sign in sign up
<<
Home , Keloid, Scleroderma, Systemic scleroderma

The Downregulation of microRNA let-7a Contributes to the Excessive Expression of Type I in Systemic and Localized This information is current as of October 1, 2021. Katsunari Makino, Masatoshi Jinnin, Ayaka Hirano, Keitaro Yamane, Mitsuhiko Eto, Takamitsu Kusano, Noritoshi Honda, Ikko Kajihara, Takamitsu Makino, Keisuke Sakai, Shinichi Masuguchi, Satoshi Fukushima and Hironobu Ihn

J Immunol published online 15 March 2013 Downloaded from http://www.jimmunol.org/content/early/2013/03/15/jimmun ol.1200822 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

• Fast Publication! 4 weeks from acceptance to publication

by guest on October 1, 2021 *average

Subscription Information about subscribing to The Journal of 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 © 2013 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published March 15, 2013, doi:10.4049/jimmunol.1200822 The Journal of Immunology

The Downregulation of microRNA let-7a Contributes to the Excessive Expression of Type I Collagen in Systemic and Localized Scleroderma

Katsunari Makino, Masatoshi Jinnin, Ayaka Hirano, Keitaro Yamane, Mitsuhiko Eto, Takamitsu Kusano, Noritoshi Honda, Ikko Kajihara, Takamitsu Makino, Keisuke Sakai, Shinichi Masuguchi, Satoshi Fukushima, and Hironobu Ihn

Systemic and localized scleroderma (SSc and LSc) is characterized by excessive deposition of collagen and tissue fibrosis in the skin. Although they have fundamental common characteristics including , little is known about the exact mechanism that mediates the excessive collagen expression in these disorders. In the current study, we tried to evaluate the possibility that micro- RNAs (miRNAs) play some roles in the pathogenesis of fibrosis seen in these diseases. miRNA expression patterns were evaluated by Downloaded from miRNA array analysis, real-time PCR, and in situ hybridization. The function of miRNAs in dermal fibroblasts was assessed us- ing miRNA inhibitors, precursors, or protectors. In the mouse model of -induced dermal sclerosis, the overexpression of miRNAs was performed by i.p. miRNA injection. We demonstrated let-7a expression was downregulated in SSc and LSc skin both in vivo and in vitro, compared with normal or skin. The inhibition or overexpression of let-7a in human or mouse skin fibroblasts affected the expression of type I collagen or luciferase activity of collagen 39-untranslated region. Also, we found let-7a was detectable and quantitative in the serum and investigated serum let-7a levels in patients with SSc or LSc. let-7a http://www.jimmunol.org/ concentration was significantly decreased in these patients, especially in LSc patients. Moreover, we revealed that the intermittent overexpression of let-7a in the skin by i.p. miRNA injection improved the skin fibrosis induced by bleomycin in mice. Investigation of more detailed mechanisms of miRNA-mediated regulation of collagen expression may lead to new therapeutic approaches against SSc and LSc. The Journal of Immunology, 2013, 190: 000–000.

ystemic scleroderma (SSc) is a disorder In contrast, are also characterized by cutaneous fibrosis that results in fibrosis of the skin and internal organs. that invade adjacent healthy tissue. However, keloid is not an auto- S Localized scleroderma (LSc) also manifests tissue fibrosis immune disease, and it is still controversial whether the collagen by guest on October 1, 2021 limited to the skin and s.c. tissue, occasionally involving the production by keloid fibroblasts is also upregulated (12–15). Thus, muscular tissues beneath the cutaneous lesions (1, 2), but the different factors are likely to mediate the tissue fibrosis in SSc/LSc presence of Raynaud’s phenomenon, acrosclerosis, and involve- and keloid, but the exact mechanism involved in each fibrotic ment of internal organs differentiates SSc from LSc (3). Abnormal condition is still unknown. collagen metabolism (4–8) and autoimmunity (9, 10) are consid- Recently, microRNAs (miRNAs) have attracted attention for its ered to be fundamental common characteristics of SSc and LSc, involvement in the pathogenesis of various human diseases such as and especially, excess collagen production by dermal fibroblasts immunological disorders, , and metabolic disorders (16–18). is thought to be caused by intrinsic activation of TGF-b signaling miRNAs are small noncoding RNAs that bind to complementary in both diseases (5, 11). sequences in the 39-untranslated regions (UTRs) of target mRNAs, resulting in inhibiting their translation into protein. Because .1000

Department of and Plastic , Faculty of Life Sciences, Kuma- of miRNAs have been identified, which corresponds to 1–5% of all moto University, Kumamoto 860-8556, Japan genes in the human genome, miRNAs are thought to be the most Received for publication March 14, 2012. Accepted for publication February 13, abundant class of regulators (19). However, little is known about the 2013. role of miRNAs in the pathogenesis of SSc and LSc. In the current This work was supported in part by a grant for scientific research from the Japanese study, we tried to evaluate the possibility that miRNAs also play some Ministry of Education, Science, Sports, and Culture, by project research on intrac- parts in the collagen metabolism and skin fibrosis of SSc/LSc. table diseases from the Japanese Ministry of Health, Labor, and Welfare, and by a research grant from Shiseido. The sequences presented in this article have been submitted to the Gene Expression Materials and Methods Omnibus microarray data repository (http://www.ncbi.nlm.nih.gov/geo/) under acces- Patient materials sion number GSE43469. Serum samples were obtained from 39 SSc patients in their first visit. These Address correspondence and reprint requests to Dr. Masatoshi Jinnin, Department of Dermatology and , Faculty of Life Sciences, Kumamoto University, patients were grouped according to the classification system proposed by 1-1-1 Honjo, Kumamoto 860-8556, Japan. E-mail address: [email protected] LeRoy et al. (20): 20 patients had diffuse cutaneous SSc (dcSSc) and 19 had limited cutaneous SSc (lcSSc), as described previously (21). Similarly, Abbreviations used in this article: Ct, threshold cycle; dcSSc, diffuse cutaneous 32 patients with LSc were classified into following three subgroups, as ; DM, ; GM, generalized ; lcSSc, lim- ited cutaneous systemic scleroderma; LS, linear scleroderma; LSc, localized sclero- described previously (22, 23): 19 patients with morphea (one or a few derma; miRNA, microRNA; SLE, systemic erythematosus; SSc, systemic circumscribed sclerotic plaque), 8 with linear scleroderma (LS, with linear scleroderma; UTR, untranslated region. distribution), and 5 with generalized morphea (GM, four or more lesions . 3 cm in diameter and involvement of two or more areas of the body out Copyright Ó 2013 by The American Association of Immunologists, Inc. 0022-1767/13/$16.00 of seven areas). When patients had both morphea and LS, they were diag-

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1200822 2 let-7a EXPRESSION IN SSc AND LSc nosed as having GM (22). We also included serum samples from other Cell lysis and immunoblotting rheumatic diseases: 8 patients with systemic (SLE) and 8 with dermatomyositis (DM). Control serum samples were obtained from Human or mouse dermal fibroblasts were washed with PBS twice and lysed 17 healthy age-/sex-matched volunteers. Skin specimens were derived from in lysis buffer (BioSource International). Aliquots of cell lysates were involved skin of 7 SSc, 7 LSc, and 5 keloid patients. These skin samples and separated by electrophoresis on 10% SDS-PAGE and transferred to poly- seven control skins were collected and fixed in formaldehyde immediately vinylidene difluoride membranes, which were blocked in blocking One P after resections. buffer (Nacalai Tesque) for 1 h and incubated overnight at 4˚C with primary Clinical and laboratory data reported in this study were obtained at the Ab for type I collagen–UNLB (Southern Biotechnology Associates). The time of serum sampling. Institutional review board approval and written membranes were washed with TBS and 0.1% TBST, probed with HRP- informed consent were obtained according to the Declaration of Helsinki. conjugated secondary Ab for 1 h, and then washed with TBST again. The detection was performed using ECL system (Thermo Scientific), according Cell cultures to the manufacturer’s recommendations. As a loading control, the same membrane was stripped and reprobed with an Ab against b-actin (Santa Human dermal fibroblasts were obtained by from the affected Cruz Biotechnology). areas of five SSc patients, LSc patients, and healthy donors. All biopsies were performed with institutional review board approval and written in- In situ hybridization formed consent according to the Declaration of Helsinki. Mouse dermal fibroblasts were obtained from the back skin of male BALB/cAJcl mice. All In situ hybridization was performed with 59-locked digoxigenin-labeled animal experimental protocols in this study were approved by the Com- nucleic acid probes complementary to human mature let-7a and scram- mittee on the Animal Research at Kumamoto University. bled negative control (Exiqon) (29, 30). Briefly, human tissues were Primary explant cultures were established in 25-cm2 culture flasks in deparaffinized and deproteinized with protease K for 5 min. Slides were modified Eagle’s medium supplemented with 10% FCS, 2 mM glutamine, then washed in 0.2% glycine in PBS and fixed with 4% paraformaldehyde. and 50 mg/ml gentamicin, as described previously (24). be- Hybridization was performed at 48˚C overnight, followed by blocking with 2% FBS and 2% BSA in PBS and 0.1% Tween 20 for 1 h. The probe– tween the third and sixth subpassages were used for experiments. Downloaded from target complex was detected immunologically by incubating with a digox- RNA extraction and real-time PCR analysis igenin Ab conjugated to alkaline phosphatase acting on the chromogen NBT/ 5-bromo-4-chloro-3-indolyl phosphate (Roche Applied Science) for 48 h. Total RNA from cultured cells was extracted using Isogen (Nippon Gene). Slides were counterstained with nuclear fast red, and examined under a First-strand cDNA was synthesized by the PrimeScript RT reagent Kit light microscope (Olympus BX50). (Takara). Quantitative real-time PCR used primers and templates mixed with the SYBR Premix Ex Taq ΙΙ Kit (Takara). Primer sets for a1(I) col- Injection of AteloGene and miRNA mixtures into mice

lagen and a2(I) collagen were from Takara. DNA was amplified for 40 abdominal cavity http://www.jimmunol.org/ cycles of denaturation for 5 s at 95˚C and annealing for 30 s at 60˚C. The transcript levels were normalized to those of GAPDH. For a purpose to deliver let-7a persistently into mice skin, the mixtures of miRNA and atelocollagen (AteloGene; Koken) were prepared. let-7a oligo miRNA extraction from skin tissue, cultured cells, and sera was designed, according to previous papers (31, 32) (Hokkaido System Science); mature chain, 59-UGAGGUAGUAGGUUGUAUAGUU-39,which The miRNA isolation from human or mouse skin tissue was performed corresponds to the mmu-let-7a, and minor chain, 59-CUAUACAAUCUA- using miRNeasy FFPE kit (Qiagen). miRNAs were obtained from the total 2 CUGUCUUUCC-39, as the passenger strand. They were annealed to each RNA of cultured cells using an RT Quantitative PCR-Grade miRNA other. Irrelevant control small dsRNA was also obtained from Hokkaido Isolation Kit (SABiosciences). Serum miRNAs were purified with miR- System Science: 59-AUCCGCGCGAUAGUACGUAUU-39 and 59-AATA- Neasy RNA isolation kit (Qiagen) following the manufacturer’s instruc- CGTACTATCGCGCGGAT-39. An equal volume of AteloGene and miRNA tions with minor modification (25). Briefly, 100 ml serum was supple- solution (40 mM in small interfering RNA buffer included in AteloGene kit) by guest on October 1, 2021 mented with 5 ml 5 fmol/ml synthetic nonhuman miRNA (Caenorhabditis was mixed by rotating for 20 min at 4˚C. The mixtures in a 50-mlvolume elegans miR-54; Takara) as controls providing an internal reference for were administered percutaneously into the abdominal cavity of the 8-wk-old normalization of technical variations between samples. After Qiazol so- male BALB/cAJcl mice (CLEA Japan) under anesthesia. Mice were housed lution (1 ml) was added and mixed well, and then, samples were incubated in a specific pathogen-free environment and did not display any evidence of at room temperature for 5 min. Aqueous- and organic-phase separation infection or disease. was achieved by the addition of chloroform. The aqueous phase was ap- plied to RNeasy spin column and RNeasy MinElute spin column. miRNA Bleomycin-induced skin fibrosis in mice was eluted from the column with nuclease-free water. Bleomycin (Nippon Kayaku) was dissolved in PBS at a concentration of miRNA PCR array 1 mg/ml and sterilized by filtration as described previously (33, 34). Bleo- mycin (100 ml) was injected intradermally into the shaved back of the For PCR array, miRNAs were reverse transcribed into first-strand cDNA 8 wk-old male BALB/cAJcl mice daily for 4 wk. The back skin, , and 2 using an RT miRNA First-Strand Kit (SABiosciences). The cDNA was liver were removed 1 d after final bleomycin injection. The removed tis- 2 mixed with RT SYBR Green/ROX Quantitative PCR Master Mix, and the sues were fixed in 4% paraformaldehyde and embedded in paraffin. Sec- 2 mixture was added into 96-well RT miRNA PCR Array that includes tions (4 mm in thickness) were stained with H&E or Masson’s trichrome. primer pairs for 88 human miRNAs (SABiosciences). PCR was performed Dermal thickness was evaluated by measuring the distance between the on Thermal Cycler Dice (TP800) (Takara) following the manufacturer’s epidermal–dermal junction and the dermal– junction in H&E sections protocol. Threshold cycle (Ct) for each miRNA was extracted using under 100-fold magnification by two investigators in a blinded manner. Thermal Cycler Dice Real-Time System version 2.10B (Takara). The raw Ct was normalized using the values of U6. The transient transfection Real-time PCR of miRNAs miRNA inhibitors and miScript target protectors for control or let-7a were purchased from Qiagen. Control or let-7a precursors were obtained from The Mir-X miRNA First-Strand Synthesis Kit (BD Clontech) was used for Life Technologies. Lipofectamine RNAiMAX (Invitrogen) was used as cDNA synthesis from miRNAs. Primers for let-7a, cel-miR-54, or U6 transfection reagent. For reverse transfection, miRNA inhibitors, miScript (Takara) and templates were mixed with SYBR Advantage Quantitative target protectors, or miRNA precursors were mixed with transfection re- PCR Premix (BD Clontech). DNA was amplified for 40 cycles of dena- agent and then added when cells were plated, followed by incubation for turation for 5 s at 95˚C and annealing for 20 s at 60˚C on Takara Thermal 48–96 h at 37˚C in 5% CO2. Control experiments showed transfection Cycler Dice (TP800). Transcript levels of let-7a were normalized to those efficiency is .80% (data not shown). of U6 or cel-miR-54 in the same sample. For the detection of mature, single-stranded form of let-7a was per- Luciferase reporter assay formed using a miScript PCR system (Qiagen): miScript II RT Kit and A luciferase reporter plasmid containing a1(I) collagen 39-UTR was miScript SYBR Green PCR Kit (26). provided by Dr. N. Mitsutake (Nagasaki University, Nagasaki, Japan) (35). Measurement of soluble IL-2R and anti-ssDNA Ab A luciferase construct containing a2(I) collagen 39-UTR was purchased from GeneCopoeia. Lipofectamine 2000 (Invitrogen) was used as trans- Soluble IL-2R and anti-ssDNA Ab were measured with ELISA, as de- fection reagent. miRNA inhibitors, synthetic oligos, and reporter plasmids scribed previously (27, 28). mixed with transfection reagent were added when cells were plated, fol- The Journal of Immunology 3

Table I. The expression profiles of miRNAs in the skin of systemic scleroderma, localized scleroderma, and keloid as measured by PCR array

SSc/Normal LSc/Normal Keloid/Normal Normal DCt SSc DDCt LSc DCt Keloid DDCt Fold Change Fold Change Fold Change let-7a 22.94 0.24 0.67 21.01 0.1 0.1 0.3 let-7b 22.43 20.07 0.06 21.02 0.2 0.2 0.4 let-7c 0.67 2.89 4.02 1.71 0.2 0.1 0.5 let-7d 4.07 4.52 7.00 6.04 0.7 0.1 0.3 let-7e 1.02 3.07 4.26 2.66 0.2 0.1 0.3 let-7f 6.25 6.18 8.86 7.98 1.0 0.2 0.3 let-7g 20.68 2.43 2.45 1.26 0.1 0.1 0.3 let-7i 0.69 2.76 2.63 0.57 0.2 0.3 1.1 miR-1 23.19 4.35 6.83 7.26 0.0 0.0 0.0 miR-7 4.65 6.57 8.22 6.07 0.3 0.1 0.4 miR-9 5.53 6.13 10.07 6.19 0.7 0.0 0.6 miR-10a 0.89 1.54 4.83 1.54 0.6 0.1 0.6 miR-10b 20.69 1.38 4.00 0.09 0.2 0.0 0.6 miR-15a 1.81 6.32 5.40 1.64 0.0 0.1 1.1 miR-15b 20.41 0.13 3.51 1.27 0.7 0.1 0.3 miR-16 24.19 22.77 0.36 22.89 0.4 0.0 0.4 miR-17 1.01 1.26 4.52 2.45 0.8 0.1 0.4 miR-18a 4.40 2.02 7.05 4.37 5.2 0.2 1.0 miR-18b 5.22 2.73 5.82 7.37 5.6 0.7 0.2 Downloaded from miR-20a 0.44 1.21 4.02 1.66 0.6 0.1 0.4 miR-20b 4.42 2.89 7.68 6.35 2.9 0.1 0.3 miR-21 24.43 22.15 21.71 26.90 0.2 0.2 5.5 miR-22 0.08 3.36 2.77 20.17 0.1 0.2 1.2 miR-23b 23.52 23.00 1.21 20.99 0.7 0.0 0.2 miR-24 22.70 21.25 0.81 21.73 0.4 0.1 0.5 2 2 2 2

miR-26a 4.51 2.56 0.19 2.40 0.3 0.1 0.2 http://www.jimmunol.org/ miR-33a 3.95 4.85 9.30 5.80 0.5 0.0 0.3 miR-92a 22.66 22.23 1.07 0.16 0.7 0.1 0.1 miR-93 2.30 0.97 5.12 3.44 2.5 0.1 0.5 miR-96 3.43 2.34 9.15 6.07 2.1 0.0 0.2 miR-99a 22.32 20.90 2.39 21.22 0.4 0.0 0.5 miR-100 22.08 20.42 2.35 20.42 0.3 0.0 0.3 miR-101 0.89 3.94 6.23 3.10 0.1 0.0 0.2 miR-103 0.10 1.62 2.81 0.95 0.3 0.2 0.6 miR-106b 2.51 2.79 5.90 3.34 0.8 0.1 0.6 miR-122 8.10 7.33 3.40 2.23 1.7 26.0 58.7

miR-124 6.65 1.50 9.02 9.98 35.5 0.2 0.1 by guest on October 1, 2021 miR-125a-5p 22.24 21.65 2.62 0.20 0.7 0.0 0.2 miR-125b 25.55 23.91 20.32 23.75 0.3 0.0 0.3 miR-126 23.38 21.81 0.45 21.96 0.3 0.1 0.4 miR-127-5p 8.27 5.81 9.54 6.54 5.5 0.4 3.3 miR-128a 1.31 9.16 5.92 3.44 0.0 0.0 0.2 miR-129-5p 6.60 0.29 8.25 9.77 79.3 0.3 0.1 miR-130a 2.67 5.24 7.29 5.46 0.2 0.0 0.1 miR-132 3.25 0.97 6.70 3.80 4.9 0.1 0.7 miR-133b 2.40 6.34 10.39 12.99 0.1 0.0 0.0 miR-134 4.48 0.37 8.06 4.36 17.3 0.1 1.1 miR-137 8.67 6.88 11.91 10.66 3.5 0.1 0.3 miR-141 1.42 4.22 6.60 3.40 0.1 0.0 0.3 miR-142-3p 2.60 5.07 3.43 3.06 0.2 0.6 0.7 miR-142-5p 7.16 7.33 8.27 9.06 0.9 0.5 0.3 miR-146a 0.18 0.74 3.44 3.14 0.7 0.1 0.1 miR-146b-5p 0.80 1.26 3.98 1.62 0.7 0.1 0.6 miR-150 20.78 0.55 1.78 1.33 0.4 0.2 0.2 miR-155 4.20 7.48 5.37 5.05 0.1 0.4 0.6 miR-181a 2.42 1.94 4.68 2.06 1.4 0.2 1.3 miR-182 3.39 2.99 7.81 7.42 1.3 0.0 0.1 miR-183 4.28 7.92 8.70 7.62 0.1 0.0 0.1 miR-185 2.34 3.78 5.58 3.68 0.4 0.1 0.4 miR-192 3.78 3.81 6.16 4.40 1.0 0.2 0.7 miR-194 3.76 3.92 6.50 4.50 0.9 0.1 0.6 miR-195 24.07 22.89 0.79 22.84 0.4 0.0 0.4 miR-196a 1.66 2.81 5.97 2.42 0.5 0.1 0.6 miR-205 23.31 21.59 0.39 21.41 0.3 0.1 0.3 miR-206 0.01 5.82 9.15 10.77 0.0 0.0 0.0 miR-208 8.10 3.21 11.99 15.27 29.7 0.1 0.0 miR-210 3.34 0.70 5.86 4.25 6.2 0.2 0.5 miR-214 0.82 0.90 3.93 0.54 0.9 0.1 1.2 miR-215 7.18 6.11 9.76 9.51 2.1 0.2 0.2 miR-218 2.72 ND 7.14 4.72 — 0.0 0.3 miR-219-5p 6.98 3.11 13.35 9.03 14.6 0.0 0.2 miR-222 1.62 4.49 3.25 2.56 0.1 0.3 0.5 (Table continues) 4 let-7a EXPRESSION IN SSc AND LSc

Table I. (Continued)

SSc/Normal LSc/Normal Keloid/Normal Normal DCt SSc DDCt LSc DCt Keloid DDCt Fold Change Fold Change Fold Change miR-223 0.04 1.18 3.01 1.73 0.5 0.1 0.3 miR-301a 6.59 4.97 9.89 7.72 3.1 0.1 0.5 miR-302a 10.04 4.55 6.98 12.04 44.9 8.3 0.3 miR-302c 12.91 6.50 10.40 11.85 85.0 5.7 2.1 miR-345 5.83 1.96 7.12 6.16 14.6 0.4 0.8 miR-370 7.54 3.97 5.78 6.59 11.9 3.4 1.9 miR-371-3p 10.58 7.01 12.50 14.52 11.9 0.3 0.1 miR-375 3.27 0.72 7.18 6.99 5.9 0.1 0.1 miR-378 0.02 4.76 5.38 3.84 0.0 0.0 0.1 miR-424 2.27 3.05 3.73 22.21 0.6 0.4 22.2 miR-452 4.57 4.77 9.19 7.02 0.9 0.0 0.2 miR-488 11.36 4.95 12.78 12.77 85.0 0.4 0.4 miR-498 13.41 8.30 17.00 19.47 34.5 0.1 0.0 miR-503 7.17 3.78 9.40 6.45 10.5 0.2 1.7 miR-518b 13.54 7.72 11.94 14.02 56.5 3.0 0.7 miR-520g 7.57 6.76 10.58 9.36 1.8 0.1 0.3 A mixture of equal amounts of miRNAs from three normal skins, three systemic scleroderma, three localized scleroderma, or three keloid were prepared, and miRNA expression profile in each disease in vivo was evaluated using PCR Array. The raw Ct was normalized using the values of small RNA housekeeping gene U6. DDCt (the raw Ct of Downloaded from each miRNA 2 Ct of U6) were shown. The fold change was calculated as 1/2(DDCt of each disease 2 DDCt of normal skin). lowed by incubation for 48 h at 37˚C. Luc-Pair miR luciferase assay were significantly lower than those in seven normal skin or five (GeneCopoeia) and FilterMax F5 microplate reader (Molecular Devices) keloid skin (Fig. 1A). The decrease of let-7a in SSc skin was milder were used to analyze the luciferase expression according to the manu- than that in LSc skin, which is consistent with the array result. facturer’s protocols. In addition, in situ hybridization of skin tissue showed that signal Statistical analysis for let-7a was evident in normal dermal fibroblasts but not in the http://www.jimmunol.org/ Statistical analysis was carried out with a Student t test for the comparison of fibroblasts of SSc and LSc (Fig. 1B). Moreover, relative let-7a two groups and with a Fisher’s exact probability test for the analysis of fre- levels in cultured dermal fibroblasts derived from SSc and LSc quency. ANOVA was used for multiple comparisons. Correlations were as- involved skin in vitro were also significantly lower than those in sessed by Pearson’s correlation coefficient. All analyses were performed with normal dermal fibroblasts (Fig. 1C). Thus, let-7a is thought to be Statcel software (OMS). The p values , 0.05 were considered significant. decreased in dermal fibroblasts of SSc and LSc both in vivo and Results in vitro. We tried to clarify the mechanism that mediates decreased miRNA expression in SSc/LSc

expression of let-7a in SSc/LSc fibroblasts. To examine the pos- by guest on October 1, 2021 As an initial experiment, to determine which miRNAs are involved sibility that the downregulation of let-7a is due to the stimulation in the pathogenesis of SSc/LSc, a mixture of equal amounts of of intrinsic TGF-b activation seen in these cell types as described miRNAs from three SSc, three LSc, three keloid, or three normal in the Introduction, normal fibroblasts were stimulated with ex- skins were prepared. We performed miRNA PCR array analysis ogenous TGF-b1. TGF-b1 downregulated let-7a expression mildly consisting of 88 miRNAs involved in human cell differentiation but significantly (Fig. 1D), suggesting that downregulation of let- and development, and the miRNA expression profile in involved 7a is consequence of activation of TGF-b signaling in SSc/LSc skin of SSc or LSc in vivo was compared with that of normal skin fibroblasts, at least partly. or keloid (the complete dataset is available at the Gene Expression Omnibus microarray data repository, http://www.ncbi.nlm.nih.gov/ The function of let-7a in the regulation of type I collagen geo/, accession number GSE43469; Table I). expression in human dermal fibroblasts We hypothesized that miRNAs overexpressed or suppressed both Next, we determined the contribution of downregulated let-7a to in SSc and LSc compared with normal skin, but not changed in the regulation of type I collagen in SSc and LSc fibroblasts. As keloid, may be important in the pathogenesis of SSc/LSc, because described in previous reports (11, 36), the expression of both a1(I) SSc/LSc and keloid are thought to have different etiology as de- and a2(I) chain, two major components of type I collagen, was scribed in Introduction. When a two-cycle difference (4-fold significantly increased in SSc (Fig. 2A) and LSc (Fig. 2B) fibro- change in the DDCT method) was considered meaningful, 10 blasts. In addition, both a1(I) and a2(I) collagen protein was in- miRNAs (let-7a, -7b, -7c, -7e, -7g, miR-10b, -15a, -22, -141, and duced by exogenous TGF-b1 in normal fibroblasts (Fig. 2C): the -302c) matched the condition (Table I). Interestingly, 5 of the 10 overexpression of type I collagen in SSc/LSc fibroblasts is thought miRNAs belonged to let-7 family, one of the first miRNA family to be due to intrinsic TGF-b1 activation (5, 11). Considering discovered in human. Furthermore, although not more than a two- above results that let-7a expression is downregulated in SSc/LSc cycle difference, let-7d and -7i, remaining members of let-7 dermal fibroblasts as well as normal fibroblasts stimulated with family, also tended to be decreased in SSc and LSc. Accord- exogenous TGF-b1 (Fig. 1), downregulation of let-7a may induce ingly, we focused on let-7a, most downregulated let-7 member in collagen expression. both SSc and LSc: Compared with normal skin, the expression of let-7a was the potent regulator of both a1(I) and a2(I) collagen, let-7a was decreased in SSc skin (DDCt 22.94 versus 0.24; 3.18- according to miRNA target gene predictions using the TargetScan cycle difference = 0.11-fold change) and in LSc skin (DDCt 22.94 (version 5.1, http://www.targetscan.org/), MiRanda (August 2010 versus 0.67; 3.61-cycle difference = 0.08-fold change). release, http://www.microrna.org/), DIANAmicroT (version 3.0; We confirmed the array result by real-time PCR with specific http://diana.cslab.ece.ntua.gr/microT/), and PicTar (http://pictar. primer for let-7a, using increased number of samples. The mean mdc-berlin.de/), leading programs in the field (37–40). Actually, let-7a levels in seven SSc skin as well as seven LSc skin in vivo transfection of the specific inhibitor of let-7a in normal fibroblasts The Journal of Immunology 5 in vitro resulted in the statistically significant increase of protein constitutively upregulated type I collagen expression in SSc/LSc expression of a1(I) and a2(I) collagen (Fig. 2D). In addition, to fibroblasts. determine direct interaction between let-7a with collagen 39-UTR, we used miRNA miScript Target Protectors (Qiagen), single- Serum levels of let-7a in SSc/LSc patients stranded, modified RNAs complementary to the let-7a binding Recently, many publications indicate serum miRNAs can be novel sites on type I collagen 39-UTR: the protectors cover the flanking biomarkers in various diseases. Serum miRNAs are rather stable, regions of the binding sites and specifically interfere with the because they are thought to be encapsulated in microvesicles and direct interaction between let-7a and collagen (http://www.qiagen. protected from RNase, extreme temperatures, extreme pHs, or com/products/miscripttargetprotectors.aspx) (41). As shown in freeze-thaw cycle (42–46). Furthermore, miRNAs in the vesicles Fig. 2E, transfection of the let-7a protector resulted in the up- can be incorporated into other cells and may alter gene expression regulation of type I collagen expression to the similar extent to let- in the recipient cells (47). Thus, serum miRNA levels may not 7a inhibitor. Furthermore, we also performed luciferase reporter merely be secreted from apoptotic cells but may exert some bio- gene assay using luciferase constructs containing the a1(I) or a2(I) logical effects. We evaluated the possibility that serum let-7a collagen 39-UTR. let-7a inhibitor induced the luciferase activi- levels can be a disease marker of SSc/LSc. ties compared with control inhibitor (Fig. 2F). In contrast, the There has been no report determining the expression of let-7a transfection of let-7a precursor into normal fibroblasts downreg- in cell-free body fluid. To confirm that let-7a was indeed present ulated protein expression of type I collagen (Fig. 2G). Taken to- in human serum, miRNA was purified from human sera and let-7a gether, our results indicated that a1(I) and a2(I) collagen is the was detected by quantitative real-time PCR using primer set specific direct target of let-7a and that let-7a plays some roles in the for let-7a. As shown in Fig. 3A, the amplification of let-7a was ob- served, and Ct values were increased by the serial dilution of Downloaded from the miRNA. Thus, let-7a was detectable and quantitative in the serum using our method. Serum samples were obtained from 17 healthy control subjects, 39 SSc patients consisted of 20 dcSSc and 19 lcSSc, and 32 LSc patients consisted of 19 morphea, 8 LS, and 5 GM. We also included serum samples from other rheumatic diseases: eight patients with http://www.jimmunol.org/ SLE and eight patients with DM. As shown in Fig. 3B, serum let-7a levels in LSc patients were significantly lower than those in healthy controls, which is consistent with the downregulation of let-7a in LSc skin. There was no significant difference in the values among the three groups of morphea, LS, and GM. In contrast, the decrease of let-7a levels in SSc sera was milder than that in LSc sera, which is consistent with the result in the skin. Although the mean serum let-7a level in dcSSc was decreased compared with that in lcSSc, by guest on October 1, 2021 we could not find statistical significance. Next, we tried to clarify the correlation between clinical/sero- logical features and serum levels of let-7a in patients with SSc. Patients with reduced serum let-7a (below the mean level of SSc patients) had significantly increased ratio of dcSSc:lcSSc, signifi- cantly higher modified Rodnan Total skin thickness score (modified Rodnan Total skin thickness score) and significantly lower frequency of anti-centromere Ab than those with elevated let-7a levels (Table II), suggesting that serum let-7a levels is inversely correlated with the severity of skin sclerosis. In contrast, correlations were assessed by Pearson’s correlation coefficient in four parameters in LSc patients: duration of disease (between symptom onset and the first visit to the hospital) reflects the subjective severity of symptom. The number of lesions (larger than 3 cm in diameter) correlates with the involved area. Soluble IL-2R is reported to be parallel with disease activity (27, 48). And ss-DNA is also a marker of disease activity, especially muscle involvement (49). As a result, there were no FIGURE 1. Expression levels of let-7a in SSc and LSc. (A) Mean rel- correlation (r , 0.4) between the serum let-7a levels and these four ative transcript levels of let-7a expression in skin tissues of seven healthy factors (Fig. 3C). Thus, decreased serum let-7a in LSc patients may controls (HC), seven SSc, seven LSc, and five keloid were determined by not be associated with specific features of this disease. real-time PCR as described in Materials and Methods.Thedataare expressed as mean 6 SD. The values in samples from HC were set at 1. Effect of let-7a on mouse dermal fibroblasts *p , 0.05 by ANOVA. (B) In situ detection of let-7a in paraffin-embedded, On the basis of the above results, we considered let-7 family, es- formalin-fixed tissues of normal skin, SSc, and LSc (n = 3). The let-7a pecially let-7a, as SSc/LSc-specific miRNA and a negative effector stained blue. Bar, 200 mm(upper panels) and = 50 mm(lower panels). (C) on fibrosis. Meanwhile, skin fibrosis induced by bleomycin injection Mean relative levels of let-7a in cultured dermal fibroblasts derived from HC skin, LSc, and SSc. Data are expressed as the mean 6 SD of three in mice is known for a murine model of SSc/LSc (33). Indeed, as independent experiments. *p , 0.05. (D) let-7a levels in normal fibroblasts seen in SSc/LSc, the levels of let-7a in bleomycin-treated fibrotic in the presence or absence of TGF-b1 (2 ng/ml) for 24 h. Data are mice skins were significantly lower than PBS-treated control mice expressed as the mean 6 SD of four independent experiments. *p , 0.05 skins (Fig. 4A). Then, we evaluated whether the supplementation of compared with untreated normal fibroblasts (1.0). let-7a in vivo could prevent from bleomycin-induced skin fibrosis. 6 let-7a EXPRESSION IN SSc AND LSc Downloaded from http://www.jimmunol.org/ by guest on October 1, 2021

FIGURE 2. The effect of let-7a on the expression of type I collagen in human dermal fibroblasts. (A) Expression of type Ι collagen protein in cultured normal and SSc dermal fibroblasts were detected by immunoblotting. Type Ι collagen protein expression levels quantitated by scanning densitometry and corrected for the levels of b-actin in the same samples are shown relative to the levels in normal fibroblasts (1.0). *p , 0.05 as compared with the values in normal fibroblasts. (B) Expression of type Ι collagen protein in cultured normal and LSc dermal fibroblasts are shown and quantitated as described in (A). *p , 0.05. (C) Type Ι collagen protein levels in normal fibroblasts treated with or without TGF-b1 (2 ng/ml) for 24 h is shown by immunoblotting. (D) Normal human fibroblasts at a density of 2 3 104 cells/well in 24-well culture plates were transfected with control inhibitor or let-7a inhibitor for 96 h. Cell lysates were subjected to immunoblotting. Type Ι collagen protein expression levels were quantitated as described in (A)(n = 3). *p , 0.05 as compared with the value in cells with control inhibitor (1.0). (E) Fibroblasts at a density of 2 3 104 cells/well in 24-well culture plates were transfected with let-7a inhibitor or miScript Target protectors for the let-7a binding sites on a1(Ι) and a2(Ι) collagen for 96 h. Cell lysates were subjected to immunoblotting. (F) Normal human fibroblasts at a density of 1 3 104 cells/well in 96-well culture plates were transfected with luciferase reporters containing 39-UTR segment of a1(I) collagen (N)ora2(I) collagen (n) and the indicated miRNA inhibitors. The bar graph shows the relative luciferase activity; *p , 0.05 as compared with the values in cells with control inhibitor (1.0). (G) Normal human fibroblasts at a density of 2 3 104 cells/well in 24-well culture plates were transfected with control miRNA precursor or let-7a precursor. Cell lysates were subjected to immunoblotting.

To our knowledge, there have been no reports on the technique of a1(I) or a2(I) collagen 39-UTR compared with control small for overexpressing miRNAs in the skin in vivo. We prepared let-7a RNA (Fig. 4E). These results confirm that let-7a oligo could regulate oligo as described previously (31, 32) and before in vivo experi- the expression of collagen directly in mouse dermal fibroblasts. ments, we evaluated whether it could reduce the collagen expres- Next, to deliver let-7a persistently into mice skin, control small sion in vitro by transfecting into cultured mouse skin fibroblasts RNA or let-7a oligo was mixed with atelocollagen (AteloGene) for and measuring collagen RNA/protein levels and 39-UTR luciferase the protection from degeneration by RNase in vivo (50). The activity. mixtures were administered percutaneously into the abdominal The induction of let-7a in mouse dermal fibroblasts upregulated cavity. Then we tried to confirm whether let-7a is overexpressed in mature let-7a in cell lysates by the mature miRNA-specific PCR anal- the skin by our method; the let-7a levels in mice back skin were ysis (Fig. 4B) (26) and reduced type I collagen protein expression significantly increased by the injection of atelocollagen and let-7a by immunoblotting (Fig. 4C). However, the mRNA expression of mixture specifically compared with the mixture of atelocollagen a1(I) or a2(I) collagen was not downregulated by let-7a induction and control small RNA (Fig. 4F). Such higher let-7a expressions (Fig. 4D), indicating that the let-7a oligo regulates the collagen continued until day 3 after injection. In contrast, the mixture of translation without changing mRNA levels. In addition, the let-7a PBS and let-7a did not increase let-7a expression in the skin (Fig. induction in mouse dermal fibroblasts reduced the luciferase activity 4F), probably because of the degradation. The Journal of Immunology 7

FIGURE 3. Serum let-7a levels in patients with SSc and LSc. (A) The let-7a was present in serum sample. Serial dilution of cDNA (as is, 10-fold dilution, 100-fold dilation, and 0) synthesized from serum miRNA was used as template for real-time PCR. Amplification curves of gene-specific transcripts are shown to illustrate the process of exponential in- crease of fluorescence. (B) Serum let-7a con- centrations measured by quantitative real- time PCR (normalized to cel-miR-54) in patients with DM, SLE, SSc, and LSc and in Downloaded from healthy control (HC) subjects are shown. let- 7a concentrations are shown on the ordinate. The horizontal bar in each group shows the mean value. The minimum value in LSc patients was set at 1. M, morphea. *p , 0.05. (C) Correlations between clinical/serological features and serum levels of let-7a in patients http://www.jimmunol.org/ with LSc. Correlations were assessed by Pearson’s correlation coefficient. Correlation coefficient (r2) and p values are shown. by guest on October 1, 2021

Systemic administration of let-7a leads to the attenuation of administration. Taken together, systemic administration of let-7a skin fibrosis induced by bleomycin treatment could lead to the let-7a overexpression in skin and the attenuation Bleomycin was locally injected in the back of the BALB/c mice of skin fibrosis by bleomycin. daily for 4 wk. At the same time, control small RNA or let-7a oligo mixed with atelocollagen was administered percutaneously into the Discussion abdominal cavity once per week (four times a month) (Fig. 5A). miRNAs, short RNA molecules on average only 22 nucleotides When PBS was injected in back skin instead of bleomycin, there long, bind to 39-UTRs of target mRNAs and lead to gene silencing. was no difference in the skin between administration of control Recent vigorous efforts of research in this field indicated that small RNA and let-7a (Fig. 5B). In contrast, in the presence of miRNAs play roles in the pathogenesis of various disorders (16– control small RNA, mice skin injected with bleomycin showed 18). This study demonstrated the role of let-7a in collagen over- dermal fibrosis with increased number of thickened collagen expression and its contribution to the pathogenesis of fibrosis in bundles and strong Masson’s trichrome staining (Fig. 5C, left SSc/LSc by several major findings. panel). Thus, the injection of control small RNAs did not have an First, we tried to identify miRNAs specifically up- or downreg- antifibrotic effect (e.g., triggering an antiviral innate immune re- ulated in SSc/LSc compared with normal or keloid skin using PCR sponse that leads to anti-fibrotic signaling pathways). However, array. Because the array analysis was performed as a single ex- administration of let-7a decreased the dermal thickness and periment, a statistical significance could not be evaluated. There- Masson’s trichrome staining in the , (Fig. 5C, right panel), fore, we confirmed the result using real-time PCR, accompanied by which suggested decreased collagen deposition. We confirmed statistical analysis: let-7a was significantly downregulated in SSc/ that the improvement of bleomycin-induced dermal thickening LSc tissue. Our study is the first, to our knowledge, to demonstrate by let-7a administration was statistically significant (Fig. 5D). the let-7a downregulation in SSc/LSc. The downregulation of let- For further investigation, we evaluated whether injection of let- 7a may result from activated endogenous TGF-b signaling. As the 7a oligo into the abdominal cavity could affect other internal negative mediator of TGF-b signaling, let-7a may play a role in organs. Bleomycin did not cause tissue fibrosis in the lung (Fig. the constitutive up-regulation of type I collagen in these cells. 5E) or liver (Fig. 5F), and there were no apparent differences in However, because the effect of TGF-b stimulation on the let-7a H&E staining between mice with control small RNA and let-7a expression in normal fibroblasts in this study was significant but 8 let-7a EXPRESSION IN SSc AND LSc

Table II. The correlation of the serum let-7a levels with the clinical and serological features in SSc patients

Patients with Elevated let-7a Patients with Reduced let-7a Clinical and Serological Features Levels (n = 18) Levels (n = 21) p Value Mean age at serum sampling, years 60.6 61.8 0.78 Mean duration of disease, month 62.5 61.4 0.98 dcSSc: lcSSc 5:13 15:6 0.01 MRSS, score 7.4 14.9 0.04 Clinical features Pitting /ulcers 33.3 47.6 0.28 Nailfold bleeding 50.0 28.6 0.15 Raynaud’s phenomenon 83.3 76.2 0.44 27.8 28.6 0.62 Contracture of phalanges 72.2 71.4 0.62 0 4.8 0.54 involvement Pulmonary fibrosis 27.8 42.9 0.26 11.1 28.6 0.17 27.8 52.4 0.11 0 9.5 0.28 27.8 9.5 0.14 ANA specificity Anti-topo I 16.7 33.3 0.21 Downloaded from Anti-centromere 66.7 28.6 0.02 Anti-U1 RNP 11.1 9.5 0.64 Unless indicated, the values are percentages. ANA, Anti-nuclear Ab; Anti-topo I, anti- I Ab; Anti-centromere, anti-centromere Ab; MRSS, modified Rodnan total skin thickness score. mild, there may be other factors mediating the let-7a downregu- Second, we found the new miRNA–mRNA target interactions: let-7 http://www.jimmunol.org/ lation in SSc. Future studies should be needed to clarify this point. and type I collagen. miRNAs have been implicated in immune re-

FIGURE 4. The effect of let-7a on type Ι col- lagen expression in mice skin fibroblasts in vitro. (A) Bleomycin (100 ml) or PBS was injected in- tradermally into the shaved back of the 8-wk-old male BALB/cAJcl mice daily for 4 wk. let-7a levels in the back skin of bleomycin- or PBS- treated mice were determined by real-time PCR by guest on October 1, 2021 as described in Materials and Methods. Data are shown on the ordinate (n = 5). Bars show means. *p , 0.05. (B–D) Cultured BALB/c mouse skin fibroblasts at a density of 2 3 104 cells/well in 24-well culture plates were transfected with control small RNA or let-7a oligo (10 nM). (B) Relative levels of mature let-7a were analyzed by real-time PCR for the mature single-stranded form of let-7a. Data are shown on the ordinate (n = 4). Bars show means. The minimum value in controls was set at 1. *p , 0.05. (C) Type Ι col- lagen protein levels in the cell lysates were shown. (D) Relative mRNA levels of a1(Ι)and a2(Ι) collagen were analyzed by real-time PCR. The value in cells with control small RNA was set at 1. (E) Cultured mouse skin fibroblasts at a density of 1 3 104 cells/well in 96-well culture plates were transfected with luciferase reporters containing the 39-UTR segment of a1(I) collagen (N)ora2(I) collagen (n) and the indicated oligos. At 24 h after the transfection, luciferase activity was measured (n =3).*p , 0.05 as compared with the value in cells with control small RNA (1.0). (F) Mean relative let-7a levels in mice back skin after let-7a injection into mice abdominal cavity. The line graphs represent; AteloGene and control small miRNA mixtures (Atelo + control; black), PBS and let-7a mixtures (PBS + let-7a; blue), and AteloGene and let-7a mixtures (Atelo + let-7a; red). Mice back skin was removed on days 1, 3, 5, and 7 after miRNA mixture injection. n = 4. *p , 0.05. The Journal of Immunology 9

FIGURE 5. systemic administration of let-7a in bleomycin-induced skin fibrosis in vivo. (A)The protocol for (B)–(F) is shown. Bleomycin was locally injected in the back of the BALB/c mice daily for 4 wk. At the same time, control small miRNA or let-7a mixed with atelocollagen were administered percu- taneously into the abdominal cavity once per week (four times a month). The back skin was obtained 1 d after final bleomycin injection. (B) H&E staining of PBS-treated mice skin with control small miRNA (left panel)orlet-7a(right panel) injection into the abdominal cavity. Scale bar, 100 mm. (C) H&E (upper panels) and Masson’s trichrome staining (lower panels) of bleomycin-treated mice skin with control

miRNA (left panels) or let-7a (right panels) injection Downloaded from into the abdominal cavity. Scale bar, 100 mm. (D) Dermal thickness was measured in bleomycin-treated mice skin with control or let-7a injection into the ab- dominal cavity. Data are shown on the ordinate (n = 6). Bars show means. *p , 0.05. (E) H&E staining of the lung in PBS- or bleomycin-treated mice with in- jection of control miRNA (left panels) or let-7a (right http://www.jimmunol.org/ panels) into the abdominal cavity. Scale bar, 100 mm. (F) H&E staining of the liver in PBS- or bleomycin- treated mice with injection of control miRNA (left panels) or let-7a (right panels) into the abdominal cavity. Scale bar, 100 mm. by guest on October 1, 2021

sponse as well as cell development, cell differentiation, proliferation, (53) have demonstrated that systemic injection of miR-146a pre- and (51). Our study suggests miRNAs are also involved in vents destruction in collagen-induced arthritic mice. the regulatory mechanisms of metabolism. In summary, our hypothetical model is shown in Fig. 6. In Also, we first investigated serum miRNA levels in SSc/LSc normal fibroblasts, let-7a has a negative effect on type Ι collagen and found let-7a can be a biomarker of SSc and LSc for the diag- expression. However, downregulated let-7a by the stimulation of nosis or the evaluation of disease activity. Contrary to our expectation, TGF-b contributes to the overexpression of type I collagen in SSc/ let-7a levels in dermal fibroblasts and sera of LSc patients were LSc fibroblasts. Although there are thought to be so many factors lower than those in SSc. This may be explained by the notion that regulating the fibrotic process in SSc and LSc, let-7a may also fibrosis in the lesion tends to be more severe in LSc compared with play some roles in the pathogenesis of these diseases. Investiga- SSc; For example, the fibrosis in LSc sometimes extends to the mus- tion of more detailed mechanisms of miRNA-mediated regulation cular tissues or bone beneath the cutaneous lesions but not in SSc. of collagen expression may lead to new therapeutic approach Thus, lower let-7a levels may reflect severer tissue fibrosis in LSc. against cutaneous fibrosis using miRNA. As described above, we considered let-7 family as the SSc/LSc- specific miRNAs. However, as seen in Table I, there were other miRNAs downregulated both in SSc and LSc, but not in keloid, including miR-10b, -15a, -22, -141, and -302c. The relation be- tween these miRNAs and SSc/LSc is needed to be examined. Last, we tried to determine the function of let-7a in vivo model. To our knowledge, this is the first report describing the method for miRNA overexpression in mice skin by i.p. miRNA injection, and our procedure may become a standard method for similar experi- ments in the future. The intermittent overexpression of let-7a by i.p. injection had therapeutic value against skin fibrosis induced by bleomycin. miRNA-targeted for human diseases is ex- FIGURE 6. Schematic model of let-7a-induced collagen overexpression pected to become a scientific breakthrough. Miravirsen (SPC3649), in SSc/LSc fibroblasts. In SSc/LSc fibroblasts, let-7a expression is con- a specific inhibitor of miR-122, has finished Phase 2a clinical trials stitutively decreased by TGF-b signaling, and subsequently collagen ex- in patients infected with C virus (52). Also, Nakasa et al. pression is induced, which results in tissue fibrosis seen in SSc/LSc. 10 let-7a EXPRESSION IN SSc AND LSc

Acknowledgments 24. Ihn, H., E. C. LeRoy, and M. Trojanowska. 1997. Oncostatin M stimulates of the human a2(I) collagen gene via the Sp1/Sp3-binding site. We thank Junko Suzuki and Chiemi Shiotsu for valuable technical assis- J. Biol. Chem. 272: 24666–24672. tance. The luciferase reporter plasmid containing a1(I) collagen 39-UTR 25. Kroh, E. M., R. K. Parkin, P. S. Mitchell, and M. Tewari. 2010. Analysis of wasprovidedbyDr.NorisatoMitsutake (Radiation Medical Sciences, circulating microRNA biomarkers in plasma and serum using quantitative re- Atomic Bomb Disease Institute, Nagasaki University, Nagasaki, Japan). verse transcription-PCR (qRT-PCR). Methods 50: 298–301. 26. Zhang, Y., J. Guo, D. Li, B. Xiao, Y. Miao, Z. Jiang, and H. Zhuo. 2010. Down- regulation of miR-31 expression in gastric tissues and its clinical sig- Disclosures nificance. Med. Oncol. 27: 685–689. 27. Ihn, H., S. Sato, M. Fujimoto, K. Kikuchi, and K. Takehara. 1996. Clinical The authors have no financial conflicts of interest. significance of serum levels of soluble interleukin-2 receptor in patients with localized scleroderma. Br. J. Dermatol. 134: 843–847. 28. Ihn, H., N. Yazawa, M. Kubo, K. Yamane, S. Sato, M. Fujimoto, K. Kikuchi, References Y. Soma, and K. Tamaki. 2000. Circulating levels of soluble CD30 are increased in patients with localized scleroderma and correlated with serological and 1. Fett, N., and V. P. Werth. 2011. Update on morphea: part I. Epidemiology, clinical features of the disease. J. Rheumatol. 27: 698–702. clinical presentation, and pathogenesis. J. Am. Acad. Dermatol. 64: 217–228, 29. Martin, M. M., J. A. Buckenberger, J. Jiang, G. E. Malana, G. J. Nuovo, quiz 229–230. M. Chotani, D. S. Feldman, T. D. Schmittgen, and T. S. Elton. 2007. The human 2. Fett, N., and V. P. Werth. 2011. Update on morphea: part II. Outcome measures and treatment. J. Am. Acad. Dermatol. 64: 231–242, quiz 243–244. angiotensin II type 1 receptor +1166 A/C polymorphism attenuates microrna- 3. Jablonska, S., and G. Rodnan. 1979. Localized forms of scleroderma. Clin. 155 binding. J. Biol. Chem. 282: 24262–24269. Rheum. Dis. 5: 215–241. 30. Nuovo, G. J. 2008. In situ detection of precursor and mature microRNAs 4. Higley, H., K. Persichitte, S. Chu, W. Waegell, R. Vancheeswaran, and C. Black. in paraffin embedded, formalin fixed tissues and cell preparations. Methods 44: 1994. Immunocytochemical localization and serologic detection of transforming 39–46. growth factor b1: association with type I procollagen and inflammatory cell 31. Kim, S. J., J. Y. Shin, K. D. Lee, Y. K. Bae, K. W. Sung, S. J. Nam, and markers in diffuse and limited systemic sclerosis, morphea, and Raynaud’s K. H. Chun. 2012. MicroRNA let-7a suppresses breast cancer cell migration and invasion through downregulation of C-C receptor type 7. Breast phenomenon. Rheum. 37: 278–288. Downloaded from 5. Kubo, M., H. Ihn, K. Yamane, and K. Tamaki. 2001. Up-regulated expression of Cancer Res. 14: R14. transforming growth factor b receptors in dermal fibroblasts in skin sections 32. Yang, Q., Z. Jie, H. Cao, A. R. Greenlee, C. Yang, F. Zou, and Y. Jiang. 2011. from patients with localized scleroderma. Arthritis Rheum. 44: 731–734. Low-level expression of let-7a in gastric cancer and its involvement in tumori- 6. Asano, Y., H. Ihn, K. Yamane, M. Jinnin, Y. Mimura, and K. Tamaki. 2004. genesis by targeting RAB40C. Carcinogenesis 32: 713–722. Phosphatidylinositol 3-kinase is involved in a2(I) collagen gene expression in 33. Yamamoto, T., S. Takagawa, I. Katayama, K. Yamazaki, Y. Hamazaki, normal and scleroderma fibroblasts. J. Immunol. 172: 7123–7135. H. Shinkai, and K. Nishioka. 1999. Animal model of sclerotic skin. I: Local 7. Asano, Y., H. Ihn, K. Yamane, M. Kubo, and K. Tamaki. 2004. Impaired injections of bleomycin induce sclerotic skin mimicking scleroderma. J. Invest. Smad7‑Smurf-mediated negative regulation of TGF-b signaling in scleroderma Dermatol. 112: 456–462. fibroblasts. J. Clin. Invest. 113: 253–264. 34. Tanaka, C., M. Fujimoto, Y. Hamaguchi, S. Sato, K. Takehara, and http://www.jimmunol.org/ 8. Asano, Y., H. Ihn, K. Yamane, M. Jinnin, Y. Mimura, and K. Tamaki. 2005. M. Hasegawa. 2010. Inducible costimulator ligand regulates bleomycin-induced Differential effects of the immunosuppressant FK-506 on human a2(I) collagen lung and skin fibrosis in a mouse model independently of the inducible gene expression and transforming growth factor b signaling in normal and costimulator/inducible costimulator ligand pathway. Arthritis Rheum. 62: 1723– scleroderma fibroblasts. Arthritis Rheum. 52: 1237–1247. 1732. 9. Kahaleh, B. 1993. Immunologic aspects of scleroderma. Curr. Opin. Rheumatol. 35. Kashiyama, K., N. Mitsutake, M. Matsuse, T. Ogi, V. A. Saenko, K. Ujifuku, 5: 760–765. A. Utani, A. Hirano, and S. Yamashita. 2012. miR-196a downregulation 10. Takehara, K., and S. Sato. 2005. Localized scleroderma is an autoimmune dis- increases the expression of type I and III in keloid fibroblasts. J. Invest. order. 44: 274–279. Dermatol. 132: 1597–1604. 11. Asano, Y., H. Ihn, M. Jinnin, Y. Mimura, and K. Tamaki. 2006. Involvement of 36. Mauch, C., E. Kozlowska, B. Eckes, and T. Krieg. 1992. Altered regulation of collagen metabolism in scleroderma fibroblasts grown within three-dimensional avb5 integrin in the establishment of autocrine TGF-b signaling in dermal fibroblasts derived from localized scleroderma. J. Invest. Dermatol. 126: 1761– collagen gels. Exp. Dermatol. 1: 185–190. by guest on October 1, 2021 1769. 37. Lewis, B. P., C. B. Burge, and D. P. Bartel. 2005. Conserved seed pairing, often 12. Ala-Kokko, L., A. Rintala, and E. R. Savolainen. 1987. Collagen gene expres- flanked by adenosines, indicates that thousands of human genes are microRNA sion in keloids: analysis of collagen metabolism and type I, III, IV, and V targets. Cell 120: 15–20. procollagen mRNAs in keloid tissue and keloid fibroblast cultures. J. Invest. 38. John, B., A. J. Enright, A. Aravin, T. Tuschl, C. Sander, and D. S. Marks. 2004. Dermatol. 89: 238–244. Human MicroRNA targets. PLoS Biol. 2: e363. 13. Smith, J. C., B. E. Boone, S. R. Opalenik, S. M. Williams, and S. B. Russell. 39. Kiriakidou, M., P. T. Nelson, A. Kouranov, P. Fitziev, C. Bouyioukos, 2008. Gene profiling of keloid fibroblasts shows altered expression in multiple Z. Mourelatos, and A. Hatzigeorgiou. 2004. A combined computational- fibrosis-associated pathways. J. Invest. Dermatol. 128: 1298–1310. experimental approach predicts human microRNA targets. Genes Dev. 18: 14. Fujiwara, M., Y. Muragaki, and A. Ooshima. 2005. Keloid-derived fibroblasts 1165–1178. show increased secretion of factors involved in collagen turnover and depend on 40. Krek, A., D. Gru¨n, M. N. Poy, R. Wolf, L. Rosenberg, E. J. Epstein, matrix metalloproteinase for migration. Br.J.Dermatol.153: 295–300. P. MacMenamin, I. da Piedade, K. C. Gunsalus, M. Stoffel, and N. Rajewsky. 15. Syed, F., E. Ahmadi, S. A. Iqbal, S. Singh, D. A. McGrouther, and A. Bayat. 2005. Combinatorial microRNA target predictions. Nat. Genet. 37: 495–500. 2011. Fibroblasts from the growing margin of keloid scars produce higher 41. Long, J. M., and D. K. Lahiri. 2011. MicroRNA-101 downregulates Alzheimer’s levels of collagen I and III compared with intralesional and extralesional sites: amyloid-b precursor protein levels in human cell cultures and is differentially clinical implications for lesional site-directed therapy. Br. J. Dermatol. 164: expressed. Biochem. Biophys. Res. Commun. 404: 889–895. 83–96. 42. Chen, X., Y. Ba, L. Ma, X. Cai, Y. Yin, K. Wang, J. Guo, Y. Zhang, J. Chen, 16. Davidson-Moncada, J., F. N. Papavasiliou, and W. Tam. 2010. MicroRNAs of the X. Guo, et al. 2008. Characterization of microRNAs in serum: a novel class immune system: roles in inflammation and cancer. Ann. N. Y. Acad. Sci. 1183: of biomarkers for diagnosis of cancer and other diseases. Cell Res. 18: 997– 183–194. 1006. 17. Furer, V., J. D. Greenberg, M. Attur, S. B. Abramson, and M. H. Pillinger. 2010. 43. Mitchell, P. S., R. K. Parkin, E. M. Kroh, B. R. Fritz, S. K. Wyman, The role of microRNA in rheumatoid arthritis and other autoimmune diseases. E. L. Pogosova-Agadjanyan, A. Peterson, J. Noteboom, K. C. O’Briant, Clin. Immunol. 136: 1–15. A. Allen, et al. 2008. Circulating microRNAs as stable blood-based markers for 18. Herrera, B. M., H. E. Lockstone, J. M. Taylor, M. Ria, A. Barrett, S. Collins, cancer detection. Proc. Natl. Acad. Sci. USA 105: 10513–10518. P. Kaisaki, K. Argoud, C. Fernandez, M. E. Travers, et al. 2010. Global 44. Hunter, M. P., N. Ismail, X. Zhang, B. D. Aguda, E. J. Lee, L. Yu, T. Xiao, microRNA expression profiles in insulin target tissues in a spontaneous rat model J. Schafer, M. L. Lee, T. D. Schmittgen, et al. 2008. Detection of microRNA of type 2 diabetes. Diabetologia 53: 1099–1109. expression in human peripheral blood microvesicles. PLoS One 3: e3694. 19. Friedman, R. C., K. K. Farh, C. B. Burge, and D. P. Bartel. 2009. Most 45. Valenti, R., V. Huber, M. Iero, P. Filipazzi, G. Parmiani, and L. Rivoltini. 2007. mammalian mRNAs are conserved targets of microRNAs. Genome Res. 19: Tumor-released microvesicles as vehicles of immunosuppression. Cancer Res. 92–105. 67: 2912–2915. 20. LeRoy, E. C., C. Black, R. Fleischmajer, S. Jablonska, T. Krieg, T. A. Medsger, 46. Chen, X., H. Liang, J. Zhang, K. Zen, and C. Y. Zhang. 2012. Horizontal transfer Jr., N. Rowell, and F. Wollheim. 1988. Scleroderma (systemic sclerosis): clas- of microRNAs: molecular mechanisms and clinical applications. Protein Cell 3: sification, subsets and pathogenesis. J. Rheumatol. 15: 202–205. 28–37. 21. Ihn, H., S. Sato, M. Fujimoto, K. Kikuchi, A. Igarashi, Y. Soma, K. Tamaki, and 47. Valadi, H., K. Ekstro¨m, A. Bossios, M. Sjo¨strand, J. J. Lee, and J. O. Lo¨tvall. K. Takehara. 1996. Measurement of anticardiolipin by ELISA using 2007. Exosome-mediated transfer of mRNAs and microRNAs is a novel b2-glycoprotein I (b2-GPI) in systemic sclerosis. Clin. Exp. Immunol. 105: 475– mechanism of genetic exchange between cells. Nat. Cell Biol. 9: 654–659. 479. 48. Uziel, Y., B. R. Krafchik, B. Feldman, E. D. Silverman, L. A. Rubin, and 22. Sato, S., M. Fujimoto, H. Ihn, K. Kikuchi, and K. Takehara. 1994. Clinical R. M. Laxer. 1994. Serum levels of soluble interleukin-2 receptor. A marker of characteristics associated with antihistone antibodies in patients with localized disease activity in localized scleroderma. Arthritis Rheum. 37: 898–901. scleroderma. J. Am. Acad. Dermatol. 31: 567–571. 49. Takehara, K., K. Kikuchi, Y. Soma, A. Igarashi, and Y. Ishibashi. 1990. Anti- 23. Falanga, V., T. A. Medsger, and M. Reichlin. 1985. High titers of antibodies to single-stranded DNA and muscle involvement in localized sclero- single-stranded DNA in linear scleroderma. Arch. Dermatol. 121: 345–347. derma. Arch. Dermatol. 126: 1368. The Journal of Immunology 11

50. Takeshita, F., L. Patrawala, M. Osaki, R. U. Takahashi, Y. Yamamoto, Th1 polarization, and the severity of delayed-type . J. Immu- N. Kosaka, M. Kawamata, K. Kelnar, A. G. Bader, D. Brown, and T. Ochiya. nol. 187: 3362–3373. 2010. Systemic delivery of synthetic microRNA-16 inhibits the growth of meta- 52. Lanford, R. E., E. S. Hildebrandt-Eriksen, A. Petri, R. Persson, M. Lindow, static prostate tumors via downregulation of multiple cell-cycle genes. Mol. Ther. M. E. Munk, S. Kauppinen, and H. Ørum. 2010. Therapeutic silencing of microRNA- 18: 181–187. 122 in primates with chronic hepatitis C virus infection. Science 327: 198–201. 51.Lu,T.X.,J.Hartner,E.J.Lim,V.Fabry,M.K.Mingler,E.T.Cole, 53. Nakasa, T., H. Shibuya, Y. Nagata, T. Niimoto, and M. Ochi. 2011. The inhib- S. H. Orkin, B. J. Aronow, and M. E. Rothenberg. 2011. MicroRNA-21 limits itory effect of microRNA-146a expression on bone destruction in collagen- in vivo immune response-mediated activation of the IL-12/IFN-g pathway, induced arthritis. Arthritis Rheum. 63: 1582–1590. Downloaded from http://www.jimmunol.org/ by guest on October 1, 2021