T Cell Clones + Βα Human the TCR of Sulfamethoxazole-Specific Interaction of Sulfonamide Derivatives With

Interaction of Sulfonamide Derivatives with the TCR of Sulfamethoxazole-Specific Human βα + T Cell Clones

This information is current as Salome von Greyerz, Martin P. Zanni, Karin Frutig, Benno of September 27, 2021. Schnyder, Christoph Burkhart and Werner J. Pichler J Immunol 1999; 162:595-602; ; http://www.jimmunol.org/content/162/1/595 Downloaded from

References This article cites 40 articles, 15 of which you can access for free at: http://www.jimmunol.org/content/162/1/595.full#ref-list-1

Why The JI? Submit online. http://www.jimmunol.org/

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on September 27, 2021

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 1999 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Interaction of Sulfonamide Derivatives with the TCR of Sulfamethoxazole-Speciﬁc Human ␣␤؉ T Cell Clones1

Salome von Greyerz, Martin P. Zanni, Karin Frutig, Benno Schnyder, Christoph Burkhart, and Werner J. Pichler2

Drugs like sulfamethoxazole (SMX) or lidocaine can be presented to specific human ␣␤؉ T cell clones (TCC) by undergoing a noncovalent association with MHC-peptide complexes on HLA-matched APCs. For a better understanding of the molecular basis of the recognition of such drugs by specific TCC, we investigated 1) the fine specificity of the recognizing TCR, 2) the dose-response relationship for the induction of proliferation or cytokine production, and 3) the mechanism of TCR triggering. For that purpose, we tested the reactivity of 11 SMX-specific CD4؉ TCC and 2 SMX-specific CD8؉ TCC to a panel of 13 different sulfonamide derivatives bearing the same core structure. Five of 13 clones recognized only SMX, while all other clones were responding to as many as 6 different compounds. Some of the compounds needed up to two orders of magnitude higher concentrations than SMX Downloaded from to stimulate TCC, thereby displaying features of weak agonists. Different clones showed clear differences in the minimal drug concentration required for the induction of a proliferative response. Therefore, weaker or stronger agonistic properties were not a characteristic of a given sulfonamide derivative but rather an intrinsic property of the reacting TCR. Finally, the number of down-regulated TCRs was a logarithmic function of the ligand concentration, implicating that specific T cells were activated by serial TCR engagement. Our data demonstrate that, despite the special way of presentation, nonpeptide Ag like drugs appear to interact with the TCR of specific T cells in a similar way as peptide Ags. The Journal of Immunology, 1999, 162: 595–602. http://www.jimmunol.org/

he T cell involvement in drug allergic reactions is well matic functional consequences for the T cell (9). These altered established (for review see Ref. 1). However, the molec- responses include cytokine production in the absence of prolifer- T ular mechanism behind the recognition of small molecu- ation, differential cytokine production, anergy, and antagonism of lar compound like drugs by the TCR has not yet been fully elu- the response to the wild-type Ag (10–13). Therefore, we were cidated. It has been thought for several years that drugs can be interested if chemical modification of nonpeptide Ag like drugs recognized by specific T cells only after covalent binding to serum can lead to “altered drug ligand”-like responses. or cellular proteins, which, after subsequent processing, are pre- The sustained signaling necessary for T cell activation re- by guest on September 27, 2021 sented by MHC class I or II molecules (2–5). This pathway of drug quires a continuous engagement of TCRs with MHC-peptide presentation requires chemically reactive compounds with the abil- complexes on APCs (14). Over time, as many as 20,000 TCRs ity to covalently modify side chains of amino acids. However, can be triggered by as few as 100 peptide-MHC complexes recent studies from our group have shown that drugs like lidocaine (15). Thus, the fast kinetics of a single TCR-ligand interaction 3 or sulfamethoxazole (SMX), which are not chemically reactive seem to be a key feature allowing multiple TCR engagements per se, could be presented to specific T cells without the need of by a single agonist. In addition, T cells appear to “count” the uptake, metabolism, and processing (6–8). Preincubation of APCs number of TCRs triggered and to respond by proliferation and with these drugs did not result in any detectable responses. There- cytokine production when a certain activation threshold is fore, we postulated that lidocaine and SMX were associating with reached (16). Also, the serial triggering model of T cell acti- MHC and/or peptide in a noncovalent way, forming a rather un- vation offers a new kinetic explanation for the phenomenon of stable MHC-peptide-drug complex. TCR antagonism and partial agonism. Ligands with lower than For peptide-specific ␣␤ϩ T cell clones (TCC), it is known that optimal stability may engage TCRs at higher rates compared the TCR can recognize ligands that are slightly altered. The rec- with agonists, thus transducing partial signals at a very high rate ognition of these altered peptide ligands (APL) can lead to dra- (17, 18). This may result in the activation of different biological responses or in an efficient antagonism, possibly due to rapid Institute of Immunology and Allergology, Inselspital, Bern, Switzerland spoiling of TCRs or rapid consumption of signaling compo- Received for publication May 29, 1998. Accepted for publication September nents (19, 20). 16, 1998. It has been shown for murine trinitrophenyl-specific, class I-re- The costs of publication of this article were defrayed in part by the payment of page stricted CTL that the MHC-restricted contacts between the hapten charges. This article must therefore be hereby marked advertisement in accordance and the corresponding TCR closely reflect those between TCR and with 18 U.S.C. Section 1734 solely to indicate this fact. nominal peptide Ags (21–24). Previous reports from our laboratory 1 This work was supported by Grant NR.31-50482.97 of the Swiss National Research demonstrated that noncovalently binding drugs can evoke a vig- Foundation (to W.J.P.) and Grant NR97.0431 of the Federal Office for Education and ϩ Science (EU-Research Programme BIOMED). orous response in specific ␣␤ TCC (6–8, 25). Therefore, we 2 Address correspondence and reprint requests to Dr. Werner J. Pichler, Institute of reasoned that the unstable trimolecular MHC-peptide drug com- Immunology and Allergology, Inselspital, Bern, Switzerland. E-mail address: plex might interact with a given TCR similar to peptide Ags. To [email protected] prove this hypothesis, we analyzed how subtle changes in drug 3 Abbreviations used in this paper: APL, altered peptide ligand; B-LCL, B lymphoblastoid cell line; CM, culture medium; SMX, Sulfamethoxazole; TCC, T cell clone; molecules like SMX could alter their recognition by a correspond- TCL, T cell line. ing TCR. For that purpose, we tested the reactivity of several

Copyright © 1999 by The American Association of Immunologists 0022-1767/99/$02.00 596 DRUG RECOGNITION BY TCR Downloaded from FIGURE 1. Chemical structures of the used sulfonamide derivatives. All sulfonamide derivatives used in this study are composed by a sulfanilamide core structure (R1) and a distinct heterocycle (R). The compounds are grouped according to the different heterocycles: 1) compounds with a ﬁve-ring-heterocycle, 2) http://www.jimmunol.org/ pyridine derivative, 3) pyrimidine derivatives containing a guanidine partial structure, 4) pyrazine derivative, and 5) pyrimidine derivatives without a guanidine partial structure. by guest on September 27, 2021

SMX-specific TCC by using a panel of different sulfonamide de- ously for the inhibitory response to mitogens (PHA) in nonallergic indi- rivatives. The obtained data indicate that the unstable trimolecular viduals. Stock solutions of each drug in RPMI 1640/0.05 M NaOH were MHC-peptide-drug complex is interacting with a specific TCR in always freshly prepared just before use. Except sulfisomidine and sulfamethazine (Aldrich Chemie, Buchs, Switzerland) and sulfadoxine (a gift a similar way as nominal peptide Ags. from Dr. L. Heidecker, Hoffmann La Roche, Basel, Switzerland), all sulfonamide derivatives used in this study were purchased from Sigma (St. Materials and Methods Louis, MO). Culture media Culture medium (CM) consisted of RPMI 1640 supplemented with 10% Generation of drug-specific TCC pooled heat-inactivated human albumin serum (Swiss Red Cross, Bern, Freshly isolated PBMC from a SMX allergic donor (6) were stimulated Switzerland), 25 mM HEPES buffer, 2 mM L-glutamine (Seromed, Fakola, with SMX (0.2–0.5 mg/ml) in CM at a cell density of 2 ϫ 106 per well in Basel, Switzerland), 25 ␮g/ml transferrin (Biotest, Dreieich, Germany), 10 ϩ ϩ a 24-well plate (No. 3047; Falcon, Lincoln Park, NJ). CM was added after mg/ml streptomycin, and 100 U/ml penicillin. The CM , used to culture 7 days of culture. After 14 days, bulk cultures were restimulated with TCC, was enriched additionally with 50 U/ml rIL-2 (obtained from Dr. D. autologous irradiated (4000 rad) PBMC plus SMX (0.2 mg/ml) and tested Wrann, Sandoz Research Institute, Vienna, Austria). for their specificity. Fourteen days later, specific T cell lines (TCL) were Drugs used for cell stimulation cloned by limiting dilution as described earlier (25). Growing TCC were expanded in CMϩ and restimulated every 14 days with allogeneic irradi- All drugs (see Fig. 1) used in these experiments have been tested previ- ated (6000 rad) PBMC plus PHA (1 ␮g/ml) (Bacto; Difco, Detroit, MI). The Journal of Immunology 597

Immunofluorescence and PCR-based TCR V␤ analysis guanidine partial structure, 4) the pyrazine-derivative sulfametho- Monoclonality of TCC used was proven either by staining with a panel of xipyridazine (SMP), and 5) pyrimidine derivatives without a gua- 21 different mAb recognizing different V␤ gene products (Immunotech, nidine partial structure. Cross-reactivity analyses include the dis- Marseille, France) or by RT-PCR based TCR-oligotyping as described pre- section of the influence of slight chemical modifications on the viously (3). interaction of the MHC-peptide-drug complex with a given TCR. Proliferation assay Thus, the following properties of the different drugs might be im- portant: capacity to undergo H bond interaction, electronic volume To determine the proliferation of the TCC to the different sulfonamide of sterically hindering groups, and dissociation constants (pKa) at derivatives, 5 ϫ 104 TCC cells were incubated with 5 ϫ 103 irradiated autologous B lymphoblastoid cell line (B-LCL) in the presence of the in- physiological pH values. dicated drug at different concentrations in 200 ␮l CM in a U-bottom mi- All compounds described above stimulated a SMX-specific croplate (Falcon No. 3077) for 48 h. Cultures were pulsed with [3H]thy- polyclonal TCL from the same donor by using autologous B-LCLs midine (0.5 ␮Ci) for the last 8 h, and cells were then harvested onto glass as APCs (data not shown). fiber disks and counted in a microplate beta-counter (Inotech Filter Count- ing System INB 384; Inotech, Dottikon, Switzerland). Cross-reactivity pattern of SMX-specific TCC Cytokine measurement To analyze the cross-reactivity pattern of the 13 clones, we used To detect cytokines produced after stimulation with the different com- the above described panel of 13 different sulfonamide derivatives. pounds, supernatants of cells stimulated as described above were collected TCC were stimulated in the continuous presence of 800 ␮M (cor- after 24 h and cytokines were quantified by a sandwich ELISA according responds to 200 ␮g/ml) of the respective sulfonamide derivative to standard protocols (PharmMingen, San Diego, CA; Ref. 26). Detection by using autologous B-LCL cells as APC. The different cross- Downloaded from limits were 10 pg/ml. reactivity pattern observed are summarized in Table I. The five Measurement of TCR down-regulation TCC in group I, including the two CD8ϩ TCC, were highly spe- TCR down-regulation induced by different compounds was determined as cific and recognized only SMX. They did not react to any other described previously (8, 15). Briefly, 2.5 ϫ 104 clone cells were mixed compound even when concentrations up to 8 mM were used (data with 5 ϫ 104 autologous B-LCL cells in 200 ␮l CM in U-bottom micro- not shown). The three TCC in group II proliferated in the presence plates in the presence or absence of the drug at the indicated concentra- of the two most similar compounds showing only moderate dif- http://www.jimmunol.org/ tions. The plates were centrifuged with 1200 rpm for 2 min to allow con- ferences in terms of electronic volume and density: SMX and sul- jugate formation and then were incubated at 37°C. After 6 h, cells were resuspended, washed in PBS containing 0.5 mM EDTA to break the con- famethizole (SMT). All 5 remaining clones recognized at least jugates, and stained with anti-CD3 (UCHT-1; Dako, Zug, Switzerland) three different compounds: SMX, SMT, and sulfathiazole (STH), followed by a FITC-labeled goat anti-mouse Ig (Dako). The samples were although STH triggered all reactive TCC with much lower effi- analyzed on an EPICS profile II flow cytometer (Coulter Immunology, ciency than the two other compounds. Group IV includes two Hialeah, FL). The absolute numbers of CD3 molecules per cell was esti- mated by reference to a standard curve of beads coated with known clones, 9.3 and 8.21, which could be stimulated by an additional amounts of mouse Ig according to the manufacturers instruction (Qifikit; compound: sulfamethoxipyridazine (SMP). TCC 9.5 (group V) Dako). Cytokine production was measured in the culture supernatant using was the only clone tested that reacted with high efficiency to the an ELISA as described above. pyridine-derivative sulfapyridine (SPD). Finally, group VI con- by guest on September 27, 2021 tains the broadly reactive TCC 8.15, which could be stimulated by Results the continuous presence of six different compounds displaying a SMX-specific TCC and chemical structures of the used quite high heterogeneity in their molecular structure. Others drugs, sulfonamide derivatives such as penicillin G or lidocaine, did not stimulate the SMX-spe- In previous studies, we have shown that specific TCC can recog- cific clones. nize per se nonreactive drugs like lidocaine or SMX presented in Five of the 13 different sulfonamide derivatives, namely sulfa- a labile association with MHC-peptide complexes (6–8). In most moxole (SMO) and the four pyrimidine derivatives containing a cases, the recognition of this trimolecular MHC-peptide-drug com- guanidine partial structure [sulfadiazine (SDZ), sulfamerazine plex is HLA-allele restricted (6, 8), although exceptions have been (SMR), sulfamethazine (SMZ), and sulfadimethoxidiazine described (7). (SMD)], were not recognized by any of the 13 clones tested. To extend our knowledge of the interaction of drugs with As it has been shown that APLs can induce cytokine production specific T cells, we analyzed the influence of slight alterations in the absence of a measurable proliferation (9), we analyzed the of the parent drug SMX on its recognition by a given TCR. For cytokine production of all TCC in response to different sulfon- this purpose, we generated a panel of 11 SMX-specific CD4ϩ amide derivatives. None of the nonstimulating compounds induced TCC and 2 SMX-specific CD8ϩ TCC from an allergic donor a detectable amount of IL-4, IL-5, or IFN-␥ in any of the clones suffering from a hypersensitivity reaction to SMX. All clones tested (data not shown). Furthermore, none of the stimulating com- recognized SMX in the context of the HLA-DRB1*1001 mol- pounds induced a cytokine pattern in the reactive TCC differing ecule or HLA-B44, respectively (Ref. 6, and data not shown). from the one obtained by stimulation with the parent drug SMX The clones were secreting a Th0-like cytokine pattern with high (data not shown). amounts of IL-5 (see Fig. 2, and data not shown). All TCC used were monoclonal; the V␤ gene usage of the clones further used in this study Compounds stimulating TCC 8.15 have stronger or weaker is indicated in Table I. agonistic properties The chemical structures of the 13 different sulfonamide deriv- Although we could not detect “altered drug ligands,” it was atives employed in these experiments are shown in Fig. 1. All already clear from the cross-reactivity experiments shown in compounds were differing only in the heterocycle bound to the Table I that the six compounds stimulating the broadly reactive N(1) of the sulfanilamide core structure. According to their het- TCC 8.15 were differing in their capacity to induce prolifera- erocycle, five different groups can be distinguished: 1) compounds tion. To further analyze the distinct abilities of these com- “bearing” a five-ring heterocycle with higher or lower similarity to pounds to activate TCC 8.15, we incubated the clone cells with the isoxazole ring of the parent drug SMX, 2) the pyridine-deriv- titrated amounts of the stimulating compounds in the presence ative sulfapyridine (SPD), 3) pyrimidine derivatives containing a of autologous B-LCL as APCs. By comparing the efficiency of 598 DRUG RECOGNITION BY TCR

Table I. Cross-reactivity pattern of 13 different SMX-speciﬁc TCCa

Group

Compound I II III IV V VI

Total no. of clonesb 53 1211 Representative clone 8.18 8.10 P21E 9.3 9.5 8.15 Phenotype CD4 CD4 CD4 CD4 CD4 CD4 V␤ usage V␤ 17 V␤ 2V␤ 6V␤ 12 V␤ 5.2 V␤ 16 Control 420c 689 430 511 473 344 Sulfamethoxazole SMXd 15,448 28,382 12,166 32,261 54,607 23,207 Sulfamethizole SMT 315 23,248 10,133 5,287 33,172 11,526 Sulfathiazole STH 417 455 3,701 8,563 8,815 2,381 Sulfamoxole SMO 240 308 400 476 365 269 Sulfapyridine SPD 447 573 510 624 23,190 491 Sulfadiazine SDZ 332 594 456 482 455 519 Sulfamerazine SMR 311 530 468 543 519 255 Sulfamethazine SMZ 419 436 487 458 499 549 Sulfadimethoxidiazine SMD 433 358 509 502 557 519 Downloaded from Sulfamethoxipyridazine SMP 328 438 468 32,988 585 316 Sulﬁsomidine SID 375 592 468 616 625 20,009 Sulfadimethoxine SDM 329 467 421 532 559 12,376 Sulfadoxine SDX 509 645 400 497 552 10,839

a TCC were analyzed for reactivity with a panel of 13 different sulfonamide derivatives (for chemical structures, see Fig. 1) by incubating the T cells in the continuous presence of 800 ␮M (200 ␮g/ml) of the indicated compounds and autologous irradiated B-LCL as APCs. Proliferation was determined http://www.jimmunol.org/ after 48 h by measurement of [3H]thymidine incorporation. b Number of clones with the same crossreactivity pattern. Further clones with the same cross-reactivity-pattern: group I) 1.3 (CD8ϩ), P26B (CD8ϩ), 9.18, 9.17; group II) S23E, 8.5; group IV) 8.21. c Values in cpm of [3H]thymidine-incorporation (the experiments were repeated at least three times). d Further used abbreviations of the compounds. the induction of a signiﬁcant proliferative response by the dif- full agonistic properties as the parent drug SMX, whereas all ferent sulfonamide derivatives, two groups of compounds could other compounds needed a 10- to 100-fold higher concentration be distinguished (Fig. 2A). Sulﬁsomidine (SID) had the same to achieve half-maximal proliferation and could therefore be by guest on September 27, 2021

FIGURE 2. Dose-response curves of the sulfonamide derivatives stimulating the broadly reactive TCC 8.15. The broadly reactive TCC 8.15 was incubated with various amounts of different stimulating sulfonamide derivatives in the presence of irradiated autologous B-LCL. A, Proliferative response measured after 48 h by [3H]thymidine incorporation. B–D, Cytokine production [IL-4, B; IL-5, C; and IFN-␥, D] measured after a culture period of 24 h by standard ELISA. Detection limit was 10 pg/ml. A representative of three independent experiments is shown. The Journal of Immunology 599

FIGURE 3. Dose-dependent TCR down- regulation in TCC 8.15 induced by different compounds. Clone 8.15 was incubated with various concentrations of the stimulating compounds in the presence of autologous B- LCL. TCR down-regulation was determined after6h(A) and is indicated as the percentage of CD3 mean ﬂuorescence calculated from values without drugs. B–D, Levels of Downloaded from IL-4 (B), IL-5 (C), and IFN-␥ (D) produced after6hasafunction of the number of TCRs down-regulated. No cytokine production was observed in cultures without drug stimulation. http://www.jimmunol.org/ by guest on September 27, 2021

classified as weak agonists. An identical pattern could be ob- TCRs 9.3 and 9.5 interact with sulfamethizole (SMT) with served when cytokine production after 24 h was analyzed (Fig. different strength 2, B–D). To our knowledge, these are the first data demonstrat- From the data obtained with clone 8.15, the question arose whether ing that unstable-presented nonpeptide Ags can have different the distinct agonistic properties of the compounds to stimulate this agonistic properties. clone were an intrinsic feature of the respective compound or if An early phenotypic event that occurs subsequent to TCR en- they reflected differences in the ability of the TCR 8.15 to interact gagement by an antigenic ligand is the down-regulation of TCR with distinct MHC-peptide-drug complexes. The comparison of expression on the surface of T cells. This process starts within the dose-response curves of sulfonamide derivatives stimulating minutes following the initial interaction and reaches a peak after a TCC 9.3 and 9.5 gave further insights in the interaction of different few hours. In our system, the plateau of TCR down-regulation was clones with compounds with distinct triggering capacities. reached after 5–6 h. To quantify the distinct interactions of strong Whereas for TCC 9.5 Ͻ20 ␮M SMT (see Fig. 4, A and C) were and weak agonists with the 8.15 TCR, we determined the extent of sufficient to induce a significant TCR down-regulation and were TCR down-regulation induced by the different compounds after additionally able to trigger a proliferative response, TCC 9.3 6 h. The sulfonamide derivatives with weak agonistic properties needed up to 2000 ␮M SMT for comparable responses (Fig. 4, B again needed up to 100 fold higher concentrations to remove a and D). In contrast, both clones required similar amounts of SMX similar percentage of TCRs from the cell surface (Fig. 3A). Non- (ϳ10 ␮M) for the induction of significant responses. The obtained stimulating compounds did not lead to any detectable TCR down- data demonstrate that different TCRs can recognize the same drug regulation (data not shown). In consequence, the question arose with different efficiencies. whether the lower response of the TCC after stimulation with the weak agonistic sulfonamide derivatives reflected a lower extent of TCR triggering or rather an altered signal leading to qualitatively SMX-specific clones are activated by serial TCR engagement different responses. Therefore, we correlated TCR down-regula- The concept of serial TCR engagement suggests that the rate of tion and cytokine production after stimulation with the strong or TCR-ligand complex dissociation has a main influence on the re- the weak agonists. As shown in Fig. 3, B–D, in spite of a decreased sulting T cell activation, because the magnitude of the response efficiency of TCR triggering, the threshold of T cell activation and depends on the frequency of TCR engagement (18, 19). The dif- the type of cytokines produced were not affected by the agonistic ferences in the agonistic properties of the compounds stimulating capacities of the different compounds. clone 8.15 and the different abilities of clones 9.3 and 9.5 to react 600 DRUG RECOGNITION BY TCR

FIGURE 4. Different SMT dose- curves observed in TCC 9.3 and 9.5. T cell clones 9.5 (A and C) and 9.3 (B and D) were incubated with different concentrations of SMX, sulfamethizole (SMT), and sulfathiazole (STH) by using autologous B-LCL cells as APCs. A and B, TCR down-regulation determined after 6 h. C and D, Proliferative response measured Downloaded from after 48 h by [3H]thymidine incorporation. The experiments were repeated at least three times. http://www.jimmunol.org/ by guest on September 27, 2021 to SMT could easily be explained by distinct strengths of the as- (SMX) on its recognition by 13 different SMX-specific TCC from sociation of the TCR with the respective MHC-peptide-drug com- an SMX-allergic donor. The chemical alteration of the different plex. Therefore, we had to prove that unstable trimolecular MHC- compounds compared with the parent drug SMX could have an peptide-drug complexes activate specific T cells by serial TCR effect on the three-dimensional surface of the antigenic MHC-pep- engagement. To do that, we analyzed the relationship between the tide-drug complex, which is accessible for the interacting TCR. number of TCR ligands expressed as the drug concentration in the However, they have no influence on the chemical reactivity of the cultures and the number of TCRs triggered after 6 h. As depicted drugs and on their inability to covalently modify side-chains of in Fig. 5A for both clones, a logarithmic function of the number of amino acids, as the P450 catalyzed reaction to nitroso derivatives, TCRs down-regulated compared with the ligand concentration of- the only reactive intermediates, is not affected by differences in the fered could be revealed, a correlation that is only compatible with substitutions at the N(1) (27, 28). the so-called “serial triggering” model, where low numbers of li- The 13 SMX-specific TCC used in this study showed a quite gands can successfully trigger high numbers of TCRs. In the case high diversity in their ability to respond to different sulfonamide of the two clones shown, the differences in their efficiency to react derivatives. On one hand, the five clones comprised in group I (see to SMX can be explained by different levels of TCR expression on Table I) were highly specific and could be stimulated by SMX both clones (TCC 8.21, ϳ90,000 TCRs/cell; TCC 9.3, ϳ140,000 only. Quite surprisingly, even a chemically very similar compound TCRs/cell). By measuring proliferation and cytokine production, a like SMT was not recognized by these five TCC. On the other similar pattern could be observed (data not shown). However, de- hand, TCC 8.15 showed a broad cross-reactivity as it was respond- spite the different prerequisites for ligand concentration both TCC ing to six different compounds sharing only a minor structural had comparable activation thresholds of approximately 15,000 similarity. TCRs (Fig. 5B). A detailed analysis of the different cross-reactivity pattern by using images of the electronic structure, obtained by molecular Discussion modeling of the different compounds, did not lead to a clear rela- Chemically per se nonreactive drugs like lidocaine or SMX can be tionship between chemical structure and the ability to activate a presented to specific TCC by undergoing a rather unstable asso- given TCC. However, some cross-reactivity pattern could be ex- ciation with the restricting MHC-peptide complex (6–8). To better plained. In the case of the clones 8.21 and 9.3, the recognition of understand this association and to analyze whether the interaction SMX, SMT, and SMP might be elucidated by the fact that all three of a corresponding TCR with a noncovalently binding drug is sim- compounds bear in the ␣-position of the first electron-negative N ilar to the interaction of TCRs with nominal peptide Ags, we an- atom of the heterocycle a second nucleophilic N or O atom, both alyzed the influence of slight modifications of the parent drug of them able to act as electron donors for H bond interactions. As The Journal of Immunology 601

induction phase of the drug allergy several quite distinct immuno- genic structures did activate the SMX-reactive T cells (33). There- fore, the recognition of the noncovalently bound drug SMX by a given TCR is not a randomly occurring event as the observed cross-reactivity pattern as well as the different strength of the TCR engagement were highly reproducible. None of the tested clones reacted to the used sulfonamide derivatives with an “altered drug ligand”-like response. However, the analysis of the dose-response relationships of broader reactive clones revealed that compounds with weaker or stronger agonistic properties could be distinguished. This phenomenon is widely known for single amino acid-substituted peptide Ags (34), but has not yet been described for nonpeptide Ags, especially for noncovalently binding drugs. One of the factors influencing the efficiency of the TCR-ligand interaction is the stability of the antigenic MHC-peptide complex, governed by the affinity of the peptide to the restricting MHC molecule (35). The unstable association of the drugs with the MHC-peptide complex did not allow the determination Downloaded from of the affinity of the different compounds to the MHC-complex. However, the comparison of the dose relationships of SMX recognition by the two clones 9.3 and 9.5 gives further insights in the possible mechanism of interaction of a given TCR with distinct but chemically minutely different compounds. Both

clones were expressing similar numbers of TCRs on their sur- http://www.jimmunol.org/ face (see below) and could be activated by SMX with the same efficiency. In contrast, the SMT dose needed to induce a significant TCR down-regulation or proliferation differed in order of two magnitudes between the two clones. These data can only be explained by different affinities of the two TCRs to the MHC- peptide-SMT complexes. Hence, these findings indicate that not distinct affinities of the different sulfonamide derivatives to the MHC-peptide-complexes influence the efficiency of the TCR- ligand interaction. Moreover, they also reveal that weaker or by guest on September 27, 2021 stronger agonistic properties are not a characteristic of a given FIGURE 5. Number of TCRs triggered is a logarithmic function of the sulfonamide derivative but rather an intrinsic feature of the re- ligand concentration. A, Clones 8.21 and 9.3 were incubated with different acting TCR. In addition, these data clearly show that the distinct concentrations of SMX in the presence of autologous B-LCL cells, and the efficiencies of the six different compounds stimulating TCC number of triggered TCRs was determined after 6 h. B, Levels of IFN-␥ produced after6hasafunction of the number of TCRs down-regulated. 8.15 might not be explained by different affinities of the sulfonamide derivatives to MHC-peptide complexes and/or different ligand densities formed on the APC, but by distinct inter- the interaction of the TCR with its ligands can involve H bond actions of the 8.15 TCR with the different MHC-peptide-drug interactions with side chains of peptide residues (29, 30), it is complexes. feasible that also in the case of MHC-peptide-drug complexes the It is known that T cell activation requires the assembly of mul- capacity to undergo H bond interactions can influence the recog- tiple layers of proteins to the phosphorylated TCR (36). A certain nition by a given TCR. amount of time is required to assemble the correct signaling com- Five sulfonamide derivatives (SMO, SDZ, SMR, SMZ, and plex, and, therefore, the formation of the complete complex will SMD) did not stimulate any of the clones tested. However, differ- require stimuli that exceed a certain threshold of strength and du- ent SMX-specific polyclonal TCL from the same donor could be ration (37, 38). This model would imply that the 8.15 TCR inter- stimulated in the presence of autologous B-LCL cells with all com- acts with weak agonistic sulfonamide derivatives with suboptimal pounds used (data not shown). Hence, we could exclude the pos- kinetics, so that only a certain percentage of interactions are stable sibility that these compounds, due to their chemical properties, enough to deliver a full signal. However, correlation of TCR could either not associate with MHC-peptide complexes or could down-regulation and cytokine production, as shown in Fig. 3, not be recognized by SMX-specific T cells. B–D, clearly demonstrates that despite of a decreased efficiency of The high heterogeneity of the observed cross-reactivity pattern TCR triggering, the threshold of T cell activation and the type of in addition to the different V␤ gene usage of the tested clones cytokines produced were not affected by the agonistic capacities of indicates that although oligoclonal outgrowth of drug-specific T the different compounds. Therefore, the lower response of the TCC cells has been described (25, 31, 32), in this patient the T cell- after stimulation with the weak agonistic sulfonamide derivatives mediated response to SMX was polyclonal and heterogeneous. reflects rather a lower extent of TCR triggering than an altered This implies that each of the TCC generated to SMX recognizes signal leading to a qualitatively different response. the drug in a different context. Therefore, it seems likely that the Our results implicate a T cell activation by serial TCR engage- per se nonreactive drug SMX has the possibility to interact with ment. In all clones used in this study, we could find a logarithmic MHC-peptide complexes in several ways, generating distinct an- correlation of the number of down-regulated TCRs compared with tigenic determinants. This would suggest that already during the the offered ligand concentration. This relationship indicates that 602 DRUG RECOGNITION BY TCR

TCR triggering is most efficient when only few antigenic com- 13. Sloan-Lancaster, J., and P. M. Allen. 1996. Altered peptide ligand-induced partial plexes were available on the APC surface and becomes less and T cell activation: molecular mechanisms and role in T cell biology. Annu. Rev. Immunol. 14:1. less efficient with increasing numbers of ligands. According to the 14. Valitutti, S., M. Dessing, K. Aktories, H. Gallati, and A. Lanzavecchia. 1995. work of the group of Lanzavecchia (15, 19), this experimental Sustained signaling leading to T cell activation results from prolonged T cell finding is difficult to reconcile with a model of TCR cross-linking receptor occupancy: role of T cell actin cytoskeleton. J. Exp. Med. 181:577. 15. Valitutti, S., S. Mu¨ller, M. Cella, E. Padovan, and A. Lanzavecchia. 1995. Serial where the number of triggered TCRs would exponentially increase triggering of many T cell receptors by a few peptide-MHC complexes. Nature with an increase in the number of ligands offered. The data ob- 375:148. 16. Viola, A., and A. Lanzavecchia. 1996. T cell activation determined by T cell tained clearly show that unstable MHC-peptide-drug complexes receptor number and tunable thresholds. Science 273:104. can activate specific TCC by serial TCR engagement. 17. Lyons, D. S., S. A. Lieberman, J. Hampl, J.J. Boniface, Y. Chien, L. J. Berg, and These data give additional information for the understanding of M. M. Davis. 1996. A TCR binds to antagonist ligands with lower affinities and the interaction of the trimolecular MHC-peptide-drug complex faster dissociation rates than to agonists. Immunity 5:53. 18. Kessler, B., D. Hudrisier, J. C. Cerottini, and I. F. Luescher. 1997. Role of CD8 with a reactive TCR. Whereas the MHC-peptide complexes are in aberrant function of cytotoxic T lymphocytes. J. Exp. Med. 186:2033. stable and can be presented for a long time on the surface of the 19. Valitutti, S., and A. Lanzavecchia. 1997. Serial triggering of TCRs: a basis for the APC (39, 40), the association of the TCR with its ligand has a sensitivity and specificity of antigen recognition. Immunol. Today 18:299. 20. Viola, A., S. Linkert, and A. Lanzavecchia. 1997. A T cell receptor (TCR) an- half-life of seconds (41). Therefore, even a very low number of tagonist competitively inhibits serial TCR triggering by low-affinity ligands but such complexes may be sufficient to engage, in several rounds of does not affect triggering by high-affinity anti-CD3 antibodies. Eur. J. Immunol. ligation, a relatively high number of TCRs, thus triggering T cell 27:3080. 21. Martin, S., A. von Bonin, C. Fessler, U. Pflugenfelder, and H. U. Weltzien. 1993. responses (14, 15). In the case of sulfonamide derivatives, the Structural complexity of antigenic determinants for class I MHC-restricted, hap- binding of the drug to MHC-peptide complexes is much more ten-specific T cells: two qualitatively differing types of H-2Kb-restricted TNP Downloaded from epitopes. J. Immunol. 151:678. unstable, as washing of drug-prepulsed APCs abrogates their abil- 22. Luescher, I. F., F. Anjue`re, M. C. Peitsch, C. V. Jongeneel, J. C. Cerottini, and ity to stimulate reactive T cells (6–8). However, the kinetics of P. Romero. 1995. Structural analysis of TCR-ligand interactions studied on a the T cell triggering resemble very closely the serial triggering H-2Kd-restricted cloned CTL specific for a photoreactive peptide derivative. Im- munity 3:51. model, indicating that the half-life time of a MHC-peptide-drug 23. Kohler, J., S. Martin, U. Pflugenfelder, H. Ruh, J. Vollmer, and H. U. Weltzien. complex must be considerably higher than the time needed to 1995. Cross-reactive trinitrophenylated peptides as antigens for class II major trigger a single TCR. histocompatibility complex-restricted T-cells and inducers of contact sensitivity http://www.jimmunol.org/ in mice: limited T cell receptor repertoire. Eur. J. Immunol. 25:92. In summary, the reported data indicate that, despite the special 24. Preckel, T., R. Grimm, S. Martin, and H. U. Weltzien. 1997. Altered hapten way of Ag presentation, noncovalently binding drugs can interact ligands antagonize trinitrophenyl-specific cytotoxic T cells and block internal- with a specific TCR similar as nominal peptide Ags. ization of hapten-specific receptors. J. Exp. Med. 185:1803. 25. Zanni, M. P., D. Mauri-Hellweg, C. Brander, T. Wendland, B. Schnyder, E. Frei, S. von Greyerz, A. Bircher, and W. J. Pichler. 1997. Characterization of lido- Acknowledgments caine-specific T cells. J. Immunol. 158:1139. 26. Elsasser-Beile, U., S. von Kleist, and H. Gallati. 1991. Evaluation of a test for We thank blood donor B.K. for his friendly collaboration, measuring cytokine production in human whole blood cell cultures. J. Immunol. E. Frei for TCR oligotyping, D. Rognon for the molecular modeling of the Methods 139:191. electronic structure of the different sulfonamide derivatives, and 27. Park, B. K., J. W. Coleman, and N. R. Kitteringham. 1987. Drug disposition and drug hypersensitivity. Biochem. Pharmacol. 36:581. by guest on September 27, 2021 T. Wendland for helpful discussion. 28. Van der Ven, A. J. A. M., M. A. Mantel, T. B. Vree, P. P. Koopmans, and J. V. M. van der Meer. 1994. Formation and elimination of sulphamethoxazole References hydroxylamine after oral administration of sulphamethoxazole. Br. J. Clin. Phar- macol. 38:147. 1. Pichler, W. J., B. Schnyder, M. P. Zanni, Y. Hari, and S. von Greyerz. 1998. Role 29. Davis, M. M., and P. J. Bjorkman. 1988. T cell antigen receptor genes and T cell of T cells in drug allergies. Allergy 53:225. recognition. Nature 334:395. 2. Brander, C., D. Mauri-Hellweg, F. Bettens, H. Rolli, M. Goldman, and 30. Jorgensen, J. L., U. Esser, B. Fazekas de St. Groth, P. A. Reay, and M. M. Davis. W. J. Pichler. 1995. Heterogeneous T cell responses to ␤-lactam-modified self- 1992. Mapping T cell receptor-peptide contacts by variant peptide immunization structures are observed in penicillin-allergic individuals. J. Immunol. 155:2670. of single-chain transgenics. Nature 355:224. 3. Mauri-Hellweg, D., F. Bettens, D. Mauri, C. Brander, T. Hunziker, and ϩ ϩ 31. Mauri-Hellweg, D., M. P. Zanni, E. Frei, F. Bettens, C. Brander, and W. J. Pichler. 1995. Activation of drug-specific CD4 and CD8 T cells in W. J. Pichler. 1996. Crossreactivity of T cell lines and clones to ␤-lactam anti- individuals allergic to sulfonamides, phenytoin and carbamazepine. J. Immunol. biotics. J. Immunol. 157:1071. 155:462. 32. Vollmer, J., M. Fritz, A. Doromy, H. U. Weltzien, and C. Moulon. 1997. Dom- 4. Weltzien, H. U., C. Moulon, S. Martin, E. Padovan, U. Hartmann, and J. Kohler. inance of the BV17 element in nickel-specific human T cell receptors relates to 1996. T cell immune response to haptens: structural models for allergic and severity of contact sensitivity. Eur. J. Immunol. 27:1865. autoimmune reactions. Toxicology 107:141. 33. Hausmann, S., and K. W. Wucherpfennig. 1997. Activation of autoreactive T 5. Padovan, E., T. Bauer, M. M. Tongia, H. Kalbacher, and H. U. Weltzien. 1997. cells by peptides from human pathogens. Curr. Opin. Immunol. 9:831. Penicilloyl peptides are recognized as T cell antigenic determinants in penicillin 34. Chen Y. Z., S. Matsushita, and Y. Nishimura. 1996. Response of a human T cell allergy. Eur. J. Immunol. 27:1303. clone to a large panel of altered peptide ligands carrying single residue substi- 6. Schnyder, B., D. Mauri-Hellweg, M. Zanni, F. Bettens, and W. J. Pichler. 1997. ␣␤ tutions in an antigenic peptide: characterization and frequencies of TCR agonism Direct, MHC-dependent presentation of the drug sulfamethoxazole to human and TCR antagonism with or without partial activation. J. Immunol. 157:3783. T cell clones. J. Clin. Invest. 100:136. 35. Jameson, S. C., and M. J. Bevan. 1995. T cell receptor antagonists and partial 7. Zanni, M. P., S. von Greyerz, B. Schnyder, T. Wendland, and W. J. Pichler. 1998. ϩ agonists. Immunity 2:1. HLA non-restricted drug recognition by ␣␤ T cell clones. Int. Immunol. 10:507. 8. Zanni, M. P., S. von Greyerz, B. Schnyder, K. A. Brander, K. Frutig, Y. Hari, 36. Chan, A. C., and A. S. Shaw. 1995. Regulation of antigen receptor signal trans- S. Valitutti, and W. J. Pichler. 1998. HLA restricted, processing- and metabolism- duction by protein tyrosine kinases. Curr. Opin. Immunol. 8:394. independent pathway of drug recognition by human ␣␤ T lymphocytes. J. Clin. 37. McKeithan, T. W. 1995. Kinetic proofreading in T cell receptor signal transduc- Invest. 102:1531. tion. Proc. Natl. Acad. Sci. USA 92:5042. 9. Kersh, G. J., and P. M. Allen. 1996. Structural basis for T cell recognition of 38. Rabinowitz, J. D., C. Beeson, D. S. Lyons, M. M. Davis, and H. M. McConnell. altered peptide ligands: a single T cell receptor can productively recognize a large 1996. Kinetic discrimination in T cell activation. Proc. Natl. Acad. Sci. USA continuum of related ligands. J. Exp. Med. 184:1259. 93:1401. 10. Evavold, B. D., and P. M. Allen. 1991. Separation of IL-4 production from Th 39. Lanzavecchia, A., P. A. Reid, and C. Watts. 1992. Irreversible association of cell proliferation by an altered T cell receptor ligand. Science 252:1308. peptides with class II MHC molecules in living cells. Nature 357:249. 11. De Magistris, M. T., J. Alexander, M. Coggeshalll, A. Altman, F. C. A. Gaeta, 40. Nelson, C. A., S. J. Petzold, and E. R. Unanue. 1994. Peptides determine the H. M. Grey, and A. Sette. 1992. Antigen analog-major histocompatibility com- lifespan of MHC class II molecules in the antigen presenting cell. Nature 37:250. plexes act as antagonists of the T cell receptor. Cell 68:625. 41. Matsui, K., J. J. Boniface, P. Steffner, P. A. Reay, and M. M. Davis. 1994. 12. Sloan-Lancaster J., B. D. Evavold, and P. M. Allen. 1993. Induction of T cell Kinetics of T cell receptor binding to peptide/I-Ek complexes: correlation of the anergy by altered T-cell-receptor ligand on live antigen-presenting cells. Nature dissociation rate with T cell responsiveness. Proc. Natl. Acad. Sci. USA 91: 363:156. 12862.