Journal of Dermatological Science 66 (2012) 64–70

Characterization of retinoic acid-inducible gene-I (RIG-I) expressioncorresponding to viral infection and UVB in human keratinocytes

Kazuyuki Kimura a, Yasushi Matsuzaki a,*, Yohei Nishikawa a, Hideo Kitamura a, Eijiro Akasaka a,Daiki Rokunohe a, Hajime Nakano a, Tadaatsu Imaizumi b, Kei Satoh b, Daisuke Sawamura a

a Department of Dermatology, Hirosaki University Graduate School of Medicine, Hirosaki, Japanb Department of Vascular Biology, Institute of Brain Science, Hirosaki University Graduate School of Medicine, Hirosaki, Japan

A R T I C L E I N F O

Article history:

Received 25 October 2011

Received in revised form 8 February 2012

Accepted 13 February 2012

Keywords:

RIG-I

IRF-1

Poly(I:C)

Keratinocytes

UVB irradiation

S U M M A R Y

Background: Retinoic acid-inducible gene-I (RIG-I) is a cytoplasmic protein that recognizes viral double-

stranded RNA to induce the type I interferon (IFN) response. In human keratinocytes, RIG-I is induced by

IFN-g and tumor necrosis factor-a stimulation, and is abundantly expressed in psoriatic keratinocytes of

the spinous and basal layers.

Objective: This study investigated the effects of extraneous stimuli including viral infection and UVB

exposure on RIG-I expression in human keratinocytes.

Methods: Human skin keratinocytes (HaCaT cells) were stimulated by polyinosinic–polycytidylic acid

(poly(I:C)), which mimics viral infection, and UVB exposure. We assessed the expression of RIG-I and

IFN-regulatory factor (IRF)-1 in HaCaT cells by RT-PCR and Western blot analysis. Moreover, we

investigated the effect of IRF-1 binding site of RIG-I gene promoter on the regulation of RIG-I expression

by luciferase promoter assay and electrophoretic mobility shift assay.

Results: Poly(I:C) induced RIG-I expression, while UVB inhibited basal RIG-I expression and the

poly(I:C)-induced RIG-I overexpression in HaCaT cells. IRF-1, which binds to a regulatory element

located on the RIG-I gene promoter, was required for both inductions of RIG-I expression. IRF-1

expression was enhanced three hours after the poly(I:C) stimulation, consistent with the RIG-I response

to poly(I:C), and thereafter was suppressed. Moreover, UVB exposure promptly decreased IRF-1

expression, resulting in decreased IRF-1 protein binding to the RIG-I promoter, and consequently,

decreased RIG-I expression.

Conclusion: Thus, suppression of RIG-I and IRF-1 expression caused by UVB exposure may partly explain

the inhibition of skin-based immune responses, leading to viral infection and recrudescence.

� 2012 Japanese Society for Investigative Dermatology. Published by Elsevier Ireland Ltd.

All rights reserved.

Contents lists available at SciVerse ScienceDirect

Journal of Dermatological Science

jou r nal h o mep ag e: w ww .e lsev ier . co m / jds

1. Introduction

Survival in an environment containing a plethora of micro-organisms depends on an organism’s ability to establish andmaintain protective mechanisms. As the outermost layer of thehuman body, skin provides the first major barrier against microbialattack and plays a critical role in the body’s immune responses toenvironmental stimuli. Keratinocytes are the major cell type inskin epidermis; they are also immune-component cells due to theirmany immunological functions including production of cytokinesand chemokines.

* Corresponding author at: Department of Dermatology Hirosaki University

Graduate School of Medicine 5 Zaifu-cho, Hirosaki, 036-8562, Japan.

Tel.: +81 172 39 5087; fax: +81 172 37 6060.

E-mail address: ymatsu@cc.hirosaki-u.ac.jp (Y. Matsuzaki).

0923-1811/$36.00 � 2012 Japanese Society for Investigative Dermatology. Published b

doi:10.1016/j.jdermsci.2012.02.006

Viral infection is a common trigger for skin damage, and causesvarious skin diseases. Viral components including viral DNA andRNA are recognized by membrane and cytoplasmic receptors incells [1]. Toll-like receptors (TLRs) are membrane-associatedreceptors on cell surfaces that bind pathogens and set off asignaling cascade to produce antiviral factors including cytokinesand to induce inflammatory and adaptive immune responses [2,3].TLR3 is one of several TLRs expressed in human keratinocytes; itrecognizes viral double-stranded RNA (dsRNA) released byinfected cells, and triggers signaling pathways leading to interfer-on regulatory factor (IRF)-3 and NF-kB activation as part of theinnate immune response [4]. In contrast with TLRs, retinoic acid-inducible gene-I (RIG-I) recognizes viral dsRNA in the cytoplasm,stimulating cytokine production [5]. RIG-I is a member of theDExH/D-box family proteins and is designated as an RNA helicase.It plays various roles in gene expression and cellular functions inresponse to many RNA viruses [6,7]. RIG-I contains an N-terminal

y Elsevier Ireland Ltd. All rights reserved.

K. Kimura et al. / Journal of Dermatological Science 66 (2012) 64–70 65

caspase recruitment domain and a C-terminal helicase domain; theN-terminal domain activates downstream signaling pathways,while the C-terminal region is important for recognizing dsRNA [7].We were the first to show that RIG-I expression in humankeratinocytes is significantly enhanced by interferon (IFN)-g andtumor necrosis factor (TNF)-a stimulation [8]. Furthermore, wedetected RIG-I immunohistochemically in psoriatic keratinocytesof the spinous and basal layers, suggesting that RIG-I mightfunction not only as a RNA helicase but also as a mediator of thecytokine network in inflammatory skin diseases [8].

UVB irradiation is another major external factor contributing toepidermal damage, including skin aging and cancer, and has beenknown to suppress immune responses in the skin [9–11]. Arepresentative event is the recrudescence associated with herpessimplex labialis infection. Herpes simplex virus type 1 (HSV-1)typically infects the orofacial region causing vesicular epidermallesions and then establishing life-long latency in the nervoussystem. Viral reactivation can occur after particular stimuli such asexposure to UV radiation, suggesting that UV exposure suppressesthe local immunity. Most likely temporary, this inhibition issufficient to allow viral replication in the epidermis, although theexact molecular events occurring in the lesions remain unclear. Wehypothesized that UVB modifies RIG-I expression in keratinocytes,leading to local immune suppression and viral reactivation. Thisstudy therefore examined the effects of viral infection and UVBirradiation on RIG-I expression in human keratinocytes.

2. Materials and methods

2.1. Cell culture and stimuli

HaCaT cells are a spontaneously immortalized, nontumorigenichuman skin keratinocyte cell line. For this study, HaCaT cells weremaintained in minimum essential medium supplemented with10% heat-inactivated fetal bovine serum, 2 mg/ml sodium hydro-gen carbonate, 100 mg/ml penicillin, 100 mg/ml streptomycin, and2.5 mg/ml amphotericin B. Cultures were maintained at 37 8C in ahumidified atmosphere of 5% CO2 and 95% air. Once 60% confluent,the HaCaT cells were washed twice with sterile phosphate-buffered saline (PBS) and maintained in the growth medium for12 h prior to addition of polyinosinic–polycytidylic acid (poly(I:C))at various final concentrations (Sigma–Aldrich, St. Louis, USA).

Cells in PBS were irradiated with UVB using two FL20S-E lamps(Toshiba, Tokyo, Japan) that emitted wavelengths of 280–320 nmwith an emission peak at 312.5 nm and had an intensity of500 mW/cm2 of the target area. The irradiation dose was 30 mJ/cm2 (25 cm distance for 1 min). After UVB irradiation, the cellswere cultured in the fresh growth medium. The irradiance of theUVB rays was determined with an UVR-3036/S2 radiometer and aUVB detector (Clinical Supply, Kakamigahara, Japan). For experi-ments involving costimulation by UVB and poly(I:C), irradiatedHaCaT cells were cultured as described in growth mediumcontaining poly(I:C) or vehicle.

2.2. RT-PCR analysis

Total RNA was isolated from cultured cells using the RNAeasytotal RNA isolation kit, as recommended by the manufacturer(Qiagen, Hilden, Germany). Extracted RNA was subjected toreverse transcription using the RNA PCR Kit (AMV) ver. 3.0(Takara, Kyoto, Japan), according to the instructions supplied bythe manufacturer. Primers for RIG-I and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were as follows: RIG-I-F (50-GCATATTGACTGGACGTGGCA-30), RIG-I-R (50-CAGTCATGGCTG-CAGTTCTGTC-30), GAPDH-F (50-CCACCCATGGCAAATTCCATGGCA-30), and GAPDH-R (50-TCTAGACGGCAGGTCAGGTCCACC-30). The

primers for RIG-I and GAPDH were designed to generate fragmentsof 644-bp and 598-bp, respectively. The amplification conditionswere 94 8C for 1 min, followed by 25 cycles of 94 8C for 1 min, 58 8Cfor 1 min, 72 8C for 1 min, and finally, 1 cycle of 72 8C for 10 min.PCR products were analyzed by electrophoresis on 1.5% agarosegels and visualized by ethidium bromide staining.

2.3. Western blot analysis

Whole cell lysates were made from HaCaT cells using RIPA lysisbuffer (1% Nonidet P-40, 0.1% sodium deoxycholate [SDS] in PBS)with 0.57 mM phenyl methyl sulfonyl fluoride (Sigma–Aldrich),1 mM sodium orthovanadate (Sigma–Aldrich), and proteaseinhibitors (Roche Diagnostics, Mannheim, Germany). The con-centrations of the extracted protein were quantified by theBradford method. Equal protein loadings (15 mg) were subjectedto 8% SDS-polyacrylamide gel electrophoresis, and electroblottedto Hybond nitrocellulose membrane (Amersham Biosciences,Chandler, USA). The membranes were incubated in blockingsolution (2% non-fat skim milk in 20 mM Tris [pH 7.6], 137 mMNaCl, and 0.1% Tween 20 [TBS-T]) for 1 h at room temperature,followed by an overnight incubation at 4 8C with rabbit anti-RIG-Iantiserum [12] or with rabbit anti-IRF-1 antibody (Santa CruzBiotechnology, Santa Cruz, USA), diluted 1:10,000 or 1:500,respectively, in blocking solution. The membranes were washedfour times in TBS-T and incubated with horseradish peroxidase-conjugated goat anti-rabbit IgG (1:2000, Amersham Biosciences) inTBS-T for 1 h at room temperature, and the bands were visualizedusing an ECL-Western blotting detection system (AmershamBiosciences). The membranes were reprobed with an antibodyagainst b-actin (Sigma–Aldrich) to verify equal protein loading.Western blots were quantitated by densitometry using ScionImage 4.0.2 analysis software.

2.4. Electrophoretic mobility shift assay (EMSA)

Nuclear fractions were separated from HaCaT cells with NE-PERreagent (Pierce, Rockford, USA) according to the manufacturer’sinstructions. The probe sequences were as follows: IRF-1E, 50-GTTGCACTTTCGATTTTCCCTT-30, mutant IRF-1E 50-GTTGCACTAAC-GACTTTCCCTT-30 (mutated nucleotides are underlined in bold)[13], and consensus IRF-1E, 50-GGAAGCGAAAATGAAATTGAC-30.The double-stranded oligonucleotides were labeled with biotin.The labeled probes were then incubated with nuclear extracts atroom temperature for 30 min. The reaction mixture consisted ofbiotin 30-labeled deoxyoligoncleotides, and 1 mg of nuclear proteinextracts with 10 mM HEPES pH 7.9, 40 mM KCl, 0.4 mM DTT, 4%glycerol, and 0.4 mM EDTA. After incubation, the reaction waselectrophoresed on a 5% nondenaturing polyacrylamide gel with0.5 � TBE (100 V for 45 min). The gel contents were transferred to anylon membrane (Pierce), and biotin-labeled DNA was detectedwith the LightShift chemiluminescent mobility shift assay kit(Pierce). Nuclear extracts were also incubated with a 25-fold molarexcess of wild-type IRF-1E competitor, 25 to 100-fold molar excessof mutant IRF-1E competitor and consensus IRF-1E competitor.

2.5. Promoter plasmid constructs

A 1298-bp DNA fragment spanning from �1141 to +153 (+1;transcriptional initiation site) of the promoter region of the humanRIG-I gene was amplified from the normal human genome using asense primer (RIG-F: 50-ATTCTCGAGTCTTCACAGTGAAAAA-CAAATT-30) containing a HindIII restriction site and an antisenseprimer (RIG-R: 50-AAAGCCAAGCTTCCTCTGCTTGCAGCTAGC-TACGT-30) containing an XhoI restriction site. The cyclingparameters were 94 8C for 4 min, 30 cycles of 94 8C for 1 min,

Fig. 1. UVB suppresses the basal RIG-I expression as well as poly(I:C)-induced RIG-I overexpression in human keratinocytes. (a and b) HaCaT cells were incubated in the

presence or absence of poly(I:C) at the indicated concentrations for 6 h. (c–h) HaCaT cells were stimulated with 104 ng/ml poly(I:C) (c and f), irradiated with 30 mJ/cm2 of UVB

(d and g), or costimulated with 104 ng/ml poly(I:C) and 30 mJ/cm2 of UVB (e and h) for the indicated time after each stimulation. (a, c–e) After stimulation, total RNA was

prepared from the cells and cDNA was synthesized followed by RT-PCR analysis for RIG-I and GAPDH. (b, f–h) Cell lysates were subjected to western blot analysis and probed

with rabbit anti-RIG-I antiserum. b-actin is shown as a loading control. Displayed data are one representative result of three independent experiments in each examination.

RIG-I protein levels quantitated by scanning densitometry and corrected for the levels of b-actin in the same sample are shown relative to that of unstimulated HaCaT cells at

each indicated time. (i) Six hours after each stimulation, the expression of RIG-I proteins was significantly modulated in comparison with unstimulated HaCaT cells. Bars are

mean � SD. **P < 0.01.

K. Kimura et al. / Journal of Dermatological Science 66 (2012) 64–7066

Fig. 2. IRF-1E on the RIG-I promoter is essential for basal and poly(I:C)-induced RIG-

I expression. The p1.3WT (a and b) and p1.3MT promoter constructs (b), mutated

within the IRF-1E of p1.3WT, were cotransfected with pGL4.74[hRluc/TK] into

HaCaT cells. At 18 h after transfection, cells were stimulated with or without 104 ng/

ml poly(I:C), irradiated with 30 mJ/cm2 of UVB, or costimulated with 104 ng/ml

poly(I:C) and 30 mJ/cm2 of UVB. After incubation for 6 h, HaCaT cells were lysed and

assayed for luciferase activity. Promoter activities were normalized against

pGL4.74[hRluc/TK] activities. The results, expressed as relative luciferase

activity, are presented as mean � SD of three independent experiments, each

performed in triplicate. Results are expressed as a ratio to luciferase activity of p1.3WT

in unstimulated cells. *P < 0.05, **P < 0.01.

K. Kimura et al. / Journal of Dermatological Science 66 (2012) 64–70 67

56 8C for 1 min, and 72 8C for 2 min; with a final extension of72 8C for 10 min. The PCR products were cloned into the XhoIand HindIII sites of pGL3-basic vector (Promega, Madison,USA) to generate the p1.3WT construct. A putative IRF-1E inthe RIG-I promoter is ‘‘ACTTTCGATTTTCC’’ (between �17 and �4)[13]. Mutagenesis of the IRF-1E spanning from �17 to �4 inp1.3WT construct was performed by site-directed mutagenesis,as described previously [14], using sense (RIG-F, IRF-1mtF:50-GTTGCACTAACGACTTTCCTT-30) and antisense primers (RIG-R, IRF-1mtR: 50-AAGGGAAAGTCGTTAGTGCAAC-30) (mutatednucleotides are underlined in bold). The final construct, desig-nated as p1.3MT, was sequenced along with p1.3WT to confirmthe correct nucleotide sequences.

2.6. Transient transfection and luciferase assay

HaCaT cells were seeded on 6-well Plates 1 day beforetransfection, and the p1.3WT plasmid (1 mg) was introduced intocells using TransIT transfection reagent (Mirus, Madison, USA).pGL4.74[hRluc/TK] luciferase plasmid (Promega) was used tocontrol transfection efficiency. Cells were maintained for 18 h, andthen twice washed with PBS 6 h before stimulation. At 24 h aftertransfection, cells were stimulated as described above. Luciferasereporter assays were performed with a Dual-Luciferase ReporterAssay System (Promega) according to the instructions provided bythe manufacturer.

2.7. Statistical analysis

Data are presented as mean � SD. The Student’s t-test was usedto assess the significance of independent experiments. The criterionP < 0.05 was used to determine statistical significance.

3. Results

3.1. Modulation of RIG-I expression by poly(I:C) and UVB in human

keratinocytes

We previously showed that HaCaT keratinocytes produce RIG-Iprotein in response to IFN-g and TNF-a stimulation, in a dose-dependent manner [8]. To evaluate whether human keratinocytesact similarly in the presence of viral infection, we treated cells withpoly(I:C) for 6 h to simulate viral infection and then measured theexpression of RIG-I. Poly(I:C) dramatically increased RIG-I mRNAlevels in the HaCaT cells at 103 and 104 ng/ml (Fig. 1a). Similarly,RIG-I protein was induced in a dose-dependent (10–104 ng/ml)manner (Fig. 1b).

Next, the RIG-I expression pattern was subjected to a time-course analysis (Fig. 1c). Expression of RIG-I mRNA was induced 3 hafter poly(I:C) treatment (104 ng/ml) and reached a maximum at6 h. In contrast to RIG-I, UVB has been implicated as animmunosuppressor in skin. To elucidate the effect of UVBirradiation on RIG-I expression in human keratinocytes, weexposed HaCaT cells to 30 mJ/cm2 UVB. As shown in Fig. 1d, thesteady-state level of RIG-I mRNA decreased at 3 h after UVBirradiation. The HaCaT cells were then treated simultaneously withUVB and poly(I:C) (Fig. 1e). In this case, UVB irradiation markedlyinhibited the effect of poly(I:C) on the expression of RIG-I mRNA,with maximal suppression occurring at 3 h.

RIG-I protein levels increased to a peak level at 12 h followingtreatment with poly(I:C), with the maximum expression approxi-mately 3-fold higher than that in unstimulated HaCaT cells(Fig. 1f). UVB stimulation suppressed the expression of RIG-Iprotein with or without poly(I:C) (Fig. 1g and h). At 6 h after eachtype of stimulation, RIG-I protein expression was significantlymodulated in comparison with unstimulated HaCaT cells (Fig. 1i).

3.2. Involvement of IRF-1E in UVB inhibition of poly(I:C)-induced RIG-

I promoter activity

Next, to determine the effect of UVB on RIG-I transcriptionalactivity, we transiently transfected the promoter-reporter geneconstruct p1.3WT, containing a 1.3-kb RIG-I gene promoter region,into HaCaT cells, and then stimulated the cells with poly(I:C) andUVB. Treatment with poly(I:C) alone significantly increased theRIG-I promoter activity up to about 4.5-fold (Fig. 2a). In contrast,costimulation with poly(I:C) and UVB dramatically decreased the

Fig. 3. UVB inhibits the basal and poly(I:C)-induced IRF-1 expression. HaCaT cells

were stimulated with 104 ng/ml poly(I:C) (a), irradiated with 30 mJ/cm2 of UVB (b), or

costimulated with 104 ng/ml poly(I:C) and 30 mJ/cm2 of UVB (c) for the indicated time

after each stimulation. Cell lysates were subjected to western blot analysis and probed

with IRF-1 antibody. b-actin is shown as a loading control. Displayed data are one

representative result of three independent experiments in each examination. IRF-1

K. Kimura et al. / Journal of Dermatological Science 66 (2012) 64–7068

poly(I:C)-induced transcriptional activity, and UVB irradiationalone decreased the basal RIG-I promoter activity. These findingsindicated that UVB downregulates the transcriptional activity ofthe RIG-I gene promoter.

A recent study suggested that IRF-1, a transcriptional factorinvolved in cytokine expression, regulates basal RIG-1 expressionas well as the IFN-b- and dsRNA-mediated induction of RIG-I [13].To assess the role of the putative IRF-1-binding site (IRF-1E)adjacent to the transcriptional initiation site on the RIG-I gene inpoly(I:C) and UVB responsiveness, discrete 3-bp mutations wereintroduced into the IRF-1E sequence within p1.3WT, designated asp1.3MT. The p1.3WT and p1.3MT plasmids were transfected intoHaCaT cells, followed by stimulation with poly(I:C) and UVB(Fig. 2b). The p1.3MT-expressing cells showed significantlyreduced basal promoter activity in comparison with p1.3WT-transfected cells. In addition, introduction of the 3-bp mutations inIRF-1E potently abolished the observed regulation of promoteractivity by poly(I:C) and UVB. These results suggest that IRF-1Eplays an essential role in the poly(I:C)-induced RIG-I expressionand in the basal transcriptional activity for this immune-defensemolecule.

3.3. UVB inhibits poly(I:C)-induced IRF-1 expression

Although UVB is a potent inducer of cytokine release bykeratinocytes, the production of IRF-1 is conversely suppressed byUVB irradiation [1]. One may therefore speculate that suppressionof the poly(I:C)-induced RIG-I gene expression by UVB occursthrough modulating IRF-1 expression in keratinocytes. To addressthis idea, the effects of poly(I:C) and UVB stimulation on IRF-1expression in human keratinocytes was examined. IRF-1 wasupregulated over 6 h after poly(I:C) treatment (104 ng/ml) (Fig. 3a).Fig. 3b shows that UVB irradiation inhibited the IRF-1 expression,confirming a previous study [15]. On the other hand, costimulationwith poly(I:C) and UVB markedly suppressed the poly(I:C)-inducedIRF-1 expression 3 h after UVB exposure (Fig. 3c). At 3 h after eachstimulation, IRF-1 expression was significantly modulated incomparison with unstimulated HaCaT cells (Fig. 3d). This resultis identical to the RIG-I response to the same stimulations shown inFig. 1i.

To further demonstrate the specificity of IRF-1 in regulatingRIG-I gene expression, we compared the binding affinity of IRF-1for the putative IRF-1E of the RIG-I gene promoter, using nuclearextracts isolated from unstimulated and stimulated HaCaT cells(Fig. 4). The putative IRF-1E oligonucleotide probe formed acomplex with nuclear extracts of unstimulated HaCaT cells, andpoly(I:C) stimulated this binding activity (Fig. 4, lane 2 vs 5). Theband almost completely disappeared following addition of a 100-fold excess of unlabeled oligomer containing a consensus IRF-1E(Fig. 4, lane 6). These results indicated that the putative IRF-1E onthe RIG-I promoter is a precise cis-element for IRF-1 and theseresults are consistent with those documented by Su et al. [13]. Incontrast, UVB irradiation decreased the complex formations (Fig. 4,lanes 3 and 4), consistent with the result of IRF-1 downregulationinduced by UVB.

Having identified that 3-bp substitutions in the IRF-1Edecreased the binding activity of IRF-1, EMSA was performedwith a biotin-labeled probe. In addition to a weak nonspecific band,a labeled DNA/protein complex was noted in electrophoresed gels

protein levels quantitated by scanning densitometry and corrected for the levels of b-

actin in the same sample are shown relative to that of unstimulated HaCaT cells at

each indicated time. (d) At 3 h after stimulation, IRF-1 expression was significantly

changed in comparison with that in unstimulated HaCaT cells. Data are mean � SD.

**P < 0.01.

Fig. 4. UVB decreases IRF-1 binding to the IRF-1E of the RIG-I promoter. HaCaT cells

were stimulated with or without 104 ng/ml poly(I:C), irradiated with 30 mJ/cm2 of

UVB, or costimulated with 104 ng/ml poly(I:C) and 30 mJ/cm2 of UVB. After

incubation for 6 h, nuclear extracts were prepared and EMSA were performed with

biotin-labeled-IRF-1E probes for the RIG-I promoter. For competition, a 100-fold

molar excess of unlabeled consensus IRF-1E oligonucleotide (cold probe) was added

(lane 6). Displayed data are one representative result of three independent

experiments. Specific DNA/protein complexes are indicated by an arrowhead.Fig. 5. Introduction of 3-bp mutations in the IRF-1E of the RIG-I promoter decreases

the binding activity of IRF-1. Nuclear extracts from HaCaT cells stimulated with

poly(I:C) for 6 h were subjected to competition assays using an oligonucleotide

corresponding to the IRF-1E sequence of the RIG-I promoter. Competition assays were

performed with a 25-fold molar excess of unlabeled wild-type IRF-1E (lane 3),

consensus IRF-1E (lane 4), or mutant IRF-1E (lane 5). Displayed data are one

representative result of three independent experiments. Specific DNA/protein

complexes are indicated by anarrowhead and nonspecific (NS)complexes byan arrow.

K. Kimura et al. / Journal of Dermatological Science 66 (2012) 64–70 69

(Fig. 5). Addition of unlabeled probes for IRF-1E and consensus IRF-1E as a competitor abolished this binding in 25-fold excess (Fig. 5,lanes 3 and 4), while the nonspecific complexes were unaffected.On the other hand, the unlabeled mutant IRF-1E probe containingthe 3-bp substitutions did not compete with wild-type probe forbinding to the nuclear protein, and similarly affected thenonspecific complexes (Fig. 5, lane 5). Accordingly, addition ofthe mutant IRF-1E competitor in a 50- or 100-fold molar excessfailed to significantly reduce the specific DNA/protein complexes(data not shown).

4. Discussion

RIG-I was characterized as a dsRNA-binding protein thatfunctions independently of TLR3 to signal IFN production againstRNA virus infection in various cells [7,16,17]. In RIG-I�/� mice,fibroblasts and dendritic cells failed to generate IFN-b upon RNAvirus infection compared to their wild-type counterparts [18]. Wefirst revealed that RIG-I is expressed in human keratinocytes andis significantly enhanced, in a dose-dependent manner, bystimulation with IFN-g and TNF-a [17]. A recent study demon-strated that RIG-I is expressed and regulated after dsRNAstimulation in normal human keratinocytes, and that recognitionof poly(I:C) promotes mainly IFN-b production through a TANK-binding kinase-1- and IRF-3-dependent pathway, and lessactivation of NF-kB-regulated genes [5]. This study showed thatpoly(I:C) is a potent dose- and time-dependent agonist for theproduction of RIG-I in keratinocytes. Moreover, UVB exposuresuppresses basal RIG-I expression as well as poly(I:C)-inducedRIG-I overexpression in human keratinocytes. The contributionsof UVB exposure to epidermal damage and immunosuppressionare best demonstrated by the UVB-mediated inhibition of cellularimmune reactions and aggravation of infectious diseases [9,19].The present results suggested that UVB may inhibit the early

response to viral infection in keratinocytes by suppressing RIG-Iexpression.

IRF-1 was originally identified as a DNA-binding protein of thepositive regulatory domain 1 element within the human IFN-bpromoter [20,21]. In addition, the IRF-1 gene is IFN-inducible, andit in turn induces other IFN-inducible genes, indicating that IRF-1 isa primary mediator of IFN action [22]. The consensus IRF-1E is theregulatory region of a variety of genes regulating diverse immunefunctions [23]. Recent work showed that both the TATA box andthe IRF-1E are required for regulating the basal RIG-I geneexpression in various cells, and that IRF-1 regulates both thisbasal expression and the IFN-b- and dsRNA-mediated induction ofRIG-I [13]. The current study also revealed that 3-bp mutationswithin IRF-1E of the 1.3-kb RIG-I gene promoter region largelyabolished not only the basal promoter activity but also theresponse to poly(I:C) irritation in human keratinocytes. Interest-ingly, IRF-1 expression was enhanced 3 h after poly(I:C) treatment,and thereafter suppressed by 12 h after stimulation. Moreover,UVB exposure promptly decreased the IRF-1 expression, resultingin decreased IRF-1protein binding to the RIG-I promoter andsuppression of RIG-I expression. IRF-1 is also clearly down-regulated by HSV-1 infection and can no longer be induced byadding IFN-g to mature dendritic cells [24]. HSV-1 is retained as alatent form in the trigeminal ganglia, from where it can reactivateand cause a recrudescent lesion in the skin. In addition, viral dsRNAaccumulates in HSV-infected cells, while RIG-I recognizes HSVinfection in fibroblasts and macrophages leading to the inductionof type I IFN [25,26]. Recrudescences of HSV-1 are triggered byvarious stimuli including exposure to sunlight. We thus elucidated

K. Kimura et al. / Journal of Dermatological Science 66 (2012) 64–7070

that UVB irradiation significantly inhibits IRF-I and RIG-I expres-sion in poly(I:C)-stimulated keratinocytes. A reasonable assump-tion from this is HSV-1 recrudescence being caused by relativedysfunction of the immune mechanisms in skin derived from HSV-1- and UVB-induced IRF-1 downregulation.

We reported previously that psoriatic keratinocytes on thespinous and basal layers strongly express RIG-I protein in thecytoplasm, compared with normal keratinocytes of the controlspecimens [8]. Recent study revealed that narrowband UVBinhibits mRNA expression of RIG-I in psoriatic lesional epidermis[27]. Psoriasis is a chronic inflammatory dermatosis characterizedby an aberrant proliferation and differentiation of keratinocytesand infiltrating activated T lymphocytes in the dermis andepidermis. In the psoriatic lesions, Th1 cytokines including IFN-g and TNF-a are predominant in comparison with Th2 cytokinessuch as IL-4 in healthy skin, thus psoriasis is considered aninflammatory disease induced by Th1 lymphocytes [28,29].Furthermore, IRF-1 is strongly induced by IFN-g and absolutelyrequired for the development of the Th1-type immune response[30], whereas IRF-2, a transcriptional repressor of IFN signaling andthereby a proxy IRF-1 antagonist, is strongly expressed in theepidermis of psoriatic skin compared with normal skin. Finally,IRF-2-deficient mice display psoriasis-like skin abnormalities[31,32], indicating that an imbalanced IRF system in the skinmight cause psoriasis. We clarified that RIG-I and IRF-1expressionin keratinocytes was significantly inhibited by UVB irradiation.UVB irradiation, also an effective therapy for psoriasis, might evenredress the IRF system imbalance in psoriatic lesions.

In conclusion, the data presented herein revealed a previouslyunknown mechanism for regulating innate immune function byRIG-I and IRF-1 in human keratinocytes. UVB exposure signifi-cantly inhibits these mechanisms, leading to destruction of thelocal immune system and viral infection or reactivation, such asthe recrudescences of HSV-1. Understanding the role of RIG-I andIRF-1 in the skin may help to develop novel and more effectiveimmunosuppressive treatments, including UV therapy for inflam-matory skin diseases.

Acknowledgments

The authors thank Ms. Yuka Toyomaki, Mrs. Yukiko Tamura,Mrs. Yuriko Takagi, and Ms. Nanako Seitoh for the excellenttechnical assistance.

References

[1] Lee MS, Kim YJ. Pattern-recognition receptor signaling initiated from extra-cellular, membrane, and cytoplasmic space. Mol Cells 2007;23:1–10.

[2] Lemaitre B, Nicolas E, Michaut L, Reichhart JM, Hoffmann JA. The dorsoventralregulatory gene cassette spatzle/Toll/cactus controls the potent antifungalresponse in Drosophila adults. Cell 1996;86:973–83.

[3] Medzhitov R, Preston-Hurlburt P, Janeway Jr CA. A human homologue of theDrosophila Toll protein signals activation of adaptive immunity. Nature1997;388:394–7.

[4] Sankar S, Chan H, Romanow WJ, Li J, Bates RJ. IKK-i signals through IRF3 andNFkappaB to mediate the production of inflammatory cytokines. Cell Signal2006;18:982–93.

[5] Kalali BN, Kollisch G, Mages J, Muller T, Bauer S, Wagner H, et al. Double-stranded RNA induces an antiviral defense status in epidermal keratinocytesthrough TLR3-, PKR-, and MDA5/RIG-I-mediated differential signaling. JImmunol 2008;181:2694–704.

[6] Meylan E, Curran J, Hofmann K, Moradpour D, Binder M, Bartenschlager R, et al.Cardif is an adaptor protein in the RIG-I antiviral pathway and is targeted byhepatitis C virus. Nature 2005;437:1167–72.

[7] Yoneyama M, Kikuchi M, Natsukawa T, Shinobu N, Imaizumi T, Miyagishi M,et al. The RNA helicase RIG-I has an essential function in double-stranded RNA-induced innate antiviral responses. Nat Immunol 2004;5:730–7.

[8] Kitamura H, Matsuzaki Y, Kimura K, Nakano H, Imaizumi T, Satoh K, et al.Cytokine modulation of retinoic acid-inducible gene-I (RIG-I) expression inhuman epidermal keratinocytes. J Dermatol Sci 2007;45:127–34.

[9] Chapman RS, Cooper KD, De Fabo EC, Frederick JE, Gelatt KN, Hammond SP,et al. Solar ultraviolet radiation and the risk of infectious disease: summary of aworkshop. Photochem Photobiol 1995;61:223–47.

[10] Fisher GJ, Datta SC, Talwar HS, Wang ZQ, Varani J, Kang S, et al. Molecular basisof sun-induced premature skin ageing and retinoid antagonism. Nature1996;379:335–9.

[11] Katsambas A, Nicolaidou E. Cutaneous malignant melanoma and sun expo-sure. Recent developments in epidemiology. Arch Dermatol 1996;132:444–50.

[12] Imaizumi T, Aratani S, Nakajima T, Carlson M, Matsumiya T, Tanji K, et al.Retinoic acid-inducible gene-I is induced in endothelial cells by LPS andregulates expression of COX-2. Biochem Biophys Res Commun 2002;292:274–9.

[13] Su ZZ, Sarkar D, Emdad L, Barral PM, Fisher PB. Central role of interferonregulatory factor-1 (IRF-1) in controlling retinoic acid inducible gene-I (RIG-I)expression. J Cell Physiol 2007;213:502–10.

[14] Ho SN, Hunt HD, Horton RM, Pullen JK, Pease LR. Site-directed mutagenesis byoverlap extension using the polymerase chain reaction. Gene 1989;77:51–9.

[15] Aragane Y, Schwarz A, Luger TA, Ariizumi K, Takashima A, Schwarz T. Ultravi-olet light suppresses IFN-gamma-induced IL-7 gene expression in murinekeratinocytes by interfering with IFN regulatory factors. J Immunol 1997;158:5393–9.

[16] Imaizumi T, Hatakeyama M, Yamashita K, Ishikawa A, Yoshida H, Satoh K, et al.Double-stranded RNA induces the synthesis of retinoic acid-inducible gene-Iin vascular endothelial cells. Endothelium 2005;12:133–7.

[17] Sumpter Jr R, Loo YM, Foy E, Li K, Yoneyama M, Fujita T, et al. Regulatingintracellular antiviral defense and permissiveness to hepatitis C virus RNAreplication through a cellular RNA helicase, RIG-I. J Virol 2005;79:2689–99.

[18] Kato H, Takeuchi O, Sato S, Yoneyama M, Yamamoto M, Matsui K, et al.Differential roles of MDA5 and RIG-I helicases in the recognition of RNAviruses. Nature 2006;441:101–5.

[19] Kripke ML. Photoimmunol Photochem Photobiol 1990;52:919–24.[20] Fujita T, Sakakibara J, Sudo Y, Miyamoto M, Kimura Y, Taniguchi T. Evidence for

a nuclear factor(s), IRF-1, mediating induction and silencing properties tohuman IFN-beta gene regulatory elements. EMBO J 1988;7:3397–405.

[21] Miyamoto M, Fujita T, Kimura Y, Maruyama M, Harada H, Sudo Y, et al.Regulated expression of a gene encoding a nuclear factor, IRF-1, that specifi-cally binds to IFN-beta gene regulatory elements. Cell 1988;54:903–13.

[22] Nakanishi G, Fujimoto W, Arata J. IRF-1 expression in normal human epider-mal keratinocytes. Arch Dermatol Res 1997;289:415–20.

[23] Tanaka N, Kawakami T, Taniguchi T. Recognition DNA sequences of Interferonregulatory factor 1 (IRF-1) and IRF-2, regulators of cell growth and theinterferon system. Mol Cell Biol 1993;13:4531–8.

[24] Eisemann J, Muhl-Zurbes P, Steinkasserer A, Kummer M. Infection of maturedendritic cells with herpes simplex virus type 1 interferes with the interferonsignaling pathway. Immunobiology 2007;212:877–86.

[25] Rasmussen SB, Jensen SB, Nielsen C, Quartin E, Kato H, Chen ZJ, et al. Herpessimplex virus infection is sensed by both Toll-like receptors and retinoic acid-inducible gene-like receptors, which synergize to induce type I interferonproduction. J Gen Virol 2009;90:74–8.

[26] Weber F, Wagner V, Rasmussen SB, Hartmann R, Paludan SR. Double-strandedRNA is produced by positive-strand RNA viruses and DNA viruses but not indetectable amounts by negative-strand RNA viruses. J Virol 2006;80:5059–64.

[27] Racz E, Prens EP, Kant M, Florencia E, Jaspers NG, Laman JD, et al. Narrowbandultraviolet B inhibits innate cytosolic double-stranded RNA receptors inpsoriatic skin and keratinocytes. Br J Dermatol 2011;164:838–47.

[28] Lew W, Bowcock AM, Krueger JG. Psoriasis vulgaris: cutaneous lymphoidtissue supports T-cell activation and Type 1 inflammatory gene expression.Trends Immunol 2004;25:295–305.

[29] Piskin G, Sylva-Steenland RM, Bos JD, Teunissen MB. T cells in psoriatic lesionalskin that survive conventional therapy with NB-UVB radiation display reducedIFN-gamma expression. Arch Dermatol Res 2004;295:509–16.

[30] Taki S, Sato T, Ogasawara K, Fukuda T, Sato M, Hida S, et al. Multistageregulation of Th1-type immune responses by the transcription factor IRF-1.Immunity 1997;6:673–9.

[31] Arakura F, Hida S, Ichikawa E, Yajima C, Nakajima S, Saida T, et al. Geneticcontrol directed toward spontaneous IFN-alpha/IFN-beta responses anddownstream IFN-gamma expression influences the pathogenesis of a murinepsoriasis-like skin disease. J Immunol 2007;179:3249–57.

[32] Hida S, Ogasawara K, Sato K, Abe M, Takayanagi H, Yokochi T, et al. CD8(+) Tcell-mediated skin disease in mice lacking IRF-2, the transcriptional attenua-tor of interferon-alpha/beta signaling. Immunity 2000;13:643–55.