Keratinocytes contribute relevantly to the pathogenesis of inflammatory skin diseases by expressing a variety of proinflammatory molecules, with T cell-derived IFN-γ being the most potent keratinocyte activator. Suppressor of cytokine signaling (SOCS)1 and SOCS3 are negative regulators of IFN-γ signaling and are induced in many cell types by IFN-γ itself or by other cytokines. We show in this work that SOCS1, SOCS2, SOCS3, and cytokine-inducible SH2-containing protein mRNA were up-regulated by IFN-γ in normal human keratinocytes, whereas only SOCS1 or SOCS1 and cytokine-inducible SH2-containing protein were induced by TNF-α or IL-4, respectively. SOCS1, SOCS2, and SOCS3 proteins were undetectable in healthy skin and highly expressed in the epidermis of psoriasis and allergic contact dermatitis, but were only weakly expressed in atopic dermatitis skin. In keratinocytes transiently transfected with SOCS1 or SOCS3 the IFN-γ-induced transactivation of an IFN-γ-responsive reporter gene was markedly inhibited. SOCS1 and SOCS3 overexpression in keratinocyte stable clones inhibited IFN-γ-induced phosphorylation of IFN-γRα and activation of STAT1 and STAT3. Furthermore, SOCS1 and, to a lesser extent, SOCS3 reduced membrane expression of ICAM-1 and HLA-DR, and release of IFN-γ-inducible protein-10, monokine induced by IFN-γ, and monocyte chemoattractant protein-1 by keratinocyte clones promoted by IFN-γ. SOCS1-expressing keratinocytes showed constitutively higher, but not IFN-γ-inducible, IL-8 levels compared with SOCS2 and SOCS3 clones, and were resistant to IFN-γ-mediated growth inhibition. Targeting keratinocyte SOCS1 may represent a novel therapeutic approach to IFN-γ-dependent skin diseases.

Chronic inflammatory skin disorders such as psoriasis, allergic contact dermatitis (ACD),4 and atopic dermatitis (AD) are characterized by an intense infiltrate of activated T lymphocytes, which release lymphokines influencing the immune functions of resident skin cells, including keratinocytes (1, 2, 3). Among T cell-derived cytokines, IFN-γ is the most potent activator of the proinflammatory functions of keratinocytes. In fact, IFN-γ-activated keratinocytes express a broad array of chemokines, cytokines, and membrane molecules that direct the recruitment, activation, and retention of specific leukocyte subpopulations in the skin (4, 5, 6). Although less potently than IFN-γ, TNF-α and IL-4 can also activate keratinocytes. The reciprocal activation of T lymphocytes and keratinocytes has a primary role in the amplification of skin inflammation during immune-mediated skin diseases.

In the last five years, a large number of studies has focused on the mechanisms by which cytokine actions are negatively regulated. Suppressors of cytokine signaling (SOCS) are a family of intracellular molecules comprising at least eight members, SOCS1 to SOCS7 and cytokine-inducible SH2-containing protein (CIS), which share structural similarities and are characterized by a central SH2 domain and a unique motif, the SOCS box, in their C-terminal region. SOCS molecules have been detected in various tissues and are produced in response to different cytokines. SOCS regulate the magnitude and duration of responses triggered by various cytokines by inhibiting their signal transduction pathway in a classic negative feedback loop (7, 8, 9). At the molecular level, SOCS/CIS bind directly to cytokine receptors or to the catalytic domain of Janus kinase (Jak) proteins and impede the recruitment and phosphorylation of STAT (10). SOCS1 induction by IFN-γ and negative regulation of the IFN-γ signaling by SOCS1 have been well documented in a wide variety of cell types, including M1, HeLa, bone marrow cells, and monocytes (11, 12, 13). In particular, SOCS1 inhibits IFN-γ signaling by binding as a pseudosubstrate to Jak1 and Jak2, which are associated, respectively, with the IFN-γR α and β subunits. Disabled Jak1 and Jak2 cannot mediate STAT1 phosphorylation, which is necessary for activation of γ-activated sequences (GAS) in the promoters of target genes (8, 10, 14). Moreover, constitutive activation of STAT1 and hyperresponsiveness to IFN-γ were found in SOCS1-deficient mice (15). Also, SOCS3 represses signaling induced by IFN-γ, although its inhibitory activity toward STAT1 activation is weaker than that exhibited by SOCS1 (12).

Although the molecular bases of SOCS molecule activity have been extensively investigated, limited information exists on inflammatory mediators affected by SOCS (16, 17) and on SOCS expression in human pathologic conditions (18). In this study, we sought to determine whether SOCS family members are expressed by normal human keratinocytes in vitro and in vivo during immune-mediated diseases. Furthermore, through establishment of SOCS1-, SOCS2-, and SOCS3-expressing keratinocyte clones, we investigated the potential role of SOCS molecules in inhibiting IFN-γ signal transduction as well as IFN-γ-induced production of chemokines and adhesion molecules.

Normal human keratinocytes were prepared from plastic surgery skin obtained from healthy individuals (n = 3), as described previously (4). Second- or third-passage keratinocytes were used in all experiments, with cells cultured in six-well plates in the serum-free medium, keratinocyte growth medium (Clonetics, San Diego, CA), for at least 3–5 days (at 60–80% confluence) before performing cytokine treatment. Stimulation with 200 U/ml human rIFN-γ, 50 ng/ml rTNF-α, or 50 ng/ml rIL-4 (R&D Systems, Abingdon, Oxon, U.K.) was performed in keratinocyte growth medium devoid of hydrocortisone and bovine pituitary extract, and supplemented with 0.1% BSA (Sigma-Aldrich, Milan, Italy). The HaCaT human keratinocyte cell line was a gift from N. E. Fusenig (Deutsches Krebsforschungszentrum, Heidelberg, Germany) and was grown in DMEM (Biochrom, Berlin, Germany) supplemented with 10% Fetalclone II serum (HyClone Laboratories, Logan, UT). When 60–80% confluence was achieved, HaCaT cells were stimulated with 200 U/ml IFN-γ.

Total RNA was extracted from cultured keratinocytes using the TRIzol reagent (Invitrogen, Carlsbad, CA). The human SOCS multiprobe template set and the complete kit for RNase protection assay were purchased from BD PharMingen (San Diego, CA). α-32P-Labeled antisense riboprobes were generated from DNA corresponding to CIS, SOCS7, SOCS6, SOCS5, SOCS3, SOCS2, and SOCS1, as well as the housekeeping genes L32 and GAPDH, and were in vitro transcribed in the presence of a GACU nucleotide pool precursors using a T7 RNA polymerase. Hybridization of riboprobes with 10 μg of each RNA sample was performed overnight, followed by digestion with RNase A and T1. The samples were treated with proteinase K, extracted with Tris-saturated phenol plus chloroform:isoamyl alcohol (50:1), and finally precipitated in the presence of ammonium acetate. Protected fragments were resolved by electrophoresis on 4.5% polyacrylamide-urea gel. For Northern analysis, 5 μg of total RNA was fractionated on 1% formaldehyde-agarose denaturing gels, blotted on nylon membrane (Amersham Pharmacia Biotech, Milan, Italy), and fixed by UV irradiation. SOCS2 probe was obtained by RT-PCR performed on RNA isolated from IFN-γ-stimulated keratinocyte cultures, and using the following primer pairs: TATCAGGATGGTACTGGGGAAGTA (5′) andCTTGTTGGTAAAGGCAGTCCCCAG (3′) (GenBank accession no. AF037989). SOCS2 amplificate was gel purified, cloned into pCR-TOPO vector (Invitrogen), and then subjected to an automated sequence analysis using a PerkinElmer Sequencer (model ABI Prism 377 XL; PerkinElmer, Roche Molecular Systems, Branchburg, NJ). SOCS2 probe was labeled with [32P]dCTP and used for hybridization conducted for 1 h at 68°C in QuickHyb solution (Stratagene, La Jolla, CA), as per the manufacturer’s protocol. The filters were washed two times at room temperature and once at 60°C under high-stringency conditions (0.1× SSC, 0.1% SDS) and finally exposed at −80°C to Kodak Biomax MS-1 films (Kodak, Rochester, NY).

Four-millimeter punch biopsies were taken from skin of adult patients with chronic plaque psoriasis (n = 5) and chronic AD (n = 5) and from healthy control subjects (n = 3). Skin biopsies were also taken from 48-h positive patch test reactions to NiSO4 from four patients with ACD to nickel. Patients were not receiving any systemic or topical therapy before sampling. Five-micrometer cryostatic sections were fixed with 5% paraformaldehyde for 10 min, treated with 0.3% hydrogen peroxide to quench endogenous peroxidase activity, incubated with normal horse serum (Vectastain ABC kit; Vector Laboratories, Burlingame, CA) for 20 min, and finally permeabilized with 0.05% Triton X-100. Staining was performed with goat polyclonal anti-SOCS1 (15 μg/ml; C-20; Santa Cruz Biotechnology, Santa Cruz, CA) and anti-SOCS2 (10 μg/ml; M-19; Santa Cruz Biotechnology), and rabbit anti-SOCS3 (5 μg/ml; Immuno-Biological Laboratories, Hamburg, Germany). Immunoreactivity was revealed using avidin-biotin-peroxidase system and 3-amino-9-ethylcarbazole chromogen (Vector Laboratories). Sections were counterstained with Mayer’s hematoxylin. As negative controls, primary Abs were omitted or replaced with control serum.

Myc-tagged full-length murine JAK-binding protein/SOCS1 and human SOCS2 or SOCS3 in pcDNA3 plasmid (pcDNA-myc/SOCS1-2-3) were a generous gift of Dr. A. Yoshimura (Kyushu University, Fukuoka, Japan). FLAG epitope-tagged murine SOCS1, SOCS2, and SOCS3 in pEF-BOS expression vector (pEF-FLAG/SOCS1-2-3) were kindly provided by Dr. D. J. Hilton (Walter and Eliza Hall Institute for Medical Research, Parkville, Victoria, Australia).

Normal human keratinocytes and HaCaT cells were transiently transfected in duplicate using Lipofectin and LipofectAMINE PLUS reagents (Invitrogen), respectively. Typically, 1.5–2 × 105 cells were seeded in six-well plates 24–48 h before transfection (60–80% confluence). For each well, 0.5 μg pCMV · SPORT-β-galactosidase plasmid (Invitrogen) and 0.5 μg pGAS-Luc or pNF-κB-Luc vectors (Stratagene) were cotransfected with increasing concentrations (0.1–2 μg) of pcDNA3-myc/SOCS1-2-3 or pEF-FLAG/SOCS1-2-3 plasmid sets. Increasing concentrations of pcDNA3 and pEF-BOS empty vectors were also used as controls. After overnight culture, the cells were incubated for 24 h with 200 U/ml IFN-γ in serum-free medium and then lysed. β-galactosidase and luciferase activities were measured using the β-Gal ELISA kit (Boehringer Mannheim, Mannheim, Germany) and luciferase assay system (Promega, Madison, WI), respectively. Luciferase activity of each sample was normalized to the β-galactosidase activity, and its basal level, in the absence of IFN-γ and the different SOCS constructs, was given the value of 1. To confirm that increasing the concentration of transfected SOCS plasmids resulted in an increased expression of SOCS proteins, lysates were analyzed by Western blot using anti-c-myc9E10 mAb (Santa Cruz Biotechnology) or anti-FLAG M2 mAb (Sigma-Aldrich). SOCS HaCaT cells were permanently transfected with pcDNA-myc/SOCS1-2-3 or empty pcDNA3 plasmids linearized by ScaI restriction endonuclease (Boehringer Mannheim). Genetycin-resistant clones were selected after ∼20 days by adding 0.4 mg/ml G418 (Invitrogen) to the culture medium. HaCaT clones expressing SOCS1-2 or -3 proteins were selected by Western blot analysis with the anti-c-myc Ab.

Protein extracts were prepared by solubilizing cells in RIPA buffer (1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS in PBS) containing a mixture of protease and phosphatase inhibitors. Proteins were subjected to SDS-PAGE and transferred to polyvinylidene difluoride membrane. Membranes were then blocked and probed with various primary Abs diluted in PBS containing 5% nonfat dried milk or 3% BSA. The latter were as follows: anti-STAT1 (E-23; Santa Cruz Biotechnology), anti-STAT3 (C20; Santa Cruz Biotechnology), anti-phosphotyrosine (PY) STAT1 (Tyr701) and STAT3 (Tyr705) (both from New England Biolabs, Beverly, MA), anti-phosphoserine (PS) STAT1 (Ser727; Upstate Biotechnology, Lake Placid, NY), anti-SOCS1 and anti-SOCS3 (both from Immuno-Biological Laboratories), anti-SOCS2 (M-19; Santa Cruz Biotechnology), and HRP-conjugated anti-c-myc (9E10; Santa Cruz Biotechnology). Filters were properly developed with anti-mouse, anti-goat, or anti-rabbit Ig Abs conjugated to HRP using the ECL-plus detection system (Amersham Pharmacia Biotech), followed by autoradiography. For immunoprecipitation, protein extracts were incubated with protein G-Sepharose beads (Amersham Pharmacia Biotech) and anti-Jak1, anti-Jak2 (Upstate Biotechnology), and anti-IFN-γRα subunit (C-20; Santa Cruz Biotechnology) polyclonal Abs. Immunoprecipitates were run on 8% SDS-PAGE and probed on polyvinylidene difluoride filters with the 4G10 anti-PY (Upstate Biotechnology) or with anti-c-myc Abs.

IFN-γ-inducible protein-10 (IP-10)/CXCL10 was assayed using the purified 4D5/A7/C5 and the biotinylated 6D4/D6/G2 Abs (BD PharMingen). Monokine induced by IFN-γ (Mig)/CXCL9 was determined using the Ab pair, 2310D mAb for coating and biotinylated B8-6 (BD PharMingen) for detection. Human recombinant chemokines (BD PharMingen) were used as standards. IL-8 (CXCL8) and monocyte chemoattractant protein (MCP)-1/CCL2 were measured with OptEIA kits (BD PharMingen), as per the manufacturer’s protocol. The plates were analyzed in an ELISA reader (model 3550 UV; Bio-Rad, Valencia, CA). Cultures were conducted in triplicate for each condition. Results are given as nanograms/106 cells ± SD.

HaCaT clones were stained with FITC-conjugated anti-CD54 (84H10; Immunotech, Marseille, France) and anti-HLA-DR (L243; BD PharMingen) mAbs. In control samples, staining was performed using isotype-matched control Abs. Cells were analyzed with a FACScan equipped with CellQuest software (BD Biosciences, Mountain View, CA).

A total of 104 cells were seeded in 24-well plates in triplicate for each condition, and after 3 days medium was changed with fresh medium with or without 200 U/ml IFN-γ. Clones were cultured for 1–7 days and the number of viable cells was determined by a trypan blue exclusion test. Experiments were performed on three SOCS1, SOCS2, SOCS3, or control clones.

Wilcoxon’s signed rank test (SigmaStat; Jandel, San Rafael, CA) was used to compare differences in luciferase activity of transiently transfected keratinocytes and chemokine release from SOCS clones. Values of p ≤ 0.05 were considered significant.

The presence of SOCS mRNA was analyzed in normal human keratinocytes in resting conditions and upon IFN-γ, TNF-α, or IL-4 treatment (Fig. 1,A). Low levels of SOCS1, SOCS2, SOCS5, SOCS6, SOCS7, and CIS mRNA were found to be constitutively expressed by keratinocytes. Expression of SOCS1, SOCS2, SOCS3, and CIS was markedly increased following IFN-γ stimulation, whereas only SOCS1 or SOCS1 and CIS were significantly up-regulated in keratinocytes exposed to TNF-α or IL-4, respectively. In contrast, SOCS5–7 were not affected by cytokine treatments. Induction of SOCS1 mRNA by IFN-γ, TNF-α, or IL-4 was increased soon after a 1-h activation and was maintained at a high level for at least 48 h. Because riboprobe for RNase protection did not permit clear evaluation of SOCS2 expression in keratinocytes, mRNA was also analyzed by Northern blot using a SOCS2-specific probe. As shown in Fig. 1 B, keratinocytes transiently up-regulated SOCS2 only after IFN-γ stimulation, with mRNA peaking at 2 h after treatment and returning to basal levels within 10 h. Similarly, IFN-γ-induced expression of SOCS3 and CIS was detectable as early as 30 min after stimulation, peaked at 2 h, and decreased thereafter. Finally, IL-4 strongly increased CIS mRNA expression, which persisted for at least 48 h upon treatment.

FIGURE 1.

SOCS mRNA induction in human keratinocytes activated with IFN-γ, TNF-α, or IL-4. Keratinocyte cultures were treated for the indicated time periods with human rIFN-γ (100 U/ml), rTNF-α, or rIL-4 (both at 50 ng/ml). Total RNA was used in RNase protection assay conducted with a human SOCS multiprobe template set as well as the housekeeping gene L32 (A). Total RNA were also analyzed by Northern blot using a SOCS2-specific probe. Eight-hour exposures of films for both RNase protection and Northern blot were conducted.

FIGURE 1.

SOCS mRNA induction in human keratinocytes activated with IFN-γ, TNF-α, or IL-4. Keratinocyte cultures were treated for the indicated time periods with human rIFN-γ (100 U/ml), rTNF-α, or rIL-4 (both at 50 ng/ml). Total RNA was used in RNase protection assay conducted with a human SOCS multiprobe template set as well as the housekeeping gene L32 (A). Total RNA were also analyzed by Northern blot using a SOCS2-specific probe. Eight-hour exposures of films for both RNase protection and Northern blot were conducted.

Close modal

To see whether keratinocytes express SOCS molecules in vivo, we examined the presence of SOCS1, SOCS2, and SOCS3 proteins in normal skin from healthy subjects and in the skin affected with psoriasis, ACD to nickel and AD. As shown in Fig. 2, AC, keratinocytes of healthy skin did not stain for SOCS1, SOCS2, or SOCS3, although rare and scattered immunoreactive cells were present in the epidermis. In contrast, psoriatic epidermis showed a diffuse and intense SOCS1, SOCS2, and SOCS3 reactivity, with keratinocytes displaying a cytoplasmic staining more evident in the basal and suprabasal epidermal layers (Fig. 2, DF). Keratinocytes in ACD reactions also stained for SOCS1, SOCS2, and SOCS3 molecules, with the staining more intense in discrete areas of the basal and suprabasal layers (Fig. 2, GI). Finally, epidermis of chronic AD skin showed only some restricted areas of faint positivity for SOCS molecules (Fig. 2, JL). A portion of leukocytes infiltrating the dermis in all three diseases was also immunoreactive for SOCS1, SOCS2, and SOCS3. Nonlesional skin from patients with psoriasis or AD did not stain for SOCS molecules (data not shown).

FIGURE 2.

SOCS expression in inflammatory skin lesions. Immunohistochemistry was performed on frozen sections from normal skin (AC), psoriasis vulgaris (DF), 48-h patch test reactions to nickel sulfate (GI), and AD lesions (JL). SOCS reactivity (red staining) was revealed by using avidin-biotin-peroxidase complexes and 3-amino-9-ethylcarbazole as substrate. Normal skin did not stain for SOCS1 (A), SOCS2 (B), or SOCS3 (C) molecules. SOCS1 positivity was found in keratinocytes of psoriasis (D), ACD (G), and, to a lesser extent, AD lesions (J). A similar pattern of expression was revealed for SOCS2 (E, H, and K) and SOCS3 (F, I, and L). Some infiltrating dermal leukocytes were also reactive for SOCS in all the diseases examined. Sections were counterstained with Mayer’s hematoxylin. Similar results were observed in skin biopsies from five patients.

FIGURE 2.

SOCS expression in inflammatory skin lesions. Immunohistochemistry was performed on frozen sections from normal skin (AC), psoriasis vulgaris (DF), 48-h patch test reactions to nickel sulfate (GI), and AD lesions (JL). SOCS reactivity (red staining) was revealed by using avidin-biotin-peroxidase complexes and 3-amino-9-ethylcarbazole as substrate. Normal skin did not stain for SOCS1 (A), SOCS2 (B), or SOCS3 (C) molecules. SOCS1 positivity was found in keratinocytes of psoriasis (D), ACD (G), and, to a lesser extent, AD lesions (J). A similar pattern of expression was revealed for SOCS2 (E, H, and K) and SOCS3 (F, I, and L). Some infiltrating dermal leukocytes were also reactive for SOCS in all the diseases examined. Sections were counterstained with Mayer’s hematoxylin. Similar results were observed in skin biopsies from five patients.

Close modal

Although SOCS molecules may also be involved in TNF-α or IL-4-mediated keratinocyte activation, we concentrated on the role of SOCS molecules in the regulation of IFN-γ signaling, because IFN-γ is the most powerful proinflammatory cytokine on keratinocytes. To this end, keratinocytes were cotransfected with the IFN-γ-inducible reporter plasmid, pGAS-Luc, in the absence or presence of increasing concentrations (0.1–2 μg/well) of pcDNA3-myc/SOCS1-2-3 or empty pcDNA3 vectors. The pGAS-Luc reporter gene contains the luciferase reporter gene driven by a basic promoter element (TATA box) joined to tandem repeats of the GAS, which are a prototypical STAT1-binding site (14). Keratinocytes transfected with the pGAS-Luc and stimulated with IFN-γ showed a 20-fold increase in luciferase activity (Fig. 3). Transfection of SOCS1 decreased the luciferase activity of pGAS-Luc plasmid very efficiently and in a dose-dependent manner (Fig. 3,B; p < 0.002). Also, SOCS3 overexpression reduced IFN-γ signaling in keratinocytes, although less potently than SOCS1 (Fig. 3,C; p < 0.03). In contrast, SOCS2 or pcDNA3 empty vector did not affect the activation of luciferase gene transcription by IFN-γ (Fig. 3, A and D). An identical inhibitory pattern by SOCS on IFN-γ signaling was obtained when the pEF-FLAG/SOCS1-2-3 plasmid set was used to transfect keratinocytes (data not shown). To test whether the SOCS1- or SOCS3-dependent inhibition of IFN-γ signaling was exerted specifically on STAT1-activated pathway, we also used the pNF-κB-Luc reporter plasmid that is activated specifically by NF-κB transcription factor. Luciferase levels resulting from transfection of IFN-γ-treated keratinocytes with pNF-κB-Luc were lower (6-fold induction) compared with that obtained with pGAS-Luc, and, more importantly, did not change in keratinocytes overexpressing SOCS1, SOCS2, or SOCS3 (Fig. 3 E). Thus, SOCS1 and SOCS3, but not SOCS2, impair the ability of IFN-γ to transactivate luciferase gene expression from STAT1- but not NF-κB-binding promoters in keratinocytes.

FIGURE 3.

SOCS1 and SOCS3 forced expression strongly reduces the IFN-γ-induced activation of a STAT1-responsive reporter plasmid in transiently transfected keratinocytes. Cultured human keratinocytes were cotransfected with increasing amounts (0–2 μg/well) of pcDNA3 (A), pcDNA3-myc-SOCS1 (B), -SOCS3 (C), or -SOCS2 (D) vectors and 0.5 μg pGAS-Luc reporter plasmid. Keratinocytes were also cotransfected with a fixed amount (1 μg/well) of pcDNA3, pcDNA3-myc-SOCS1, -SOCS2, and -SOCS3, and 0.5 μg pNF-κB-Luc reporter plasmid (E). After overnight culture, cells were incubated for 24 h with 200 U/ml IFN-γ and then lysed to determine β-galactosidase and luciferase activities. Luciferase activity of each sample was normalized to the β-galactosidase activity, and its basal level, in the absence of IFN-γ and the different SOCS constructs, was given the value of 1.

FIGURE 3.

SOCS1 and SOCS3 forced expression strongly reduces the IFN-γ-induced activation of a STAT1-responsive reporter plasmid in transiently transfected keratinocytes. Cultured human keratinocytes were cotransfected with increasing amounts (0–2 μg/well) of pcDNA3 (A), pcDNA3-myc-SOCS1 (B), -SOCS3 (C), or -SOCS2 (D) vectors and 0.5 μg pGAS-Luc reporter plasmid. Keratinocytes were also cotransfected with a fixed amount (1 μg/well) of pcDNA3, pcDNA3-myc-SOCS1, -SOCS2, and -SOCS3, and 0.5 μg pNF-κB-Luc reporter plasmid (E). After overnight culture, cells were incubated for 24 h with 200 U/ml IFN-γ and then lysed to determine β-galactosidase and luciferase activities. Luciferase activity of each sample was normalized to the β-galactosidase activity, and its basal level, in the absence of IFN-γ and the different SOCS constructs, was given the value of 1.

Close modal

To study the functional role of SOCS proteins in IFN-γ signaling in keratinocytes, the keratinocyte-like cell line HaCaT was stably transfected with SOCS1, SOCS2, or SOCS3 cDNAs. HaCaT clones permanently expressing SOCS were generated by transfecting cells with linearized pcDNA3-myc/SOCS1-2-3 and, as control, empty pcDNA3 vector. Genetycin-resistant clones were screened for SOCS expression by Western blot analysis with the anti-c-myc Ab, and 20 SOCS1 and SOCS2 clones and nine SOCS3 clones were obtained and included in this study. Before establishing clones, HaCaT cells were analyzed and compared with normal keratinocytes in SOCS expression and IFN-γ-induced Jak/STAT signaling pathways. IFN-γ promoted the same pattern of SOCS gene expression in HaCaT cells and normal keratinocytes, and IFN-γ signaling was inhibited in transiently transfected HaCaT by SOCS1 and SOCS3, but not by SOCS2 (data not shown). Like normal keratinocytes, 5 min after IFN-γ treatment HaCaT cells showed Jak1, Jak2, and IFN-γRα subunit phosphorylated in tyrosine residues (Fig. 4). In both keratinocytes and HaCaT cells, STAT1 and, to a lesser degree, STAT3 proteins became phosphorylated 15 min after IFN-γ stimulation (Fig. 4). Therefore, the proximal steps of IFN-γ signaling appear to be identical in normal keratinocytes and HaCaT cells.

FIGURE 4.

The proximal steps of IFN-γ signaling are identical in normal keratinocytes and HaCaT cells. Cultured keratinocytes and HaCaT cells were left untreated or stimulated with 200 U/ml IFN-γ for 5–15 min. Lysates were immunoprecipitated with anti-Jak1, -Jak2, and -IFN-γRα Abs and Western blotted with anti-PY Ab. Lysates were also subjected to Western blot analysis performed with anti-PY-STAT1, -PS-STAT1, and -PY-STAT3 Abs. The positions of activated Jak1 (130 kDa), Jak2 (130 kDa), IFN-γRα (90 kDa), STAT1 (91 and 84 kDa), and STAT3 (92 kDa) are indicated.

FIGURE 4.

The proximal steps of IFN-γ signaling are identical in normal keratinocytes and HaCaT cells. Cultured keratinocytes and HaCaT cells were left untreated or stimulated with 200 U/ml IFN-γ for 5–15 min. Lysates were immunoprecipitated with anti-Jak1, -Jak2, and -IFN-γRα Abs and Western blotted with anti-PY Ab. Lysates were also subjected to Western blot analysis performed with anti-PY-STAT1, -PS-STAT1, and -PY-STAT3 Abs. The positions of activated Jak1 (130 kDa), Jak2 (130 kDa), IFN-γRα (90 kDa), STAT1 (91 and 84 kDa), and STAT3 (92 kDa) are indicated.

Close modal

SOCS1 and SOCS3 have been demonstrated to inhibit signaling induced by IFN-γ by inactivating Jak-mediated phosphorylation and homodimerization of STAT proteins (19, 20). After a 5-min stimulation with IFN-γ, Jak1 and Jak2 became phosphorylated in control keratinocyte clones, as assessed by immunoprecipitation experiments (Fig. 5). Unexpectedly, only a slight reduction of Jak1 phosphorylation was observed in SOCS1 and SOCS3 clones compared with controls or SOCS2 clones (Fig. 5,A). Consistently, Jak2 phosphorylation was unaffected in SOCS1 and SOCS3 clones (Fig. 5,B). Although we did not observe a significant down-regulation of Jak activation, SOCS molecules bound tightly to Jak1 and Jak2 because they were coimmunoprecipitated by anti-Jak1 or Jak2 Abs (Fig. 5). The molecular mechanisms by which SOCS inhibit Jak activity are not fully understood. It has been hypothesized that SOCS molecules can function as pseudosubstrates for Jaks, which in turn cannot phosphorylate their natural substrates (20). A strong inhibition of IFN-γRα phosphorylation was observed in SOCS1 and, to a lower extent, in SOCS3 clones. In contrast, we could not detect any changes in the phosphorylation of IFN-γRα of SOCS2 clones (Fig. 6). These data were confirmed in four SOCS1 and three SOCS2 or SOCS3 clones. As a consequence of the inhibition of IFN-γRα phosphorylation, STAT1 and STAT3 proteins cannot be recruited and activated in IFN-γ-stimulated SOCS1 and SOCS3 transfectants (Fig. 7). Following IFN-γ stimulation, control and SOCS2 keratinocyte clones phosphorylated latent STAT1 at both tyrosine 701 and serine 727 residues. While tyrosine phosphorylation of STAT1 occurred in keratinocytes only upon treatment with IFN-γ (Fig. 7,A), serine phosphorylation of STAT1 was constitutive and up-regulated by IFN-γ (Fig. 7,B). Low amounts of tyrosine phosphorylated STAT3 were also detected in keratinocytes following IFN-γ stimulation (Fig. 7,C). The IFN-γ-induced STAT1 and STAT3 activation was greatly inhibited in all keratinocyte clones expressing SOCS1 (Fig. 7, AC). Six of nine SOCS3 stable clones showed a significant reduction of STAT1 and STAT3 phosphorylation in response to IFN-γ, whereas no differences with controls were observed for cells permanently transfected with SOCS2 (Fig. 7, AC). Keratinocyte clones expressed comparable levels of SOCS1, SOCS2, and SOCS3 proteins (Fig. 7 D). Taken together, these results indicate that SOCS1 and, less efficiently, SOCS3 block the IFN-γ-dependent phosphorylation of IFN-γRα by binding to Jak1 and Jak2, and thus prevent STAT1 and STAT3 activation in keratinocytes.

FIGURE 5.

Jak1 and Jak2 phosphorylation status does not significantly change in IFN-γ-stimulated keratinocytes permanently transfected with SOCS1, SOCS2, or SOCS3. pcDNA3, SOCS1, SOCS2, and SOCS3 clones were left untreated or stimulated with 200 U/ml IFN-γ for 5 min. Lysates were immunoprecipitated with the anti-Jak1 (A) or anti-Jak2 (B) Abs and subjected to Western blot analysis with anti-PY Ab. To confirm the presence of similar amounts of Jak proteins in all samples, anti-PY-stained filters were stripped and reprobed with anti-Jak1 or anti-Jak2 Abs. SOCS1, SOCS2, and SOCS3 binding to Jak proteins were assessed by probing filters of blotted Jak1 and Jak2 immunoprecipitates also with the anti-myc Ab. Myc-tagged SOCS1, SOCS2, and SOCS3 showed molecular masses of 35, 34, and 40 kDa, respectively. Similar results were obtained with four SOCS1 and three SOCS2 or SOCS3 clones.

FIGURE 5.

Jak1 and Jak2 phosphorylation status does not significantly change in IFN-γ-stimulated keratinocytes permanently transfected with SOCS1, SOCS2, or SOCS3. pcDNA3, SOCS1, SOCS2, and SOCS3 clones were left untreated or stimulated with 200 U/ml IFN-γ for 5 min. Lysates were immunoprecipitated with the anti-Jak1 (A) or anti-Jak2 (B) Abs and subjected to Western blot analysis with anti-PY Ab. To confirm the presence of similar amounts of Jak proteins in all samples, anti-PY-stained filters were stripped and reprobed with anti-Jak1 or anti-Jak2 Abs. SOCS1, SOCS2, and SOCS3 binding to Jak proteins were assessed by probing filters of blotted Jak1 and Jak2 immunoprecipitates also with the anti-myc Ab. Myc-tagged SOCS1, SOCS2, and SOCS3 showed molecular masses of 35, 34, and 40 kDa, respectively. Similar results were obtained with four SOCS1 and three SOCS2 or SOCS3 clones.

Close modal
FIGURE 6.

Keratinocytes permanently transfected with SOCS1 and SOCS3, but not with SOCS2, do not phosphorylate the IFN-γRα subunit in response to IFN-γ. Clones were left untreated or stimulated with 200 U/ml IFN-γ for 5 min before lysis. Immunoprecipitation of samples with the IFN-γRα chain Ab was followed by Western blot analysis with anti-PY Ab, and the membranes were stripped and reprobed with the anti-IFN-γRα subunit Ab.

FIGURE 6.

Keratinocytes permanently transfected with SOCS1 and SOCS3, but not with SOCS2, do not phosphorylate the IFN-γRα subunit in response to IFN-γ. Clones were left untreated or stimulated with 200 U/ml IFN-γ for 5 min before lysis. Immunoprecipitation of samples with the IFN-γRα chain Ab was followed by Western blot analysis with anti-PY Ab, and the membranes were stripped and reprobed with the anti-IFN-γRα subunit Ab.

Close modal
FIGURE 7.

Inhibition of STAT1 and STAT3 activation induced by IFN-γ in keratinocytes permanently transfected with SOCS1 and SOCS3, but not with SOCS2. Control and SOCS clones were untreated or stimulated with 200 U/ml IFN-γ for 15 min. Western blot analyses were performed on lysates using anti-PY-STAT1 (A), anti-PS-STAT1 (B), and anti-PY-STAT3 (C). Each filter was stripped and reprobed with Abs against the nonphosphorylated forms of STAT1 and STAT3. SOCS levels in SOCS1, SOCS2, and SOCS3 clones were detected using the anti-c-myc 9E10 Ab (D).

FIGURE 7.

Inhibition of STAT1 and STAT3 activation induced by IFN-γ in keratinocytes permanently transfected with SOCS1 and SOCS3, but not with SOCS2. Control and SOCS clones were untreated or stimulated with 200 U/ml IFN-γ for 15 min. Western blot analyses were performed on lysates using anti-PY-STAT1 (A), anti-PS-STAT1 (B), and anti-PY-STAT3 (C). Each filter was stripped and reprobed with Abs against the nonphosphorylated forms of STAT1 and STAT3. SOCS levels in SOCS1, SOCS2, and SOCS3 clones were detected using the anti-c-myc 9E10 Ab (D).

Close modal

IFN-γ-activated keratinocytes express membrane Ags, such as ICAM-1 and HLA-DR (5, 21), and a broad array of chemokines, including MCP-1, IP-10, Mig, and IL-8 (4, 5, 6). Because many of the IFN-γ-induced genes in keratinocytes are transcriptionally regulated by STAT1 (22, 23) and STAT1 is disabled in SOCS1 and SOCS3 keratinocyte clones, we tested whether SOCS clones could still express inflammatory mediators in response to IFN-γ. Flow cytometry analysis revealed that all 20 SOCS1 clones examined showed a significant reduction (50–65%) of ICAM-1 and an almost complete abrogation of HLA-DR expression (Fig. 8). Although less efficiently, ICAM-1 and HLA-DR were reduced also in some (four of nine) SOCS3 clones, whereas SOCS2 expressed ICAM-1 and HLA-DR levels comparable to that of control (Fig. 8). Interestingly, in some SOCS2-expressing clones (4 of 20) a superinduction of ICAM-1 and HLA-DR was revealed (data not shown). Similar results were obtained when we tested chemokine content in supernatants from IFN-γ-activated keratinocyte clones. IP-10, Mig, and MCP-1 release was greatly impaired in all SOCS1 clones, with 86–90% reduction for IP-10 and 75–80% reduction for Mig and MCP-1 (Fig. 9; p < 0.002). A 50–80% and 44–80% reduction of IP-10 and Mig, respectively, was observed in five of nine IFN-γ-stimulated SOCS3 clones (p < 0.05). In contrast, MCP-1 release by SOCS3 clones was similar to that of controls. Chemokine release was not affected in the majority of IFN-γ-stimulated SOCS2 clones (Fig. 9). Similar to what we observed for ICAM-1 and HLA-DR expression, four SOCS2-expressing clones showed an enhanced release of IP-10, Mig, and MCP-1 compared with controls (data not shown). SOCS clones not exposed to IFN-γ secreted very limited amounts of IP-10, Mig, and MCP-1 (data not shown). Unexpectedly, IL-8 release was higher in IFN-γ-stimulated SOCS1 clones compared with control and SOCS2 or SOCS3 clones (Fig. 10,A). However, a more detailed study of IL-8 revealed that unstimulated SOCS1 clones secreted higher levels of IL-8 compared with control, SOCS2, or SOCS3 clones (3.2–6 vs0.6–1.6 ng/106 cells; p < 0.01). After IFN-γ stimulation, IL-8 secretion was not up-regulated in SOCS1 clones, whereas it was significantly enhanced in control, SOCS2, or SOCS3 clones (Fig. 10 B). Therefore, SOCS1 overexpression in keratinocytes can enhance IL-8 production through a STAT1-independent mechanism, and, at the same time, can block the IFN-γ-induced IL-8 up-regulation in a STAT1-dependent manner.

FIGURE 8.

Keratinocytes permanently transfected with SOCS1 and SOCS3, but not with SOCS2, express reduced ICAM-1 and HLA-DR following IFN-γ stimulation. Control and SOCS1, SOCS2, or SOCS3 clones were analyzed for ICAM-1 and HLA-DR expression by flow cytometry 48 h after treatment with medium alone or 200 U/ml IFN-γ. A, Histogram plots relative to ICAM-1 and HLA-DR expression by representative control and SOCS clones. Thin lines represent staining with matched isotype Ig, whereas gray and bold lines represent staining of unstimulated and IFN-γ-treated cells, respectively. The x-axis and the y-axis indicate the relative cell number and fluorescence intensity, respectively. The numbers indicate the net mean fluorescence intensity of IFN-γ-treated samples. B, Net mean fluorescence intensities (ΔMFI) for ICAM-1 and HLA-DR expression of two control, four SOCS1, three SOCS2, and three SOCS3 clones after IFN-γ treatment.

FIGURE 8.

Keratinocytes permanently transfected with SOCS1 and SOCS3, but not with SOCS2, express reduced ICAM-1 and HLA-DR following IFN-γ stimulation. Control and SOCS1, SOCS2, or SOCS3 clones were analyzed for ICAM-1 and HLA-DR expression by flow cytometry 48 h after treatment with medium alone or 200 U/ml IFN-γ. A, Histogram plots relative to ICAM-1 and HLA-DR expression by representative control and SOCS clones. Thin lines represent staining with matched isotype Ig, whereas gray and bold lines represent staining of unstimulated and IFN-γ-treated cells, respectively. The x-axis and the y-axis indicate the relative cell number and fluorescence intensity, respectively. The numbers indicate the net mean fluorescence intensity of IFN-γ-treated samples. B, Net mean fluorescence intensities (ΔMFI) for ICAM-1 and HLA-DR expression of two control, four SOCS1, three SOCS2, and three SOCS3 clones after IFN-γ treatment.

Close modal
FIGURE 9.

Keratinocyte clones overexpressing SOCS1 show an impaired IFN-γ-induced release of IP-10, Mig, and MCP-1. Supernatants from IFN-γ-treated keratinocyte cultures were tested by ELISA for IP-10 (A), Mig (B), and MCP-1 (C) content. Data are expressed as mean nanograms per 106 cells ± SD of triplicate cultures.

FIGURE 9.

Keratinocyte clones overexpressing SOCS1 show an impaired IFN-γ-induced release of IP-10, Mig, and MCP-1. Supernatants from IFN-γ-treated keratinocyte cultures were tested by ELISA for IP-10 (A), Mig (B), and MCP-1 (C) content. Data are expressed as mean nanograms per 106 cells ± SD of triplicate cultures.

Close modal
FIGURE 10.

IL-8 constitutive release is enhanced in keratinocyte clones overexpressing SOCS1 and is not up-regulated by IFN-γ. A, IL-8 release was examined in supernatants from SOCS clones treated for 48 h with 200 U/ml IFN-γ by ELISA. B, IL-8 secretion was also tested in supernatants from control (circles), SOCS1 (squares), SOCS2 (triangles), and SOCS3 (diamonds) clones treated with medium alone (open symbols) or 200 U/ml IFN-γ (filled symbols) for the indicated time periods. Data are expressed as mean nanograms per 106 cells ± SD of triplicate cultures.

FIGURE 10.

IL-8 constitutive release is enhanced in keratinocyte clones overexpressing SOCS1 and is not up-regulated by IFN-γ. A, IL-8 release was examined in supernatants from SOCS clones treated for 48 h with 200 U/ml IFN-γ by ELISA. B, IL-8 secretion was also tested in supernatants from control (circles), SOCS1 (squares), SOCS2 (triangles), and SOCS3 (diamonds) clones treated with medium alone (open symbols) or 200 U/ml IFN-γ (filled symbols) for the indicated time periods. Data are expressed as mean nanograms per 106 cells ± SD of triplicate cultures.

Close modal

Other than inducing proinflammatory genes, IFN-γ exerts an antiproliferative activity on a number of cell types including keratinocytes, and STAT1 activation is required for mediating this effect (24). The antiproliferative activity of IFN-γ was completely abolished in SOCS1 clones. In fact, the number of unstimulated and IFN-γ-stimulated SOCS1-expressing keratinocytes was comparable over all the time points of experiments (Fig. 11). In contrast, a dramatic and progressive reduction of cell number was observed in control, SOCS2, and SOCS3 clones following IFN-γ stimulation (Fig. 11).

FIGURE 11.

The antiproliferative effect of IFN-γ is abrogated in keratinocyte clones overexpressing SOCS1. Control (•), SOCS1 (▪), SOCS2 (▴), and SOCS3 (♦) clones were cultured for the indicated time periods in medium with or without 200 U/ml IFN-γ. Viable cells were determined by trypan blue exclusion test. Data are expressed as the ratio between the number of IFN-γ-treated and untreated cells (relative cell number) ± SD. Experiments were conducted in triplicate for each condition.

FIGURE 11.

The antiproliferative effect of IFN-γ is abrogated in keratinocyte clones overexpressing SOCS1. Control (•), SOCS1 (▪), SOCS2 (▴), and SOCS3 (♦) clones were cultured for the indicated time periods in medium with or without 200 U/ml IFN-γ. Viable cells were determined by trypan blue exclusion test. Data are expressed as the ratio between the number of IFN-γ-treated and untreated cells (relative cell number) ± SD. Experiments were conducted in triplicate for each condition.

Close modal

In this study we focused on the role of the cytokine negative regulators SOCS in IFN-γ-driven activation of keratinocytes. Understanding the mechanisms by which IFN-γ signaling is switched off in keratinocytes is fundamental because this cytokine potently transforms keratinocytes in strong producers of inflammatory mediators during immune-mediated skin diseases (4, 5, 6). That IFN-γ activates keratinocytes in a dominant fashion compared with TNF-α and IL-4 becomes evident also from SOCS mRNA analysis upon cytokine treatment. IFN-γ induced SOCS1, SOCS2, SOCS3, and CIS in keratinocytes, whereas TNF-α promoted only SOCS1 and IL-4 promoted only SOCS1 and CIS. A high expression of SOCS1, SOCS2, and SOCS3 molecules was also found in epidermal keratinocytes of psoriasis and ACD lesions, whereas chronic AD skin showed only a faint epidermal staining. These differences are likely due to the prominent infiltration of IFN-γ-producing type 1 lymphocytes in ACD and psoriasis compared with AD (1, 3). Indeed, studies performed on IFN-γ-stimulated keratinocytes cultured from normal-appearing skin of healthy subjects and patients with psoriasis or AD demonstrated no significant differences in SOCS1, SOCS2, and SOCS3 mRNA induction (data not shown). Therefore, the reduced keratinocyte expression of SOCS molecules in the epidermis of AD compared with psoriasis or ACD cannot be attributed to altered properties of keratinocytes to respond to IFN-γ but rather to the lower amount of IFN-γ produced locally by activated T lymphocytes.

Key events in IFN-γ signal transduction in keratinocytes involve Jak1 and Jak2 activation followed by IFN-γRα phosphorylation and triggering of STAT1 and small amounts of STAT3. Once phosphorylated in tyrosine 701 and serine 727 residues, STAT1 is fully active and can thus induce a variety of inflammatory genes. For instance, ICAM-1, class II transactivator, IP-10, Mig, and MCP-1 genes are all strongly expressed by IFN-γ-treated keratinocytes and tightly regulated at the transcriptional level by STAT1 (22, 23, 25, 26, 27). Regulation by STAT1 was also observed for several cell cycle and proapoptotic genes, including cyclin kinase inhibitor p21 (WAF1), c-myc, and caspase 1 (28, 29, 30), modulating the IFN-γ-induced antiproliferative effect. In transiently transfected keratinocytes, IFN-γ-induced STAT1-dependent gene activation was impaired in the presence of SOCS1 and SOCS3, but not of SOCS2. SOCS1- and SOCS3-dependent inhibition of IFN-γ signaling was exerted specifically on the STAT1 and not the NF-κB pathway. A discrete portion of SOCS1 and SOCS3, called kinase inhibitory region, contributes to high-affinity binding to the Jak2 tyrosine kinase domain and is required for the inhibition of Jak2 activity by preventing the access of substrates and/or ATP in the catalytic pocket (20, 31). As a result of Jak1 and Jak2 inactivation, the phosphorylation in tyrosine 440 residue of IFN-γRα subunit induced by IFN-γ was impaired in keratinocyte clones overexpressing SOCS1 and reduced in SOCS3, but not in SOCS2 clones. Therefore, we could not observe a direct reduction of Jak1 and Jak2 phosphorylation, but rather a decrease of Jak1 and Jak2 intrinsic activities in IFN-γ-treated SOCS1 and SOCS3 clones. In contrast, SOCS1 binds to the catalytic domain of Jak2 specifically in the phosphorylated tyrosine 1007 residue, known as one of the major sites of autophosphorylation of Jak2 (20). The reduced IFN-γRα phosphorylation in SOCS1 and SOCS3 keratinocyte clones led to a substantial inhibition of STAT1 and STAT3 activation. In particular, SOCS1 overexpression in keratinocytes markedly inhibited the IFN-γ-induced STAT1 phosphorylation on both tyrosine 701 and serine 727 residues. Although the mechanisms through which STAT1 is phosphorylated on serine are not yet fully understood, it is known that Jak2 is required for STAT1 serine phosphorylation in response to IFN-γ (32). Therefore, the inhibition of the IFN-γ-induced serine phosphorylation of STAT1 by SOCS1 and SOCS3 can occur indirectly through the abrogation of Jak2 activity. Inhibition of STAT1 and STAT3 phosphorylation was observed in all the SOCS1-expressing keratinocytes clones, in six of nine SOCS3 clones, and in none of the 20 SOCS2 clones examined, although the expression of SOCS2 protein was comparable with those of SOCS1 and SOCS3. Compared with SOCS1, SOCS3 appears to be a weaker inhibitor of IFN-γ signaling, possibly because SOCS3 is not very efficient in binding and inhibiting the catalytic domain of Jak proteins (20). As a direct consequence of inhibition of STAT1 activation, we found a significant reduction of ICAM-1 and HLA-DR expression and IP-10 and Mig production in IFN-γ-activated SOCS1 clones and in four of nine SOCS3 clones. Moreover, SOCS1, but not SOCS3, overexpression inhibited IFN-γ-induced MCP-1 release in keratinocytes. Consistent with these results, mice lacking SOCS1 die of a complex disease characterized by a massive T cell, macrophage, and eosinophil infiltration of visceral organs and the skin (15, 33). Membrane molecule and chemokine production was not affected by SOCS2 constitutive expression. Interestingly, in a small percentage of SOCS2 clones (4 of 20) a superinduction of ICAM-1, HLA-DR, IP-10, Mig, and MCP-1 was observed. These findings can be explained by considering SOCS2 as a negative regulator of SOCS1 and SOCS3. Indeed, previous works showed the SOCS2 capacity of restoring sensitivity of 293 cells to prolactin and growth hormone suppressed by SOCS1 (34, 35). Because SOCS2 protein does not contain kinase inhibitory region domain (10) but can interact with Jak1 and Jak2 (Fig. 5), it is possible that SOCS2 functions as an inefficacious pseudosubstrate for Jak1 and Jak2 or as a competitor of SOCS1 or SOCS3. Another interesting feature of SOCS1 clones was their capacity to produce higher amounts of IL-8 compared with SOCS2 and SOCS3 clones and to actively proliferate even in the presence of IFN-γ. Despite the higher constitutive production of IL-8, SOCS1 clones did not up-regulate this chemokine in response to IFN-γ. We are currently investigating the mechanisms by which SOCS1 influences IL-8 production, and whether enhanced IL-8 production and abrogated IFN-γ-induced growth inhibition in SOCS1 clones are related.

Targeting of SOCS molecules can be an important strategy for the control of cytokine-induced disorders. Indeed, SOCS3 overexpression with blockade of the IL-6/STAT3 pathway improved experimental intestinal inflammation and arthritis (18, 36). Our study provides the first evidences of SOCS expression in human keratinocytes and in immune-mediated skin diseases. Keratinocytes overexpressing SOCS1 are particularly resistant to the proinflammatory effect of IFN-γ. Interestingly, this resistance is achieved at the most proximal step of IFN-γ pathway, namely at the receptor level, and it is effective in preventing downstream expression of relevant inflammatory genes. These findings identify SOCS1 as a potential molecular target for the treatment of IFN-γ-dependent skin diseases.

We gratefully thank Marco A. Cassatella, Flavia Bazzoni (Department of Pathology, General Pathology Unit, University of Verona, Verona, Italy), and Saveria Pastore for helpful suggestions.

1

This work was supported by grants from the Italian Ministry of Health.

4

Abbreviations used in this paper: ACD, allergic contact dermatitis; AD, atopic dermatitis; PS, phosphoserine; MCP, monocyte chemoattractant protein; Mig, monokine induced by IFN-γ; SOCS, suppressor of cytokine signaling; CIS, cytokine-inducible SH2 protein; Jak, Janus kinase; GAS, γ-activated sequence; IP-10, IFN-γ-inducible protein-10; PY, phosphotyrosine.

1
Cavani, A., C. Albanesi, C. Traidl, S. Sebastiani, G. Girolomoni.
2001
. Effector and regulatory T cells in allergic contact dermatitis.
Trends Immunol.
24
:
217
2
Girolomoni, G., S. Sebastiani, C. Albanesi, A. Cavani.
2001
. T-cell subpopulations in the development of atopic and contact allergy.
Curr. Opin. Immunol.
13
:
733
3
Bos, J. D., M. A. De Rie.
1999
. The pathogenesis of psoriasis: immunological facts and speculations.
Immunol. Today
20
:
40
4
Albanesi, C., C. Scarponi, S. Sebastiani, A. Cavani, M. Federici, S. Sozzani, G. Girolomoni.
2001
. A cytokine-to-chemokine axis between T lymphocytes and keratinocytes can favor Th1 cell accumulation in chronic inflammatory skin diseases.
J. Leukocyte Biol.
70
:
617
5
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 antagonistic effects with IFN-γ and TNF-α.
J. Immunol.
162
:
494
6
Albanesi, C., C. Scarponi, S. Sebastiani, A. Cavani, M. Federici, O. De Pità, P. Puddu, G. Girolomoni.
2000
. IL-4 enhances keratinocyte expression of CXCR3 agonistic chemokines.
J. Immunol.
165
:
1395
7
Alexander, W. S., R. Starr, D. Metcalf, S. E. Nicholson, A. Farley, A. G. Elefanty, M. Brysha, B. T. Kile, R. Richardson, M. Baca, et al
1999
. Suppressors of cytokine signaling (SOCS): negative regulators of signal transduction.
J. Leukocyte Biol.
66
:
588
8
Yasukawa, H., A. Sasaki, A. Yoshimura.
2000
. Negative regulation of cytokine signaling.
Annu. Rev. Immunol.
18
:
143
9
Chen, X. P., J. A. Losman, P. Rothman.
2000
. SOCS proteins, regulators of intracellular signaling.
Immunity
13
:
287
10
Gadina, M., D. Hilton, J. A. Johnston, A. Morinobu, A. Lighvani, Y. Zhou, R. Visconti, J. J. O’Shea.
2001
. Signaling by type I and II cytokine receptors: ten years after.
Curr. Opin. Immunol.
13
:
363
11
Sakamoto, H., H. Yasukawa, M. Masuhara, S. Tanimura, A. Sasaki, K. Yuge, M. Ohtsubo, A. Ohtsuka, T. Fujita, T. Ohta, et al
1998
. A Janus kinase inhibitor, JAB, is an interferon-γ-inducible gene and confers resistance to interferons.
Blood
92
:
1668
12
Song, M. M., K. Shuai.
1998
. The suppressor of cytokine signaling (SOCS)1 and SOCS3 but not SOCS2 proteins inhibit interferon-mediated antiviral and antiproliferative activities.
J. Biol. Chem.
273
:
35056
13
Starr, R., T. A. Willson, E. M. Viney, L. J. L. Murray, J. R. Rayner, B. J. Jenkins, T. J. Gonda, W. S. Alexander, D. Metcalf, N. A. Nicola, D. J. Hilton.
1997
. A family of cytokine-inducible inhibitors of signaling.
Nature
387
:
917
14
Stark, G. R., I. M. Kerr, B. R. G. Williams, R. H. Silverman, R. D. Schreiber.
1998
. How cells respond to interferons.
Annu. Rev. Biochem.
67
:
227
15
Alexander, W. S., R. Starr, J. E. Fenner, C. L. Scott, E. Handman, N. S. Sprigg, J. E. Corbin, A. L. Cornish, R. Darwiche, C. M. Owczarek, et al
1999
. SOCS1 is a critical inhibitor of interferon γ signaling and prevents the potentially fatal neonatal actions of this cytokine.
Cell
98
:
597
16
Cottet, S., P. Dupraz, F. Hamburger, W. Dolci, M. Jaquet, B. Thorens.
2001
. SOCS-1 protein prevents Janus kinase/STAT-dependent inhibition of β cell insulin gene transcription and secretion in response to interferon-γ.
J. Biol. Chem.
276
:
25862
17
Karlsen, A. E., S. G. Rønn, K. Lindberg, J. Johannesen, E. D. Galsgaard, F. Pociot, J. H. Nielsen, T. Mandrup-Poulsen, J. Nerup, N. Billestrup.
2001
. Suppressor of cytokine signaling 3 (SOCS-3) protects β-cells against interleukin-1β- and interferon-γ-mediated toxicity.
Proc. Natl. Acad. Sci. USA
98
:
12191
18
Suzuki, A., T. Hanada, K. Mitsuyama, T. Yoshida, S. Kamizono, T. Hoshino, M. Kubo, A. Yamashita, M. Okabe, K. Takeda, et al
2001
. CIS3/SOCS3/SSI3 plays a negative regulatory role in STAT3 activation and intestinal inflammation.
J. Exp. Med.
193
:
471
19
Endo, T. A., M. Masuhara, M. Yokouchi, R. Suzuki, H. Sakamoto, K. Mitsui, A. Matsumoto, S. Tanimura, M. Ohtsubo, H. Misawa, et al
1997
. A new protein containing an SH2 domain that inhibits JAK kinases.
Nature
387
:
921
20
Yasukawa, H., H. Misawa, H. Sakamoto, M. Masuhara, A. Sasaki, T. Wakioka, S. Ohtsuka, T. Imaizumi, T. Matsuda, J. N. Ihle, A. Yoshimura.
1999
. The JAK-binding protein JAB inhibits Janus tyrosine kinase activity through binding in the activation loop.
EMBO J.
18
:
1309
21
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
22
Walter, M. J., D. C. Look, R. M. Tidwell, W. T. Roswit, M. J. Holtzman.
1997
. Targeted inhibition of interferon-γ-dependent intercellular adhesion molecule-1 (ICAM-1) expression using dominant-negative Stat1.
J. Biol. Chem.
272
:
28582
23
Muhlethaler-Mottet, A., W. Di Berardino, L. A. Otten, B. Mach.
1998
. Activation of the MHC class II transactivator CIITA by interferon-γ requires cooperative interaction between Stat1 and USF-1.
Immunity
8
:
157
24
Bromberg, J. F., C. M. Horvath, Z. Wen, R. D. Schreiber, J. E. Darnell.
1996
. Transcriptionally active Stat1 is required for the antiproliferative effects of both interferon α and interferon β.
Proc. Natl. Acad. Sci. USA
93
:
7673
25
Majumder, S., L. Z. Zhou, P. Chaturvedi, G. Babcock, S. Aras, R. M. Ransohoff.
1998
. p48/STAT1-α-containing complexes play a predominant role in induction of IFN-γ-inducible protein, 10 kDa (IP-10) by IFN-γ alone or in synergy with TNF-α.
J. Immunol.
161
:
4736
26
Mahalingam, S., G. Chaudhri, C. L. Tan, A. John, P. S. Foster, G. Karupiah.
2001
. Transcription of the interferon γ (IFN-γ)-inducible chemokine Mig in IFN-γ-deficient mice.
J. Biol. Chem.
276
:
7568
27
Valente, A. J., J. Xie, M. A. Abramova, U. O. Wenzel, H. E. Abboud, D. T. Graves.
1998
. A complex element regulates IFN-γ-stimulated monocyte chemoattractant protein-1 gene transcription.
J. Immunol.
161
:
3719
28
Chin, Y. E., M. Kitagawa, W. S. Su, Z. You, Y. Iwamoto, X. Fu.
1996
. Cell growth arrest and induction of cyclin-dependent kinase inhibitor p21WAF1/CIP1 mediated by STAT1.
Science
272
:
719
29
Ramana, C., N. Grammatokakis, M. Chernov, H. Nguyen, K. C. Goh, B. R. G. Williams, G. R. Stark.
2000
. Regulation of c-myc expression by IFN-γ through Stat1-dependent and -independent pathways.
EMBO J.
19
:
263
30
Chin, Y. E., M. Kitagawa, K. Kuida, R. A. Flavell, X. Fu.
1997
. Activation of the STAT signaling pathway can cause expression of caspase 1 and apoptosis.
Mol. Cell. Biol.
17
:
5328
31
Sasaki, A., H. Yasukawa, A. Suzuki, S. Kamizono, T. Syoda, I. Kinjyo, M. Sasaki, J. A. Johnston, A. Yoshimura.
1999
. Cytokine-inducible SH2 protein-3 (CIS3) inhibits Janus tyrosine kinase by binding through the N-terminal kinase inhibitory region as well as SH2 domain.
Genes Cells
4
:
339
32
Xuejun, Z., Z. Wen, L. Z. Xu, J. E. Darnell.
1997
. Stat1 serine phosphorylation occurs independently of tyrosine phosphorylation and requires an activated Jak2 kinase.
Mol. Cell. Biol.
17
:
6618
33
Metcalf, D., L. Di Rago, S. Mifsud, L. Hartley, W. S. Alexander.
2000
. The development of fatal myocarditis and polymyositis in mice heterozygous for IFN-γ and lacking SOCS-1 gene.
Proc. Natl. Acad. Sci. USA
97
:
9174
34
Pezet, A., H. Favre, P. A. Kelly, M. Edery.
1999
. Inhibition and restoration of prolactin signal transduction by suppressors of cytokine signaling.
J. Biol. Chem.
274
:
24497
35
Favre, H., A. Benhamou, J. Finidori, P. A. Kelly, M. Edery.
1999
. Dual effects of suppressor of cytokine signaling (SOCS-2) on growth hormone signal transduction.
FEBS Lett.
453
:
63
36
Shouda, T., T. Yoshida, T. Hanada, T. Wakioka, M. Oishi, K. Miyoshi, S. Komiya, K. Kosai, Y. Hanakawa, K. Hashimoto, et al
2001
. Induction of the cytokine signal regulator SOCS3/CIS3 as a therapeutic strategy for treating inflammatory arthritis.
J. Clin. Invest.
108
:
1781