Activation and skin-selective homing of T cells and their effector functions in the skin represent sequential immunological events in the pathogenesis of atopic dermatitis (AD). Apoptosis of keratinocytes, induced mainly by T cells and mediated by IFN-γ and Fas, is the essential pathogenetic event in eczema formation. Keratinocyte apoptosis appears as activation-induced cell death in AD. By IFN-γ stimulation, chemokines such as IFN-γ-inducible protein 10, monokine induced by IFN-γ, and IFN-γ-inducible α-chemoattractant are strongly up-regulated in keratinocytes. These chemokines attract T cells bearing the specific receptor CXCR3, which is highly expressed on T cells isolated from skin biopsies of AD patients. Accordingly, an increased T cell chemotaxis was observed toward IFN-γ-treated keratinocytes. Supporting these findings, enhanced IFN-γ-inducible protein 10, monokine induced by IFN-γ, and IFN-γ-inducible α-chemoattractant expression was observed in lesional AD skin by immunohistochemical staining. These results indicate a second step of chemotaxis inside the skin after transendothelial migration of the inflammatory cells. Keratinocytes undergoing apoptosis in acute eczematous lesions release chemokines that attract more T cells toward the epidermis, which may further augment the inflammation and keratinocyte apoptosis.

Chemokines are a superfamily of small structurally related proteins that regulate the traffic of lymphocytes, dendritic cells, monocytes, neutrophils, and eosinophils. They are classified into C, CC, CXC, and CX3C subfamilies, based on the position of characteristic structure determining cysteine residues within the N-terminal part of the proteins (1, 2, 3). Another classification distinguishes chemokines in inflammatory (inducible) and homeostatic (constitutive, housekeeping or lymphoid) chemokines (4, 5, 6, 7, 8, 9). Homeostatic chemokines are constitutively produced in discrete microenvironments within lymphoid or nonlymphoid tissues including skin and mucosa. They are responsible for maintaining physiological traffic and positioning of cells that mainly belong to the adaptive immune system. Inflammatory chemokines are expressed after stimulation by proinflammatory cytokines or after contact with pathogenic agents. This causes the recruitment of effector cells, such as monocytes, granulocytes, and effector T cells (4). IFN-γ-inducible protein 10 (IP-10,3 CXCL10), monokine induced by IFN-γ (Mig, CXCL9), and IFN-γ-inducible α-chemoattractant (iTac, CXCL11) belong to the group of inflammatory chemokines. They are all induced by IFN-γ, which inhibits the expression of most other chemokines (10, 11), and these three chemokines share a common receptor, CXCR3 (12).

Atopic dermatitis (AD) is a chronic inflammatory skin disease that frequently predates the development of allergic rhinitis or asthma (13, 14). Elevated IgE levels and eosinophilia in AD suggest increased expression of Th2-type cytokines. The majority of allergen-specific T cells derived from skin lesions that have been provoked in AD patients by epicutaneous allergen application or peripheral blood skin- homing T cells produce predominantly Th2 cytokines such as IL-4, IL-5, and IL-13 (15, 16, 17, 18). Previously, such a polarized Th2 cytokine pattern was regarded as a specific feature reflecting immune dysregulation in AD. However, recent studies have demonstrated that IFN-γ predominates over IL-4 in chronic skin lesions and older patch test reactions in AD, whereas IL-5 and IL-13 still remain at high levels (19, 20, 21, 22).

Lesional AD skin is characterized with dermal mononuclear cell infiltrate and spongiosis in the epidermis. Apoptosis of keratinocytes induced by T cells and mediated by IFN-γ and Fas (CD95) is a crucial event in the transition from activation of the immune system to the manifestation of eczematous dermatitis (23). Apparently, keratinocyte apoptosis is an activation-induced cell death, because IFN-γ up-regulates Fas, ICAM-1, and HLA-DR and renders keratinocytes susceptible to apoptosis (24). Induction of keratinocyte apoptosis by skin-infiltrating T cells, subsequent cleavage of E-cadherin, and resisting desmosomal cadherins represent molecular events in spongiosis formation (25).

We investigated whether T cell-keratinocyte interaction contributes to the intensity of skin inflammation and tissue injury mechanisms. Here, we demonstrate that keratinocytes undergoing T cell- and particularly IFN-γ-mediated activation-induced cell death release chemokines such as IP-10, Mig, and iTac. These chemokines play a role in a second step of chemotaxis from the dermis to the epidermis that increases the severity of tissue injury and keratinocyte apoptosis in the AD skin.

Nineteen patients with AD (19–68 years of age) were diagnosed according to standard criteria of Hanifin and Rajka (26). They were all allergic to at least three environmental allergens, and serum IgE was >500 U/ml. Ten healthy subjects (22–42 years of age) had no allergic disease and had normal levels of serum IgE. The study was approved by the Ethical Committee of Davos.

All fluorescent- or biotin-labeled mAbs for flow cytometry analysis were purchased from Immunotech (Marseilles, France) or BD PharMingen (San Diego, CA). Anti-CD45RO, anti-CD45RA magnetic microbeads for MACS were from Miltenyi Biotec (Bergisch Gladbach, Germany). Human IFN-γ, IL-2, and IL-4 were provided by Novartis (Basel, Switzerland). Ethidium bromide was from Sigma-Aldrich (St. Louis, MO). Soluble Fas ligand and TNF-α were from Alexis Biochemicals (San Diego, CA). Abs to human IP-10, Mig, and iTac for immunohistochemistry were purchased from PeproTech EC (London, U.K.).

Primary human keratinocytes (pooled normal human epidermal keratinocytes from neonatal skin) were purchased from BioWhittaker (Walkersville, MD) and grown in keratinocyte growth medium (KGM-2) from the same company supplemented with bovine insulin, hydrocortisone, human recombinant epidermal growth factor, 30 μg/ml bovine pituitary extract, 100 μg/ml gentamicin, 1000 ng/ml amphotericin B, epinephrine, transferrin, and calcium. Hydrocortisone was not added in experiments.

Human HaCaT keratinocytes, a gift from N. E. Fusenig (German Cancer Research Center, Heidelberg, Germany), (27), were grown in RPMI 1640 supplemented with 1 mM l-glutamine, sodium pyruvate, nonessential amino acids, and 10% FCS (all from Life Technologies, Basel, Switzerland).

Total RNA from keratinocytes was extracted using the RNeasy kit (Qiagen, Basel, Switzerland) according to the manufacturer’s instructions. During this procedure, a DNase digestion with the RNase-free DNase set for use with RNeasy/QIAamp columns (Qiagen) was used. RNA was reverse transcribed using Omniscript reverse transcriptase (Qiagen). The resulting cDNA was analyzed by real time PCR (Light Cycler; Roche Diagnostics, Mannheim, Germany). The probes for IP-10 were: Red640LC-labeled (LC), 5′-LC Red640-act gga ggt tcc tct gct gta ggc tc p; and fluorescein-labeled (FL), 5′-aat cgc agt ttg att cat ggt gct ga x. The primers for IP-10 were: forward, 5′-gac att cct caa ttg ctt aga cat a; and reverse, 5′-aat gat ctc aac acg tgg aca a. The probes for Mig were: LC, 5′-LC Red640-cct aca caa ttc act gaa cct ccc ctg p; and FL, 5′-ggt tta aat tct ggc cac aga caa cct c x. The primers for Mig were: reverse, 5′-tag ata aga cgt tcg ggt ggg at and forward, 5′-cgg tta gtg gaa gca tga ttg g. The probes for iTac were: LC, 5′-LC Red640-gac agc gtc ctc ttt tga aca tgg g p; and FL, 5′-ctg ctt tta ccc cag ggc cta tgc x. The primers for iTac were: forward, 5′-tgg ctg tga tat tgt gtg cta ca; and reverse, 5′-ctt cga ttt ggg att tag gca. All primers and probes were designed and obtained from TIB MOLBIOL (Berlin, Germany). For amplification of the PCR product, the Light Cycler-Fast Start DNA Master Hybridization Probes (Roche Diagnostics) were used according to the manufacturer’s instructions. The forward primer sequence for GAPDH was 5′-gct gat gat ctt gag gct gtt g-3′, and the reverse primer sequence was 5′-ctt cgc tct ctg ctc ctc ct-3′. The GAPDH was amplified in the real time PCR using the Light Cycler-Fast Start DNA Master SYBR Green I (Roche Diagnostics).

Quantification of the amplicons in each well was determined according to the comparative threshold cycle number (Ct) method as previously described (Applied Biosystems, Foster City, CA). Briefly, the formula for each sample is 2E − (CttargetCtstandard). This yields a quantification of the target PCR products relative to the PCR products for the internal calibration (GAPDH) and unstimulated control keratinocytes. The results were plotted on a log scale.

Total RNA was extracted using a guanidine isothiocyanate-acid phenol protocol. For simultaneous RNase protection assay of RANTES (CCL5), IP-10, monocyte chemoattractant protein (MCP-1, CCL2), and IL-8, the multiprobe template set hCK5 was used; hCR-6 was used for the detection of CXCR1, CXCR2, CXCR3, and CXCR4 (BD PharMingen) according to the manufacturer’s instructions. The unprotected RNA was digested using a RNase T1/A mix. The digested RNA was purified and loaded on a 6% urea gel. The dried radioactive gels were exposed to imager plates and visualized by an imager system (FLA-3000; Raytest, Urdorf, Switzerland). RNase protection assay bands were quantified by using the AIDA software (Raytest).

PBMC were isolated by Ficoll (Biochrom, Berlin, Germany) density gradient centrifugation of peripheral venous blood of normal donor and cord blood (28). Cells were washed three times and resuspended in DMEM (Life Technologies, Switzerland), supplemented as described (16). CD45RO+ and CD45RA+ cells were negatively selected with the MACS magnet activated cell separation system (Miltenyi Biotech) (16, 17). In brief, anti-CD14 and anti-CD19 depleted cells were incubated with MACS microbead-conjugated anti-CD45RO and anti-CD16 mAbs and anti-CD8. For the differentiation of Th1 and Th2 cells, freshly isolated cord blood CD4+CD45RA+ T cells were suspended in RPMI 1640 medium supplemented as described (16). The cells were cultured in 48-well plates at a cell density of 105/ml. They were stimulated with the combination of mAbs to T cell surface molecules anti-CD2 (two mAbs, 4B2 and 6G4 each 0.5 μg/ml), anti-CD3 (1 μg/ml), and anti-CD28 (1 μg/ml), and IL-2 (20 ng/ml). For Th1 differentiation, human IL-12 (10 ng/ml) and neutralizing anti-IL-4 mAb (10 μg/ml) were added to individual wells (29). For Th2 differentiation, human IL-4 (25 ng/ml) and neutralizing anti-IL-12 (10 μg/ml) were used (29). The growing cell cultures were expanded with fresh culture medium containing human IL-2. After 12 days, the cells were harvested, washed, and then restimulated with mAbs to CD2/CD3/CD28, and cytokine patterns of differentiated cells were determined by flow cytometry and ELISA.

To isolate T cells from epidermis of lesional biopsies of AD patients, epidermis pieces (0.5–1 mm) of the skin biopsy specimens from 3- to 4-day-old lesions of four chronic AD patients were minced with two scalpels (no proteolytic enzymes were used). Pieces of epidermis were stimulated in complete RPMI 1640 with 25 U/ml IL-2 and anti-CD2, anti-CD3, and anti-CD28 mAb. Growing T cells were expanded in medium containing IL-2 (purity was 100% as assessed by flow cytometry). After 7–10 days, the cytokine profile and chemokine receptors were characterized (19). The supernatants were collected 48 h after anti-CD2, anti-CD3, and anti-CD28 mAb stimulation; frozen; and used for the chemokine expression experiments or ELISA measurements of cytokines.

The cytokine profiles of the T cells that were isolated from skin biopsies were evaluated in a sandwich ELISA (16, 28) by measuring IL-4, IL-5, IL-13, and IFN-γ. The sensitivity of IFN-γ ELISA was ≤10 pg/ml (mAb and IFN-γ standard were gifts from Dr. S. S. Alkan, Novartis). The sensitivity of IL-4 ELISA was ≤20 pg/ml (mAb and IL-4 standard were provided by Dr. C. H. Heusser, Novartis). The detection limit of the IL-5 ELISA was 50 pg/ml (mAb and IL-5 standard were from BD PharMingen). The sensitivity of the IL-13 ELISA was 100 pg/ml (mAb and IL-13 standard were from BD PharMingen).

CD45RO+ T cells (1 × 106)/ml were stimulated with a combination of soluble anti-CD2, anti-CD3, and anti-CD28 mAb. The cells were stained with Cell Tracker Green CMFDA (Molecular Probes, Eugene, OR; 5 μM in medium) for 45 min under growth conditions.

Keratinocytes were grown in a 24-well plate until a confluence of ∼80% was reached. A fluoro-block insert (BD Biosciences, Basel, Switzerland) with a pore size of 3 μm was used to separate the fluorescence-labeled T cells from the keratinocytes at the bottom of the well. Keratinocytes were pretreated for 12 h with 10 ng/ml IFN-γ or 10 ng/ml TNF-α. Migration occurred through the pores of the membrane and was measured by the amount of fluorescence in the lower part of the well with an imager system, FLA-3000 (Raytest).

Biopsy specimens were taken from 3- to 4-day-old lesions of chronic AD patients or from normal skin of healthy individuals. Frozen 4-μm skin sections were fixed with 4% paraformaldehyde for 20 min. Endogenous peroxidases were quenched with a 0.3% H2O2 solution in methanol. Nonspecific binding was blocked with 10% normal goat serum (Chemicon, Temecula, CA) and then incubated for 2 h with the appropriate dilution of Abs, anti-human CD4 (BD PharMingen), anti-human IP-10, anti-human iTac, or anti-human Mig (all from PeproTech EC) or rabbit IgG as isotype control. Sections were then treated with biotin-labeled goat anti-rabbit IgG and stained with avidin-biotin-peroxidase (Vector, Burlingame, CA), followed by 3,3′-diaminobenzidine (Sigma-Aldrich) development.

Keratinocytes or T cells were incubated with appropriate FITC- or PE-labeled mAb for 30 min at 4°C and then fixed in 2% paraformaldehyde (pH 7.4). Negative control cells were prepared in a similar fashion with isotype control IgG1-FITC/PE mAb. Fluorescence intensity was measured by flow cytometry (EPICSTMXL-MCL; Beckmann Coulter, Nyon, Switzerland). KC viability was monitored by ethidium bromide (1 μM; Sigma-Aldrich) uptake and flow cytometry as previously described (30).

Student’s t test was used for statistical analysis. Data are presented as the mean ± SD of triplicate cultures.

Primary human keratinocytes or HaCaT keratinocytes were cultured with IFN-γ, TNF-α, soluble Fas (sFas) ligand, and anti-CD2/CD3/CD28 mAb stimulated supernatant of T cells isolated from AD skin, and cell viability was measured (Fig. 1). A significantly increased keratinocyte death was induced after 5 days by supernatants of stimulated T cells isolated from AD skin and conditions, which contain IFN-γ (p < 0.001). TNF-α induced a relatively low level and late keratinocyte death. Triggering of Fas alone on keratinocytes did not induce keratinocyte death. In our previous studies, IFN-γ and sFas-ligand were identified as keratinocyte apoptosis-inducing factors within the activated T cell supernatants. The type of T cell-induced keratinocyte death was also shown to be apoptosis by using several different apoptosis-specific methods (23, 25).

FIGURE 1.

Role of T cells and cytokines on keratinocyte death. Primary human keratinocytes were incubated with 10 ng/ml IFN-γ, 10 ng/ml TNF-α, 10 ng/ml sFas ligand (sFasL), and 50% (v/v) T cell supernatant (sup., 48-h anti-CD2/CD3/CD28 mAb-activated AD skin T cell supernatant) and left unstimulated (u.s.). KC viability was monitored by ethidium bromide exclusion and flow cytometry. One representative of five experiments with primary human keratinocytes is shown. Same results were obtained with HaCaT keratinocytes. Standard deviation of triplicate cultures is shown. ∗∗, p < 0.001; ∗, p < 0.05.

FIGURE 1.

Role of T cells and cytokines on keratinocyte death. Primary human keratinocytes were incubated with 10 ng/ml IFN-γ, 10 ng/ml TNF-α, 10 ng/ml sFas ligand (sFasL), and 50% (v/v) T cell supernatant (sup., 48-h anti-CD2/CD3/CD28 mAb-activated AD skin T cell supernatant) and left unstimulated (u.s.). KC viability was monitored by ethidium bromide exclusion and flow cytometry. One representative of five experiments with primary human keratinocytes is shown. Same results were obtained with HaCaT keratinocytes. Standard deviation of triplicate cultures is shown. ∗∗, p < 0.001; ∗, p < 0.05.

Close modal

Because T cells are essential players in keratinocyte apoptosis, factors that attract T cells toward and into the epidermis of lesional AD skin are further investigated. Because IFN-γ and TNF-α are two major proinflammatory cytokines in eczematous skin, we investigated whether they affect release of chemokines by keratinocytes. In RNase protection assays analyzing several chemokines, IL-8 mRNA was constitutively expressed, whereas IP-10 mRNA was up-regulated by IFN-γ in human primary keratinocytes (Fig. 2 A).

FIGURE 2.

IFN-γ induces IP-10, Mig, and iTac in human keratinocytes. A, RNase protection assay of cultured primary keratinocytes shows constitutive expression of IL-8 mRNA and induction of IP-10 mRNA after treatment with 10 ng/ml IFN-γ or IFN-γ and an additional 10 ng/ml TNF-α. B, Real time RT-PCR shows up-regulation of IP-10, iTac, and Mig mRNA after treatment of primary keratinocytes with 50% (v/v) T cell supernatant (sup.; 48 h anti-CD2/CD3/CD28 mAb stimulated AD skin T cell supernatant). IFN-γ, TNF-α, and sFas ligand (sFasL) were used at 10 ng/ml. The target PCR products are quantified relative to the PCR products for the internal calibration (GAPDH) and unstimulated control keratinocytes. One representative is shown. Same results were obtained in two experiments from two different donors. MIP, macrophage-inhibitory protein; Ltn., lymphotactin; L32, housekeeping gene that encodes L32 ribosomal protein.

FIGURE 2.

IFN-γ induces IP-10, Mig, and iTac in human keratinocytes. A, RNase protection assay of cultured primary keratinocytes shows constitutive expression of IL-8 mRNA and induction of IP-10 mRNA after treatment with 10 ng/ml IFN-γ or IFN-γ and an additional 10 ng/ml TNF-α. B, Real time RT-PCR shows up-regulation of IP-10, iTac, and Mig mRNA after treatment of primary keratinocytes with 50% (v/v) T cell supernatant (sup.; 48 h anti-CD2/CD3/CD28 mAb stimulated AD skin T cell supernatant). IFN-γ, TNF-α, and sFas ligand (sFasL) were used at 10 ng/ml. The target PCR products are quantified relative to the PCR products for the internal calibration (GAPDH) and unstimulated control keratinocytes. One representative is shown. Same results were obtained in two experiments from two different donors. MIP, macrophage-inhibitory protein; Ltn., lymphotactin; L32, housekeeping gene that encodes L32 ribosomal protein.

Close modal

Increased mRNA expression of IP-10, Mig, and iTac was observed by IFN-γ stimulation of primary human keratinocytes. TNF-α had a synergistic effect on the up-regulation of all three chemokines by IFN-γ, without showing any effect alone. Anti-CD2/CD3/CD28 mAb-stimulated AD skin T cell supernatants also up-regulated these three chemokines. Triggering of the Fas by sFasL did not stimulate chemokine expression alone or in addition to IFN-γ and TNF-α (Fig. 2 B).

To confirm the mRNA expression profiles, we investigated the IP-10, Mig, and iTac protein expression on primary human keratinocytes by immunohistology. An up-regulation of IP-10, Mig, and iTac was observed after stimulation of keratinocytes with supernatant of T cells isolated from AD skin and IFN-γ. TNF-α alone or together with Fas-ligand did not show any effect (Fig. 3).

FIGURE 3.

IP-10, Mig, and iTac are expressed by IFN-γ stimulation in keratinocytes. Immunohistochemical staining of cultured keratinocytes show up-regulation of IP-10, Mig, and iTac after incubation with T cell supernatant and IFN-γ. Primary human keratinocytes were treated with 50% (v/v) T cell supernatant (sup.; 48 h anti-CD2/CD3/CD28 mAb-stimulated AD skin T cell supernatant), IFN-γ, TNF-α, and sFas ligand (sFasL); u.s., unstimulated keratinocytes. IFN-γ, TNF-α, and sFas ligand were all used at 10 ng/ml. All original magnifications are ×200. Same results were obtained in four independent experiments.

FIGURE 3.

IP-10, Mig, and iTac are expressed by IFN-γ stimulation in keratinocytes. Immunohistochemical staining of cultured keratinocytes show up-regulation of IP-10, Mig, and iTac after incubation with T cell supernatant and IFN-γ. Primary human keratinocytes were treated with 50% (v/v) T cell supernatant (sup.; 48 h anti-CD2/CD3/CD28 mAb-stimulated AD skin T cell supernatant), IFN-γ, TNF-α, and sFas ligand (sFasL); u.s., unstimulated keratinocytes. IFN-γ, TNF-α, and sFas ligand were all used at 10 ng/ml. All original magnifications are ×200. Same results were obtained in four independent experiments.

Close modal

IP-10, Mig, and iTac share the same receptor, CXCR3. We investigated the expression of CXCR3 on T cells isolated from lesional skin of AD patients, peripheral blood T cells of healthy individuals, and AD patients as well as in vitro differentiated Th1 and Th2 cells. Freshly purified CD45RO+ cells and in vitro differentiated cord blood CD4+ Th1 and Th2 cells expressed CXCR3. Higher expression of CXCR3 was observed on Th1 than on Th2 cells (Fig. 4,A). We determined CXCR3 expression on peripheral blood T cells of healthy individuals and atopic dermatitis patients by double staining together with an anti-CD3 mAb. CXCR3 was expressed on 63.0 ± 9.3% of T cells from healthy individuals and 70.2 ± 7.2% on T cells of AD patients. There was no significant difference (Fig. 4,B). Within several chemokine receptors, CXCR3 mRNA was expressed in addition to CXCR4, CXCR5, and CCR7 mRNA on T cells, which were isolated and cultured from skin biopsy samples of AD patients (Fig. 4 C).

FIGURE 4.

CXCR3 is expressed on T cells. A, FACS analysis showing that CXCR3 was expressed on CD4 Th1 cells and CD45RO+ cells. CD4 Th2 cells showed a lower expression of CXCR3. One of two experiments with similar results is shown. The percent of CXCR3+ T cells is shown on the right side of each histogram. B, FACS analysis of CXCR3 and CD3 on PBMC of seven AD patients and six healthy individuals. One demonstrative FACS datum shown with mean ± SD of all individuals on the upper right quadrant. IC, isotype control Ab. C, RNase protection assay showed expression of CXCR3, CXCR4, CXCR5, and CCR7 mRNA in anti-CD2/CD3/CD28 mAb-stimulated skin T cells isolated from lesional epidermis of two different AD patients. L32, housekeeping gene that encodes L32 ribosomal protein.

FIGURE 4.

CXCR3 is expressed on T cells. A, FACS analysis showing that CXCR3 was expressed on CD4 Th1 cells and CD45RO+ cells. CD4 Th2 cells showed a lower expression of CXCR3. One of two experiments with similar results is shown. The percent of CXCR3+ T cells is shown on the right side of each histogram. B, FACS analysis of CXCR3 and CD3 on PBMC of seven AD patients and six healthy individuals. One demonstrative FACS datum shown with mean ± SD of all individuals on the upper right quadrant. IC, isotype control Ab. C, RNase protection assay showed expression of CXCR3, CXCR4, CXCR5, and CCR7 mRNA in anti-CD2/CD3/CD28 mAb-stimulated skin T cells isolated from lesional epidermis of two different AD patients. L32, housekeeping gene that encodes L32 ribosomal protein.

Close modal

The histological hallmark of eczematous lesions in the AD skin is dermal mononuclear cell infiltration and spongioform morphology in the epidermis. As shown in CD4+ T cell staining in Fig. 5,A, some of the CD4+ T cells have migrated inside the epidermis. Several CD4+ T cells have mainly concentrated in the epidermal area, which shows substantial spongiosis. Because differential chemotactic features of Th1 and Th2 cells have been demonstrated, we analyzed the cytokine profile of T cells, which infiltrate the epidermis of AD lesions. Epidermal T cells were cultured from different AD patients, and their cytokine profile was characterized. Epidermal T cells showed a Th0 cytokine profile with high quantities of IFN-γ, IL-5, and IL-13 and relatively lower amounts of IL-4 (Fig. 5 B).

FIGURE 5.

Enhanced chemotaxis of CD4+ T cells toward IFN-γ-stimulated keratinocytes. A, CD4+ T cells infiltrating the spongiotic epidermis. ×200. Magnification of the boxed area is ×400. B, Epidermal T cells showed a Th0 cytokine profile with high quantities of IFN-γ, IL-5, and IL-13 and relatively lower amounts of IL-4. ○, ⋄, and □, unstimulated (u.s.); •, ♦, and ▪, anti-CD2/CD3/CD28 mAb stimulated. C, In vitro differentiated CD4+ Th1 and Th2 cells both showed high migration toward IFN-γ- but not TNF-α-stimulated keratinocytes. IFN-γ-stimulated human keratinocytes induced T cell chemotaxis in three experiments with three different T cell sources. SD of duplicate cultures is shown.

FIGURE 5.

Enhanced chemotaxis of CD4+ T cells toward IFN-γ-stimulated keratinocytes. A, CD4+ T cells infiltrating the spongiotic epidermis. ×200. Magnification of the boxed area is ×400. B, Epidermal T cells showed a Th0 cytokine profile with high quantities of IFN-γ, IL-5, and IL-13 and relatively lower amounts of IL-4. ○, ⋄, and □, unstimulated (u.s.); •, ♦, and ▪, anti-CD2/CD3/CD28 mAb stimulated. C, In vitro differentiated CD4+ Th1 and Th2 cells both showed high migration toward IFN-γ- but not TNF-α-stimulated keratinocytes. IFN-γ-stimulated human keratinocytes induced T cell chemotaxis in three experiments with three different T cell sources. SD of duplicate cultures is shown.

Close modal

To analyze the role of IFN-γ on T cell chemotaxis, we used a Transwell system, in which fluorescence-labeled T cells migrated toward primary human keratinocytes separated with a 3-μm pore size filter. Enhanced migration of T cells toward IFN-γ-treated keratinocytes in comparison with unstimulated or TNF-α treated keratinocytes was observed. There was no difference between CD4+ Th1 and CD4+ Th2 cells. TNF-α-pretreated keratinocytes did not induce either CD4+ Th1 or CD4+ Th2 cell chemotaxis (Fig. 5 C).

The in vivo relevance of the studied hypothesis was investigated by demonstration of the IFN-γ-induced chemokines in AD skin (Fig. 6). IP-10, Mig, and iTac were stained in lesional AD and normal skin of healthy individuals by immunohistochemistry. Enhanced IP-10 distribution was detected in the suprabasal layer of the epidermis in lesional acute AD skin biopsies. In healthy skin, IP-10 expression was restricted to basal keratinocytes, whereas suprabasal epidermis did not show any expression. In AD skin, Mig and iTac expression was more distributed within the whole epidermis. Similar to IP-10, Mig expression was restricted to basal keratinocytes in healthy skin, whereas stronger expression was observed in suprabasal layers of lesional AD skin. In contrast, iTac was expressed both in basal keratinocytes and in suprabasal keratinocytes in normal and AD skin. The expression of IP-10, Mig, and iTac was mainly focused to epidermis. All three chemokines were not expressed in dermis, demonstrating a chemotaxis gradient from dermis to epidermis.

FIGURE 6.

Increased IP-10, Mig, and iTac expression in AD skin. Representative images from two healthy donors and eight different AD patients are shown. ×200. Arrows indicate IP-10 and Mig expression on basal layer keratinocytes in healthy epidermis and enhanced Mig, IP-10, and iTac expression on suprabasal keratinocytes in AD.

FIGURE 6.

Increased IP-10, Mig, and iTac expression in AD skin. Representative images from two healthy donors and eight different AD patients are shown. ×200. Arrows indicate IP-10 and Mig expression on basal layer keratinocytes in healthy epidermis and enhanced Mig, IP-10, and iTac expression on suprabasal keratinocytes in AD.

Close modal

Multiple molecules, including families of adhesion molecules and chemokines, provide signals for the dynamic trafficking of T cells into inflammatory tissues. Transendothelial migration and influx into skin represent the first phase leading to dermal perivascular infiltration by T cells in AD. Interestingly, a second step of chemotaxis, as demonstrated in this study, takes place in the migration of T cells closer to and into the epidermis where they augment T cell-mediated effector functions.

Eczematous skin lesions with distinct etiology are associated with T cell infiltration in the dermis leading to an interaction between T cells and keratinocytes and a marked keratinocyte pathology (31). Several studies have demonstrated that IFN-γ is one of the most active cytokines during T cell-keratinocyte interaction (32, 33). IFN-γ-producing T-cells are present in chronic AD lesions (21). Keratinocyte activation-induced cell death by IFN-γ in the activation phase of keratinocytes might strongly contribute to inflammation in AD. Relatively low concentrations of IFN-γ can induce keratinocyte apoptosis and spongioform morphology as an essential mechanism in the pathogenesis of eczematous dermatitis (25, 30). ICAM-1, which is crucial for T cell retention in the epidermis, is expressed on keratinocytes after exposure to IFN-γ (34, 35). IFN-γ up-regulates MHC class I molecules and Fas and induces de novo synthesis of MHC class II molecules on keratinocytes (36, 37). IFN-γ also induces the expression of several cytokines such as IL-1α, IL-1 receptor agonist, TNF-α and GM-CSF (37, 38). TNF-α and IL-17 are able to enhance the efficiency of IFN-γ in activating keratinocytes (35, 39). In addition, IFN-γ was demonstrated to induce squamous differentiation in epidermal keratinocytes. Apoptosis induced by IFN-γ in the present study might be the consequence of terminal differentiation and growth arrest (40). Moreover, whether suppressive cytokines such as IL-10 may inhibit IFN-γ secretion inside the skin during the development of eczema lesions and during the healing process remains to be elucidated (28).

The present study demonstrates that Mig, iTac, and IP-10 mRNAs are expressed in higher amounts after exposure of keratinocytes to IFN-γ. This leads to the migration of T cells from the dermis to the epidermis and represents a second step of chemotaxis after the recruitment of T cells from blood. Arrest and activation of leukocyte integrins via locally expressed chemokines are important steps for the transmigration of T cells through the vascular barrier. Cutaneous lymphocyte-associated Ag, CD45RO, LFA-1 and different chemokine receptors are expressed on the surface of T cells. Beside ICAM-1, surface molecules like CD34, VCAM-1, and E-selectin are expressed on endothelial cells and play an important role in diapedesis together with the receptors on T cells (10). In allergic contact dermatitis, T lymphocytes transmigrate into the dermis driven by an early chemokine gradient. Chemokines, such as RANTES, MCP-1, thymus, and activation-regulated chemokine (CCL17), pulmonary and activation-regulated chemokine (CCL18), and macrophage-derived chemokine (CCL22) are reported to be involved in this process (41). In contrast to chemokines and adhesion molecules in T cell transendothelial migration, the chemokines IP-10, Mig, and iTac are expressed on late stages of contact dermatitis (41). In addition, IP-10 is up-regulated in the epidermis in inflammatory skin diseases like psoriasis or allergic contact dermatitis (42, 43).

CXCR3, the specific receptor for these chemokines, is expressed on T cells, preferentially on Th1 cells (44, 45, 46). We detected the expression of CXCR3 on in vitro differentiated CD4+ Th1 and Th2 cells with Th1 predominance as well as on T cells isolated from epidermis. In addition, epidermal T cells rather showed a Th0 cytokine profile with increased IFN-γ, IL-5, and IL-13 production. After migration from dermis to epidermis, IFN-γ released from epidermal T cells may further induce the expression of more IP-10, Mig, and iTac on keratinocytes and further induce keratinocyte apoptosis. Epidermal T cells also expressed CXCR4, CXCR5, and CCR7, which bind to homeostatic chemokines (4). Interestingly, CCR7 as a central memory T cell chemokine receptor was expressed on epidermal effector T cells (47). The reasons for the expression of these chemokine receptors in epidermal T cells remain to be elucidated. Beside the function of lymphocytes as immunological effector cells, they also produce growth factors such as basic fibroblast growth factor and leukocyte-derived growth factor (48, 49, 50). Their later appearance in the eczematous lesions may additionally favor tissue formation and remodeling of the tissue components. Accordingly, Mig and IP-10 are expressed in a later stage of wound healing and are essential for the persistence of lymphocyte recruitment (50). These studies on chemokines and their receptors demonstrate a complex system of different chemokines and chemokine receptors inside the skin during different migration stages of T cells.

The present study demonstrates an active interaction between effector T cells and target keratinocytes in the AD skin. Apparently, during the activation phase of T cell- and IFN-γ-mediated activation-induced keratinocyte death, epidermal keratinocytes release chemokines that attract T cells from dermis to closer to and into the epidermis. The second step of chemotaxis of T cells from the dermis to the epidermis may play an essential role in the augmentation of keratinocyte apoptosis, enhanced infiltration of T cells into the dermis, remodeling of the skin, and chronicity of AD lesions and disease.

We thank Andreas Karatsaidis for technical help in immunohistochemistry and Prof. Dr. N. E. Fusenig (Heidelberg, Germany) for providing HaCaT cells.

1

This study was supported by Swiss National Foundation Grant 32.65661.01.

3

Abbreviations used in this paper: IP-10, IFN-γ-inducible protein 10 (CXCL10); Mig, monokine induced by IFN-γ (CXCL9); iTac, IFN-γ-inducible α-chemoattractant (CXCL11); AD, atopic dermatitis; FL, fluorescein labeled; LC, Red640LC labeled; MCP, monocyte chemoattractant protein; sFas, soluble Fas.

1
Baggiolini, M., B. Dewald, B. Moser.
1997
. Human chemokines: an update.
Annu. Rev. Immunol.
15
:
675
.
2
Luster, A. D..
1998
. Chemokines: chemotactic cytokines that mediate inflammation.
N. Engl. J. Med.
338
:
436
.
3
Tensen, C. P., J. Flier, E. M. Van Der Raaij-Helmer, S. Sampat-Sardjoepersad, R. C. Van Der Schors, R. Leurs, R. J. Scheper, D. M. Boorsma, R. Willemze.
1999
. Human IP-9: a keratinocyte-derived high affinity CXC-chemokine ligand for the IP-10/Mig receptor (CXCR3).
J. Invest. Dermatol.
112
:
716
.
4
Moser, B., P. Loetscher.
2001
. Lymphocyte traffic control by chemokines.
Nat. Immunol.
2
:
123
.
5
Loetscher, P., B. Moser, M. Baggiolini.
2000
. Chemokines and their receptors in lymphocyte traffic and HIV infection.
Adv. Immunol.
74
:
127
.
6
Sallusto, F., C. R. Mackay, A. Lanzavecchia.
2000
. The role of chemokine receptors in primary, effector, and memory immune responses.
Annu. Rev. Immunol.
18
:
593
.
7
von Andrian, U. H., C. R. Mackay.
2000
. T-cell function and migration. Two sides of the same coin.
N. Engl. J. Med.
343
:
1020
.
8
Cyster, J. G..
1999
. Chemokines and cell migration in secondary lymphoid organs.
Science
286
:
2098
.
9
Campbell, J. J., E. C. Butcher.
2000
. Chemokines in tissue-specific and microenvironment-specific lymphocyte homing.
Curr. Opin. Immunol.
12
:
336
.
10
Baggiolini, M..
1998
. Chemokines and leukocyte traffic.
Nature
392
:
565
.
11
Baggiolini, M., B. Dewald, B. Moser.
1994
. Interleukin-8 and related chemotactic cytokines: CXC and CC chemokines.
Adv. Immunol.
55
:
97
.
12
Loetscher, M., B. Gerber, P. Loetscher, S. A. Jones, L. Piali, I. Clark-Lewis, M. Baggiolini, B. Moser.
1996
. Chemokine receptor specific for IP10 and Mig: structure, function, and expression in activated T-lymphocytes.
J. Exp. Med.
184
:
963
.
13
Leung, D. Y. M., M. Tharp, M. Boguniewicz.
1998
. Atopic dermatitis. M. E. A. Z., and I. M. Freeberg, and K. Wolff, and F. K. Austen, and L. A. Goldsmith, and S. I. Katz, and T. Fritzpatrick, eds.
Dermatology in General Medicine
5th Ed.
1464
. McGraw-Hill, New York.
14
Leung, D. Y..
2000
. Atopic dermatitis: new insights and opportunities for therapeutic intervention.
J. Allergy Clin. Immunol.
105
:
860
.
15
Santamaria Babi, L. F., L. J. Picker, M. T. Perez Soler, K. Drzimalla, P. Flohr, K. Blaser, C. Hauser.
1995
. Circulating allergen-reactive T cells from patients with atopic dermatitis and allergic contact dermatitis express the skin-selective homing receptor, the cutaneous lymphocyte-associated antigen.
J. Exp. Med.
181
:
1935
.
16
Akdis, M., C. A. Akdis, L. Weigl, R. Disch, K. Blaser.
1997
. Skin-homing, CLA+ memory T cells are activated in atopic dermatitis and regulate IgE by an IL-13-dominated cytokine pattern: IgG4 counter-regulation by CLA memory T cells.
J. Immunol.
159
:
4611
.
17
Akdis, M., H. U. Simon, L. Weigl, O. Kreyden, K. Blaser, C. A. Akdis.
1999
. Skin homing (cutaneous lymphocyte-associated antigen-positive) CD8+ T cells respond to superantigen and contribute to eosinophilia and IgE production in atopic dermatitis.
J. Immunol.
163
:
466
.
18
Hamid, Q., M. Boguniewicz, D. Y. Leung.
1994
. Differential in situ cytokine gene expression in acute versus chronic atopic dermatitis.
J. Clin. Invest.
94
:
870
.
19
Akdis, C. A., M. Akdis, D. Simon, B. Dibbert, M. Weber, S. Gratzl, O. Kreyden, R. Disch, B. Wuthrich, K. Blaser, H. U. Simon.
1999
. T cells and T cell-derived cytokines as pathogenic factors in the nonallergic form of atopic dermatitis.
J. Invest. Dermatol.
113
:
628
.
20
Werfel, T., A. Morita, M. Grewe, H. Renz, U. Wahn, J. Krutmann, A. Kapp.
1996
. Allergen specificity of skin-infiltrating T cells is not restricted to a type-2 cytokine pattern in chronic skin lesions of atopic dermatitis.
J. Invest. Dermatol.
107
:
871
.
21
Grewe, M., C. A. Bruijnzeel-Koomen, E. Schopf, T. Thepen, A. G. Langeveld-Wildschut, T. Ruzicka, J. Krutmann.
1998
. A role for Th1 and Th2 cells in the immunopathogenesis of atopic dermatitis.
Immunol. Today
19
:
359
.
22
Thepen, T., E. G. Langeveld-Wildschut, I. C. Bihari, D. F. van Wichen, F. C. van Reijsen, G. C. Mudde, C. A. Bruijnzeel-Koomen.
1996
. Biphasic response against aeroallergen in atopic dermatitis showing a switch from an initial Th2 response to a Th1 response in situ: an immunocytochemical study.
J. Allergy Clin. Immunol.
97
:
828
.
23
Trautmann, A., M. Akdis, D. Kleemann, F. Altznauer, H. U. Simon, T. Graeve, M. Noll, E. B. Bröcker, K. Blaser, C. A. Akdis.
2000
. T cell-mediated Fas-induced keratinocyte apoptosis plays a key pathogenetic role in eczematous dermatitis.
J. Clin. Invest.
106
:
25
.
24
Federici, M., M. L. Giustizieri, C. Scarponi, G. Girolomoni, C. Albanesi.
2002
. Impaired IFN-γ-dependent inflammatory responses in human keratinocytes overexpressing the suppressor of cytokine signaling 1.
J. Immunol.
169
:
434
.
25
Trautmann, A., F. Altznauer, M. Akdis, H. U. Simon, R. Disch, E. B. Bröcker, K. Blaser, C. A. Akdis.
2001
. The differential fate of cadherins during T-cell-induced keratinocyte apoptosis leads to spongiosis in eczematous dermatitis.
J. Invest. Dermatol.
117
:
927
.
26
Hanifin, J. M., G. Rajka.
1980
. Diagnostic features of atopic dermatitis.
Acta Dermatol. Venerol.
92
:
44
.
27
Boukamp, P., R. T. Petrussevska, D. Breitkreutz, J. Hornung, A. Markham, N. E. Fusenig.
1988
. Normal keratinization in a spontaneously immortalized aneuploid human keratinocyte cell line.
J. Cell Biol.
106
:
761
.
28
Akdis, C. A., T. Blesken, M. Akdis, B. Wüthrich, K. Blaser.
1998
. The role of IL-10 in specific immunotherapy.
J. Clin. Invest.
102
:
98
.
29
Sallusto, F., C. R. Mackay, A. Lanzavecchia.
1997
. Selective expression of the eotaxin receptor CCR3 by human T helper 2 cells.
Science
277
:
2005
.
30
Akdis, C. A., M. Akdis, A. Trautmann, K. Blaser.
2000
. Immune regulation in atopic dermatitis.
Curr. Opin. Immunol.
12
:
641
.
31
Akdis, M., A. Trautmann, K. Blaser, C. A. Akdis.
2002
. Mechanisms of allergic skin inflammation. D. Y. M. Leung, ed.
Atopic Dermatitis
145
. Marcel Dekker, New York.
32
Nickoloff, B. J., Y. Naidu.
1994
. Perturbation of epidermal barrier function correlates with initiation of cytokine cascade in human skin.
J. Am. Acad. Dermatol.
30
:
535
.
33
Grabbe, S., T. Schwarz.
1998
. Immunoregulatory mechanisms involved in elicitation of allergic contact hypersensitivity.
Immunol. Today
19
:
37
.
34
Albanesi, C., S. Pastore, E. Fanales-Belasio, G. Girolomoni.
1998
. Cetirizine and hydrocortisone differentially regulate ICAM-1 expression and chemokine release in cultured human keratinocytes.
Clin. Exp. Allergy
28
:
101
.
35
Albanesi, C., A. Cavani, G. Girolomoni.
1999
. IL-17 is produced by nickel-specific T lymphocytes and regulates ICAM-1 expression and chemokine production in human keratinocytes: synergistic or antagonist effects with IFN-γ and TNF-α.
J. Immunol.
162
:
494
.
36
Albanesi, C., A. Cavani, G. Girolomoni.
1998
. Interferon-γ-stimulated human keratinocytes express the genes necessary for the production of peptide-loaded MHC class II molecules.
J. Invest. Dermatol.
110
:
138
.
37
Traidl, C., S. Sebastiani, C. Albanesi, H. F. Merk, P. Puddu, G. Girolomoni, A. Cavani.
2000
. Disparate cytotoxic activity of nickel-specific CD8+ and CD4+ T cell subsets against keratinocytes.
J. Immunol.
165
:
3058
.
38
Pastore, S., S. Corinti, M. La Placa, B. Didona, G. Girolomoni.
1998
. Interferon-γ promotes exaggerated cytokine production in keratinocytes cultured from patients with atopic dermatitis.
J. Allergy Clin. Immunol.
101
:
538
.
39
Albanesi, C., C. Scarponi, S. Sebastiani, A. Cavani, M. Federici, O. De Pita, P. Puddu, G. Girolomoni.
2000
. IL-4 enhances keratinocyte expression of CXCR3 agonistic chemokines.
J. Immunol.
165
:
1395
.
40
Saunders, N., A. Dahler, S. Jones, R. Smith, A. Jetten.
1996
. Interferon-γ as a regulator of squamous differentiation.
J. Dermatol. Sci.
13
:
98
.
41
Goebeler, M., A. Trautmann, A. Voss, E. B. Bröcker, A. Toksoy, R. Gillitzer.
2001
. Differential and sequential expression of multiple chemokines during elicitation of allergic contact hypersensitivity.
Am. J. Pathol.
158
:
431
.
42
Giustizieri, M. L., F. Mascia, A. Frezzolini, O. De Pita, L. M. Chinni, A. Giannetti, G. Girolomoni, S. Pastore.
2001
. Keratinocytes from patients with atopic dermatitis and psoriasis show a distinct chemokine production profile in response to T cell-derived cytokines.
J. Allergy Clin. Immunol.
107
:
871
.
43
Sebastiani, S., C. Albanesi, P. O. De, P. Puddu, A. Cavani, G. Girolomoni.
2002
. The role of chemokines in allergic contact dermatitis.
Arch. Dermatol. Res.
293
:
552
.
44
Qin, S., J. B. Rottman, P. Myers, N. Kassam, M. Weinblatt, M. Loetscher, A. E. Koch, B. Moser, C. R. Mackay.
1998
. The chemokine receptors CXCR3 and CCR5 mark subsets of T cells associated with certain inflammatory reactions.
J. Clin. Invest.
101
:
746
.
45
Bonecchi, R., G. Bianchi, P. P. Bordignon, D. D’Ambrosio, R. Lang, A. Borsatti, S. Sozzani, P. Allavena, P. A. Gray, A. Mantovani, F. Sinigaglia.
1998
. Differential expression of chemokine receptors and chemotactic responsiveness of type 1 T helper cells (Th1s) and Th2s.
J. Exp. Med.
187
:
129
.
46
Ebert, L. M., S. R. McColl.
2001
. Coregulation of CXC chemokine receptor and CD4 expression on T lymphocytes during allogenic activation.
J. Immunol.
166
:
4870
.
47
Sallusto, F., D. Lenig, R. Forster, M. Lipp, A. Lanzavecchia.
1999
. Two subsets of memory T lymphocytes with distinct homing potentials and effector functions.
Nature
401
:
708
.
48
Iida, N., M. Haisa, A. Igarashi, D. Pencev, G. R. Grotendorst.
1996
. Leukocyte-derived growth factor links the PDGF and CXC chemokine families of peptides.
FASEB J.
10
:
1336
.
49
Blotnick, S., G. E. Peoples, M. R. Freeman, T. J. Eberlein, M. Klagsbrun.
1994
. T lymphocytes synthesize and export heparin-binding epidermal growth factor-like growth factor and basic fibroblast growth factor, mitogens for vascular cells and fibroblasts: differential production and release by CD4+ and CD8+ T cells.
Proc. Natl. Acad. Sci. USA
91
:
2890
.
50
Engelhardt, E., A. Toksoy, M. Goebeler, S. Debus, E. B. Bröcker, R. Gillitzer.
1998
. Chemokines IL-8, GROα, MCP-1, IP-10, and Mig are sequentially and differentially expressed during phase-specific infiltration of leukocyte subsets in human wound healing.
Am. J. Pathol.
153
:
1849
.