Advertisement
Research ArticleArticle

Hypersensitivity Reactions to Carbamazepine: Characterization of the Specificity, Phenotype, and Cytokine Profile of Drug-Specific T Cell Clones

D. J. Naisbitt, M. Britschgi, G. Wong, J. Farrell, J. P. H. Depta, D. W. Chadwick, W. J. Pichler, M. Pirmohamed and B. K. Park
Molecular Pharmacology March 2003, 63 (3) 732-741; DOI: https://doi.org/10.1124/mol.63.3.732

Abstract

Administration of carbamazepine (CBZ) causes hypersensitivity reactions clinically characterized by skin involvement, eosinophilia, and systemic symptoms. These reactions have an immune etiology; however, the role of T cells is not well defined. The aim of this study was to characterize the specificity, phenotype, and cytokine profile of CBZ-specific T cells derived from hypersensitive individuals. Proliferation of blood lymphocytes was measured using the lymphocyte transformation test. CBZ-specific T cell clones were generated by serial dilution and characterized in terms of their cluster of differentiation and T cell receptor Vβ phenotype. Proliferation, cytotoxicity, and cytokine secretion were measured by [3H]thymidine incorporation, 51Cr release, and enzyme-linked immunosorbent assay, respectively. HLA blocking antibodies were used to study the involvement of antigen-presenting cells. The specificity of the drug T cell receptor interaction was studied using CBZ metabolites and other structurally related compounds. Lymphocytes from hypersensitive patients (stimulation index: 32.1 ± 24.2 [10 μg ml−1]) but not control patients (stimulation index: 1.2 ± 0.4 [10 μg ml−1]) proliferated upon stimulation with CBZ. Of 44 CBZ-specific T cell clones generated, 10 were selected for further analysis. All 10 clones were either CD4+ or CD4+/CD8+, expressed the αβ T cell receptor, secreted IFN-γ, and were cytotoxic. T-cell recognition of CBZ was dependent on the presence of HLA class II (DR/DQ)-matched antigen-presenting cells. The T cell receptor of certain clones could accommodate some CBZ metabolites, but no cross-reactivity was seen with other anticonvulsants or structural analogs. These studies characterize drug-specific T cells in CBZ-hypersensitive patients that are phenotypically different from T cells involved in other serious cutaneous adverse drug reactions.

Administration of carbamazepine (CBZ), which is commonly administered for the treatment of epilepsy, causes hypersensitivity reactions characterized by skin involvement and systemic manifestations such as hepatitis and eosinophilia (Leeder, 1998). Reactions vary in severity and may cause death. The frequency of CBZ hypersensitivity is between 1 in 1000 and 1 in 10,000 new exposures to the drug (Vitorrio and Mulglia, 1995; Tennis and Stern, 1997). Taken from clinical manifestations, CBZ hypersensitivity reactions are believed to have an immune etiology. In accordance with this, a sensitization period of 3 to 4 weeks is required for the initial development of clinical symptoms, whereas rechallenge results in the recurrence of symptoms much sooner than with primary exposure (Leeder, 1998). The mechanism by which CBZ induces a hypersensitivity reaction is unclear. Although CBZ undergoes bioactivation to toxic arene oxide and quinone metabolites (Madden et al., 1996; Ju and Uetrecht, 1999), definite evidence linking this to immunogenicity is lacking (Park et al., 1998).

The generation and characterization of drug-specific T cell lines and clones from drug-allergic patients with severe skin reactions has recently led to the classification of cutaneous eruptions in terms of their cellular and molecular pathophysiology (Pichler et al., 2002) (Fig. 1). Skin-infiltrating CD4+ T cells, which secrete IL-5, perforin, and granzyme, are characteristic of maculopapular exanthemas (Pichler et al., 1997; Schnyder et al., 1998;Yawalker et al., 2000a,b), whereas bullous reactions such as Stevens-Johnson syndrome and toxic epidermal necrolysis are caused by cytotoxic CD8+ T cells (Hertl et al., 1995; Hari et al., 2001; Nassif et al., 2002). Recently, Britschgi et al. (2001) have shown that CD4+ T cells from patients with pustular reactions such as generalized exanthematous pustulosis secrete high levels of the neutrophil attracting chemokine IL-8.

Figure 1
Figure 1

Scheme depicting our current understanding of cellular and molecular pathophysiology of severe skin reactions (adapted from Pichler et al., 2002).

Laboratory investigations have shown that blood mononuclear cells (PBMC) from hypersensitive patients but not from drug-naı̈ve patients or patients administered CBZ without detectable adverse effects proliferate on exposure to CBZ (Zakrzewska and Ivanyi, 1988;Mauri-Hellweg et al., 1995; Brown et al., 1999; Hari et al., 2001). Cells from CBZ-hypersensitive patients have been shown to produce variable but occasionally high levels of IFN-γ and IL-5 (Koga et al., 2000; Leyva et al., 2000; Villada et al., 1992; Mauri-Hellweg et al., 1995; Schnyder et al., 1998), which is consistent with the pathophysiology of the adverse reaction. However, the precise role of T cells in CBZ hypersensitivity is not known.

To provide a definition of the role of T cells in CBZ hypersensitivity and further laboratory evidence of the immune pathogenesis, we cloned drug-specific T cells from the circulation of CBZ-hypersensitive patients and analyzed their phenotype, specificity, and cytokine profile in vitro. To investigate the structural requirements of T cell receptor activation, CBZ-specific T cell clones were stimulated with nine structurally related compounds.

Materials and Methods

Donor Characteristics.

PBMCs were obtained from five CBZ-hypersensitive patients, four patients who had been administered CBZ for at least 12 months without adverse effects, and four healthy volunteers with no history of previous exposure to CBZ. Approval for the study was obtained from the local ethics committee, and informed consent was obtained from each participant. The clinical details of the CBZ-hypersensitive patients are shown in Table1.

View this table:
Table 1

Details of the carbamazepine-hypersensitive patients

Culture Medium and Chemicals.

Culture medium consisted of RPMI 1640 supplemented with 10% pooled heat-inactivated human blood type AB serum, HEPES buffer (25 mM), l-glutamine (2 mM), transferrin (25 μg·ml−1), streptomycin (100 μg·ml−1) and penicillin (100 U·ml−1). For culture of the T cell lines and clones, the medium was enriched with human recombinant IL-2 (60 U·ml−1). The above reagents were obtained from Sigma Chemical (Poole, Dorset, UK).

Epstein-Barr virus-transformed B-lymphoblastoid cell lines (B-LCL) were generated from the drug-allergic donors by transformation of PBMC with supernatant from the Epstein-Barr virus-producing cell line B9–58 (obtained from Dr. D. Neumann-Haefelin, University of Freiburg, Freiburg, Germany) and cultured in RPMI 1640 medium supplemented with 10% fetal calf serum, HEPES buffer (25 mM), andl-glutamine (2 mM).

Phenylurea, ethylurea, dicyclohexylurea, diphenylurea, iminostilbene, dihydro-CBZ, CBZ-10,11-epoxide, phenytoin (all from Sigma Chemical) and lamotrigine (GlaxoSmithKline, Uxbridge, Middlesex, UK) were used at concentrations that did not inhibit the proliferative response to the mitogen phytohemagglutinin (1 μg·ml−1) for T cell culture. Stock solutions (10 mg·ml−1) were prepared in a mixture of culture media and dimethyl sulfoxide (4:1 v/v) and diluted before use. Tetanus toxoid (TT; Serum and Vaccine Institute, Bern, Switzerland) was used as a control protein antigen. All general reagents were purchased from Sigma Chemical (Poole, Dorset, UK) and were of the best available grade.

Determination of Carbamazepine-Specific Proliferation In Vitro.

Proliferation of PBMC from patients and control volunteers was measured using the lymphocyte transformation test as described previously (Nyfeler and Pichler, 1997). Briefly, freshly isolated PBMC (1.5 × 105; total volume, 0.2 ml) was incubated with CBZ (1–100 μg·ml−1), CBZ-10,11-epoxide (1–100 μg·ml−1; hypersensitive patients 1 and 2 only), or TT (0.1 and 1 μg·ml−1) in 96-well U-bottomed tissue culture plates for 6 days (37°C; 5% CO2). Proliferation was determined by the addition of [3H]thymidine (0.5 μCi) for the final 16 h of the incubation period. Cells were harvested, and [3H]thymidine incorporation was measured as cpm on a β-counter (PerkinElmer Life Sciences, Boston, MA). Proliferative responses were calculated as stimulation indices (SI; cpm in drug-treated cultures/cpm in cultures with dimethyl sulfoxide alone).

After a 6-day incubation with CBZ (10 μg·ml−1), PBMC from hypersensitive patients 1 and 2 were stained with phosphatidylethanolamine-labeled anti-CD4+/CD8+ and fluorescein isothiocyanate-labeled anti-HLA-DR antibodies (Serotec, Oxford, UK). Cells were stained in phosphate-buffered saline (50 μl) containing fetal calf serum (1%) and NaN3 (0.02%) at 4°C. After 30 min, the cells were washed, and fluorescence was measured on a Coulter EPICS XL flow cytometer (Beckman Coulter Inc., Fullerton, CA).

Determination of Carbamazepine-Specific Cytotoxic Activity In Vitro.

To enrich the quantity of CBZ-specific T cells, PBMC (5 × 106; total volume, 1 ml) from hypersensitive patients 1 and 5 were incubated with CBZ (25 μg·ml−1) in 24-well tissue culture plates (37°C; 5% CO2); on day 4, IL-2 (60 U·ml−1) was added to preserve the CBZ-specific response. After 7 days, growing cells were restimulated with irradiated (45 Gy) autologous PBMC (1.5 × 105/well), CBZ (25 μg·ml−1), and IL-2 (60 U·ml−1). This procedure was repeated for a further 3 weeks, and cytotoxic activity was measured on weeks 3 and 4 using a standard 51Cr release assay (Brunner et al., 1968). Autologous B-LCLs (1 × 106) were labeled with 51Cr (50 μl; room temperature). After 1 h, the B-LCLs were washed (3 × 50 ml) to remove free 51Cr and then incubated (2.5 × 103; total volume, 0.2 ml) with T cell-enriched PBMC (0.6 × 105) in the presence and absence of CBZ (10 and 50 μg·ml−1). Specific lysis was calculated as 100 × (experimental release − spontaneous release)/maximal release − spontaneous release). Direct CBZ-specific toxicity was investigated by incubating CBZ (50 μg·ml−1) with 51Cr-loaded B-LCL (2.5 × 103) for 4 h, in the absence of T cells.

The 51Cr release assay also was used to assess the cytotoxic activity of three CBZ-specific T cell clones. In these experiments, the concentration of CBZ used was 50 μg·ml−1 and the effector/target ratios were 3:1, 10:1, and 30:1.

Generation of Carbamazepine-Specific T Cell Lines and Clones.

For the generation of T cell lines, PBMC (2 × 106; total volume, 1 ml) from hypersensitive patients 1 and 2 were incubated with CBZ (25 μg·ml−1) in 24-well tissue culture plates. On days 6 and 9, IL-2 (60 U·ml−1) was added to maintain drug-specific proliferation. After 14 days, T cell lines were cloned by serial dilution, as described previously (Schnyder et al., 1997). In brief, cells from the above cultures were diluted to 3 to 30 cells·ml−1, dispensed (100 μl/well) into 96-well U-bottomed tissue culture plates, and stimulated with irradiated (45 Gy) allogenic PBMC (0.5 × 105/well) and the mitogen phytohemagglutinin (5 μg·ml−1). Growing cells were expanded in culture media containing IL-2 (60 U·ml−1) and restimulated on day 14.

To test the specificity of the clones (28 days after serial dilution), T cell clones (0.5 × 105; total volume, 0.2 ml) were incubated with autologous-irradiated (60 Gy) B-LCL (0.1 × 105) and CBZ (25 μg·ml−1) in U-bottomed 96-well tissue culture plates. After 48 h, [3H]thymidine was added, and proliferation was measured 16 h later by scintillation counting as described above. Cell cultures with an SI of greater than or equal to 2.5 were expanded further.

Characterization of Carbamazepine-Specific T Cell Clones.

CD phenotype of the T cell clones was measured by flow cytometry using fluorescent-labeled anti-CD3, anti-CD4, and anti CD8 antibodies. Monoclonality was demonstrated by T cell receptor Vβ-chain staining using an anti-CD3 antibody and a panel of 25 antibodies that detect more than 75% of all known Vβ families (Immunotech, Marseille, France) (Zanni et al., 1997). Cells were stained as described above, and fluorescence was measured by flow cytometry.

To investigate the specificity of the interaction between CBZ and the T cell receptor, T cell clones were incubated with CBZ (1–100 μg·ml−1) and nine structurally related compounds (10 and 50 μg·ml−1) (Fig.2). These compounds include the widely used anticonvulsants phenytoin and lamotrigine and CBZ-10,11-epoxide, a major metabolite formed in humans after administration of CBZ (Maggs et al., 1997). Proliferation was determined after 48 h, as described above.

Figure 2
Figure 2

Chemical structures of CBZ, CBZ metabolites, CBZ analogs, and structurally unrelated anticonvulsants. Each compound has been given a number, which is used in Fig. 7.

Restriction of the CBZ-specific T cell clones was determined: first, by the addition of specific anti-HLA blocking antibodies (anti-class 1, anti-DR, anti-DP, and anti-DQ [all obtained from Dr. E. Padovan, University of Basel, Basel, Switzerland]) to the proliferation assay at concentrations known to inhibit MHC-restricted stimulations of CD8+ and CD4+ T cell clones (Schnyder et al., 1997); and second, by measuring proliferation with partly HLA-matched and unmatched B-LCLs as antigen-presenting cells. To exclude self-presentation by HLA-DR+ T cells, certain incubations contained T cells and CBZ in the absence of antigen-presenting cells. For matching HLA alleles, HLA-typing was performed by PCR in our tissue-typing laboratories.

Measurement of Cytokine Levels in Cell Culture Supernatant.

Cell culture supernatant from the proliferation assay was taken after 48 h to measure cytokine levels by ELISA. Supernatant of incubations containing T cells and B-LCLs in the absence of CBZ was taken as a control. The following ELISA kits were used: IL-4, IL-10, IFN-γ (Diaclone, Besancon, France), and IL-5 (BD PharMingen, San Diego, CA). Samples were diluted in duplicate (1:10, 1:100, and 1:1000; v/v), and cytokine levels were measured according to standard protocols of the corresponding ELISA kit. The detection limits were 1 pg·ml−1 for IL-4, 8 pg·ml−1 for IL-5, and 13 pg·ml−1 for IL-10 and IFN-γ.

Statistical Anlaysis.

All data are expressed as mean ± S.D. The Mann-Whitney test was used for comparison of control and test values, accepting P < 0.05 as significant.

Results

In Vitro Stimulation of PBMC by Carbamazepine.

Incubation of CBZ with PBMC from hypersensitive patients, but not from the two control groups (SI: 1.3 ± 0.4 for CBZ-exposed nonhypersensitive; SI: 1.1 ± 0.3 for nonexposed; 10 μg·ml−1), caused a strong proliferative response, as measured by the incorporation of [3H]thymidine. Proliferation was seen with 1 to 50 μg·ml−1 CBZ (Fig.3a), whereas concentrations of 75 μg·ml−1 and greater inhibited proliferation without causing direct toxicity (determined by trypan blue dye exclusion; results not shown). Flow cytometric analysis of CBZ-stimulated PBMC on day 6 revealed an up-regulation in the expression of HLA-DR molecules on CD4+ and CD8+ T cell subsets from hypersensitive patients 1 (3.8 ± 0.5% CD4+ and 7.7 ± 1.2% CD8+) and 2 (6.2 ± 0.4% CD4+ and 0.8 ± 0.2% CD8+). PBMC also proliferated in the presence of the major metabolite of CBZ, CBZ-10,11-epoxide (Fig. 3b). No significant difference in TT (1 μg·ml−1)-specific proliferation was seen when patient and control PBMCs were compared (P > 0.05).

Figure 3
Figure 3

Demonstration of dose-dependent (a) CBZ-specific and (b) CBZ-10,11-epoxide-specific PBMC proliferation in vitro. PBMC (1.5 × 105/well) from hypersensitive patients were incubated with CBZ (1–100 μg·ml−1) or CBZ-10,11-epoxide (1–100 μg·ml−1) for 6 days, and proliferation was measured by [3H]thymidine incorporation. Results are expressed as mean SI ± S.D. of triplicate cultures. The cpm in control incubations was consistently less than 1500. Statistical analysis was performed by comparing incubations in the presence of drug with those in the absence of drug (*, P < 0.05).

The proliferative response of PBMC to CBZ from hypersensitive patients was reproduced in further experiments using fresh PBMC approximately 6 months after the first blood donation (results not shown).

Carbamazepine-Specific T Cells Display Cytotoxic Activity Against Autologous B-LCL.

T-cell–enriched PBMC from CBZ hypersensitive patients showed a dose-dependent cytotoxic response against autologous B-LCLs (Fig. 4a). In addition, two of three CBZ-specific T cell clones incubated with CBZ were able to kill51Cr-loaded B-LCL with an effector/target ratio of 10:1 and greater (Fig. 4b). Incubation of CBZ with51Cr-loaded B-LCL alone did not cause significant51Cr release (P > 0.05).

Figure 4
Figure 4

Cytotoxic activity of (a) T cell-enriched PBMC (PBMC stimulated for 3–4 weeks with irradiated autologous PBMC and CBZ) and (b) T cell clones stimulated with CBZ. Cell death was measured in a 4-h51Cr-release assay. T cells were incubated with CBZ (10–50 μg·ml−1) and 51Cr-loaded B-LCL (2.5 × 103) at the indicated dilutions. Specific lysis was calculated as 100 × (experimental release − spontaneous release)/maximal release − spontaneous release). The value of spontaneous release from 51Cr-loaded B-LCL was between 124 and 860 cpm, and the ratio of maximal release to spontaneous release ranged between 3.1 and 5.8. Statistical analysis was performed by comparing incubations in the presence of drug to those in the absence of drug (*, P < 0.05). The coefficient of variation was consistently less than 20%.

Phenotype and Specificity of Carbamazepine-Specific T Cell Clones.

More than 500 T cell clones were generated from PBMCs of hypersensitive patients 1 and 2. Thirty-nine CBZ-specific T cell clones were obtained from patient 1 (SI: 11.2 ± 15.7; range, 2.3–80.3), and 5 were obtained from patient 2 (SI: 20.1 ± 11.2; range, 3.0–31.4); 10 wells growing clones were chosen for further analysis. Seven clones were of the CD4+ phenotype, whereas three clones stained double-positive (i.e., the same cells expressed both CD4+ and CD8+); all clones expressed the αβ T cell receptor. Of the 10 clones, 7 expressed a single Vβ chain (Fig. 5), whereas the Vβ chain of the other clones was not detected with our panel of 25 antibodies. Four clones expressed the T cell receptor Vβ 5.1, which suggests a preferential usage in CBZ hypersensitivity reactions.

Figure 5
Figure 5

Phenotype and dose-dependent proliferation of CBZ-specific T cell clones. Each T cell clone is presented by the clone name and CD phenotype followed by the Vβ phenotype in parentheses. CD and T cell receptor Vβ phenotype were measured by flow cytometry using fluorescent-labeled antibodies. To measure proliferation, T cell clones (0.5 × 105/well) were incubated with CBZ (1–100 μg·ml−1) and irradiated (60 Gy) autologous B-LCL (0.1 × 105) for 48 h as described underMaterials and Methods. Results are given as mean [3H]thymidine incorporation of duplicate cultures. Statistical analysis was performed by comparing incubations in the presence of drug to those in the absence of drug (*,P < 0.05). The coefficient of variation was consistently less than 20%.

All of the clones responded to CBZ in a dose-dependent manner (Fig. 5). The concentration required for maximal proliferation ranged between 5 and 50 μg·ml−1; proliferation was also seen with 100 μg·ml−1, a concentration that inhibited proliferation of PBMC (Fig. 3). Concentrations of CBZ greater than 100 μg·ml−1 inhibited proliferation (results not shown).

T Cell Receptor Recognition of Carbamazepine by Antigen-Presenting Cells is HLA-DR– and HLA-DQ–Restricted.

There was no proliferation when the T cell clones were stimulated with CBZ in the absence of B-LCL. These data indicate that T cells cannot present CBZ to each other (data not shown). In the presence of autologous B-LCL, the response to CBZ was HLA class II-restricted. Antibodies against HLA-DR and HLA-DQ, but not HLA-DP, inhibited proliferation. CBZ was presented to eight clones on HLA-DR and two clones on HLA-DQ. Figure6 shows a panel of representative T cell clones.

Figure 6
Figure 6

HLA restriction of CBZ-specific T cell clones. T cell clones (0.5 × 105/well) were incubated with irradiated (60 Gy) autologous B-LCL (0.1 × 105) and CBZ (10 μg·ml−1) in the presence of monoclonal antibodies against MHC class I or MHC class II (DR, DP, and DQ). Results are given as mean [3H]thymidine incorporation of duplicate cultures. Statistical analysis was performed by comparing incubations in the presence of drug to those in the absence of drug (*,P < 0.05). The coefficient of variation was consistently less than 20%.

Previous studies have demonstrated that approximately 75% of HLA-DR–restricted drug-specific T cell clones (von Greyerz et al., 2001) recognize drug antigens in the presence of HLA-matched antigen-presenting cells, and the other clones proliferated irrespective of the HLA phenotype of the antigen presenting cell used. In these studies, hypersensitive patient 1 was positive for HLA-DRB1*15. All of the HLA-DR–restricted T cell clones from this patient showed a CBZ-specific proliferative response in the presence of allogenic HLA-matched B-LCL, whereas no proliferation was seen with HLA-unmatched B-LCL. Hypersensitive patient 2 was positive for HLA-DRB1*04 and *09. CBZ was not presented by unmatched B-LCL or B-LCL partly matched with either allele.

Cytokine Production by Carbamazepine-Specific T Cell Clones.

Analysis of CBZ-specific T cell clones revealed three distinct cytokine patterns: four clones produced low to moderate levels of IL-4, -5, -10, and IFN-γ; three clones produced particularly high amounts of the proinflammatory cytokine IFN-γ; and two clones produced moderate levels of IFN-γ, but no IL-4, -5, or -10 (Table2). High levels of IFN-γ were seen with T cell clones from hypersensitive patients 1 and 2.

View this table:
Table 2

Cytokine secretion from CBZ-stimulated T cell clones

Cross-Reactivity of Carbamazepine-Specific T Cell Clones.

To study the structural requirements of drug-antigen recognition by the MHC-restricted T cell receptor, we analyzed the cross-reactivity pattern of CBZ-specific T cell clones with nine structurally related compounds (Fig. 2). Cross-reactivity was seen with dihydro-CBZ (structure 3) and CBZ-10,11-epoxide (structure 2); however, any further structural modification inhibited proliferation. Four clones proliferated in the presence of dihydro-CBZ, five clones proliferated in the presence of CBZ-10,11-epoxide, and three clones were cross-reactive. Four clones did not recognize any compounds other than CBZ. Figure 7 shows the cross-reactivity profile of three representative T cell clones.

Figure 7
Figure 7

Stimulation of T cell clones by CBZ, CBZ metabolites, structural analogs, and structurally unrelated anticonvulsants. Proliferation was measured after 48 h by [3H]thymidine incorporation. Results are given as mean [3H]thymidine incorporation ± S.D. of duplicate cultures. Numbers refer to the structures displayed in Fig. 2. Statistical analysis was performed by comparing incubations in the presence of drug to those in the absence of drug (*,P < 0.05). The coefficient of variation was consistently less than 20%.

With the exception of clone CBZ 25, the capacity of dihydro CBZ and CBZ-10,11-epoxide to stimulate proliferation was significantly lower than that seen with CBZ itself. Clone CBZ 25, which responded weakly with CBZ (results represent the mean of three separate experiments), proliferated more vigorously in the presence of CBZ-10,11-epoxide (Fig.8).

Figure 8
Figure 8

Dose-dependent proliferation of two representative T cell clones stimulated with CBZ, CBZ-10,11-epoxide, dihydro-CBZ, and iminostilbene. Proliferation was measured after 48 h by [3H]thymidine incorporation. Results are given as mean [3H]thymidine incorporation of duplicate cultures. Statistical analysis was performed by comparing incubations in the presence of drug with those in the absence of drug (*,P < 0.05). The coefficient of variation was consistently less than 20%.

Discussion

Although relatively rare, CBZ hypersensitivity reactions are a cause of significant patient morbidity and mortality and can prevent effective drug therapy, especially if the drug has to be withdrawn (Leeder, 1998). Hypersensitivity reactions occur as a result of an individual's immune system responding to an inappropriate stimulus, which may take the form of the parent drug or a drug-modified self-protein subsequent to drug bioactivation.

Zakrzewska and Ivanyi (1988) and others (Mauri-Hellweg et al., 1995;Brown et al., 1999; Hari et al., 2001) have shown CBZ-specific proliferation of PBMC isolated from the peripheral blood of hypersensitive patients. Our data extend these initial findings: PBMC from five hypersensitive patients proliferated vigorously when challenged in vitro with therapeutic CBZ concentrations that are estimated to be in the range of 5 to 10 μg·ml−1 (Tomson et al., 1980) (SI: 32.1 ± 24.2; CBZ 10 μg·ml−1) (Fig. 3). Proliferation was dose-dependent, and both CD4+ and CD8+ T cells expressed elevated levels of the activation marker HLA-DR. PBMCs were also stimulated by CBZ-10,11-epoxide (25–75 μg·ml−1), a major metabolite formed after administration of CBZ to patients (Madden et al., 1996). Previous work failed to detect T cells responsive to the epoxide metabolite (Zakrzewska and Ivanyi, 1988); however, Zakrzewska and Ivanyl used only three concentrations of CBZ-10,11-epoxide (1, 10, and 100 μg·ml−1), which in our study caused nominal proliferation.

In a retrospective study derived from the clinical history of 923 drug allergic patients, the lymphocyte transformation test was shown to have a sensitivity (i.e., frequency of a positive result in drug-allergic patients) and specificity (i.e., frequency of a negative result in nonallergic patients) of 78% and 85%, respectively (Nyfeler and Pichler, 1997). A recent prospective study of 22 patients by Hari et al. (2001) produced similar findings, with the lymphocyte transformation test being positive in 67% of drug-allergic patients. In our small cohort of CBZ-hypersensitive patients and controls, the lymphocyte transformation test was positive in only the hypersensitive patients. The potential use of this test in a clinical situation requires further investigation, but it could include the identification of the culprit drug in a patient on multiple drugs and determination of cross-reactivity.

To further investigate the nature of the cellular immune response, drug-specific T cells were cloned and characterized. CD4+ and CD4+/CD8+ T cell clones were generated, and drug stimulation triggered proliferation and cytotoxicity (Figs. 3 and 4). All of the CBZ-specific clones expressed the αβ T cell receptor, and the use of at least four different Vβ genes shows that the response to CBZ was polyclonal. Four clones expressed Vβ chain 5.1 (three from hypersensitive patient 1, one from hypersensitive patient 2), which suggests a preferential use in CBZ-hypersensitivity reactions. It is possible that clones which express the same Vβ chain derive from a single cell, with its outgrowth being a consequence of the cloning procedure; this, however, seems unlikely because CBZ was presented on HLA-DR and HLA-DQ, and cross-reactivity and cytokine profiles were heterogeneous (see below).

A high level of the proinflammatory cytokine IFN-γ was detected in the supernatant of several CBZ-stimulated T cell clones. These data are in line with previous investigations by Leyva et al. (2000) showing the overexpression of IFN-γ mRNA in PBMC isolated from CBZ-hypersensitive patients, and they explain the enhanced expression of HLA-DR on epidermal keratinocytes at the time of the hypersensitivity reaction (Friedmann et al., 1994). In mice, IFN-γ production by CD4+ T cells is involved in delayed-type cutaneous reactions observed after the topical exposure of dinitrochlorobenzene (Dearman et al., 1996a,b) and in hypersensitivity reactions solely affecting the liver (Ohta et al., 2000).

A prominent clinical feature of CBZ hypersensitivity is an increase in the number of circulating eosinophils (Leeder, 1998). IL-5 is essential for the growth and activation of eosinophils, and most drug-specific CD4+ T cells from patients with drug-induced maculopapular eruptions (the eosinophil count is also increased in these reactions) release high levels of IL-5 (Pichler et al., 1997; Yawalker et al., 2000b). Our panel of CBZ-specific T cell clones secreted low to moderate amounts of IL-5; this may have been adequate to account for the eosinophilia observed in these reactions. Clearly, it is also possible that there may be other clones, which we did not characterize, that are particularly high secretors of IL-5.

Although most T cells recognize drug antigens in the context of MHC expressed on the surface of antigen-presenting cells (von Greyerz et al., 1999), chemicals such as isopentenyl pyrophosphate and fluorescein activate T cells via a direct interaction with the T cell receptor (Diamond et al., 1991; Morita et al., 1995). In our studies, there was no proliferation when the T cell clones were incubated with CBZ in the absence of antigen-presenting cells. These data indicate that T cells do not present CBZ to each other. Unlike previous studies investigating the involvement of T cells in drug-mediated allergic disease, in which certain clones responded in an MHC-restricted but MHC-allele unrestricted way (von Greyerz et al., 2001), no stimulation was seen when the T cell clones were incubated with CBZ- and MHC-mismatched antigen-presenting cells. Most clones recognized CBZ in the context of HLA-DR; however, CBZ was presented to two clones on HLA-DQ. These data are in accordance with the recently observed MHC gene polymorphism associated with severe CBZ-hypersensitivity reactions (Pirmohamed et al., 2001).

In CBZ hypersensitivity, the nature of the drug antigen presented on MHC (i.e., CBZ or a CBZ metabolite peptide conjugate) to T cells is not known. It is possible that CBZ (and CBZ-10,11-epoxide) is metabolized to a reactive intermediate in situ in the proliferation assay; however, using the bioanalytical techniques that are currently available, it is not possible to detect such low levels of metabolic activation. Therefore, to attain a greater understanding of the structural requirements of T cell receptor activation, CBZ-specific T cell clones were stimulated with a panel of nine structurally related compounds, which included the commonly administered anticonvulsants lamotrigine and phenytoin. Structural modification alters the pharmacokinetic properties of CBZ, which may in turn affect its subcellular distribution within the MHC-restricted T cell receptor and/or the active site of cytochrome P450. Our data show that the drug-T cell receptor interaction is highly regulated (i.e., small structural changes inhibited proliferation) and are in accordance with previous studies with sulfamethoxazole (Schnyder et al., 1997; von Greyerz et al., 1999). Modification of the nonaromatic double bond on the CBZ molecule (i.e., dihydro-CBZ [structure 3] and CBZ-10,11-epoxide [structure 2]) could be accommodated by the MHC T cell receptor binding site; however, further structural modification inhibited the interaction with either the MHC or T cell receptor. Removal of the amide side chain (iminostilbene [4], a known CBZ metabolite) (Csetenyi et al., 1973), which is not necessary for the interaction of CBZ with the active site of cytochrome P450 (Riley et al., 1989), or the aromatic ring completely abolished T cell receptor activation. The apparent direct stimulation of T cells by parent drug suggests that metabolic activation is not a prerequisite for T cell receptor activation, at least in vitro. However, further experiments using specific inhibitors of drug metabolism and/or antigen processing alongside more sophisticated and sensitive bioanalytical methodology is required to confirm this concept. Because the T cell clones were generated by culturing PBMC with soluble CBZ, these studies do not exclude the possibility that hypersensitive patients have T cells that recognize a CBZ metabolite protein conjugate; in ongoing investigations, we are trying to select clones derived from proliferation with CBZ metabolites and/or CBZ-modified protein. One clone that proliferated weakly with CBZ showed a more vigorous proliferation in the presence of CBZ-10,11-epoxide (Fig. 8). These data suggest that the patient may have T cells that respond to both CBZ and CBZ-10,11-epoxide in vivo.

In conclusion, CBZ was presented on MHC class II expressed on the surface of antigen-presenting cells to the T cell receptor of CD4+ T cells. T cell-receptor activation resulted in proliferation, cytotoxicity, and secretion of high levels of IFN-γ. These studies show that patients with CBZ have a higher proportion of IFN-γ–secreting cells, which might be relevant for the more general clinical symptoms seen in patients with CBZ hypersensitivity, particularly the accompanying hepatitis. Additionally, the cells had a phenotype that was qualitatively different from those seen in other serious cutaneous adverse drug reactions, such as bullous and pustular rashes.

Acknowledgments

The authors thank the Departments of Immunology and Tropical Medicine, The University of Liverpool, Liverpool, United Kingdom, for use of equipment.

Footnotes

    • Received June 11, 2002.
    • Accepted December 10, 2002.

  • These studies were supported by The Wellcome Trust. D.J.N. is a Wellcome Trust Research Career Development Fellow. Data were presented in part as an oral communication entitled “Naisbitt DJ, Britschgi M, Wong G, Farrell J, Chadwick DW, Pichler WJ, Pirmohamed M, Park BK. Anticonvulsant hypersensitivity syndrome: characterization of the phenotype and cytokine profile of carbamazepine-specific T cell clones.Br J Clin Pharmacol (in press)” at the British Pharmacological Society Meeting, London, December 2001.

Abbreviations

CBZ
carbamazepine
B-LCL
B-lymphoblastoid cell line
PBMC
peripheral blood mononuclear cells
SI
stimulation index
TT
tetanus toxoid
CD
circular dichroism
HLA
human leukocyte antigen
IFN-γ
interferon-γ
IL
interleukin
MHC
major histocompatibility complex
ELISA
enzyme-linked immunosorbent assay

References

  1. ↵
    1. Britschgi M,
    2. Steiner UC,
    3. Schmid S,
    4. Depta JP,
    5. Senti G,
    6. Bircher A,
    7. Burkhart C,
    8. Yawalkar N, and
    9. Pichler WJ
    (2001) T-cell involvement in drug-induced acute generalized exanthematous pustulosis. J Clin Investig 107:1433–1441.
    OpenUrlCrossRefPubMed
  2. ↵
    1. Brown KL,
    2. Henderson DC,
    3. Nadel S,
    4. Tanveer A, and
    5. Booy R
    (1999) Carbamazepine hypersensitivity and the use of lymphocyte proliferation responses. Dev Med Child Neurol 41:267–269.
    OpenUrlCrossRefPubMed
  3. ↵
    1. Brunner KT,
    2. Mauel J,
    3. Cerottini JC, and
    4. Chapuis B
    (1968) Quantitative assay of the lytic action of immune lymphoid cells on 51Cr-labelled allogenic target cells in vitro: inhibition by isoantibody and by drugs. Immunology 14:181–196.
    OpenUrlPubMed
  4. ↵
    1. Csetenyi J,
    2. Baker KM,
    3. Frigerio A, and
    4. Morselli PL
    (1973) Iminostilbene—a metabolite of carbamazepine isolated from rat urine. J Pharm Pharmacol 25:340–341.
    OpenUrlPubMed
  5. ↵
    1. Dearman RJ,
    2. Basketter DA, and
    3. Kimber I
    (1996a) Characterization of chemical allergens as a function of divergent cytokine secretion profiles induced in mice. Toxicol Appl Pharmacol 138:308–316.
    OpenUrlCrossRefPubMed
  6. ↵
    1. Dearman RJ,
    2. Moussavi A,
    3. Kemeny DM, and
    4. Kimber I
    (1996b) Contribution of CD4+ and CD8+ T lymphocytes to the cytokine secretion patterns induced in mice during sensitization to contact and respiratory chemical allergens. Immunology 89:502–510.
    OpenUrlCrossRefPubMed
  7. ↵
    1. Diamond DJ,
    2. Szalay P,
    3. Symer D,
    4. Hao P,
    5. Shin HS,
    6. Dintzis RZ,
    7. Dintzis HM,
    8. Reinherz EL, and
    9. Siliciano RF
    (1991) Major histocompatibility complex independent T cell receptor-antigen interaction: functional analysis using fluorescein derivatives. J Exp Med 174:229–241.
    OpenUrlAbstract/FREE Full Text
  8. ↵
    1. Friedmann PS,
    2. Strickland I,
    3. Pirmohamed M, and
    4. Park BK
    (1994) Investigation of mechanisms in toxic epidermal necrolysis induced by carbamazepine. Arch Dermatol 130:598–604.
    OpenUrlCrossRefPubMed
  9. ↵
    1. Hari Y,
    2. Frutig-Schnyder K,
    3. Hurni M,
    4. Yawalkar N,
    5. Zanni MP,
    6. Schnyder B,
    7. Kappeler A,
    8. von Greyerz S,
    9. Braathen LR, and
    10. Pichler WJ
    (2001) T cell involvement in cutaneous drug eruptions. Clin Exp Allergy 31:1398–1408.
    OpenUrlCrossRefPubMed
  10. ↵
    1. Hertl M,
    2. Jugert F, and
    3. Merk HF
    (1995) CD8+ dermal T cells from a sulphamethoxazole-induced bullous exanthem proliferate in response to drug-modified liver microsomes. Br J Dermatol 132:215–220.
    OpenUrlPubMed
  11. ↵
    1. Ju C and
    2. Uetrecht JP
    (1999) Detection of 2-hydroxyiminostilbene in the urine of patients taking carbamazepine and its oxidation to a reactive iminoquinone intermediate. J Pharmacol Exp Ther 288:51–56.
    OpenUrlAbstract/FREE Full Text
  12. ↵
    1. Koga T,
    2. Kubota Y, and
    3. Nakayama J
    (2000) Interferon-gamma production in the peripheral lymphocytes of a patient with carbamazepine hypersensitivity syndrome. Acta Derm Venereol 80:73.
    OpenUrlCrossRefPubMed
  13. ↵
    1. Leeder JS
    (1998) Mechanisms of idiosyncratic hypersensitivity reactions to antiepileptic drugs. Epilepsia 39:S8–S16.
    OpenUrlCrossRefPubMed
  14. ↵
    1. Leyva L,
    2. Torres MJ,
    3. Posadas S,
    4. Blanca M,
    5. Besso G,
    6. O'Valle F,
    7. del Moral RG,
    8. Santamaria LF, and
    9. Juarez C
    (2000) Anticonvulsant-induced toxic epidermal necrolysis: monitoring the immunologic response. J Allergy Clin Immunol 105:157–165.
    OpenUrlCrossRefPubMed
  15. ↵
    1. Madden S,
    2. Maggs JL, and
    3. Park BK
    (1996) Bioactivation of carbamazepine in the rat in-vivo—evidence for the formation of reactive arene oxide(s). Drug Metab Dispos 24:469–479.
    OpenUrlAbstract
  16. ↵
    1. Maggs JL,
    2. Pirmohamed M,
    3. Kitteringham NR, and
    4. Park BK
    (1997) Characterization of the metabolites of carbamazepine in patient urine by liquid chromatography/mass spectrometry. Drug Metab Dispos 25:275–280.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    1. Mauri-Hellweg D,
    2. Bettens F,
    3. Mauri D,
    4. Brander C,
    5. Hunziker T, and
    6. Pichler WJ
    (1995) Activation of drug-specific CD4+ and CD8+ T cells in individuals allergic to sulfonamides, phenytoin and carbamazepine. J Immunol 155:462–472.
    OpenUrlAbstract
  18. ↵
    1. Morita CT,
    2. Beckman EM,
    3. Bukowski JF,
    4. Tanaka Y,
    5. Band H,
    6. Bloom BR,
    7. Golan DE, and
    8. Brenner MB
    (1995) Direct presentation of nonpeptide prenyl pyrophosphate antigens to human gamma delta T cells. Immunity 3:495–507.
    OpenUrlCrossRefPubMed
  19. ↵
    1. Nassif A,
    2. Bensussan A,
    3. Dorothee G,
    4. Mami-Chouaib F,
    5. Bachot N,
    6. Bagot M,
    7. Boumsell L, and
    8. Roujeau JC
    (2002) Drug specific cytotoxic T-cells in the skin lesions of a patient with toxic epidermal necrolysis. J Investig Dermatol 118:728–733.
    OpenUrlCrossRefPubMed
  20. ↵
    1. Nyfeler B and
    2. Pichler WL
    (1997) The lymphocyte transformation test for the diagnosis of drug allergy: sensitivity and specificity. Clin Exp Allergy 27:175–181.
    OpenUrlCrossRefPubMed
  21. ↵
    1. Ohta A,
    2. Sekimoto M,
    3. Sato M,
    4. Koda T,
    5. Nishimura S,
    6. Iwakura Y,
    7. Sekikawa K, and
    8. Nishimura T
    (2000) Indispensable role for TNF-alpha and IFN-gamma at the effector phase of liver injury mediated by Th1 cells specific to hepatitis B virus surface antigen. J Immunol 165:956–961.
    OpenUrlAbstract/FREE Full Text
  22. ↵
    1. Park BK,
    2. Pirmohamed M, and
    3. Kitteringham NR
    (1998) Role of drug disposition in drug hypersensitivity: a chemical, molecular and clinical perspective. Chem Res Toxicol 9:969–988.
    OpenUrl
  23. ↵
    1. Pichler WJ,
    2. Yawalker N,
    3. Britschgi M,
    4. Depta J,
    5. Strasse I,
    6. Schmid S,
    7. Kuechle P, and
    8. Naisbitt D
    (2002) Cellular and molecular pathophysiology of cutaneous drug reactions. Am J Clin Dermatol 3:229–238.
    OpenUrlCrossRefPubMed
  24. ↵
    1. Pichler WJ,
    2. Zanni M,
    3. von Greyerz S,
    4. Schnyder B,
    5. Mauri-Hellweg D, and
    6. Wendland T
    (1997) High IL-5 production by human drug-specific T cell clones. Int Arch Allergy Immunol 113:177–180.
    OpenUrlPubMed
  25. ↵
    1. Pirmohamed M,
    2. Lin K,
    3. Chadwick D, and
    4. Park BK
    (2001) TNF-alpha promoter region gene polymorphisms in carbamazepine-hypersensitive patients. Neurology 56:890–896.
    OpenUrlAbstract/FREE Full Text
  26. ↵
    1. Riley RJ,
    2. Kitteringham NR, and
    3. Park BK
    (1989) Structural requirements for bioactivation of anticonvulsants to cytotoxic metabolites in vitro. Br J Clin Pharmacol 28:482–487.
    OpenUrlPubMed
  27. ↵
    1. Schnyder B,
    2. Frutig K,
    3. Mauri-Hellweg D,
    4. Limat A,
    5. Yawalkar N, and
    6. Pichler WJ
    (1998) T-cell-mediated cytotoxicity against keratinocytes in sulfamethoxazole-induced skin reaction. Clin Exp Allergy 28:1412–1417.
    OpenUrlCrossRefPubMed
  28. ↵
    1. Schnyder B,
    2. Mauri-Hellweg D,
    3. Zanni M,
    4. Bettens F, and
    5. Pichler WJ
    (1997) Direct, MHC-dependent presentation of the drug sulfamethoxazole to human alphabeta T cell clones. J Clin Investig 100:136–141.
    OpenUrlCrossRefPubMed
  29. ↵
    1. Tennis P and
    2. Stern RS
    (1997) Risk of serious cutaneous disorders after initiation of use of phenytoin, carbamazepine, or sodium valproate: a record linkage study. Neurology 49:542–546.
    OpenUrlAbstract/FREE Full Text
  30. ↵
    1. Tomson T,
    2. Tybring G,
    3. Bertilsson L,
    4. Ekbom K, and
    5. Rane A
    (1980) Carbamazepine therapy in trigeminal neuralgia: clinical effects in relation to plasma concentration. Arch Neurol 37:699–703.
    OpenUrlCrossRefPubMed
  31. ↵
    1. Villada G,
    2. Roujeau JC,
    3. Clerici T,
    4. Bourgault I, and
    5. Revuz J
    (1992) Immunopathology of toxic epidermal necrolysis. Arch Dermatol 128:50–53.
    OpenUrlCrossRefPubMed
  32. ↵
    1. Vitorrio CC and
    2. Muglia JJ
    (1995) Anticonvulsant hypersensitivity syndrome. Arch Intern Med 155:2285–2290.
    OpenUrlCrossRefPubMed
  33. ↵
    1. von Greyerz S,
    2. Bultemann G,
    3. Schnyder K,
    4. Burkhart C,
    5. Lotti B,
    6. Hari Y, and
    7. Pichler WJ
    (2001) Degeneracy and additional alloreactivity of drug-specific human alpha beta+ T cell clones. Int Immunol 13:877–885.
    OpenUrlAbstract/FREE Full Text
  34. ↵
    1. von Greyerz S,
    2. Zanni MP,
    3. Frutig K,
    4. Schnyder B,
    5. Burkhart C, and
    6. Pichler WJ
    (1999) Interaction of sulfonamide derivatives with the TCR of sulfamethoxazole-specific human alpha beta+ T cell clones. J Immunol 162:595–602.
    OpenUrlAbstract/FREE Full Text
  35. ↵
    1. Yawalkar N,
    2. Hari Y,
    3. Frutig K,
    4. Egli F,
    5. Wendland T,
    6. Braathen LR, and
    7. Pichler WJ
    (2000a) T cells isolated from positive epicutaneous test reactions to amoxicillin and ceftriaxone are drug specific and cytotoxic. J Investig Dermatol 115:647–652.
    OpenUrlCrossRefPubMed
  36. ↵
    1. Yawalkar N,
    2. Shrikhande M,
    3. Hari Y,
    4. Nievergelt H,
    5. Braathen LR, and
    6. Pichler WJ
    (2000b) Evidence for a role for IL-5 and eotaxin in activating and recruiting eosinophils in drug-induced cutaneous eruptions. J Allergy Clin Immunol 106:1171–1176.
    OpenUrlCrossRefPubMed
  37. ↵
    1. Zakrzewska JM and
    2. Ivanyi L
    (1988) In vitro lymphocyte proliferation by carbamazepine, carbamazepine-10, 11-epoxide and oxcarbazepine in the diagnosis of drug-induced hypersensitivity. J Allergy Clin Immunol 82:110–115.
    OpenUrlCrossRefPubMed
  38. ↵
    1. Zanni MP,
    2. Mauri-Hellweg D,
    3. Brander C,
    4. Wendland T,
    5. Schnyder B,
    6. Frei E,
    7. von Greyerz S,
    8. Bircher A, and
    9. Pichler WJ
    (1997) Characterization of lidocaine-specific T cells. J Immunol 158:1139–1148.
    OpenUrlAbstract
    1. Zanni MP,
    2. von Greyerz S,
    3. Schnyder B,
    4. Brander KA,
    5. Frutig K,
    6. Hari Y,
    7. Valitutti S, and
    8. Pichler WJ
    (1998) HLA-restricted, processing- and metabolism-independent pathway of drug recognition by human alpha beta T lymphocytes. J Clin Investig 102:1591–1598.
    OpenUrlCrossRefPubMed
Back to top

In this issue

Download PDF
Article Alerts
Email Article
Citation Tools
Research ArticleArticle

Hypersensitivity Reactions to Carbamazepine: Characterization of the Specificity, Phenotype, and Cytokine Profile of Drug-Specific T Cell Clones

D. J. Naisbitt, M. Britschgi, G. Wong, J. Farrell, J. P. H. Depta, D. W. Chadwick, W. J. Pichler, M. Pirmohamed and B. K. Park
Molecular Pharmacology March 1, 2003, 63 (3) 732-741; DOI: https://doi.org/10.1124/mol.63.3.732
del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

Jump to section