Th17 cytokines interleukin (IL)-17 and IL-22 modulate distinct inflammatory and keratinocyte-response pathways

Skin samples

Skin punch biopsies (6 mm diameter) were obtained from normal volunteers (n = 5) and patients with moderate-to-severe chronic plaque psoriasis (n = 16) under a Rockefeller University Institutional Review Board-approved protocol. The biopsy specimens were frozen in OTC (Sakura, Torrance, CA, U.S.A.) and stored at −80°C for immunohistochemistry and immunofluorescence and flash frozen in liquid nitrogen for RNA extraction and analysis. Dermal single-cell suspensions were obtained from abdominoplasty by overnight incubation in dispase (Invitrogen, Carlsbad, CA, U.S.A.) and collagenase (Roche, Indianapolis, IN, U.S.A.) 1 mg mL⁻¹ at 4°C, peeling off the epidermis, and culturing the dermis for 36–48 h at 37°C in RPMI 1640 (Gibco, Carlsbad, CA, U.S.A.) supplemented with 10% pooled human serum (Mediatech Inc., Manassas, VA, U.S.A.), 0.1% gentamicin reagent solution (Gibco) and 1% 1 mol L⁻¹ HEPES buffer (Sigma, St Louis, MO, U.S.A.). Epidermal single-cell suspensions were obtained by incubation in 0.25% trypsin/ethylenediamine tetraacetic acid (EDTA) (Gibco) for 10 min at 37°C, then in RPMI 1640 with 10% pooled human serum, 0.1% gentamicin reagent solution and 1% 1 mol L⁻¹ HEPES buffer overnight.

Peripheral blood samples

Peripheral blood mononuclear cells from freshly drawn blood of normal volunteers (n = 3) were purified by gradient centrifugation with Ficoll-Paque Plus (Pharmacia, Piscataway, NJ, U.S.A.), collected at the interface, then washed with phosphate-buffered saline (PBS) prior to fluorescence-activated cell sorting (FACS) analysis.

Primary keratinocyte cultures

Primary pooled human keratinocytes (n = 3) were obtained from Yale Skin Diseases Research Center core facility and cultured in RPMI 1640 with 10% pooled human serum, 0.1% gentamicin reagent solution and 1% 1 mol L⁻¹ HEPES buffer at 37°C as above. Once 80% confluent, the medium was supplemented with or without the cytokines recombinant human (rh)-IL-17 (R&D Systems, Minneapolis, MN, U.S.A.) 200 ng mL⁻¹, rh-IL-22 (Peprotech Inc., Rocky Hill, NJ, U.S.A.) 200 ng mL⁻¹ or rh-IFN-γ (R&D Systems) 20 ng mL⁻¹ for 24 h before harvesting for other analyses.

Human full-thickness skin model

Full-thickness human skin models (MatTek Corp., Ashland, MA, U.S.A.) were incubated in assay media (MatTek Corp.) supplemented with or without the cytokines rh-IL-17 200 ng mL⁻¹ or rh-IL-22 200 ng mL⁻¹ for 4 days (n = 3). Media, with or without supplemention, were changed every 48 h. On days 2 and 4, the skin models were harvested for histological and RNA analyses.

Antibodies

All antibodies used for immunofluorescence and FACS are listed in Table 1 and Table 2.

Immunohistochemistry

Tissue sections were stained with haematoxylin (Fisher Scientific, Pittsburgh, PA, U.S.A.) and eosin (Shandon, Pittsburg, PA, U.S.A.), or with purified mouse antihuman monoclonal antibodies listed in Table 1. Biotin-labelled horse antimouse antibodies (Vector Laboratories, Burlingame, CA, U.S.A.) were amplified with avidin-biotin complex (Vector Laboratories) and developed with chromogen 3-amino-9-ethylcarbazole (Sigma Aldrich, St Louis, MO, U.S.A.). Appropriate negative controls were used.

Immunofluorescence

Skin sections were stained as previously described¹⁹ using antibodies listed in Table 1. Images were acquired using appropriate filters of a Zeiss Axioplan 2I microscope with Plan Apochromat 20 × 0.7 numerical aperture lens and a Hagamatsu orca ER-cooled charge-coupled device camera, controlled by METAVUE software (Molecular Devices, Sunnyvale, CA, U.S.A.).

Reverse transcriptase–polymerase chain reaction

RNA was extracted from primary human keratinocytes, full-thickness skin equivalents and human skin using the RNeasy Mini Kit (Qiagen, Valencia, CA, U.S.A.). Reverse transcriptase (RT)–polymerase chain reaction (PCR) was performed using EZ PCR core reagents, primers and probes (Applied Biosystems, Foster City, CA, U.S.A.) as previously published.³³ Sequences of primers and probes used in this study were as follows: CCL20 (Hs00171125_m1), CXCL1 (Hs00236937_m1), CXCL3 (Hs00171061_m1), CXCL5 (Hs00171085_m1), CXCL6 (Hs00237017_m1), CXCL8 (Hs00174103_m1), S100A7 (HS00161488_m1) and DEFB4 (Hs00175474_m1). The data were analysed and samples quantified by software provided with Applied Biosystems PRISM 7700 (Sequence Detection Systems, ver. 1.7). Data were normalized to HARP housekeeping gene.

Gene array

RNA was extracted using the RNeasy Mini Kit (Qiagen), and DNA was removed with on-column DNAse digestion using RNAse-free DNAse Set (Qiagen), and used for either RT-PCR or gene array. For each Affymetrix genechip, 4 µg total RNA was reverse transcribed, amplified, and labelled as described previously using BioArray High Yield RNA Transcription Labeling Kit (Enzo Biochem Inc., Farmingdale, NY, U.S.A.).³⁴ Fifteen micrograms of the biotinylated cRNA was then hybridized to Affymetrix Human Genome U133A 2.0 Array (14 500 probe sets) (Affymetrix, Santa Clara, CA, U.S.A.). The chips were washed, stained with streptavidin-phycoerythin, and scanned with an Hewlett-Packard HP GeneArray Scanner (Hewlett-Packard, Palo Alto, CA, U.S.A.).

Preprocessing and statistical analysis was conducted in R (http://www.rproject.org/). GeneChip CEL files were scrutinized for spatial artifacts using Harshlight package (http://asterion.rockefeller.edu/Harshlight/index2.html).³⁵ Row intensities values (CEL files) were preprocessed to obtained expression values using GCRMA algorithm.

Fluorescence-activated cell sorting analysis

Cells were stained with the antibodies listed in Table 2. Briefly, cells were stained for cell surface molecules for 20 min at 4°C, washed with FACSwash (PBS, 0.1% sodium azide and 2% fetal bovine serum) (BD Biosciences, San Jose, CA, U.S.A.) and resuspended in FACSwash (BD Biosciences). For intracellular cytokine staining assays, cells were activated for 4 h using 25 ng mL⁻¹ phorbol myristate acetate and 2 µg mL⁻¹ ionomycin, in the presence of 10 µg mL⁻¹ brefeldin A (all Sigma Aldrich) at 37°C. Unactivated controls were treated with brefeldin A only. EDTA 2 mmol L⁻¹ (Fisher Scientific) was added for 10 min at 37°C to stop activation. Cells were then incubated in aqua marina live/dead dye (Invitrogen) for 30 min for dead cell discrimination then fixed with 4% paraformaldehyde (BD Biosciences) for 20 min. The cells were washed, blocked in 1:100 mouse serum (BD Biosciences), permeabilized in FACSPerm (BD Biosciences), incubated for 30 min with fluorochrome-conjugated monoclonal antibodies to cell surface molecules and intracellular cytokines, washed, and collected. Samples were acquired by an LSR-II flow cytometer (BD Biosciences) and analysed with FlowJo software (Treestar, Ashland, OR, U.S.A.). Keratinocyte single-cell suspensions were analysed after gating out CD45+ and CD11c+ cells. All incubation steps were done on ice. Appropriate isotype controls were used.

Statistical analysis

Two-tailed paired t-test was used to compare psoriasis nonlesional and lesional RT-PCR data. Two-tailed Student’s t-test was used to analyse RT-PCR data between keratinocyte treatment groups, and between full-thickness skin models. The two-tailed P-values are designated as P < 0.05 (*), P < 0.01 (**) and P < 0.005 (***). For gene array analysis, probe sets with SD > 0.1 and at least one sample with expression > 2 were included. Moderated paired t-test available at limma package from R was performed. The resulting P-value was adjusted for multiple hypotheses testing, controlling the false discovery rate (FDR) using the Benjamini–Hochberg procedure. Genes with FDR < 0.1 and more than 1.5-fold change were considered significant.

Distinct populations of interleukin (IL)-17 and IL-22-producing T cells are present in normal human dermis and peripheral blood

We have previously shown that there are discrete populations of IL-17 and IFN-γ-producing T-helper cells in psoriatic and normal dermis and peripheral blood.¹⁸ As murine Th17 cells can produce both IL-17 and IL-22,¹⁵ we wanted to investigate whether IL-17-producing cells in normal human skin and peripheral blood could also produce IL-22.

Using intracellular cytokine staining and eight-colour flow cytometry, we simultaneously analysed the expression of IFN-γ, IL-17 and IL-22 within live CD3+ CD4+ CD8− T cells. We found that approximately 2% of CD4+ T cells in peripheral blood, and 1% in normal dermis, produced IL-17 – defined as Th17 cells (Fig. 1, second and third columns). A larger population of CD4+ T cells, 16% in the blood and 7% in the dermis, produced IFN-γ but not IL-17 (Fig. 1, third column). These cells were defined as Th1 cells.

Approximately 1.2–1.7% of all CD4+ T cells in peripheral blood produced only IL-22, independent of IFN-γ or IL-17 production (Fig. 1, second and fourth columns). In normal dermis, 1.8% of all CD4+ T cells produced only IL-22, while 0.4% produced both IL-22 and IL-17, and 0.8% produced IL-22 and IFN-γ (Fig. 1, lower panels).

In order to determine the percentage of IL-22-producing cells that also produce IL-17, we gated on IL-22+ IFN-γ+ or IL-22− IFN-γ− cells (Fig. 1, lower right panel) and plotted IL-22 vs. IL-17 expression. We found that while some of the IL-22+ IFN-γ+ cells and IL-22+ IFN-γ− cells produced IL-17 (13% and 21%, respectively), the majority did not (87% and 79%, respectively).

Interleukin (IL)-17 and IL-22 receptors are constitutively expressed by keratinocytes

We assessed expression and localization of receptors for IL-17 (IL-17R) and IL-22 (IL-22R1) on cells in normal human skin using immunohistochemical staining (n = 5). Both IL-17R and IL-22R1 were strongly expressed on cell surfaces of viable keratinocytes throughout the epidermis (Fig. 2a, b; inset is negative control). Keratinocyte surface expression of these receptors was confirmed by flow cytometry on single-cell suspensions of normal epidermis (Fig. 2c, d).

We also noted expression of IL-17R and IL-22R1 on scattered dermal cells. To determine which dermal cells expressed these receptors, we performed double-label immunofluorescence: IL-17R was expressed by some CD11c+ dendritic cells (DCs) in the upper dermis (Fig. 2e) but not on CD3+ T cells (Fig. 2g) or CD163+ macrophages (data not shown). In contrast, IL-22R1 was not coexpressed on CD11c+ DCs (Fig. 2f), CD3+ T cells (Fig. 2h) or CD163+ macrophages (data not shown). However, vimentin+ fibroblasts in the dermis expressed IL-22R1 (data not shown).

Th1 and Th17 cytokines induce different keratinocyte gene signatures

To determine the effects of IL-17, IL-22 and IFN-γ on keratinocyte gene expression, we cultured primary human keratinocytes with IL-17 (200 ng mL⁻¹), IL-22 (200 ng mL⁻¹), IFN-γ (20 ng mL⁻¹) or media alone for 24 h (n = 3). Interferons induce the transcription of large sets of known genes by activating STAT1 and ISGF3γ³⁶ Hence, we assessed the response of keratinocytes to IFN-γ as both a positive control for the induction of expected genes, and to compare the gene sets that are regulated by Th1 vs. Th17 T-cell cytokines. We then performed microarray analysis filtering on genes that had a fold difference of > 1.5 compared with control with P < 0.05.

IFN-γ treatment induced keratinocyte expression of a large number of genes, with ~800 genes induced > 2-fold (Supplementary Table S1; see Supplementary material). The chemokines CXCL9, CXCL10 and CXCL11, all considered key IFN-γ-regulated chemokines, were induced > 7000 fold compared with control keratinocytes (Fig. 3). These three chemokines bind to CXCR3-bearing, activated T cells²⁹ and are thought to be involved in T-cell trafficking to psoriatic dermis and overlying epidermis.³⁷

We observed 28 keratinocyte genes that were regulated by IL-17 but not IL-22 nor IFN-γ, notably β-defensin 2 (DEFB4) and the neutrophil chemoattractants CXCL1, CXCL3, CXCL5, CXCL6 and CXCL8 (Fig. 3 and Supplementary Table S2; see Supplementary material). IL-17 also induced the expression of CCL20, a chemokine that preferentially attracts memory T cells and DCs,^38,39 and is reportedly upregulated in psoriatic skin lesions.⁴⁰

In contrast, the genes modulated by IL-22 but not IL-17 are part of the keratinocyte terminal differentiation pathway: keratin 1, filaggrin and CALML5 that were all downregulated by IL-22 (Fig. 3 and Supplementary Table S3; see Supplementary material).

Genes regulated by both IL-17 and IL-22 included the antimicrobial S100A7 (upregulated), and loricrin (downregulated) (Fig. 3).

Interleukin-17 upregulates the expression of immune cell chemoattractants that are increased in psoriatic lesions

To verify the microarray results, we quantified chemokine gene expression in treated keratinocytes by real-time RT-PCR, and normalized these expression values to the housekeeping gene HARP. We then assessed chemokine gene expression by treated keratinocytes as fold change difference from control keratinocyte values.

We confirmed that CXCL1, CXCL3, CXCL5, CXCL6 and CXCL8 were significantly induced by IL-17 treatment compared with IL-22 and IFN-γ (Fig. 4a). CCL20 gene expression was also found to be highly upregulated in IL-17-treated keratinocytes and was absent in cultures treated with either IL-22 or IFN-γ (Fig. 4a).

To determine the extent to which these IL-17-induced chemokines were upregulated in psoriasis lesions, we performed quantitative RT-PCR on paired samples of lesional and nonlesional psoriatic skin (n = 16). We found that gene expression of all six chemokines was significantly increased in psoriatic lesions (Fig. 4b), paralleling results seen in IL-17-treated keratinocytes.

Neutrophils bearing CXCR1 and CXCR2 are present in psoriatic epidermis

CXCL1, CXCL3, CXCL5, CXCL6 and CXCL8 all activate CXCR2 receptor, and CXCL8 can also activate CXCR1.³⁰ We confirmed by immunohistochemistry that neutrophils found accumulating within subcorneal microabscesses bear both CXCR1 and CXCR2 receptors (Fig. 5). Our data suggest that these cells are present in psoriatic epidermis in response to ELR+ chemokines that are induced by IL-17.

Interleukin-22 alters phenotype and gene expression of full-thickness skin equivalents

To determine the effect of IL-17 and IL-22 on human skin, we treated full-thickness skin equivalents with IL-17 (200 ng mL⁻¹), IL-22 (200 ng mL⁻¹), IFN-γ (20 ng mL⁻¹) or control (n = 3). We found that the IL-22-treated skin equivalents developed acanthosis (Fig. 6c), downward epidermal projections (Fig. 6c) and parakeratotic areas (Fig. 6d) within 4 days, changes that were not observed with IL-17-treated (Fig. 6b) or control (Fig. 6a) equivalents. We then verified by quantitative RT-PCR that the genes we found regulated by IL-17 or IL-22 in cultured keratinocytes were similarly regulated in full-thickness skin equivalents. Indeed, as in cultured keratinocytes, IL-17 induced the expression of CCL20, β-defensin 2 (DEFB4) and CXCL8 in the skin equivalents, while both IL-17 and IL-22 upregulated S100A7 (Fig. 6e).