To define the surface characteristics of the two different types of Tcell clones represented by C37 and C38, we
2300
T. Wyss-Coray. C. Brander, K. Frutig and W. J. Pichler
Eur. J. Immunol. 1992. 22: 2295-2302
followcd some activation markers and adhesion molecules expressed on the cell surface from day 3 to day 12 after restimulation with PBMC plus P30. Table 2 shows surface expression measured by cytofluorometry at day 9 after rcstimulation, when cells were mainly used for experiments.
individual. Whether the observed higher frequency of T-responder to T-nonresponder clones was due to our culture conditions and specifically the choice of a TT peptide instead of the whole protein as stimulus is presently under investigation.
Among the activation markers, the 55-kDa component of the IL-2 receptor (CD25) disappears in both clones similarly,while DP is constantly expressed. Down-regulation of the TcR/CD3 complcx, which had been recently ascribed to be responsible for antigen nonresponsiveness in some cells [32].is no more relevant > 5 days after restimulation, since CD3 was constantly expressed at a high level in both clones. Of the adhesion molecules tested, CD28 and CD5 were found to be most different between clone C37 and C38. CD5 was detected on 83Y0 of C37 cells but only on 30% of C38 cells. Similarly, CD28 was expressed on 46% of C37 cells but only on 7% of C38 cells at day 9. ICAM-1 also showed a different expression on the two clones, whereas LFA-1, LFA-2 and LFA-3 as well as the B7 marker were more or less similarly expressed on the two clones. In general, all T-nonresponder clones tested had a high expression of CD28 (range between 21% and 73%) at day 9, whereas most T-responder clones had a low CD28 expression with < 7% positive cells. Four T-responder clones, however, stained as strong as the T-nonresponder clones for the CD28 molecule. The expression of CD5 seem? to be less restricted to one subset (data not shown).
4 Discussion T cells can be distinguished on the basis of different T cell rcccptors for antigen (af~or y6), distinct effector functions, diffcrent adhesion molecules (CD4 or CD8) or different patterns of cytokine secretion [1-31. Here we present data showing that CD4+ T cell clones can be differentiated by their reactivity to distinct APC and that the pattern of adhesion molecules may differ between these clones.
The clones were found to be specific for aa 947-967 of tetanus toxin presented on the DP4 molecule and recognize the recently described DP-restricted fragment aa 947-960 of P30.The fine specificityof the clones might in part differ, since not all clones respond to native tetanus toxin processed by autologous B cells or PBMC (data not shown). But these differences do not correlate with the reactivity to Tcell presentation. Moreover, fixation of APC prior to addition of the peptide (Fig. 2) did not changeT-responder or T-nonresponder discrimination. Therefore, one can rule out that the discrimination between T-responder and T-nonresponder clones is due to an additional T o r B cellspecific processing of the peptide. Different avidities of theTcR for the P30/HLA-DP complex are also not responsible for the distinction of Tcell clones. At present, theTcR avidity of a given clone can be measured only by dose-response curves for the specific peptide. With this assay, no correlation between TcR avidity and the capacity to proliferate to T-APC was observed. T-responder clones proliferate not only to MHC class IIexpressing Tcells pulsed with P30 but also to P30 added directly to the clone without additional APC. This stimulation is not due to classical APC remaining from the restimulation of the clone during the regular cell culture (Fig. 4), but to the fact that the clones themselves express class I1 in very high densities and can thus present P30. Indeed, the use of PBMC as APC, which contain normally only about 20% class II-expressing cells, yielded often lower proliferative responses than T-APC with > 80% class II-expressing cells.
T-nonresponder clones never proliferated to P30 without the addition of classical APC. Nevertheless, they recognized the peptide on T-APC as shown by Ca2+ influx induced by P30 added to the T-nonresponder clone C37. This Ca2+ influx is not due to peptide binding to MHC Based on our data we discriminate between two types of class I1 molecules, as this would require cross-linking of the T cell clones: “T-nonresponder” clones proliferate to pep- MHC class I1 molecules [33], which does not occur with tide presentation by EBV-B cells and PBMC (classical soluble peptide. In addition, peptide P2, which is not APC) but not to presentation by MHC class 11-expressing recognized by our clones but binds to their DR4 molecules, T cells, while “T-responder” clones proliferate to classical does not induce Ca2+ influx. Moreover, T-nonresponder APC but also toT cells as APC.This distinction between the clones can function as APC, as the T-nonresponder clone cloncs was stable throughout an observation period of > 4 C37 (as well as others, not shown) could present P30 to the months, as a T-responder clone never becamc a T-nonres- T-responder clone C38 (Table 1).Thus, the discrimination ponder or vice versu. The description of these clones may of peptide-specific CD4+ T cell clones intoT-responder and explain the contradictory data reported earlier with human T-nonresponder clones is a feature of the clones themselves ‘Tccll clones, some of which were reported to proliferate to and not of the APC. It is thus feasible that also other exogenously added free antigen or peptide [9, 121, while non-classical APC besides activated, MHC class IIother cloncs failed to react to soluble peptide [31]. expressing T cells are suitable to discriminate between the two types of T cell clones. This study is based on 22 Tcell clones, which were generated by in vitro stimulation of PBMC from one The finding that T-nonresponder clones can recognize and individual with the synthetic tetanus toxin peptide P30. present the peptide, but are nevertheless not stimulated to T-responder and T-nonresponder clones could also be proliferate upon T cell presentation suggests that the Tobserved in earlier studies with another individual, using responder and T-nonresponder clones differ in their the tetanus toxin epitope P2 (aa 830-843),which shows that requirement to be fully activated. T-responder clones may this subdivision is not restricted to the peptide or a certain require fewer signals for activation, as they already prol-
Eur. J. Immunol. 1YY2. 22: 2205-2302
Human CD4 Tccll clones discriminatcd by diffcrent APC
ifcratc to pcptidc presentation by Tcells. In contrast, T-nonresponder cells. although some of them appear to have a very high avidity for the peptide/MHC complex. require more signals to bc activated. which are obviously not provided if peptide presentation occurs o n Tcells. As the addition of IL-1. 1L-2. 1L-4 or IFN-y could not overcome the unresponsiveness to T cell presentation (Fig. 3). interactions with adhcsion molecules (or eventually other cytokines) might be neccsary to inducc proliferation in T-nonresponder cells. Since T-nonresponder clones do not react to peptide presentation by T-responder clones. the expression of adhesion molcculcs on thc rcactive but not on the presenting Tcells seems to be decisive. In this context. it is interesting that we found a different pattcrn of adhesion molecules on C37 and C38 Tcell clones. which are representative for T-nonrespondcr or T-rcsponder clones. Of special interest may be the higher lcvcl of CD28 o n T-nonresponder cells. as this surface molcculc is known to transducc a co-stimulatory signal to theTcell that is distinct from the TcRKD.3 signal [26]. The counterpart of CD28. t h e B7 or BBl antigen [20]. is not present onTcells which indicates that this signal might not be transmitted inT-Tcell interactions or that another ligand for CD28 exists. Eventually. only T-nonresponder clones require the signal through the CD28 molecules. whilcT-rcspondcr cloncs may not nccd this interaction to be activated. The finding of these Tcells in humans is rcminiscent of recently dcscribcd murine Tcell hybrids. which differ in thcir rcquirements for activation [Wl. Onc T hybrid was activated by rccognizing the MHC-peptide complex only. while thc activation of the other Tcell hybrid was dependent on an additional LFA-1-ICAM-1 intcraction 1341. ICAM- 1 was also cnhanced in the T-nonresponder clone C.37 in comparison with the T-responder C.38. It has to be tested whether this molcculc is involved in the observed phenomenon. Therc is much controversy about the function of class I1 molecules on Tcells since they appear to he present only in very low numbers on murine Tcells 13.5. 361 and almost nothing is known about their function in vivo. In contrast. human Tcells can express high levels of MHC class 11 molecules. Somc MHC class I I + Tcells are normally found o n freshly isolated human cells cithcr from blood. skin o r lung lavagc and their number might be dramatically augmcnted in diseases.
I n our expcriments the use of Tcells as APC allows for the discrimination betwccn the two types ofclones.This implies that Tcells are not classical APC. Nevertheless. they are functionally distinct from other cells with aberrant exprcssion of MHC class I1 molcculcs 14-61. which were shown to inducc ancrgy. Pcptidc presentation by Tcells did not induce anergy as restimulation with peptide on B cells was possiblc (Fig. 6) and addition of IL-2. which was shown to prevent anergy induction 161. had no effect. Our data emphasize that Tcclls can function as APC and can induce a prolifcrativc immune response in certain (T-respondcr) Tccll clones. Pulsing of Tcells with peptidc also stimulatcs rcsting. freshly isolated PBMC (own obscr-
2301
vation). At first sight this responder Tcell reaction. elicited by Tcells as APC. cannot be distinguished from the Tcell reaction to classical APC. However. it is impossible to maintain T-responder clones without regular addition of feeder cells. containing classical APC (own obscrvation). Furthermore. stimulation of T-responder or T-nonresponder clones on P30 pulsed T-APC does not induce apoptosis as judged by DNA-fragmentation analysis (F. Bettens. personal communication). Thus, also in responder Tcell clones the Tcell APC induce discrete hut decisively different responses. which nccd further clarification.
In conclusion. we describe two types of CD4+ Tcell clones. which can be distinguished by their reactivity with different APC. Our data suggest that T-rcsponder and T-nonresponder cells differ in their ability to react to antigen, which is presented by either B cells or Tcells. The description of these two types of Tcell cloncs may be helpful in analyzing the optimum rcquircments for Tcell stimulation and activation. as well as for the understanding of the regulation of T cell immunity. specifically if thcsc differently reacting Tcells can also he dcmonstrated in t-ivo.We are currently following these questions.
5 References I Brodsky. F. M . and Guagliardi. L. E.. .4nnu. Krr: ltnrnuriol. I Y Y I . 9: 707. 2 Weaver.T.W. and Unnnue. E. R.. Imrnunol. 1990. 11: 4Y. 3 Yewdell. J. W. and Bcnnink. J. R.. c‘c~ll1990. 62: 203. 4 Ba1.V.. Mclndoe. A . . Denton. G.. Hudson. D.. Lomhardi. G.. Lamh. J. and Lcchlcr. K . . Eur. J. Irnmunol. IYYO. 20: 18Y3.
5 Markmann. J.. Lo. D.. Naji. D.. Palmiter. R. D.. Brinster. R. L. and tiebcr-Katz. E.. Nrrrz4r.c. 1988. .?.?6: 476. 6 Jcnkins. M. K . and Schwartz. R. H.. .I. Exp. Mrci. 1987. 16-5:
302.
D. R . . Uhrdahl. K. B. and Jenkins. M. K . . J. lmrnunol. 1YY 1. 147. 3261. 8 Evans. R. L. Fa1dctta.T. J.. Humphreys. R. E.. Pratt. D. M.. Yunis. E. J. and Schlossman. S. F..J. Exp. M i d . 1978.148: 1440. Y Hcwitt. C . K. A. and Fcldmann. M.. J. Imrnunol. IY89. 14?:
7 DeSilva.
762. 10 Frdnco. A . . Paroli. M..Tcsta. U.. Benvenuto. R.. Peschle. C.. Balsoni. F. and Barnaha.V.. J. E.vp Mrd. 1092. 175: 1195. 1 1 Lanzavecchia. A . . Roosncck. E.. (3regory.T.. Berman. P and Abrignani. S.. Nururc. IYXX. M4: 530. 12 Wyss-Cor;iy.T.. Brander. C.. Bettcns. F.. Mijic. D. and Pichler. W. J.. Ci4. lmmunol. I Y Y I . 139: 268. 13 Weiss. S. nnd Bogen. B.. (PI1 I90 1. 64: 767. 14 Panina-Bordignon. P. Corradin. G.. Roosnek. E.. Sette. A . and Lanzavecchin. A , . Science I Y Y 1. 252: 1548. 15 Rudensky. A.Y.. Prcston-Hurlhurt. P. Hong. S. C.. Barlow. A . and Jancway Jr. C . A . . Nururc. 1YYl. .?53: 622. 16 Lamb. J. R.. Skidmorc. B. J.. Green N.. Chiller. J. M. and Feldmann. M.. J. Exp. Mcd. 1983. 167: 1434. 17 Gajewski. T. F.. Pinnas. M.. Wing. T. and Fitch. F. W.. J. lmmunol. 1901. 146: 1750. 1x Patarca. R..Wei. F.Y.. Iregui. M.V. and Cantor. H . . Proc. Nut/. Acacl. Sci. USA I Y Y I . 88: 2736.
T. Wyss-Coray, C. Brander, K. Frutig and W. J. Pichler
Eur. J. Immunol. 1992. 22: 2295-2302
19 Wyss, T., Bettens, F., Walker, Ch. and Pichler, W. J., Cell. lmmunol. 1990. 126: 91. 20 Evavold, B. D. and Allen, F! M., Science 1991. 252: 1308. 21 Mosmann,T. R., Cherwinski, H., Bond, M.W., Giedlin, M. A. and Coffman, R. L., J. Immunol. 1986. 136: 2348. 22 Bendelac, A. and Schwartz, R. H., Nature 1991. 353: 68. 23 Kapsenberg, M. L. ,Wierenga, E. A., Bos, J. D. and Jansen, H. M., Immunol. Today. 1991. 11: 392. 24 DeFranco, A. L., Nature 1991. 351: 603. 25 De Soma, M., Tilney, N. L. and Kupiec-Weglinski, J. W., Immunol. Today 1991. 12: 262. 26 Ledbetter, J. A., Parsons, M., Martin, P. J., Hansen, J. A., Rabinovitch, P. S. and June, C. H., J. Immunol. 1986.137: 3299. 27 Demotz, S., Lanzavecchia, A., Eisel, U., Niemann, H . , Widmann, C. and Corradin, G., J. Imrnunol. 1989. 142: 394. 28 Panina-Bordignon, F!, Tan, A., Termijtelen, A., Demotz, S., Corradin, G. and Lanzavecchia, A., Eur. J. Immunol. 1989.19: 2237.
29 Linsley, P. S., Clark, E. A. and Ledbetter, J. A., Proc. Natl. Acad. Sci. U S A 1990. 87: 5031. 30 VonTscharner,V., Deranleau, D. A. and Baggiolini, M., J. Biol. Chem. 1986. 261: 10163. 31 Ho, F! C., Mutch, D. A.,Winkel, K. D., Saul, A. J., Jones, G. L., Doran,T. L. and Rzepczyk, C. M., Eur. J. Immunol. 1990. 20: 477. 32 Schonrich, G., Kalinke, U., Momburg, F., Malissen, M., Schmitt-Verhulst, A. M., Malissen, B., Hammerling, G. J. and Arnold, B., Cell 1991. 65: 293. 33 Bdum, N., Martin, P. J., Schieven, G. J., Hansen, J. A. and Ledbetter, J. A., Eur. J. Immunol. 1991. 21: 123. 34 St.-Pierre,Y and Watts,T. H., J. Immunol. 1991. 147: 2875. 35 Reske-Kunz, A. B., Reske, K. and Rude, E., J. Immunol. 1986. 136: 2033. 36 Heuer, J. and Kolsch, E., J. Immunol. 1985. 134: 4031.
2302
