The Journal of Allergy and Clinical Immunology
Volume 125, Issue 3 , Pages 703-710.e8, March 2010

CD94/NKG2C is a killer effector molecule in patients with Stevens-Johnson syndrome and toxic epidermal necrolysis

  • Esther Morel, PhD

      Affiliations

    • Research Unit, Hospital Universitario La Paz–FIBHULP, Madrid, Spain
    • ,
  • Salvador Escamochero, BS

      Affiliations

    • Research Unit, Hospital Universitario La Paz–FIBHULP, Madrid, Spain
    • ,
  • Rosario Cabañas, MD, PhD

      Affiliations

    • Allergy Service, Hospital Universitario La Paz, Madrid, Spain
    • ,
  • Rosa Díaz, MD, PhD

      Affiliations

    • Dermatology Service, Hospital Universitario La Paz, Madrid, Spain
    • ,
  • Ana Fiandor, MD

      Affiliations

    • Allergy Service, Hospital Universitario La Paz, Madrid, Spain
    • ,
  • Teresa Bellón, PhD

      Affiliations

    • Research Unit, Hospital Universitario La Paz–FIBHULP, Madrid, Spain
    • Corresponding Author InformationReprint requests: Teresa Bellón, PhD, Research Unit, Hospital Universitario “La Paz,” P° Castellana 261, 28046 Madrid, Spain.

Received 22 March 2009; received in revised form 9 September 2009; accepted 19 October 2009. published online 04 February 2010.

Article Outline

Background

Toxic epidermal necrolysis (TEN) and Stevens-Johnson syndrome (SJS) are severe, bullous cutaneous diseases with uncertain pathogenesis, although cytotoxic T cells seem to be involved. Natural killer (NK)–like activity has been found in blister infiltrates. Cytotoxic T lymphocytes (CTLs) with NK-like activity (NK-CTLs) have been shown to express T-cell receptors restricted by the HLA-Ib molecule HLA-E. Alternatively, the HLA-E–specific activating receptor CD94/NKG2C can trigger T-cell receptor–independent cytotoxicity in CTLs.

Objective

Our aim was to test whether HLA-E expression sensitizes keratinocytes to killing by CTLs with NK-like activity and to explore the expression of activating receptors specific for HLA-E in blister cytotoxic lymphocytes.

Methods

We used flow cytometry and immunohistochemistry to analyze HLA-E expression in keratinocytes from affected skin in patients with SJS, TEN, and other less severe drug-induced exanthemas. The expression of CD94/NKG2C was analyzed by means of flow cytometry in PBMCs and blister cells from patients. PBMCs and blister cells were analyzed for their ability to kill HLA-E–expressing cells. Involvement of CD94/NKG2C in triggering degranulation of cytolytic cells was explored by means of CD107a mobilization assays and standard cytotoxicity chromium release assays.

Results

We found that keratinocytes from affected skin expressed HLA-E and that cell-surface HLA-E sensitizes keratinocytes to killing by CD94/NKG2C+ CTLs. Frequencies of CD94/NKG2C+ peripheral blood T and NK cells were increased in patients with SJS and TEN during the acute phase. Moreover, activated blister T and NK lymphocytes expressed CD94/NKG2C and were able to degranulate in response to HLA-E+ cells in an NKG2C-dependent manner.

Conclusion

CD94/NKG2C might be involved in triggering cytotoxic lymphocytes in patients with SJS and TEN.

Key words: Stevens-Johnson syndrome, toxic epidermal necrolysis, drug allergy, NK-CTLs, CD94/NKG2C, HLA-E

Abbreviations used: BFC, Blister fluid cell, CTL, Cytotoxic T lymphocyte, FasL, Fas ligand, FITC, Fluorescein isothiocyanate, KIR, Killer immunoglobulin-like receptor, MPE, Maculopapular exanthema, NK, Natural killer, NKR, Natural killer receptor, PE, Phycoerythrin, rh, Recombinant human, SJS, Stevens-Johnson syndrome, TCR, T-cell receptor, TEN, Toxic epidermal necrolysis

 

Cutaneous reactions are the most frequent manifestations of delayed drug-induced hypersensitivity.1 They comprise a broad spectrum of clinical features spanning benign diseases, such as maculopapular exanthema (MPE), to life-threatening severe reactions. Among them, Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are the most severe forms of drug-induced skin diseases and are now considered variants of the same disease.2 SJS and TEN are rare diseases with a prevalence of 1 to 2 cases per million persons per year in white subjects and a mortality rate ranging from 2% to 10% in patients with SJS and 30% in patients with TEN. The most common feature of these diseases is the formation of subepidermal blisters and detachment of the epidermis, which appears as scalded skin. Necrosis of the full thickness of the epidermis is the pathognomonic finding in this entity.3 Blisters under the necrotic epidermis contain T lymphocytes, which can be regarded as the effectors of the immune reaction. T cells within blisters showed high expression of CD56,4 which has been associated with the acquisition of cytolytic effector functions by cytotoxic T lymphocytes (CTLs).5 Moreover, these cells express high levels of natural killer (NK) cell receptors (NKRs) and showed NK-like cytotoxic activity.6

T cells exhibiting NK-like activity have been named NK-CTLs and are characterized as T cells expressing CD56 and NKRs.7 It has been reported that T-cell receptors (TCRs) from NK-CTLs are HLA-E restricted, and therefore these cells kill targets through the recognition of the nonclassical MHC class Ib molecule HLA-E by specific TCRs.7, 8 Additionally, two HLA-E–specific NKRs (CD94/NKG2A and CD94/NKG2C) have been identified in human subjects.9 Both are expressed in subpopulations of NK and T cells. These heterodimers belong to the C-type lectin family and consist of an invariant CD94 chain covalently linked to NKG2A or NKG2C. Although the extracellular regions of NKG2A and NKG2C are highly homologous, the transmembrane and intracytoplasmic domains define opposite functions. CD94/NKG2A constitutes an inhibitory receptor because of the presence of two immunotyrosine-based inhibitory motifs in the NKG2A cytoplasmic tail. However, CD94/NKG2C also constitutes an activating receptor because NKG2C contains a charged residue in its transmembrane region that allows its association with the immunoreceptor tyrosine-based activation motif–containing adaptor molecule DNAX activating protein of 12 kDa (DAP12).10 It has recently been reported that CTLs might use CD94/NKG2C to kill HLA-E–expressing cells.11, 12

Because blister lymphocytes express cell-surface molecules similar to those described for NK-CTLs and are endowed with NK-like activity, we hypothesized that they could also be activated by HLA-E expressed on target cells.

Back to Article Outline

Methods 

A list of materials and reagents used in this study is found in this article's Methods section in the Online Repository available at www.jacionline.org.

Human samples 

Patients meeting the criteria for drug-induced delayed generalized cutaneous reactions1, 13 were evaluated in the Allergy Service of Hospital La Paz. All patients had negative serologic results for HIV or mycoplasma infection, no hematologic disorders were involved, and none of the patients were receiving corticosteroids before blood sampling. Fifteen patients with TEN or SJS and 38 patients with MPE associated to various drugs participated in the study (see Table E1, Table E2 in this article's Online Repository at www.jacionline.org). Peripheral blood was drawn from healthy donors or from patients at the moment of admission to the hospital (acute samples) or on complete remission of the clinical symptoms, typically at least 1 month after hospital discharge in patients with SJS/TEN (resolution samples). PBMCs were isolated by means of Ficoll/Hypaque (GE Healthcare, Uppsala, Sweden) density gradient centrifugation of blood samples. Blister fluids were obtained from tense blisters by means of puncture aspiration into a syringe. Blister fluid cells (BFCs) were collected by means of centrifugation for functional assays, and blister fluids were harvested and frozen at −80°C until use.

Cell suspensions were obtained by using 2 steps of 20 minutes' digestion with 0.1% Trypsin and 5 mmol/L EDTA of skin biopsy specimens. Cultures of primary human keratinocytes were generated as previously described14 and are explained in the Methods section in this article's Online Repository.

Informed consent was obtained from all participants or their legal representatives. The study was approved by the Ethics Committee of Hospital La Paz and conducted according to Declaration of Helsinki principles.

Cell lines, mAbs, and flow cytometry 

Cell lines, mAbs, and flow cytometry are detailed in the Methods section in this article's Online Repository.

Immunohistochemistry 

Immunohistochemistry was performed as described in the Methods section in this article's Online Repository by using the anti–HLA-E–specific mAb MEM-E/02.

Western blotting 

Protein content in sera and blister fluids was quantified, and equal amounts of protein (50 μg) were resolved by 10% SDS-PAGE and transferred onto polyvinylidene difluoride membranes (Bio-Rad Laboratories, Hercules, Calif). Equal volumes of culture supernatants from 721.221 and 221.AEH cell lines (20 μL) were loaded as control samples. Filters were probed with MEM-E/02 mAb and anti-mouse IgG horseradish peroxidase–linked antibodies (Chemicon International, Temecula Calif). Immunoreactivity was detected with the ECL Advance Western Blotting detection kit (GE Healthcare).

Isolation, magnetic separation, and cloning of CD56+ lymphocytes 

PBMCs from healthy donors were incubated in RPMI-1640 with 10% heat-inactivated FCS for collecting nonadherent cells. Afterward, CD56+ cells were isolated by means of positive selection with CD56 magnetic microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany). CD56+ purified cells were seeded under limiting dilution conditions in round-bottom microtiter plates in RPMI-1640 with 20% heat-inactivated FCS, irradiated (6000 rads) allogeneic PBMCs (3 × 104 cells/well), 20 U/mL recombinant human (rh) IL-2 (Peprotech, London, United Kingdom), and 1 μg/mL PHA (Murex, London, United Kingdom). The generated clones were analyzed by means of flow cytometry, and CD3+CD8+CD56+ clones were selected for subsequent cytotoxicity assays.

Cytotoxicity assays 

Cell-mediated cytotoxicity was measured in a standard 4-hour 51Cr release assay (PerkinElmer, Boston, Mass) against HaCaT cells and human primary keratinocytes pretreated or not with rhIFN-γ and 721.221 and 221.AEH cell lines. Assays were conducted in RPMI plus 10% heat-inactivated FCS in the presence or absence of saturating concentrations (10 μg/mL) of the specific mAbs 3D12, W6/32, and/or anti-NKG2C. CD56+CD8+ T-cell clones, PBMCs, and BFCs were used as effector cells at a 50:1 (PBMCs) or 20:1 (BFCs) effector/target ratio. Assays were conducted in triplicates by using U-bottom 96-well microtiter plates. Specific lysis was calculated as follows:

In every case spontaneous release was less than 20% of the maximum lysis.

CD107a mobilization assays 

BFCs or PBMCs from patients' acute samples were cultured or not at a 2:1 (efector:target) ratio for 4 hours with the 721.221 and 221.AEH cell lines or with autologous keratinocytes, pretreated or not with 10 ng/mL rhIFN-γ for 48 hours, and in the presence or absence of specific mAbs anti-NKG2C (10 μg/mL). Phycoerythrin (PE)–conjugated anti-human CD107a and Monensin A (2 μmol/L; Sigma-Aldrich, Lyon, France) were added during the coculture period. After this time, cells were collected and stained with the specific mAbs CD3–peridinin-chlorophyll-protein (PerCP), CD8–fluorescein isothiocyanate (FITC), or anti-NKG2C. A secondary FITC-conjugated rabbit anti-mouse F(ab′)2 was used for detection of NKG2C. Assays were performed in triplicates. CD107a+ cells were analyzed by means of flow cytometry on selected populations within the lymphocyte gate.

Statistical analysis 

Statistical analyses used in this study are described in the Methods section in this article's Online Repository.

Back to Article Outline

Results 

Expression of NKRs in lymphocytes from patients with SJS and TEN 

To confirm previously reported data, we performed a flow cytometric analysis of cell-surface expression of various NK receptors in BFCs of patients with SJS/TEN. Additionally, the phenotype of peripheral blood lymphocytes was studied in PBMCs drawn simultaneously to the blister cells (acute samples) and on complete remission of the clinical symptoms (resolution samples). Most blister lymphocytes were CD8+ and were enriched in cells expressing CD56, CD94, and killer immunoglobulin-like receptors (KIRs). Moreover, the proportion of cells expressing these receptors was increased in peripheral blood during the acute phase of SJS/TEN (see Fig E1 in this article's Online Repository at www.jacionline.org). Table I shows significantly increased CD8+ T cells in patients with SJS/TEN compared with numbers seen in healthy donors, which is in agreement with previous reports.15, 16 The expression of other NKRs, such as ILT2/CD85j and KIR2DL2/L3/S2, was also significantly increased in patients with drug-induced hypersensitivity reactions, whereas no significant differences were found in the percentages of cells expressing CD94, KIR2DL1/S1/S4, and KIR3D receptors.

Table I. Expression of cell-surface receptors on T lymphocytes from peripheral blood
Cell-surface receptorPatients with TEN/SJSPatients with MPEHealthy donors
CD8
Percent∗43.4 ± 2.539 ± 1.734.9 ± 1
N†163856
P value‡.003.04
ILT2/CD85j
Percent17.6 ± 5.117 ± 2.17.1 ± 0.7
N143449
P value.02.0001
CD94
Percent11.5 ± 2.611.2 ± 18.6 ± 0.6
N103656
P value.37.18
KIR2DL2/L3/S2
Percent10 ± 35.7 ± 0.83.3 ± 0.4
N143543
P value.007.002
KIR2DL1/S1/S4
Percent3.2 ± 0.83.5 ± 0.52.9 ± 0.3
N143253
P value.57.31
KIR3DL1/L2/2DS4
Percent2.8 ± 0.54.4 ± 0.92.3 ± 0.4
N153554
P value.22.05
KIR3DL1
Percent4.4 ± 2.42.4 ± 0.72.3 ± 0.3
N163250
P value.98.38

∗CD8+, ILT2/CD85j+, CD94+, and KIR+ cell subsets were defined by means of flow cytometry in gated CD3+ lymphocytes. Data are expressed as means ± SEMs.

†Number of samples.

‡Statistical analysis of PBMCs from patients with TEN/SJS and MPE in the acute phase versus healthy donors (Mann-Whitney U test).

High levels of soluble Fas ligand (FasL) were found in blister fluids from patients with SJS/TEN. IFN-γ levels were significantly higher in sera from patients, and slightly higher concentrations were detected in blisters (see Fig E2 in this article's Online Repository at www.jacionline.org).

HLA-E is expressed in keratinocytes from affected skin 

Analysis of cell suspensions from affected skin biopsy specimens revealed increased HLA-I and HLA-E cell-surface expression in CD45− cells (Fig 1, A and B). These data suggested an upregulation of HLA-E cell-surface expression in epidermal keratinocytes from patients with drug-induced delayed hypersensitivity reactions. This was confirmed by means of immunohistochemistry in skin biopsy specimens of patients with MPE and TEN (Fig 1, C). Moreover, although soluble HLA-E was not found in sera from patients and healthy donors or in blister fluids from mechanical trauma, immunoblot assays demonstrated that soluble HLA-E is released into the blister fluid in patients with SJS/TEN (Fig 1, D), which further supports the expression of HLA-E by epidermal cells in these patients.

  • View full-size image.
  • Fig 1. 

    HLA-E expression in affected skin. A, Flow cytometric analysis of HLA-I (solid histograms) and HLA-E (open histograms) expression in keratinocytes from a representative patient with MPE and a healthy donor (HD). B, Mean fluorescence intensity (MFI) of HLA-I and HLA-E expression in keratinocytes from 5 healthy donors (HD) and 8 patients (P). Horizontal bars correspond to the mean values. C, Immunoperoxidase staining of HLA-E in skin biopsy specimens from 3 patients. Original magnification 200× (panels a-f) and 400× (panels g-h). Arrowheads in panel h show specific staining in keratinocytes within necrotic skin. D, Immunoblot analysis of soluble HLA-E (sHLA-E) in fluids and sera from acute or resolution samples in 3 representative patients with SJS/TEN, healthy donors, and patients with mechanical trauma (MT). Culture supernatants from the HLA-I− cell line 721.221 and its HLA-E transfectant, 221.AEH, were used as negative and positive controls.

HLA-E cell-surface expression in human keratinocytes sensitizes them to killing by CD56+ CTLs 

Flow cytometric analysis revealed no detectable expression of HLA-E either in the cell membrane of in vitro cultured primary keratinocytes or in the human keratinocyte cell line HaCaT. However, IFN-γ stimulation induced cell-surface expression of HLA-E in primary cultures of human keratinocytes, as well as in HaCaT cells (Fig E3, A, in this article's Online Repository at www.jacionline.org). CD56+CD8+ T-cell clones generated from healthy donors and selected based on increased cytotoxicity against the HLA-E–transfected cell line (221.AEH)17 compared with the HLA-I–deficient parental cell line 721.221 were also able to kill HaCaT cells previously stimulated with IFN-γ (see Fig E3, B) in an HLA-E–dependent manner, as demonstrated when cytotoxicity assays were performed in the presence of the blocking antibodies anti-HLA class I or anti-HLA-E (Fig E3, C). Altogether, these data show that HLA-E cell-surface expression in keratinocytes sensitizes them to killing by CD56+CD8+ CTLs restricted or activated by HLA-E.

Lymphocytes expressing HLA-E–specific activating receptors are found in PBMCs from patients with SJS and TEN 

PBMCs from healthy donors and patients during the acute phase of the disease were used as effector cells in 51Cr release cytotoxicity assays against HLA-I–deficient and HLA-E+ cells. The expression of HLA-E in target cells significantly inhibited cell lysis by PBMCs drawn from patients with MPE or healthy donors, which is in agreement with the high frequency of lymphocytes expressing the HLA-E–specific inhibitory receptor CD94/NKG2A in healthy subjects.18 In contrast, PBMCs from patients with SJS/TEN showed similar cytolytic activity against HLA-I–deficient and HLA-E+ targets (Fig 2, A). Because no differences were found in CD94/NKG2A expression in patients and healthy donors (see Fig E4 in this article's Online Repository at www.jacionline.org), these results suggest an increased ratio of effector cells expressing activating versus inhibitory receptors specific for HLA-E in peripheral blood from patients with SJS/TEN. Similarly, when BFCs were used as effector cells, we did not find inhibition of the percentage of lysis by the expression of HLA-E in targets (Fig 2, B). Moreover, BFCs showed increased cytolytic activity against IFN-γ–treated primary keratinocytes (Fig 2, C).

  • View full-size image.
  • Fig 2. 

    Cytotoxic lymphocytes are activated by HLA-E in peripheral blood and blister fluid from patients with SJS/TEN. A, 51Cr release assays showing lysis of PBMCs from healthy donors and patients with MPE and SJS/TEN against the 721.221 and 221.AEH cell lines. Means and SEMs of independent experiments are shown. B and C, BFCs from patients with SJS/TEN were used as effector cells in 51Cr release assays. Fig 2, B, Mean values and SEMs of percentages of lysis against 721.221 and 221.AEH cells. Fig 2, C, Lysis of primary allogenic keratinocytes (KC) pretreated or not with IFN-γ by BFCs from 2 representative patients.

High frequencies of cytotoxic lymphocytes expressing the HLA-E–specific activating receptor CD94/NKG2C are found in patients with SJS/TEN 

Two activating receptors expressed in T cells, NK cells, or both might be involved in the recognition of HLA-E in keratinocytes. HLA-E–restricted TCRs have been described in NK-CTLs.7, 19 Additionally, it has recently been reported that CTLs might also use the activating NKR CD94/NKG2C to kill HLA-E+ target cells.12, 19 The availability of a specific mAb against CD94/NKG2C allowed us to study the expression of this receptor in PBMCs and compare it with its inhibitory counterpart, CD94/NKG2A, in the same donors. Although in healthy donors the proportion of CD94/NKG2A+ NK and T cells was found to be significantly higher than that of CD94/NKG2C+ NK and T cells, no differences were found in the frequencies of activating and inhibitory receptors in patients with SJS/TEN. CD94/NKG2C, which is expressed at low frequency in T cells from healthy subjects (mean ± SD, 2.6% ± 2.0%), was found on the cell surface of a substantial proportion of T cells from patients with SJS/TEN (mean ± SD, 4.7% ± 4.04%), although no statistical significance was achieved (see Fig E4, A). Additionally, significantly higher frequencies of CD94/NKG2C+CD3− NK cells were found in peripheral blood of patients with SJS/TEN during the acute phase (20.16% ± 17.6% in patients vs 2.15% ± 2.0% in healthy donors). As previously described,12 CD94/NKG2C was found mainly in the CD8+ subpopulation of CD3+ lymphocytes (Fig 3, A). Notably, after recovery, patients experienced a sharp decrease in the frequency of peripheral CD94/NKG2C+CD8+CD3+ or CD94/NKG2C+CD3− lymphocytes (Fig 3, A, and Fig 4, B). Higher frequencies of CD94/NKG2C+ cells were found also in CD3+CD8+ and CD3− cells from patients with SJS/TEN when compared with those seen in patients with drug-induced MPE (see Fig E4, B).

  • View full-size image.
  • Fig 3. 

    CD94/NKG2C expression in lymphocytes or blasts from patients with SJS/TEN. A, Frequencies of CD94/NKG2C+ cells in CD3+CD8+, CD3+CD8−, and CD3− peripheral blood lymphocytes from 3 representative acute samples (SJS8, SJS2, and TEN2) and 1 resolution sample (TEN2 resolution). B, Frequencies of CD94/NKG2C+ cells in blister lymphocytes or blasts in 3 representative patients. PerCP, Peridinin-chlorophyll-protein.

  • View full-size image.
  • Fig 4. 

    CD94/NKG2C expression on peripheral blood lymphocytes and blasts from patients with SJS/TEN. A, Frequencies of CD94/NKG2C+ cells in peripheral blood lymphocytes or blast subpopulations in acute samples from patients with SJS/TEN. B, Frequencies of CD94/NKG2C+ cells in peripheral blood lymphocyte (L) or blast (B) subpopulations in acute and resolution samples from 3 patients with SJS/TEN.

The expression of CD94/NKG2C was also analyzed in BFCs from patients with bullous reactions. According to flow cytometric morphologic criteria (forward light scatter), an important proportion of blasts was found within blisters. High frequencies of CD94/NKG2C+ cells were detected in CD8+CD3+ cells, particularly when blasts were analyzed. In addition to CD3+CD8+ cytotoxic T cells, CD94/NKG2C+CD3−CD56+ NK cells were found in blister fluids from patients with SJS/TEN (Table II and Fig 3, B).

Table II. Expression of cell-surface receptors on lymphocytes from blister fluids∗
Blister samplesCD3+CD8+†CD3+CD8+ CD94/NKG2C+‡CD3−CD56+CD3−CD94/NKG2C+
SJS158.9NT41NT
SJS246.211.434.87.2
SJS438.710.12221.6
SJS583.332.32423.1
SJS649.55.563.824.4
TEN180.42.36.55
TEN269.5NT81NT

NT, Not tested.

∗CD3+CD8+, CD3+CD8+CD94/NKG2C+, CD3−CD56+, and CD3−CD94/NKG2C+ blister lymphocytes were defined by means of flow cytometry. Data are expressed as percentages of cell subpopulations.

†Analysis according to total CD3+ lymphocytes.

‡Analysis according to total CD3+CD8+ lymphocytes.

Analysis according to total CD3− lymphocytes.

In light of the high frequencies of CD94/NKG2C+CD8+ T-cell blasts found in blisters, flow cytometric data were further analyzed to study the blast cell population in peripheral blood. Increased frequencies of CD94/NKG2C were detected within this fraction. Additionally, CD3−CD56+ NK cells were found as well within the blast region in some patients and were also enriched in CD94/NKG2C+ cells (see Fig E5 in this article's Online Repository at www.jacionline.org and Fig 4, A). The population of CD94/NKG2C+ blasts was dramatically decreased on complete remission of clinical symptoms (Fig 4, B). Altogether, these data show that CD94/NKG2C is overexpressed in cytotoxic cells and that CD94/NKG2C+ cytolytic cells are activated during SJS/TEN.

CD94/NKG2C–HLA-E interactions trigger cytotoxicity in peripheral blood and blister lymphocytes from patients with SJS/TEN 

Cytotoxicity assays against 721.221 and 221.AEH cell lines were conducted in which PBMCs from acute samples were used as effector cells in the presence of blocking anti-NKG2C–specific antibodies. Although the cytotoxic activity against 721.221 cells was not affected, lysis of 221.AEH cells was substantially reduced (Fig 5, A), indicating that CD94/NKG2C+ cells were responsible for the increased cytolytic activity against HLA-E+ cells (see also Fig 2, A). Degranulation assays were performed and surface expression of CD107a was analyzed in specific subpopulations to further analyze the contribution of CD94/NKG2C+ CTLs and CD94/NKG2C+ NK cells. CD94/NKG2C+CD3+ (T cells) and CD94/NKG2C+CD3− (NK cells) upregulated cell-surface CD107a on encounter with HLA-E+ cells (Fig 5, B), indicating a contribution of both cell lineages to target lysis.

  • View full-size image.
  • Fig 5. 

    Peripheral blood NK and T lymphocytes from patients with SJS/TEN recognize HLA-E through the activating receptor CD94/NKG2C. A, Cytotoxicity mediated by PBMCs from patients' acute samples and samples from healthy donors against 721.221 and 221.AEH cell lines in the presence or absence of blocking anti-NKG2C mAbs. Means and SEMs of independent experiments carried out with 3 patients with SJS are shown. B, CD107a mobilization assays. Graphs show the percentage of CD107a+ cells in CD3+ or CD3− lymphocyte subpopulations expressing or not CD94/NKG2C in a representative patient with SJS on coculture with 721.221 and 221.AEH cell lines.

BFCs were also used as effector cells in CD107a degranulation assays against HLA-I− and HLA-E+ cell lines. We found that blister fluid T and NK cells were able to degranulate on exposure to HLA-E+ cells in a CD94/NKG2C-dependent manner because it was hampered in the presence of blocking NKG2C-specific antibodies (Fig 6, A and B). In line with these findings, degranulation of BFCs was also increased in a CD94/NKG2C-dependent way in the presence of autologous keratinocytes previously stimulated with IFN-γ to induce HLA-E cell-surface expression (Fig 6, C, upper graph). Interestingly, when CD3− cells were analyzed separately, we found degranulation in the presence of unstimulated autologous keratinocytes, which was further increased in an NKG2C-dependent manner if IFN-γ–stimulated keratinocytes were used as targets (Fig 6, C, lower graph). Altogether, the data support an involvement of the activating receptor CD94/NKG2C, which is expressed in CTLs and NK cells, in the recognition and killing of HLA-E+ keratinocytes by blister cytotoxic lymphocytes in patients with SJS/TEN.

  • View full-size image.
  • Fig 6. 

    CD94/NKG2C-dependent degranulation in blister lymphocytes. A and B, BFCs from patients SJS6 and SJS4 were cultured alone or with 721.221 and 221.AEH cells in the absence or presence of anti-NKG2C mAbs. Fig 6, A, shows percentages of CD3+CD8+CD107a+ lymphocytes in a representative experiment. Fig 6, B, CD3−CD107a+ blister fluid lymphocytes in patient SJS6 after coculture in the absence or presence of 721.221 or 221.AEH cells and blocking anti-NKG2C mAbs. C, Induction of surface CD107a in total and CD3− BFCs from patient SJS4 after coculture in the absence or presence of autologous keratinocytes (KC) pretreated or not with IFN-γ and with or without specific anti-NKG2C mAbs. Mean percentages of CD107a+ lymphocytes in triplicate cultures are shown.

Back to Article Outline

Discussion 

SJS and TEN are severe hypersensitivity reactions to drugs characterized by the widespread destruction of the epithelium of the skin and mucous membranes, which occurs through massive apoptosis of keratinocytes.20 Formation of subepidermal bullae is characteristic, and necrosis of the full thickness of the epidermis is the pathognomonic feature of SJS/TEN.3 Several mechanisms have been postulated to trigger this process.21 It has been proposed that the death receptor Fas/CD95 plays a key role in the apoptosis of keratinocytes that leads to epidermal necrolysis.22 Furthermore, cell-free blister fluids are cytotoxic (unpublished data).23 Nonetheless, mononuclear cells are present in the blister fluid, principally CD8+ T lymphocytes, which are supposed to be the effector cells responsible for keratinocyte death either by means of direct killing or secretion of soluble cytotoxic proteins. Cytotoxic cells use at least 2 pathways to induce cell death. A nonsecretory pathway involves the interaction of cell-surface receptor-ligand pairs. FasL on effector cells interacts with Fas/CD95 on target cells, inducing target cell death by activating a caspase cascade. Additionally, a secretory pathway involves the directional release of cytolytic granules containing perforin, granzymes, and granulysin. Recently, it has been suggested that granulysin can play an important role in inducing keratinocyte death in patients with SJS and TEN,23 although other soluble mediators, such as TNF-α,24 and soluble FasL,25 which can be released by cytotoxic lymphocytes on cell activation, might contribute to keratinocyte apoptosis. Nonetheless, the release of cytolytic granules and soluble cytotoxic factors implies triggering of effector cells through ligand recognition on targets. Published data suggest that blister T lymphocytes might exhibit a drug-specific MHC class I–restricted cytotoxicity against autologous cells.26 Previously, it was published that blister T cells expressed NKRs and showed NK-like activity.6 The recognition of the nonclassical HLA-Ib molecule HLA-E by specific TCRs has recently been reported to be the molecular mechanism responsible for NK-like activity in T cells, which was previously described as HLA-I unrestricted, and the term NK-CTL was proposed.7 An alternative pathway to activate CTL lysis of HLA-E+ targets through the activating receptor CD94/NKG2C was also described.12 In the present work we found an upregulation of HLA-E expression in keratinocytes and of CD94/NKG2C in cytotoxic T and NK cells from patients with SJS/TEN. In addition, we have demonstrated that HLA-E expression sensitizes keratinocytes to killing by HLA-E–activated CTLs and BFCs. In T cells CD94 is acquired after antigen stimulation, and CD94/NKG2C expression is restricted to highly differentiated effector cells.27, 28 Functionally active CD94/NKG2C+ cells were detected not only in blisters but also in peripheral blood from patients during the acute phase. The simultaneous presence of HLA-E–activated cytotoxic cells and HLA-Ia–restricted specific T cells is plausible and could explain our data and previous reports showing drug-independent lysis of allogenic keratinocytes23 and drug-dependent lysis of autologous cells.26, 29

Shedding of HLA-E has been described in IFN-γ– and TNF-α–activated cells.30, 31 The presence of soluble HLA-E within blister fluids supports the expression of cell-surface HLA-E by activated keratinocytes in patients with SJS/TEN. High levels of IFN-γ and TNF-α, which has been reported in blister fluids from some patients,32 could stimulate surface expression of HLA-E in keratinocytes from affected skin and the release of soluble molecules into the blister fluid.31 It has been proposed that soluble HLA-E can protect target cells from lysis through its interaction with the inhibitory receptor CD94/NKG2A.31 Likewise, it could act as an agonist to activate cytotoxicity through its interaction with activating receptors, such as CD94/NKG2C, HLA-E–restricted TCRs, or both. This issue deserves further research.

High concentrations of granulysin, a molecule contained within the cytolytic granules of CTLs and NK cells, have been found in blister fluids.23 The same authors reported high percentages of granulysin-expressing CTLs and NK cells in blisters. Consistent with these observations, our study supports a role for both CTLs and NK cells in the induction of keratinocyte apoptosis in patients with SJS/TEN because we found that CD94/NKG2C+ blister NK cells, as well as CTLs, were able to degranulate in the presence of HLA-E+ targets. Moreover, high expression of granulysin has been reported in NKG2C+ CTLs.11

Similar to our results in patients with SJS/TEN, a predominance of cells expressing CD94/NKG2C over CD94/NKG2A has also been found in HIV-infected subjects,33 and strikingly, the incidence of SJS/TEN augments in HIV-infected patients.3 On the other hand, the involvement of CD94/NKG2C in tissue damage has been previously reported in patients with acute celiac disease, in whom CD94/NKG2C+ CTLs have been shown to kill HLA-E+ enterocytes,11 and expansion of CD94/NKG2C+ T cells and NK-CTLs have been detected in CMV-seropositive subjects,18, 34 suggesting that these subpopulations might be involved in the immunopathology of various chronic infectious and inflammatory diseases.11

In conclusion, our results suggest that the activating receptor CD94/NKG2C might be involved in triggering cytotoxic lymphocytes in patients with SJS/TEN and that CD94/NKG2C–HLA-E ligation could be a mechanism involved in the physiopathology of these diseases. Targeting this pathway could be of interest in designing future treatments for these patients.

Clinical implications

Novel therapies targeting CD94/NKG2C–HLA-E interactions might be applicable for the treatment, prevention, or both of SJS and TEN.

Back to Article Outline

Methods 

Cell lines 

The HaCaT human keratinocyte cell line was grown in Dulbecco's modified Eagle medium with 5% (vol/vol) heat-inactivated FCS and in the presence or absence of 2.5 ng/mL rhIFN-γ (Peprotech) for 48 hours at 37°C and 5% CO2. The HLA class I–defective B-cell line 721.221 was grown in RPMI-1640 medium supplemented with 5% heat-inactivated FCS. The 221.AEH cell line, a gift from Dr Geraghty (University of Washington, Seattle, Wash), was cultured in the presence of 300 μg/mL hygromycin B (Calbiochem, La Jolla, Calif).

mAbs 

The following mAbs were used: HP-1F7 (anti-HLA class I) and HP-3B1 (anti-CD94), kindly provided by Dr López-Botet (Universitat Pompeu Fabra, Barcelona, Spain), and 3D12 (anti-HLA-E), kindly provided by Dr Geragthy (University of Washington). CD3–peridinin-chlorophyll-protein, CD158b-FITC (anti-KIR2DS2, KIR2DL2, and KIR2DL3), HLA-DR–PE, CD8-FITC, CD56-FITC, CD25-FITC, HLA-DR–FITC, and CD107a-PE were purchased from BD Biosciences (San Jose, Calif). CD45-allophycocyanain and CD8-PE were from Miltenyi Biotec. MEM-E/02 (anti–HLA-E specific) was from Exbio (Praha, Czech Republic), anti-NKG2C was from R&D Systems (Minneapolis, Minn), and W6/32 (anti-HLA class I) was from Sigma-Aldrich.

Flow cytometry 

Flow cytofluorimetric analyses were performed on a FACSCalibur cytofluorimeter with CellQuest software (BD Biosciences). Surface immunofluorescence staining was conducted according to standard procedures by using the appropriate labeled or unlabeled mAbs. In the latter case, staining was followed by an FITC-conjugated polyclonal rabbit anti-mouse F(ab′)2 from Dako Cytomation (Glostrup, Denmark).

The expression of HLA molecules in patient and control keratinocytes was performed by excluding dead cells from the analysis by means of propidium iodide (0.5 μg/mL) staining in cell suspensions obtained from affected skin biopsy specimens and from surgical operations of control donors. HLA expression was determined in live CD45− cells with specific mAbs, and background mean fluorescence intensity derived from staining with control IgG1 mAb was subtracted from the mean fluorescence intensity determined for the experimental mAb (W6/32 or 3D12).

FasL and IFN-γ quantification 

The concentration of FasL and IFN-γ in sera and blister fluids from patients with SJS and TEN and healthy donors was measured by means of flow cytometry with human FasL and IFN-γ cytometric bead array Flex Set Kits (BD Biosciences), respectively, and according to the manufacturer's instructions.

Keratinocyte cell culture 

Cell suspensions were obtained by means of 2 steps of 20 minutes' digestion with 0.1% Trypsin and 5 mmol/L EDTA of punch biopsy specimens from acute skin eruptions or biopsy specimens from healthy donors. Cultures of primary human keratinocytes were generated by seeding these cell suspensions on irradiated feeder layers of 3T3 fibroblasts. The keratinocyte growth medium was Dulbecco modified Eagle medium supplemented with 10% FCS, glutamine, hydrocortisone (0.4 μg/mL), insulin (5 μg/mL), triiodothyronine (2 nmol/L), epidermal growth factor (10 ng/mL) and cholera toxin (0.1 nmol/L). Epidermal growth factor was not included in the medium at seeding. Cultures were fed with keratinocyte medium 3 times per week. Keratinocytes were routinely subcultured onto fresh feeder cells after reaching 80% confluence after the selective removal of the feeders with EDTA (0.02% in PBS) and trypsinization.

IFN-γ treatment in keratinocytes 

The HaCaT cell line and keratinocytes from skin biopsy specimens from healthy donors were treated with 2.5 or 10 ng/mL rhIFN-γ (Peprotech), respectively, for 48 hours.

Immunohistochemistry 

Paraffin sections (5 μm thick) were mounted on pretreated slides, deparaffinized with xylol, rehydrated through a graded series of ethanol, and rinsed in distilled water. The endogen peroxidase activity was blocked with 3% H2O2 for 20 minutes, and tissue sections were then boiled with citrate buffer (pH 6.0). Nonspecific binding was prevented by applying 1% BSA (Sigma-Aldrich), human IgG (50 μg/mL), and 0.3% Triton X-100 for 3 hours before staining with 10 μg/mL of the anti–HLA-E–specific mAb MEM-E/02 for 72 hours at 4°C. The sections were incubated with horseradish peroxidase–conjugated anti-mouse antibody (Chemicon International) for 1 hour at room temperature. Immunostaining was visualized with the DAB Substrate Kit (Invitrogen, Carlsbad, Calif), and tissues were counterstained with hematoxylin.

Statistical analysis 

Cell-surface receptor data from patients with MPE and SJS/TEN; CD94/NKG2C expression data in CD3+, CD3+CD8+, and/or CD3– peripheral blood lymphocytes from healthy donors' and patients' samples; HLA-E expression data from keratinocytes from healthy donors and patients; and FasL and IFN-γ concentrations in sera from patients and healthy donors were analyzed for statistical significance by using the Mann-Whitney U test for unpaired samples. The Wilcoxon matched paired test was used to compare NKG2A expression, NKG2C expression, or both in paired samples from patients and healthy donors; cytotoxic activity of PBMCs and BFCs against 721.221 and 221.AEH cells; and FasL and IFN-γ concentration in sera and blister fluids from patients. In all cases data are expressed as means and SEMs. Statistical analysis was performed by using the GraphPad Prism software package (GraphPad Software, Inc, San Diego, Calif).

Back to Article Outline

Fig E1. 

  • View full-size image.

  • Flow cytometric analysis of cell-surface receptor expression on T lymphocytes from peripheral blood and blister fluids from patients with SJS/TEN. Panels show frequencies of CD8+, CD94+, CD56+, KIR2DL2/L3/S2+, HLA-DR+, and CD25+ cells in CD3+ blister fluid and peripheral blood lymphocytes from acute and resolution samples in a representative patient with TEN.

Back to Article Outline

Fig E2. 

  • View full-size image.

  • FasL and IFN-γ levels in sera and blister fluids from patients with SJS/TEN. Graphs show picograms of FasL and IFN-γ per milliliter of serum or blister fluid from patients' acute samples and healthy donors (HD). Horizontal bars correspond to mean values. ∗P value, Wilcoxon matched pair test; #P value, Mann-Whitney U test.

Back to Article Outline

Fig E3. 

  • View full-size image.

  • HLA -E–specific CD56+CD8+ T-cell clones are cytotoxic against IFN-γ–stimulated keratinocytes expressing HLA-E. A, HLA-I and HLA-E expression in primary keratinocytes and HaCaT cells stimulated or not with IFN-γ. Staining with control IgG1 and specific mAbs is shown. B and C, Cytotoxicity mediated by CD56+CD8+ T-cell clones against HaCaT keratinocytes pretreated or not with IFN-γ. The 721.221 cell line and its HLA-E transfectant, 221.AEH, were used as negative and positive controls, respectively. Fig E3, B, Percentages of lysis from 3 representative clones of 5 tested. Fig E3, C, Lysis in the presence of anti HLA-E and anti-HLA-I mAbs.

Back to Article Outline

Fig E4. 

  • View full-size image.

  • Flow cytometric analysis of receptor expression in PBMCs. A, Percentages of CD94+, CD94/NKG2A+, and CD94/NKG2C+ cells in CD3+ and CD3− peripheral blood lymphocytes from patients with SJS/TEN (acute samples) and healthy donors (HD). B, Percentages of CD94/NKG2C+ cells (means ± SEMs) in CD3+CD8+ and CD3− peripheral blood lymphocytes from acute samples in patients with SJS/TEN and MPE . ∗P value, Wilcoxon matched pair test; #P value, Mann-Whitney U test.

Back to Article Outline

Fig E5. 

  • View full-size image.

  • CD94/NKG2C expression on peripheral blood lymphocytes and blasts in acute and resolution samples from patients with drug-induced cutaneous reactions. A, CD3 and CD8 expression and forward scatter (FSC) analysis in total PBMCs from a representative patient. Red- and blue-colored populations correspond to lymphocytes and blasts, respectively. B, Panels show percentages of CD3+CD8+, CD94/NKG2C+, and CD3−CD94/NKG2C+ cells in peripheral blood lymphocytes (red gates) or blasts (blue gates) in 3 patients.

Back to Article Outline

Table E1. 

Patients with bullous reactions
Patient IDSexAge (y)Drug or drugs involved
TEN 1M37Ibuprofen
TEN 2F28Erythromycin, acetaminophen
TEN 3M55allopurinol
SJS 1F27Spiramycin, metronidazole, ibuprofen
SJS 2F39Amoxicillin, clavulanic acid, ibuprofen
SJS 3M67Phenytoin
SJS 4F82Undetermined
SJS 5F45Undetermined
SJS 6F64Undetermined
SJS 7F74Amoxicillin, clavulanic acid
SJS 8F38Ibuprofen
SJS 9F17Anticonvulsant
SJS 10M8Anticonvulsant
SJS 11F73Anticonvulsant
SJS 12M62Isoniazid
SJS 13M2Cefotaxime, vancomycin, metronidazole
SJS 14M2Cefotaxime, clindamycin
SJS 15F54Doxorubicin

F, Female; M, male.

Back to Article Outline

Table E2. 

Patients with drug hypersensitivity reactions included in the study

Solid squares indicate figures in which samples from each patient were included.

Back to Article Outline

References 

  1. Roujeau JC. Clinical heterogeneity of drug hypersensitivity. Toxicology. 2005;209:123–129
  2. Allanore L, Roujeau JC. Clinic and pathogenesis of severe bullous skin reactions: Stevens-Johnson syndrome, toxic epidermal necrolysis. In:  Pichler WJ editors. Drug hypersensitivity. Basel: Karger; 2007;p. 267–277
  3. Pereira FA, Mudgil AV, Rosmarin DM. Toxic epidermal necrolysis. J Am Acad Dermatol. 2007;56:181–200
  4. Nassif A, Bensussan A, Dorothee G, Mami-Chouaib F, Bachot N, Bagot M, et al. Drug specific cytotoxic T-cells in the skin lesions of a patient with toxic epidermal necrolysis. J Invest Dermatol. 2002;118:728–733
  5. Pittet MJ, Speiser DE, Valmori D, Cerottini JC, Romero P. Cutting edge: cytolytic effector function in human circulating CD8+ T cells closely correlates with CD56 surface expression. J Immunol. 2000;164:1148–1152
  6. Le Cleach L, Delaire S, Boumsell L, Bagot M, Bourgault-Villada I, Bensussan A, et al. Blister fluid T lymphocytes during toxic epidermal necrolysis are functional cytotoxic cells which express human natural killer (NK) inhibitory receptors. Clin Exp Immunol. 2000;119:225–230
  7. Pietra G, Romagnani C, Falco M, Vitale M, Castriconi R, Pende D, et al. The analysis of the natural killer-like activity of human cytolytic T lymphocytes revealed HLA-E as a novel target for TCR alpha/beta-mediated recognition. Eur J Immunol. 2001;31:3687–3693
  8. Moretta L, Mingari MC, Bottino C, Pende D, Biassoni R, Moretta A. Cellular and molecular basis of natural killer and natural killer-like activity. Immunol Lett. 2003;88:89–93
  9. Braud VM, Allan DS, O'Callaghan CA, Soderstrom K, D'Andrea A, Ogg GS, et al. HLA-E binds to natural killer cell receptors CD94/NKG2A, B and C. Nature. 1998;391:795–799
  10. López-Botet M, Bellón T, Llano M, Navarro F, Garcia P, de Miguel M. Paired inhibitory and triggering NK cell receptors for HLA class I molecules. Hum Immunol. 2000;61:7–17
  11. Meresse B, Curran SA, Ciszewski C, Orbelyan G, Setty M, Bhagat G, et al. Reprogramming of CTLs into natural killer-like cells in celiac disease. J Exp Med. 2006;203:1343–1355
  12. Gumá M, Busch LK, Salazar-Fontana LI, Bellosillo B, Morte C, Garcia P, et al. The CD94/NKG2C killer lectin-like receptor constitutes an alternative activation pathway for a subset of CD8+ T cells. Eur J Immunol. 2005;35:2071–2080
  13. Bastuji-Garin S, Rzany B, Stern RS, Shear NH, Naldi L, Roujeau JC. Clinical classification of cases of toxic epidermal necrolysis, Stevens-Johnson syndrome, and erythema multiforme. Arch Dermatol. 1993;129:92–96
  14. Rheinwald JG. Methods for clonal growth and serial cultivation of normal human epidermal keratinocytes and mesothelial cells. In:  Baserga R editors. Cell growth and division. A practical approach. Oxford (UK): IRL Press; 1989;p. 81–94
  15. Correia O, Delgado L, Ramos JP, Resende C, Torrinha JA. Cutaneous T-cell recruitment in toxic epidermal necrolysis. Further evidence of CD8+ lymphocyte involvement. Arch Dermatol. 1993;129:466–468
  16. Correia O, Delgado L, Barbosa IL, Domingues JC, Azevedo R, Vaz CP, et al. CD8+ lymphocytes in the blister fluid of severe acute cutaneous graft-versus-host disease: further similarities with toxic epidermal necrolysis. Dermatology. 2001;203:212–216
  17. Lee N, Goodlett DR, Ishitani A, Marquardt H, Geraghty DE. HLA-E surface expression depends on binding of TAP-dependent peptides derived from certain HLA class I signal sequences. J Immunol. 1998;160:4951–4960
  18. Gumá M, Angulo A, Vilches C, Gómez-Lozano N, Malats N, López-Botet M. Imprint of human cytomegalovirus infection on the NK cell receptor repertoire. Blood. 2004;104:3664–3671
  19. García P, Llano M, De Heredia AB, Willberg CB, Caparros E, Aparicio P, et al. Human T cell receptor-mediated recognition of HLA-E. Eur J Immunol. 2002;32:936–944
  20. Paul C, Wolkenstein P, Adle H, Wechsler J, Garchon HJ, Revuz J, et al. Apoptosis as a mechanism of keratinocyte death in toxic epidermal necrolysis. Br J Dermatol. 1996;134:710–714
  21. Roujeau JC, Gelard K, Bensussan A. [Epidermal necrolysis: mechanisms of keratinocyte apoptosis]. Med Sci (Paris). 2006;22:188–191
  22. Viard I, Wehrli P, Bullani R, Schneider P, Holler N, Salomon D, et al. Inhibition of toxic epidermal necrolysis by blockade of CD95 with human intravenous immunoglobulin. Science. 1998;282:490–493
  23. Chung WH, Hung SI, Yang JY, Su SC, Huang SP, Wei CY, et al. Granulysin is a key mediator for disseminated keratinocyte death in Stevens-Johnson syndrome and toxic epidermal necrolysis. Nat Med. 2008;14:1343–1350
  24. Hunger RE, Hunziker T, Buettiker U, Braathen LR, Yawalkar N. Rapid resolution of toxic epidermal necrolysis with anti-TNF-alpha treatment. J Allergy Clin Immunol. 2005;116:923–924
  25. Abe R, Shimizu T, Shibaki A, Nakamura H, Watanabe H, Shimizu H. Toxic epidermal necrolysis and Stevens-Johnson syndrome are induced by soluble Fas ligand. Am J Pathol. 2003;162:1515–1520
  26. Nassif A, Bensussan A, Boumsell L, Deniaud A, Moslehi H, Wolkenstein P, et al. Toxic epidermal necrolysis: effector cells are drug-specific cytotoxic T cells. J Allergy Clin Immunol. 2004;114:1209–1215
  27. Mingari MC, Schiavetti F, Ponte M, Vitale C, Maggi E, Romagnani S, et al. Human CD8+ T lymphocyte subsets that express HLA class I-specific inhibitory receptors represent oligoclonally or monoclonally expanded cell populations. Proc Natl Acad Sci U S A. 1996;93:12433–12438
  28. Arlettaz L, Villard J, De Rham C, Degermann S, Chapuis B, Huard B, et al. Activating CD94:NKG2C and inhibitory CD94:NKG2A receptors are expressed by distinct subsets of committed CD8+ TCR alphabeta lymphocytes. Eur J Immunol. 2004;34:3456–3464
  29. Schnyder B, Frutig K, Mauri-Hellweg D, Limat A, Yawalkar N, Pichler WJ. T-cell-mediated cytotoxicity against keratinocytes in sulfamethoxazol-induced skin reaction. Clin Exp Allergy. 1998;28:1412–1417
  30. Derre L, Corvaisier M, Charreau B, Moreau A, Godefroy E, Moreau-Aubry A, et al. Expression and release of HLA-E by melanoma cells and melanocytes: potential impact on the response of cytotoxic effector cells. J Immunol. 2006;177:3100–3107
  31. Coupel S, Moreau A, Hamidou M, Horejsi V, Soulillou JP, Charreau B. Expression and release of soluble HLA-E is an immunoregulatory feature of endothelial cell activation. Blood. 2007;109:2806–2814
  32. Nassif A, Moslehi H, Le Gouvello S, Bagot M, Lyonnet L, Michel L, et al. Evaluation of the potential role of cytokines in toxic epidermal necrolysis. J Invest Dermatol. 2004;123:850–855
  33. Mela CM, Burton CT, Imami N, Nelson M, Steel A, Gazzard BG, et al. Switch from inhibitory to activating NKG2 receptor expression in HIV-1 infection: lack of reversion with highly active antiretroviral therapy. AIDS. 2005;19:1761–1769
  34. Mazzarino P, Pietra G, Vacca P, Falco M, Colau D, Coulie P, et al. Identification of effector-memory CMV-specific T lymphocytes that kill CMV-infected target cells in an HLA-E-restricted fashion. Eur J Immunol. 2005;35:3240–3247

 Supported by grant PI 06/0441 from the Ministerio de Sanidad/ISCIII, Spain, to T.B.. E.M. is the recipient of a fellowship from the FIBHULP.

 Disclosure of potential conflict of interest: The authors have declared that they have no conflict of interest.

PII: S0091-6749(09)01566-8

doi:10.1016/j.jaci.2009.10.030

The Journal of Allergy and Clinical Immunology
Volume 125, Issue 3 , Pages 703-710.e8, March 2010

Article Tools

* Image Usage & Resolution