Inflammatory Break Down Intrinsic Immunological Tolerance of Human Primary Keratinocytes to Cytosolic DNA

This information is current as Srikanth Chiliveru, Stine H. Rahbek, Simon K. Jensen, Sofie of September 24, 2021. E. Jørgensen, Sara K. Nissen, Stig H. Christiansen, Trine H. Mogensen, Martin R. Jakobsen, Lars Iversen, Claus Johansen and Søren R. Paludan J Immunol published online 31 January 2014

http://www.jimmunol.org/content/early/2014/01/31/jimmun Downloaded from ol.1302120

Supplementary http://www.jimmunol.org/content/suppl/2014/01/31/jimmunol.130212 Material 0.DCSupplemental http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists by guest on September 24, 2021 • Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2014 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published January 31, 2014, doi:10.4049/jimmunol.1302120 The Journal of Immunology

Inflammatory Cytokines Break Down Intrinsic Immunological Tolerance of Human Primary Keratinocytes to Cytosolic DNA

Srikanth Chiliveru,*,†,1 Stine H. Rahbek,*,†,1 Simon K. Jensen,* Sofie E. Jørgensen,*,† Sara K. Nissen,‡ Stig H. Christiansen,* Trine H. Mogensen,†,‡ Martin R. Jakobsen,*,† Lars Iversen,x Claus Johansen,x and Søren R. Paludan*,†

Keratinocytes are involved in protecting the body from infections and environmental challenges, but also in inflammatory conditions like psoriasis. DNA has emerged as a potent stimulator of innate immune responses, but there is largely no information of how keratinocytes respond to cytosolic DNA. In this study, we report that human keratinocytes are tolerant to cytoplasmic DNA. How- ever, if treated with inflammatory cytokines, keratinocytes gained the capacity to respond to DNA through a mechanism antag- Downloaded from onized by the antimicrobial peptide LL37, proposed to be involved in activation and regulation of skin inflammation. The DNA sensor IFN-inducible 16 (IFI16) colocalized with DNA and the signaling molecule stimulator of IFN (STING) in the cytoplasm only in -stimulated cells, correlating with recruitment of the essential kinase TANK-binding kinase 1. Moreover, IFI16 was essential for DNA-driven innate immune responses in keratinocytes. Finally, IFI16 was upregulated in psoriasis skin lesions and localized to the cytoplasm in a subpopulation of cells. Collectively, this work suggests that inflammatory environments in the skin can lead to breakdown of tolerance for DNA in keratinocytes, which could contribute to the development of inflammatory http://www.jimmunol.org/ diseases. The Journal of Immunology, 2014, 192: 000–000.

he skin responds to a variety of environmental, chemical, ogenic development of inflammation (2, 3). Moreover, expression and physical stimuli by expressing a plethora of membrane of type I IFN (IFN-a and -b) and the IFN-producing plasmacytoid T and soluble proinflammatory molecules. Epidermal ker- dendritic cells (pDCs) are increased and activated in early psori- atinocytes play an active role in the initiation and persistence of atic lesions (4). skin inflammation (1). Nevertheless, strong inflammatory reactions DNA is a potent inducer of innate immune responses and has

affecting the skin can be linked to diseases, as observed in in- been proposed to be involved in the pathogenesis of psoriasis (5, 6). by guest on September 24, 2021 flammatory conditions such as erysipelas, atopic dermatitis, and For instance, pDCs can be activated through DNA, hence stimu- markedly in psoriasis. Psoriasis is a common immune-mediated lating TLR9-driven IFN-a production (5). In psoriasis, there is inflammatory disease affecting the skin and joints. There is ample extensive DNA replication in keratinocytes, which is associated evidence of a pathogenic role of both the innate and adaptive with accumulation of dsDNA breaks (7). It has previously been immune system in psoriasis (2). Most notably, psoriasis is char- reported that DNA replication byproducts as a result of uncon- acterized by infiltration of Th17 cells, which stimulate the path- trolled DNA replication end up in the cytoplasm and stimulate DNA-driven immune stimulation (8). In support of a role for cy- toplasmic DNA in psoriasis pathogenesis, one report (6) has *Department of Biomedicine, Aarhus University, Aarhus DK-8000, Denmark; demonstrated that cytoplasmic DNA stimulates inflammasome †Aarhus Research Center for Innate Immunology, Aarhus University, Aarhus DK- activation and IL-1b production in keratinocytes. With respect to 8000, Denmark; ‡Department of Infectious Diseases, Aarhus University Hospital- Skejby, Aarhus DK-8000, Denmark; and xDepartment of Dermatology, Aarhus Uni- the cellular systems sensing intracellular DNA and stimulating versity Hospital, Aarhus DK-8000, Denmark immune responses, several have been identified (9). This includes 1S.C. and S.H.R. contributed equally to this work. IFN-inducible protein 16 (IFI16), DDX41, cyclic GMP-AMP syn- Received for publication August 9, 2013. Accepted for publication January 5, 2014. thase (cGAS), and DNA-dependent protein kinase (10–13). Com- This work was supported by grants from The Danish Medical Research Council (09- mon for these DNA sensors is the need for the molecule stimulator 072636 and 12-124330), The Lundbeck Foundation (R83-A7598 and R100-A9403), of IFN genes (STING) to transduce signals downstream to the tran- The Novo Nordisk Foundation, the Kathrine og Vivo Skovgaards Fond, the Aase og scription machinery (14). Ejnar Danielsens Fond, and the Aarhus University Research Foundation. S.C. is a recipient of a Ph.D. scholarship from the Faculty of Health Sciences, Aarhus Antimicrobial peptides are produced at damaged epithelial University. surfaces, where they prevent microbial invasion of the host by Address correspondence and reprint requests to Dr. Søren R. Paludan, Department directly killing pathogens. A common feature of antimicrobial of Biomedicine, Aarhus University, Wilhelm Meyers Alle´ 4, DK-8000 Aarhus C, peptides is their cationic and amphiphatic structure (15). The major Denmark. E-mail address: [email protected] antimicrobial peptides in human skin are the defensins comprising The online version of this article contains supplemental material. multiple cysteine-rich b-sheet peptides and the cathelicidins, a Abbreviations used in this article: cGAS, cyclic GMP-AMP synthase; DAMP, danger- a associated molecular pattern; IFI16, IFN-inducible protein 16; IKK, IkBkinase;IRF, family having only one member with helical structure, called IFN regulatory factor; MDM, monocyte-derived macrophage; pDC, plasmacytoid den- LL37 (16). The knowledge on biological functions of antimicro- dritic cell; RT-qPCR, quantitative RT-PCR; siRNA, small interfering RNA; STING, bial peptides has expanded in recent years beyond direct microbial stimulator of IFN genes; TBK1, TANK-binding kinase 1. killing to also include immune modulation. For instance, LL37 is Copyright Ó 2014 by The American Association of Immunologists, Inc. 0022-1767/14/$16.00 responsible for directing DNA to the endosomal compartment in

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1302120 2 DNA-MEDIATED IMMUNE ACTIVATION IN KERATINOCYTES pDCs, where TLR9-mediated recognition takes place (5). LL37 Genomic DNA isolation and sonication production is elevated in psoriasis skin compared with normal skin Keratinocytes were washed and trypsinized. Genomic DNA was purified and has been proposed to be involved in disease progression. using DNA Isolation kit for cells and tissues (Roche Applied Science). Cells However, this does not fit with the finding that treatments ame- were counted and resuspended in PBS to wash the cells and centrifuged at liorating psoriasis symptoms, like vitamin D analogs, narrow-band 1500 rpm for 10 min. Cellular lysis buffer was added and homogenized to UVB light, and fumaric acid esters, induce expression of LL37 a fine suspension. Proteinase K was added and incubated at 65˚C for 2 h. Further RNase solution was added and incubated at 37˚C for 15 min. (17–19). This indicates that LL37 has properties to dampen in- were precipitated, and the supernatant was separated. The geno- flammation in the skin, and one report (6) has demonstrated that mic DNA was precipitated by using isopropanol. The isolated genomic this antimicrobial peptide decreases activation of the IL-1b–acti- DNA was diluted and sonicated into small fragments using Sonifier B-12 vating inflammasome in keratinocytes. However, there is no (Branson Sonic Power, Dietzenbach, Germany). Sonication produced a concoction of random small fragments of DNA of varied lengths. knowledge on whether LL37 is able to control the responses stimulating antimicrobial and inflammatory expression in Small interfering RNA transfections keratinocytes. In this work, we report that keratinocytes in contrast to mac- Primary keratinocytes were seeded 24 h prior to transfection. Cell culture medium was changed to keratinocyte basal medium before the transfection. rophages are largely tolerant to cytoplasmic DNA. However, when Small interfering RNAs (siRNAs; Life Technologies; silencing select: IFI16 treated in combination with either of the inflammatory cytokines siRNA catalog number s7138; control siRNA catalog number 4390846) and TNF-a or IL-1b, the keratinocytes displayed a potent response to Lipofectamine 2000 (Invitrogen, Carlsbad, CA) were incubated with ker- synthetic and purified human DNA. Interestingly, we found that atinocyte basal medium for 20 min to allow complex formation. The siRNA–Lipofectamine complexes were added to the cells and incubated Downloaded from LL37 modulated the DNA-driven innate responses in cytokine- for 6 h before the medium was changed to keratinocyte serum-free me- treated keratinocytes. This correlated with mobilization of IFI16 dium. The cells were allowed to silence IFI16 for 48 h before verifying to the areas containing DNA in the cytoplasm and recruitment of knockdown efficiency by quantitative RT-PCR (RT-qPCR) and further STING and TANK binding kinase 1 (TBK1). Finally, we dem- analysis. onstrate that IFI16 is upregulated in psoriasis skin and can be RNA isolation found in the cytoplasm in a subpopulation of keratinocytes in

psoriatic lesions. In summary, we suggest that an inflammatory Keratinocytes were washed with ice-cold PBS. RNA was purified with http://www.jimmunol.org/ environment in the skin can render keratinocytes sensitive to cy- a High Pure RNA Isolation kit (Roche Applied Science). Cells were resuspended in PBS, and Roche RNA lysis buffer was added to the cells. toplasmic DNA, which could initiate an uncontrolled inflamma- The lysates were transferred to high-pure filter tube assembly and centri- tory response eventually promoting the pathological Th17 immune fuged at 8000 3 g for 15 s. The subsequent filtrate was subjected to DNase response. treatment for 15 min. The DNase-treated samples were washed by a series of buffers and centrifuged at 8000 3 g for 15 s in each step. An extra centrifugation at 13,000 3 g for a 2-min step was performed to remove the Materials and Methods residual buffer. Finally, 50–100 ml elution buffer was added to the filter tube to elute high-pure RNA and stored at 280˚C until further use. Cells

Normal adult human keratinocytes were obtained by trypsinization of skin RNA analysis by guest on September 24, 2021 samples from patients undergoing plastic surgery as previously described RNA for human CCL20, CXCL10, IFN-b, IFI16, and b-actin was analyzed (20). Second-passage keratinocytes were grown in serum-free keratinocyte using SYBR Green-based expression analysis with Brilliant II QRT-PCR 1- medium supplemented with human recombinant epidermal growth factor, step kit (Agilent Technologies) and the following primers: CCL20 forward, bovine pituitary extract, and gentamicin (all from Life Technologies/ 9 9 9 Invitrogen Carlsbad, CA). Cells were grown at 37˚C and 5% CO in 5 -TACTCCACCTCTGCGGCGAATCAGAA-3 and reverse 5 -GTGAA- 2 ACCTCCAACCCCAGCAAGGTT-39;IFI16:forward,59-TAGGCCC- a humidified incubator. We have previously reported that this protocol 9 9 routinely results in very pure keratinocyte cultures with .95% keratin 14– AGCTGAGAGCCATCC-3 and reverse, 5 -TGAGGTCACTCTGGGCA- CTGTCTT-39 (both from DNA Technology); and CXCL10 Qiagen positive cells (21). Fresh medium was added every second day. By the time QuantiTect primer assay Hs_CXCL10_1_SG, and b-actin Qiagen Quanti Tect the cells achieved 60–80% confluency, the medium was changed to growth primer assay Hs_ACTB_2_SG (Qiagen). RNA expression was normalized to factor–free keratinocyte medium, and the cells were incubated for various b stimulations. To generate monocyte-derived macrophages (MDMs), buffy -actin expression and to expression in the relevant untreated control sample. coats from Aarhus University Hospital blood bank were used to collect IFIT1 and GAPDH quantification analysis of the RNA was performed by PBMCs by Ficoll-Paque (GE Healthcare) gradient centrifugation. Mono- a two-step process involving cDNA synthesis by using M-MLV RT (Invi- 8 trogen) followed by QuantiFast SYBR Green Kit (Qiagen) with IFIT1 for- cytes were purified by plastic adherence using 1 3 10 PBMCs seeded in ward primer 59-CCTCCTTGGGTTCGTCTACA-39 and reverse primer 59- 6-cm UpCell plates (Nunc, Thermo-Scientific) precoated with poly-L-ly- GGCTGATATCTGGGTGCCTA-39 and GAPDH forward primer 59-TCT- sine (0.01% w/v; Cultrex) and allowed to stabilize overnight in IMDM m TTTGCGTCGCCAGCCGAG-39 and reverse primer 59-ACCAGGCGCCC- supplemented with 10% FCS, 600 g/ml glutamine, 200 IU/ml penicillin, 9 and 100 mg/ml streptomycin. The following day, unattached cells were AATACGACCA-3 (DNA Technology). For detection of IFI16 mRNA in the washed away, and adherent cells were differentiated into MDMs by cul- psoriasis biopsies, TaqMan primer-probe sets from Life Technologies were turing for an additional 5 d in culture media: IMDM supplemented with used (catalog number Hs00194261_m1 for IFI16 and catalog number + Hs99999902_m1 for RPLP0 used for normalization). 10% (v/v) pooled AB human sera (Invitrogen), 600 mg/ml glutamine, 200 IU/ml penicillin, 100 mg/ml streptomycin (Life Technologies), and ELISA 20 ng/ml M-CSF (Sigma-Aldrich). Human CXCL10 protein was measured on culture supernatants using the Reagents Duoset ELISA Development System (R&D systems) and following the instructions of the manufacturer. Cells were transfected with oligo dsDNA and FITC-dsDNA (2 mg/ml) synthesized chemically (DNA Technology, Risskov, Denmark) and stim- Western blotting ulated with either TNF-a (50 ng/ml) or IL-1b (50 ng/ml), both from R&D Systems (Abingdon, U.K.). The transfection was done by Lipofectamine Cytosolic and nuclear fractions of keratinocytes were separated and isolated 2000 (Invitrogen) and LL37 (Innovagen, Lund, Sweden). ODN2216 and using the protocol previously described (22). For SDS-PAGE, 20 mg protein ODN4084F were obtained from InvivoGen. In some experiments, kerati- was loaded per sample and run on Bio-Rad premade Criterion gels (Bio- nocytes were pretreated with an endosomal acidification inhibitor Bafilo- Rad) and then transferred to polyvinylidene difluoride membranes. For mycin A1 (100 nM) and TBK1/IkB kinase (IKK) ε Inhibitor BX-795 Western blotting, the following specific Abs were used: IFI16 (1G7; Santa (1 mM) (both from Invivogen, San Diego, CA), IΚΚb inhibitor Manumy- Cruz Biotechnology, Santa Cruz, CA), cGAS (C6orf150; Sigma-Aldrich), cin A (10 mM), and a p38 MAPK inhibitor SB202190 (20 mM) (both from RCC1 (sc-1162; Santa Cruz Biotechnology), and b-actin (AC-15 HRP; Sigma-Aldrich) were also employed. Abcam). The Journal of Immunology 3

Confocal microscopy we examined the expression of chemokines in response to TNF-a Cells were fixed at the indicated time point posttreatment, permeabilized and LL37. Surprisingly, DNA delivered into the cytoplasm of with methanol for 5 min at 220˚C, labeled with Abs against IFI16 (sc-8023; keratinocytes did not evoke significant expression of CCL20 or Santa Cruz Biotechnology), STING (Imgenex IMG-6485A), or TBK1 (sc- CXCL10 (Fig. 1B, 1C), whereas TNF-a stimulation did lead to 9910; Santa Cruz Biotechnology). Images were acquired on a Zeiss LSM expression of the two chemokines. Expression of IFN-b was also 3 710 confocal microscope using a 63 1.4 oil-immersion objective (Zeiss). not induced by DNA, and the highly induced IFN-stimulated gene Image processing was performed using Zen 2010 (Zeiss) and ImageJ (National Institutes of Health). All images are representative of at least two IFIT1 was induced only marginally (Supplemental Fig. 1A, 1B). or three independent experiments. Treatment with LL37 alone did modestly stimulate expression of CCL20 and CXCL10. To examine whether the observed lack of Immunofluorescence analysis response to DNA in keratinocytes was due to the use of a too low Four-micrometer paraffin-embedded sections of lesional and nonlesional concentration of DNA or harvest of RNA at the wrong time point, psoriatic skin were deparaffinized and rehydrated. Ag retrieval was per- we performed dose-response and kinetics experiments and also formed by boiling in TEG buffer (pH 9) in a microwave. Blockage of the nonspecific Ab binding sites was achieved by incubating the samples for collected supernatants after 24 h of treatment to allow accumu- 30 min with Image-iT FX Signal Enhancer (Invitrogen). Sections were then lation of chemokine in the supernatant to occur. DNA stimulation incubated overnight at 4˚C with mouse monoclonal anti-IFI16 (sc-8023; did not evoke a chemokine response at any of the DNA concen- Santa Cruz Biotechnology), diluted in TBS buffer with 5% goat serum. trations tested (Fig. 1D), and up to 24 h of treatment did not lead Secondary staining was obtained incubating with Alexa Fluor 488 Dye to production of CXCL10 mRNA or protein (Fig. 1E, 1F). By (Invitrogen; 1:300) diluted in TBS buffer with 5% goat serum for 1 h. Finally, the samples were washed, and nuclear staining was performed by contrast, primary human macrophages responded to DNA stimu- embedding samples in Prolong Gold antifade reagent with DAPI (Invi- lation with strong induction of CXCL10 at both the RNA and Downloaded from trogen). Samples were evaluated by confocal microscopy. As a negative protein level (Fig. 1D, 1E). In contrast to the modest response of control, sections were incubated with blocking buffer without primary Ab the keratinocytes to DNA stimulation, we found that RNA deliv- and as an isotype control with normal mouse IgG (Santa Cruz Biotech- nology) instead of primary Ab. ery into the cytoplasm of keratinocytes, by use of a virus that stimulates though the RNA-sensing RIG-I pathway (26), led to Biopsies potent induction of CXCL10 (Fig. 1F, 1G).

Biopsies were obtained from nonlesional and lesional psoriatic skin. Despite the lack of response to cytoplasmic DNA in keratino- http://www.jimmunol.org/ Keratome and 4-mm punch biopsies from untreated lesional plaque-type cytes, these cells did express the DNA sensors IFI16 and cGAS psoriatic skin were taken from the center of a chronic plaque from ei- (Fig. 1H), with cGAS localizing exclusively to the cytoplasm and ther the upper or lower extremities. For each patient, biopsies were taken IFI16 localizing mainly to the nucleus but with a distinct and from only one anatomical site, and the biopsies from nonlesional skin were taken at a distance of at least 5 cm from a lesional plaque. For quantitative easily detectable pool of IFI16 localizing to the cytoplasm. For PCR, 4-mm punch biopsies were immediately snap-frozen in liquid ni- AIM2, we were able to detect the mRNA in the keratinocytes, but trogen and stored until further use. For immunofluorescence and immu- did not detect AIM2 protein (data not shown), similar to what has nohistochemical analysis, biopsies were embedded in paraffin. The study been reported previously (6, 27). Thus, DNA alone is a weak in- was conducted according to the Declaration of Helsinki Principles. The regional ethical committee, Region Midtjylland, approved all described ducer of innate immune responses in keratinocytes. by guest on September 24, 2021 studies, and signed informed consent was obtained from each patient a b (permission numbers M-20090102 and M-20110027). DNA synergizes with TNF- and IL-1 to induce in keratinocytes Donor variations Cytosolic DNA has been reported to be an important danger- All in vitro work with primary human keratinocytes was done with cells associated molecular pattern (DAMP) in keratinocytes (6). In a from two to three different donors. Data included in the article are all from series of experiments in which the conclusions drawn were independent recent study, it was reported that skin inflammation induced by of the donor used. synergistic action of IL-17A, IL-22, IL-1a, and TNF-a recapit- ulates some features of psoriasis (28). To address the potential role Statistical analysis of cytosolic DNA and its synergy with proinflammatory cytokines The data are shown as mean 6 SD. The statistical significance was de- in inflamed keratinocytes, we stimulated keratinocytes with termined by two-tailed Student t test or Wilcoxon rank sum test. The p transfected synthetic dsDNA together with TNF-a or IL-1b. In- , values 0.05 were considered to be statistically significant. terestingly, this dual treatment led to very potent induction of the chemokines CCL20 and CXCL10 (Fig. 2A–D) as well as IFN-b Results and IFIT1 (Supplemental Fig. 1A, 1B). Interestingly, the strong Intracellular DNA alone is a weak inducer of innate immune synergistic induction of CXCL10 expression after stimulation responses in keratinocytes with DNA and inflammatory cytokines was observed irrespective To test whether synthetic DNA localizes to the cytoplasm in pri- of whether the cytokine stimulation was given before or after mary human epidermal keratinocytes, we transfected such cells DNA (Supplemental Fig. 1C, 1D), although we observed the with a dsDNA oligonucleotide labeled with FITC. As shown in synergy to be strongest when cytokine stimulation was given after Fig. 1A, synthetic dsDNA localized to the cytoplasm when trans- DNA transfection. The observed chemokine response was inde- fected by either Lipofectamine or LL37, and for both DNA de- pendent of endosomal acidification (Supplemental Fig. 1E, 1F), as livery methods, .95% of the DNA-positive spots localized to the known for the TLR9 pathway (29). Furthermore, a TLR9 agonist cytoplasm. We also observed that the DNA localized to specific induced only limited gene expression, and the response evoked by foci in the cytoplasm, consistent with what has previously been dsDNA in cytokine-treated cells was not affected by cotreatment found in macrophages (10, 23). with the TLR9 antagonist ODN4084F (Supplemental Fig. 1G, To evaluate the cellular response to cytosolic DNA, we measured 1H).Bytheuseofaseriesofwell-described small-molecule expression of the chemokines CCL20 and CXCL10. CCL20 is inhibitors of the signaling pathways through IKK–NF-kB, TBK1– reported to play a key role in recruitment of Th17 CD4+ T cells IFN regulatory factor 3 (IRF3), and MAPK, we found that cytokine/ (24), and CXCL10 is a signature of IFN-stimulated genes, also DNA-driven expression of CCL20 was dependent on the IKK– reported to be involved in skin inflammation (25). For comparison, NF-kB pathway but not the TBK1, IRF3, and MAPK pathways, 4 DNA-MEDIATED IMMUNE ACTIVATION IN KERATINOCYTES Downloaded from http://www.jimmunol.org/ by guest on September 24, 2021

FIGURE 1. Intracellular DNA alone is a weak inducer of innate immune responses in keratinocytes. (A) Human primary keratinocytes were treated with FITC-dsDNA (delivered with Lipofectamine [Lipo] or LL37, 2 mg/ml). Two hours posttreatment, the cells were fixed, stained with DAPI, and analyzed by confocal microscopy. Scale bar, 10 mm. (B–D) Human primary keratinocytes were treated with dsDNA (Lipo 2 mg/ml, or the indicated concentrations in C), TNF-a (50 ng/ml), or LL37 (5 mg/ml) as indicated for 6 h. Total RNA was isolated, and mRNA levels of CCL20 and CXCL10 were quantified by RT-qPCR. In (D), RNA was also harvested from MDMs treated with Lipofectamine-delivered DNA for 6 h and analyzed for levels of CXCL10 mRNA. (E) Human primary keratinocytes and MDMs were treated as indicated for 24 h using the same concentrations as in (A). Culture supernatants were harvested, and CXCL10 protein levels were determined by ELISA. (F) Human primary keratinocytes were treated with DNA (2 mg/ml, delivered with Lipofectamine or LL37) or infected with Sendai virus (multiplicity of infection of 1). Total RNA was harvested at the indicated time points posttreatment and analyzed for levels of CXCL10 mRNA. All PCR data in this figure were normalized to b-actin and are presented as mean fold induction of triplicate cultures 6 SD. (G) Schematic illustration of the pathways activated by cytosolic RNA and DNA. (H) Western blotting of cytosolic (Cyt) and nuclear (Nuc) fractions of keratinocytes. Conc, Concentration; NR, normalized ratio; UT, untreated. whereas induction of CXCL10 was dependent on all pathways strong synergy between the stimuli (Fig. 2E, 2F). Altogether, these (Supplemental Fig. 2). results suggest that inflammatory cytokines enable keratinocytes to To investigate whether human genomic DNA could induce anal- respond to cytoplasmic DNA with a potent innate immune response. ogous synergistic responses, we isolated genomic DNA from primary keratinocytes and transfected keratinocytes cultures with this DNA. LL37 selectively inhibits DNA-driven gene expression Keratinocytes transfected with genomic human DNA and stimulated It has been previously reported that LL37 transports extracellular with TNF-a induced significant chemokine responses with a very self-DNA fragments into TLR9-containing endosomal compart- The Journal of Immunology 5

FIGURE 2. DNA synergizes with TNF-a and IL-1b to induce gene expression in keratinocytes. Human primary keratinocytes were trea- m A D ted with synthetic dsDNA (Lipofectamine [Lipo], 2 g/ml) ( – ), Downloaded from human keratinocyte genomic DNA (gDNA; Lipo, 2 mg/ml) (E, F), TNF-a (50 ng/ml) (A, B, E, F), or IL-1b (10 ng/ml) (C, D) as indicated for 6 h. Total RNA was isolated, and mRNA levels of CCL20 and CXCL10 were quantified by RT-qPCR. The data were normalized to b-actin and are presented as mean fold induction of triplicate cultures 6 SD relative to untreated (UT). Similar results were obtained in at least three independent experiments. *p , 0.05. NR, Normalized ratio. http://www.jimmunol.org/ by guest on September 24, 2021

ments of pDCs, in turn inducing type I IFN production (5). To test responses induced by the combined treatment of Lipofectamine- the effect of LL37-mediated DNA delivery into the cytosol on delivered DNA and TNF-a were found to be significantly inhibited chemokine responses, we used the primary epidermal keratino- by LL37 (Fig. 3C, 3D). To assess the LL37 activity on self-DNA, cytes and compared with the response evoked by DNA delivered genomic DNA was used as stimulus. Similar to what was found with Lipofectamine. In contrast to what was observed after when stimulating with synthetic DNA, LL37 inhibited chemokine DNA delivery with Lipofectamine, cytokine-treated keratino- expression induced by combined stimulation with Lipofectamine- cytes transfected with synthetic DNA using LL37 as delivery delivered genomic DNA and TNF-a (Fig. 3E, 3F). The ability of agent induced only marginal expression of CCL20 and CXCL10 LL37 to inhibit DNA-driven gene expression was not dependent on (Fig. 3A, 3B). the antimicrobial peptide being given at the same time as the DNA, LL37 couples with DNA and converts it into a potent DAMP, because LL37 treatment 4 h post–DNA stimulation still led to a trigger of type I IFN in pDCs (5). A recent study also reported strong inhibition of the cellular response (Supplemental Fig. 3). We LL37-mediated DNA delivery to be able to activate type I IFN also investigated the effect of LL37 on the modest gene induction responses in monocytes (30). On the contrary, Dombrowski et al. stimulated by TNF-a alone to exclude the cytokine being the target of (6) suggested an anti-inflammatory function of LL37 by interfer- LL37. As expected, LL37 did not modulate TNF-a–induced gene ing with DNA-triggered inflammasome activation. Based on these expression, hence demonstrating DNA or DNA-stimulated activities to data, we were interested in evaluating whether the ability of LL37 be the target of LL37 (Fig. 3G, 3H). to inhibit DNA-driven gene expression in keratinocytes was de- pendent on the antimicrobial peptide being the DNA transfection Assembly of the DNA-activated signalsome in keratinocytes is agent. To address this, we transfected DNA into keratinocytes supported by cytokine treatment using Lipofectamine and stimulated in parallel with LL37, which The chemokine expression data presented above suggested that had not been preincubated with DNA. Interestingly, chemokine stimulation with inflammatory cytokines endowed keratinocytes 6 DNA-MEDIATED IMMUNE ACTIVATION IN KERATINOCYTES Downloaded from

FIGURE 3. LL37 selectively inhibits DNA-driven gene expression. Human primary keratinocytes were treated with synthetic dsDNA (delivered with Lipofectamine [Lipo] or LL37, 2 mg/ml) (A–D), hu- man keratinocyte genomic DNA (gDNA; Lipo, 2 mg/ml), LL37 (5 mg/ ml), or TNF-a (50 ng/ml) (E, F), as indicated for 6 h. Total RNA was http://www.jimmunol.org/ isolated, and mRNA levels of CCL20 and CXCL10 were quantified by RT-qPCR. The data were normalized to b-actin and are presented as mean fold induction of triplicate cultures 6 SD relative to untreated (UT). Similar results were obtained in at least three independent experiments. *p , 0.05. NR, Normalized ratio. by guest on September 24, 2021

with the ability to respond to intracellular DNA and that LL37 with cytokine did we observe mobilization of IFI16 to the DNA inhibited this response. To get molecular information on what may and colocalization among DNA, IFI16, STING, and TBK1, in- be underlying these observations, we fixed keratinocytes treated dicative of assembly of a functional signalsome (Fig. 4A, Sup- with FITC-dsDNA (Lipofectamine transfection) and TNF-a for 2 plemental Fig. 4). At 6 h posttreatment, the DNA foci in the or 6 h and stained with Abs specific for IFI16, STING, and TBK1. cytoplasm in the cytokine-treated cells remained positive for IFI16 In the untreated cells, STING was broadly distributed in the cy- and STING (Fig. 4B). Moreover, abundant colocalization between toplasm, and IFI16 was primarily found in the nucleus (Fig. 4A), STING and TBK1 was observed, often in foci not containing consistent with the data presented in Fig. 1H. At 2 h posttreatment, DNA or IFI16. If the Lipofectamine-delivered DNA was given to STING was found to be mobilized to specific foci colocalizing cells also receiving LL37, colocalization between DNA and IFI16 with DNA in the cytoplasm (Fig. 4A). This was strongly stimu- in the cytoplasm was still observed in TNF-a–treated cells (Fig. lated by TNF-a treatment. However, only in keratinocytes treated 4B). However, we found significantly less colocalization between The Journal of Immunology 7

FIGURE 4. Assembly of the DNA-activated signalsome in kerati- nocytes is supported by cytokine treatment. (A) Human primary ker- Downloaded from atinocytes were treated with FITC-dsDNA (delivered with Lipofectamine [Lipo], 2 mg/ml), LL37 (5 mg/ml), or TNF-a (50 ng/ml), as indicated. The cells were fixed after 2 h, stained for IFI16, STING, and TBK1, and analyzed by confocal microscopy. White box indicates area displayed in zoom column. Scale bar, 10 mm. (B) The cells were treated as in (A), but fixed after 6 h of treatment for subsequent staining and analysis. Similar results were obtained in at least three independent experiments. (C)Ker- http://www.jimmunol.org/ atinocytes were left untreated (UT) or stimulated with 50 ng/ml of TNF-a or IL-1b for 2 h. Cytoplasmic extracts were analyzed for IFI16 and b-actin by Western blotting. by guest on September 24, 2021

DNA and STING and between STING and TBK1 (Fig 4B). Given ciation of DNA with IFI16 and formation of STING foci, we were the correlation between IFI16–DNA association and DNA-driven interested in evaluating whether IFI16 was essential for DNA- gene induction (Figs. 2, 4A, Supplemental Fig. 4), we were also stimulated gene expression in cytokine-treated cells. For this interested in evaluating how cytokine treatment may allow the purpose, we performed siRNA-mediated knockdown of IFI16 association between IFI16 and DNA. To address this, we treated expression in primary human keratinocytes and were able to reduce keratinocytes with TNF-a and IL-1b for 2 h and isolated cyto- levels of IFI16 mRNA by ∼70% (Fig. 5A). Importantly, the strong plasmic extracts. Interestingly, cytosolic extracts from cells stim- DNA-mediated induction of CCL20 and CXCL10 in TNF-a– ulated with IL-1b, and to a lesser extent also TNF-a, contained treated cells was significantly reduced in cells with knockdown of more IFI16 that extracts from untreated keratinocytes (Fig. 4C). IFI16 (Fig. 5B, 5C). These data demonstrate an essential role for Collectively, these data suggest that treatment of keratinocytes IFI16 in DNA-driven innate immune responses in human primary with inflammatory cytokines induces translocation of a small pool keratinocytes. of IFI16 into the cytoplasm, which enables the cells to assemble signaling complexes including IFI16, STING, and TBK1 and IFI16 exhibits subcellular localization in psoriasis lesions suggests that LL37 inhibits assembly of a functional STING sig- similar to what is observed after DNA transfection of nalsome. cytokine-treated keratinocytes IFI16 is essential for DNA-driven innate immune responses in With the aim to examine the potential relevance of the findings cytokine-treated keratinocytes. Given the correlation between asso- described above in human skin diseases, we isolated skin biopsies 8 DNA-MEDIATED IMMUNE ACTIVATION IN KERATINOCYTES

FIGURE 5. DNA-driven gene expression in cyto- kine-treated keratinocytes is dependent on IFI16. (A) Human primary keratinocytes were transfected with control (Ctrl) and IFI16-specific siRNA. Total RNA was harvested 48 h later, and levels of IFI16 mRNA were measured by RT-qPCR. (B and C) Keratinocytes treated with siRNA for 48 h were stimulated with synthetic dsDNA (Lipofectamine [Lipo], 2 mg/ml) or TNF-a (50 ng/ml) as indicated for 6 h. Total RNAwas isolated, and mRNA levels of CCL20 and CXCL10 were quantified by RT-qPCR. The data were normal- ized to b-actin and are presented as mean fold in- duction of triplicate cultures 6 SD. Similar results were obtained in at least three independent experi- ments. *p , 0.05. NR, Normalized ratio. from nonlesional and lesional psoriasis skin and stained tissue due to the protocol used, including deparaffinization and rehy- sections for IFI16 and DNA. As expected, and in contrast to the drated of paraffin-embedded sections. In the lesional psoriasis nonlesional skin, the lesional psoriasis skin contained a thickened skin, IFI16 was upregulated at both the mRNA and protein level cell layer with nucleated cells in the epidermis (Fig. 6A). IFI16 was (Fig. 6A, 6B). More importantly, in between 5 and 8% of the cells, Downloaded from expressed at low levels in the nonlesional skin and localized pri- IFI16 could be found in the cytoplasm, where it was observed marily to specific areas resembling nucleoli in the nuclei in the to either distribute throughout the cytoplasm or to localize to keratinocytes (Fig. 6A). However, because we have previously specific perinuclear foci (Fig. 6A, zoom panels). The latter pattern observed that extensive sample handling and harsh protocols (like exhibited remarkable similarity to what was observed after DNA e.g., fluorescence in situ hybridization) affect the apparent local- transfection of cytokine-treated keratinocytes in vitro (Fig 4). ization of IFI16 as measured by immunofluorescence (31), it is not Unfortunately, attempts to stain the tissue sections for STING http://www.jimmunol.org/ clear at this stage whether this is a real biological phenomenon showed that the Ab was not compatible with the protocol used. by guest on September 24, 2021

FIGURE 6. IFI16 is upregulated in psoriasis lesions and exhibits a cellular localization pattern similar to keratinocytes receiving dual stimulation with DNA and inflammatory cytokines. (A) Tissue sections of punch biopsies from nonlesional and lesional psoriatic skin were stained for IFI16 and DNA (DAPI) and analyzed by confocal microscopy. Arrows indicate cytosolic IFI16 foci. Scale bar, 100 mm. (B) RNA was isolated from nonlesional (NLS) and lesional psoriatic skin (LS), and levels of IFI16 mRNA were quantified by RT- qPCR. Data shown as means of normalized values relative to nonlesional psoriasis skin 6 SD. n =9.(C) Tissue section of punch biopsies from lesional psoriatic skin stained for DNA (DAPI) and analyzed by confocal microscopy. Arrows indicate DAPI staining (DNA) in the cytosol. Scale bar, 10 mm. *p , 0.05. DIC, Dif- ferential interference contrast; NR, Normalized ratio. The Journal of Immunology 9

Finally, we used the DAPI-stained tissue sections to look for addressing this key issue had been included. This is now an area of the potential presence of DNA in the cytosol. Interestingly, we focus in the laboratory. The concept of IFI16 redistribution in observed DAPI-positive foci outside the nucleus in between response to inflammatory stimuli is in agreement with a previous 3 and 5% of the skin cells, but with no significant difference publication by Landolfo and associates (32), demonstrating that between cells from psoriasis lesions and nonlesional skin exposure of human keratinocytes and human skin explants for (Fig. 6C). high doses of UVB irradiation induced IFI16 localization to the Collectively, the presented data demonstrate that keratinocytes cytoplasm. It has been reported that acetylation of two arginine do not evoke innate immune responses to cytoplasmic DNA residues in the nuclear localization signal of IFI16 inhibits the unless present in an inflammatory environment. The antimicrobial nuclear localization of the protein (33), and in this respect, it is peptide LL37, the expression of which is upregulated by several interesting that inflammatory cytokines induce acetyl transferase treatments known to ameliorate psoriasis symptoms (17–19), activity (34), which could provide a mechanistic explanation for inhibits DNA-driven gene expression and interferes with as- the observed phenomenon. sembly of functional signaling complexes downstream of DNA The antimicrobial peptide LL37 has intrinsic DNA-binding sensing. In cells treated with TNF-a, the DNA sensor IFI16 function and has been reported to be able to deliver DNA into colocalizes with DNA, which correlates with DNA-driven gene endosomes and the cytoplasm (5, 6). Moreover, LL37 is upregu- expression. Finally, IFI16 is upregulated in lesional psoriasis lated in psoriasis skin (5), which could indicate a role for LL37 in skin and can be found to localize to specific spots in the cyto- promotion of disease. Paradoxically, LL37 expression is further plasm, similar to what is observed after stimulation with syn- upregulated in response to some treatments known to ameliorate thetic DNA in vitro. psoriasis (17–19). With focus on chemokines of relevance for type Downloaded from I IFN function and Th17 cell recruitment, we found that DNA delivered with LL37 did not stimulate expression of CCL20 and Discussion CXCL10 in keratinocytes and that the antimicrobial peptide also Skin inflammation can be caused by autoimmune reactions, infec- inhibited gene expression induced by Lipofectamine-delivered tions, or physical injury. DNA has recently emerged as an important DNA,evenwhengiven4hafterDNAtreatment.Thepeptide

stimulator of innate immune responses (9) and is known to evoke did not affect chemokine expression induced by TNF-a. Regard- http://www.jimmunol.org/ immune responses in keratinocytes (6). Given the abundant expo- ing the mechanism of action of LL37, our data demonstrate that sure of keratinocytes to DNA-associated danger signals, it is im- treatment with LL37 has no impact on recruitment of IFI16 to portant that immune activation by DNA is tightly regulated in these cytosolic DNA STING-positive foci, but does inhibit recruitment cells. In this study, we report that immunological reactions of human of TBK1 to the signaling complex. This could indicate that LL37 primary keratinocytes after DNA exposure require parallel stimu- inhibits DNA signaling by acting at a level downstream of IFI16 lation with inflammatory cytokines. This stimulation enables and STING but upstream of TBK1. The idea of LL37 acting keratinocytes to assemble functional signaling complexes acting within the cells is consistent with previous data demonstrating upstream of gene expression. The antimicrobial peptide LL37 LL37 to inhibit inflammasome activation by DNA delivered with interferes with this process, hence specifically inhibiting DNA- lipofection (6). It should be mentioned that the mode of action of by guest on September 24, 2021 driven immune activation. Finally, we provide data on material LL37 may be cell-type specific, because pDCs and monocytes from psoriasis patients demonstrating that IFI16 is upregulated have been reported to respond potently to LL37-delivered DNA, in psoriasis lesions and that IFI16 and DNA under such conditions through the TLR9–MyD88 and STING–TBK1 pathways, respec- can be found in the cytoplasm of keratinocytes in specific peri- tively (5, 30). Collectively, the data could indicate a regulatory nuclear foci, similar to what is seen after DNA stimulation of function for LL37 in keratinocytes by controlling DNA-driven keratinocytes in vitro. amplification of several inflammatory pathways involved in the A key finding of the present project is that human keratinocytes pathogenesis of psoriasis. normally do not respond to cytoplasmic DNA with an innate im- Psoriasis has for many years been considered a possible auto- mune response and that stimulation with inflammatory cytokines immune-driven inflammatory skin disease, but an endogenous enables the cells to induce gene expression in response to this trigger of the inflammatory responses has never been identified (2). DAMP. This indicates that cytokine treatment activates an essential The disease is characterized by hyperproliferation of keratinocytes factor in the DNA sensing pathway acting upstream of STING, and an excessive Th17 response (2). In addition, expression of hence breaking down the intrinsic immunological tolerance of IFN-stimulated genes is elevated in early psoriasis plaques (4). keratinocytes for cytosolic DNA. We found that the DNA sensor The data presented in this work point toward a role for the DNA- IFI16 was primarily expressed in the nucleus in keratinocytes not sensing pathway in evoking some of these responses and could treated with cytokines, and this was not affected by delivery of also indicate that DNA may serve as an autoantigen in psoriasis. DNA into the cytoplasm. By contrast, in keratinocytes treated with We found that IFI16 was upregulated in psoriasis lesions and was TNF-a, IFI16 was found to localize with DNA in the cytoplasm, localized to specific foci in the cytoplasm in a fraction of cells, which correlated with DNA-induced gene expression. Interest- reminiscent of what is observed after immunostimulatory DNA ingly, although DNA treatment stimulated some degree of STING recognition in vitro (10). Interestingly, it has been suggested that redistribution to the well-described punctuate foci (14), even in the aberrant DNA replication intermediates, as may be found in hyper- absence of cytokine treatment, recruitment of the IRF3 kinase proliferating keratinocytes, can trigger innate immune responses (35). TBK1 was seen only in cytokine-treated cells in which IFI16 was Physical injury, which is a classical activator of cutaneous inflam- also recruited to the STING foci. We also observed that cytokine mation in psoriasis and known as the Ko¨bner phenomenon, has also treatment led to accumulation of IFI16 in the cytoplasm. This been reported to induce release of DNA to the cytoplasm in kerati- could indicate that redistribution of a small pool of cellular IFI16 nocytes (6). Thus, it is possible that release of DNA into the cyto- from the nucleus to the cytoplasm is a critical event in stimulation plasm, induced by physical damage, DNA replication byproducts, or of DNA sensitivity in keratinocytes. The authors do acknowledge even infection, could amplify inflammatory reactions in keratinocytes that this conclusion could have been strengthened with higher in which tolerance to DNA has been broken due to low-grade con- quality data on the distribution of IFI16 and also if more data stitutive cytokine expression. Although the focus of this study has 10 DNA-MEDIATED IMMUNE ACTIVATION IN KERATINOCYTES been on cytosolic DNA sensing, it should be noted that the elevated 17. Va¨ha¨vihu, K., M. Ala-Houhala, M. Peric, P. Karisola, H. Kautiainen, T. Hasan, E. Snellman, H. Alenius, J. Schauber, and T. Reunala. 2010. Narrowband ul- expression of IFI16 could also impact on exacerbation of psoriasis by traviolet B treatment improves vitamin D balance and alters antimicrobial other mechanisms. Most notably, nuclear IFI16 has been proposed to peptide expression in skin lesions of psoriasis and atopic dermatitis. Br. J. stimulate inflammasome activation (36) and modulate expression of Dermatol. 163: 321–328. 18. Peric, M., S. Koglin, Y. Dombrowski, K. Gross, E. Bradac, A. Buchau,€ a range of inflammatory genes (37). A. Steinmeyer, U. Zugel,€ T. Ruzicka, and J. Schauber. 2009. Vitamin D analogs Altogether, we report that human keratinocytes have intrinsic differentially control antimicrobial peptide/”alarmin” expression in psoriasis. immunological tolerance to cytoplasmic DNA. However, treatment PLoS ONE 4: e6340. 19. Gambichler, T., F. G. Bechara, N. Scola, S. Rotterdam, P. Altmeyer, and with inflammatory cytokines leads to breakdown of this tolerance. M. Skrygan. 2012. Serum levels of antimicrobial peptides and proteins do not This correlates with cytosolic colocalization between DNA and the correlate with psoriasis severity and are increased after treatment with fumaric acid esters. Arch. Dermatol. Res. 304: 471–474. DNA sensor IFI16 and assembly of a STING-TBK1–containing 20. Kragballe, K., L. Desjarlais, and C. L. Marcelo. 1985. Increased DNA synthesis signaling complex. The antimicrobial peptide LL37 inhibits DNA- of uninvolved psoriatic epidermis is maintained in vitro. Br. J. Dermatol. 112: driven innate immune responses and acts by preventing assembly 263–270. 21. Otkjaer, K., K. Kragballe, C. Johansen, A. T. Funding, H. Just, U. B. Jensen, of a functional signaling complex. Finally, in psoriasis skin lesions, L. G. Sørensen, P. L. Nørby, J. T. Clausen, and L. Iversen. 2007. IL-20 gene IFI16 is upregulated and can be found in the cytoplasm in foci re- expression is induced by IL-1beta through mitogen-activated protein kinase and sembling DNA recognition/signaling complexes. Collectively, these NF-kappaB-dependent mechanisms. J. Invest. Dermatol. 127: 1326–1336. 22. Paludan, S. R., S. Ellermann-Eriksen, V. Kruys, and S. C. Mogensen. 2001. data have revealed the immunological DNA-sensing pathway as Expression of TNF-alpha by herpes simplex virus-infected macrophages is a regulatory switch in keratinocytes and suggest a role for host or regulated by a dual mechanism: transcriptional regulation by NF-kappa B and activating 2/Jun and translational regulation through the AU- microbial DNA as danger signals in skin inflammation. rich region of the 39 untranslated region. J. Immunol. 167: 2202–2208. 23. Horan, K. A., K. Hansen, M. R. Jakobsen, C. K. Holm, S. Søby, L. Unterholzner, Downloaded from M. Thompson, J. A. West, M. B. Iversen, S. B. Rasmussen, et al. 2013. Pro- Acknowledgments teasomal degradation of herpes simplex virus capsids in macrophages releases We thank Kirsten Stadel Petersen for technical assistance. DNA to the cytosol for recognition by DNA sensors. J. Immunol. 190: 2311– 2319. 24. Esplugues, E., S. Huber, N. Gagliani, A. E. Hauser, T. Town, Y. Y. Wan, Disclosures W. O’Connor, Jr., A. Rongvaux, N. Van Rooijen, A. M. Haberman, et al. 2011. The authors have no financial conflicts of interest. Control of TH17 cells occurs in the small intestine. Nature 475: 514–518.

25. Albanesi, C., C. Scarponi, S. Pallotta, R. Daniele, D. Bosisio, S. Madonna, http://www.jimmunol.org/ P. Fortugno, S. Gonzalvo-Feo, J. D. Franssen, M. Parmentier, et al. 2009. Chemerin expression marks early psoriatic skin lesions and correlates with References plasmacytoid dendritic cell recruitment. J. Exp. Med. 206: 249–258. 1. Stamatas, G. N., A. P. Morello, III, and D. A. Mays. 2013. Early inflammatory 26. Kato, H., O. Takeuchi, S. Sato, M. Yoneyama, M. Yamamoto, K. Matsui, processes in the skin. Curr. Mol. Med. 13: 1250–1269. S. Uematsu, A. Jung, T. Kawai, K. J. Ishii, et al. 2006. Differential roles of MDA5 2. Nestle, F. O., D. H. Kaplan, and J. Barker. 2009. Psoriasis. N. Engl. J. Med. 361: and RIG-I helicases in the recognition of RNA viruses. Nature 441: 101–105. 496–509. 27. Reinholz, M., Y. Kawakami, S. Salzer, A. Kreuter, Y. Dombrowski, S. Koglin, 3. Miossec, P., and J. K. Kolls. 2012. Targeting IL-17 and TH17 cells in chronic S. Kresse, T. Ruzicka, and J. Schauber. 2013. HPV16 activates the AIM2 inflammation. Nat. Rev. Drug Discov. 11: 763–776. inflammasome in keratinocytes. Arch. Dermatol. Res. 305: 723–732. 4. Nestle, F. O., C. Conrad, A. Tun-Kyi, B. Homey, M. Gombert, O. Boyman, 28. Guilloteau, K., I. Paris, N. Pedretti, K. Boniface, F. Juchaux, V. Huguier, G. Burg, Y. J. Liu, and M. Gilliet. 2005. Plasmacytoid predendritic cells initiate G. Guillet, F.-X. Bernard, J.-C. Lecron, and F. Morel. 2010. Skin inflammation a psoriasis through interferon-alpha production. J. Exp. Med. 202: 135–143. induced by the synergistic action of IL-17A, IL-22, oncostatin M, IL-1 ,and by guest on September 24, 2021 5. Lande, R., J. Gregorio, V. Facchinetti, B. Chatterjee, Y. H. Wang, B. Homey, TNF-a recapitulates some features of psoriasis. J. Immunol. 184: 5263–5270. W. Cao, Y. H. Wang, B. Su, F. O. Nestle, et al. 2007. Plasmacytoid dendritic cells 29. Ahmad-Nejad, P., H. Ha¨cker, M. Rutz, S. Bauer, R. M. Vabulas, and H. Wagner. sense self-DNA coupled with antimicrobial peptide. Nature 449: 564–569. 2002. Bacterial CpG-DNA and lipopolysaccharides activate Toll-like receptors 6. Dombrowski, Y., M. Peric, S. Koglin, C. Kammerbauer, C. Go¨ss, D. Anz, at distinct cellular compartments. Eur. J. Immunol. 32: 1958–1968. M. Simanski, R. Gla¨ser, J. Harder, V. Hornung, et al. 2011. Cytosolic DNA 30. Chamilos,G.,J.Gregorio,S.Meller,R.Lande,D.P.Kontoyiannis,R.L.Modlin,and triggers inflammasome activation in keratinocytes in psoriatic lesions. Sci. M. Gilliet. 2012. Cytosolic sensing of extracellular self-DNA transported into Transl. Med. 3: 82ra38. monocytes by the antimicrobial peptide LL37. Blood 120: 3699–3707. 7. Kawashima, K., H. Doi, Y. Ito, M. A. Shibata, R. Yoshinaka, and Y. Otsuki. 31. Jakobsen, M. R., R. O. Bak, A. Andersen, R. K. Berg, S. B. Jensen, T. Jin, 2004. Evaluation of cell death and proliferation in psoriatic epidermis. J. Der- A. Laustsen, K. Hansen, L. Ostergaard, K. A. Fitzgerald, et al. 2013. IFI16 matol. Sci. 35: 207–214. senses DNA forms of the retroviral replication cycle and controls HIV-1 repli- 8. Yang, Y. G., T. Lindahl, and D. E. Barnes. 2007. Trex1 exonuclease degrades ssDNA cation. Proc. Natl. Acad. Sci. USA 110: E4571–E4580. to prevent chronic checkpoint activation and autoimmune disease. Cell 131: 873–886. 32. Costa, S., C. Borgogna, M. Mondini, M. De Andrea, P. L. Meroni, E. Berti, 9. Paludan, S. R., and A. G. Bowie. 2013. Immune sensing of DNA. Immunity 38: M. Gariglio, and S. Landolfo. 2011. Redistribution of the nuclear protein IFI16 870–880. into the cytoplasm of ultraviolet B-exposed keratinocytes as a mechanism of 10. Unterholzner, L., S. E. Keating, M. Baran, K. A. Horan, S. B. Jensen, S. Sharma, autoantigen processing. Br. J. Dermatol. 164: 282–290. C. M. Sirois, T. Jin, E. Latz, T. S. Xiao, et al. 2010. IFI16 is an innate immune 33. Li, T., B. A. Diner, J. Chen, and I. M. Cristea. 2012. Acetylation modulates sensor for intracellular DNA. Nat. Immunol. 11: 997–1004. cellular distribution and DNA sensing ability of interferon-inducible protein 11. Zhang, Z. Q., B. Yuan, M. S. Bao, N. Lu, T. Kim, and Y. J. Liu. 2011. The IFI16. Proc. Natl. Acad. Sci. USA 109: 10558–10563. helicase DDX41 senses intracellular DNA mediated by the adaptor STING in 34. Ruiz, P. A., A. Braune, G. Ho¨lzlwimmer, L. Quintanilla-Fend, and D. Haller. dendritic cells. Nat. Immunol. 12: 959–965. 2007. Quercetin inhibits TNF-induced NF-kappaB transcription factor recruit- 12. Sun, L., J. Wu, F. Du, X. Chen, and Z. J. Chen. 2013. Cyclic GMP-AMP syn- ment to proinflammatory gene promoters in murine intestinal epithelial cells. J. thase is a cytosolic DNA sensor that activates the type I interferon pathway. Nutr. 137: 1208–1215. Science 339: 786–791. 35. Stetson, D. B., J. S. Ko, T. Heidmann, and R. Medzhitov. 2008. Trex1 prevents 13. Ferguson, B., D. Mansur, N. Peters, H. Ren, and G. L. Smith. 2012. DNA-PK is cell-intrinsic initiation of autoimmunity. Cell 134: 587–598. a DNA sensor for IRF-3-dependent innate immunity. eLIFE 1: e00047. 36. Li, J., S. Hu, L. Zhou, L. Ye, X. Wang, J. Ho, and W. Ho. 2011. Interferon 14. Ishikawa, H., Z. Ma, and G. N. Barber. 2009. STING regulates intracellular lambda inhibits herpes simplex virus type I infection of human astrocytes and DNA-mediated, type I interferon-dependent innate immunity. Nature 461: 788–792. neurons. Glia 59: 58–67. 15. Zasloff, M. 2002. Antimicrobial peptides of multicellular organisms. Nature 37. Baggetta, R., M. De Andrea, G. R. Gariano, M. Mondini, M. Ritta`,P. Caposio, 415: 389–395. P. Cappello, M. Giovarelli, M. Gariglio, and S. Landolfo. 2010. The interferon- 16. Morizane, S., and R. L. Gallo. 2012. Antimicrobial peptides in the pathogenesis inducible gene IFI16 secretome of endothelial cells drives the early steps of the of psoriasis. J. Dermatol. 39: 225–230. inflammatory response. Eur. J. Immunol. 40: 2182–2189.