TNF Expression Noncoding Rnas and LRRFIP1 Regulate

Noncoding RNAs and LRRFIP1 Regulate TNF Expression Lihua Shi, Li Song, Michael Fitzgerald, Kelly Maurer, Asen Bagashev and Kathleen E. Sullivan This information is current as of October 1, 2021. J Immunol published online 24 February 2014 http://www.jimmunol.org/content/early/2014/02/21/jimmun ol.1302063 Downloaded from Supplementary http://www.jimmunol.org/content/suppl/2014/02/21/jimmunol.130206 Material 3.DCSupplemental

Why The JI? Submit online. http://www.jimmunol.org/

• 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

*average by guest on October 1, 2021 Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2014 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published February 24, 2014, doi:10.4049/jimmunol.1302063 The Journal of Immunology

Noncoding RNAs and LRRFIP1 Regulate TNF Expression

Lihua Shi, Li Song, Michael Fitzgerald, Kelly Maurer, Asen Bagashev,1 and Kathleen E. Sullivan

Noncoding RNAs have been implicated in the regulation of expression of numerous genes; however, the mechanism is not fully understood. We identiﬁed bidirectional, long noncoding RNAs upstream of the TNF gene using ﬁve different methods. They arose in a region where the repressors LRRFIP1, EZH2, and SUZ12 were demonstrated to bind, suggesting a role in repression. The noncoding RNAs were polyadenylated, capped, and chromatin associated. Knockdown of the noncoding RNAs was associated with derepression of TNF mRNA and diminished binding of LRRFIP1 to both RNA targets and chromatin. Overexpression of the noncoding RNAs led to diminished expression of TNF and recruitment of repressor proteins to the locus. One repressor protein, LRRFIP1, bound directly to the noncoding RNAs. These data place the noncoding RNAs upstream of TNF gene as central to the transcriptional regulation. They appear to serve as a platform for the assembly of a repressive complex. The Journal of Immunology, 2014, 192: 000–000. Downloaded from

umor necrosis factor is a member of a family of proteins TNF mRNA turnover (32–34). Collectively, these studies dem- that regulate immunologically competent cells. It is pre- onstrate that the regulation of TNF is rigorous and redundant, T dominantly produced by myeloid cells, activated T cells, presumably to limit the adverse consequences related to under- or and NK cells. The major roles of TNF include killing of tumor overexpression. cells, the induction of adhesion molecule expression at sites of In our studies of chromatin at the TNF locus, we identified inflammation, stimulation of bone resorption, induction of fever, a region 300 bp upstream of the transcriptional start site (TSS) http://www.jimmunol.org/ and activation of B cells, neutrophils, and monocytes (1–3). TNF where the majority of the transcriptionally relevant histone mod- inhibition is used therapeutically for arthritis and inflammatory ifications were found (19). We also identified a transcriptional bowel disease, and inhibition is associated with an increased risk repressor called LRRFIP1 (previously called GCF2) (35). Further of infection (4–8). Conversely, overexpression of TNF in murine evidence that this region might be important in the regulation of models is associated with pathologic inflammation (9–14). Thus, TNF came from a study of patients with systemic lupus erythe- regulation of TNF is of paramount importance. matosus, which found that the histone modifications at this site TNF is regulated at the level of chromatin, transcription, splicing, were different in patients compared with controls (36). This led us message turnover, and cleavage from the membrane (15–19). DNA to examine the upstream region of the TNF promoter more care- by guest on October 1, 2021 methylation inhibitors and histone deacetylase inhibitors can in- fully. We found significant levels of noncoding RNAs (ncRNAs) duce TNF expression, supporting a role for chromatin in the that mapped to this region. regulation of TNF transcription (19–21). Priming of TNF tran- ncRNAs are common in the genome, with ∼8000 identified (37). scription requires PU.1 and CEBP proteins, as is true for most In general, their abundance, conservation, and correlation with lineage-specific transcripts in monocytes (22–25). After stimula- transcription have argued for functionality, but there are relatively tion, NF-kB, AP1, and ETS family members bind to specific few specific examples known (38–41). The best-known ncRNA promoter motifs and drive active elongation (26–28). As is true for that regulates chromatin conformation is Xist, which coats the many highly inducible genes, message turnover is highly regulated X-chromosome destined for inactivation (42, 43). Long noncoding with tristetraprolin and TIA-1, predominantly responsible for RNAs have been implicated in pluripotency and innate immune destabilizing and repressing translation, respectively (10, 29–31). responses (44–46). Several studies have found that chromatin- HuR, TIAR, KSRP, and microRNAs have also been implicated in associated RNAs are bound to chromatin-modifying complexes on chromatin, suggesting a role in epigenetic regulation (47–49). In general, the ncRNA is thought to confer locus specificity and Division of Allergy Immunology, Children’s Hospital of Philadelphia, University of Pennsylvania School of Medicine, Philadelphia, PA 19104 alter local histone modifications, but the specific mechanisms for 1Current address: Molecular Studies of Neurodegenerative Diseases Laboratory, Fels each gene appear to be diverse and are largely not understood Institute for Cancer Research and Molecular Biology, School of Medicine, Temple (50, 51). University, Philadelphia, PA. Several groups have manipulated ncRNAs in an effort to dissect Received for publication August 9, 2013. Accepted for publication January 12, 2014. their exact function. The most common model is one in which the This work was supported by National Institutes of Health Grants 1RO1 AI 51323 and ncRNAs regulate H3K9me2 and H3K27me3 marks in cis and R21 AI090914 and the Wallace Chair of Pediatrics. mediate transcriptional repression (52–58). Nevertheless, a recent Address correspondence and reprint requests to Dr. Kathleen E. Sullivan, Division study found that many ncRNAs regulate gene expression in trans, of Allergy Immunology, Children’s Hospital of Philadelphia, 3615 Civic Center Boulevard, Philadelphia, PA 19104. E-mail address: [email protected] suggesting that there are many more mechanisms yet to be iden- The online version of this article contains supplemental material. tified (45). Despite the rapid increase in our understanding of Abbreviations used in this article: ChIP, chromatin immunoprecipitation; IRF1, IFN RNA-mediated transcriptional repression, much remains to be regulatory factor 1; LTA, lipoteichoic acid; ncRNA, noncoding RNA; qRT-PCR, learned regarding the mechanisms of repression. quantitative RT-PCR; RIP, receptor interacting protein; RNA-IP, RNA immunopre- This study was undertaken to examine ncRNAs upstream of the cipitation; shRNA, short hairpin RNA; TSS, transcriptional start site. TNF gene. We found a tightly linked choreography of ncRNAs Copyright Ó 2014 by The American Association of Immunologists, Inc. 0022-1767/14/$16.00 and repressors on intergenic chromatin upstream of TNF. Further-

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1302063 2 NONCODING RNA REGULATION OF TNF EXPRESSION more, we identified a novel function of the transcriptional repressor K562 RNA, according to the manufacturer’s recommendation. The RNA LRRFIP1. extracted from the supernatant or wash buffer was used as a negative control. Controls using 18S (noncapped, nonpolyadenylated) and IL-1b (capped, polyadenylated) were used to confirm the appropriate recovery. Materials and Methods Cells, transfections, and reagents Northern blot, run-on assay, and RNA-binding assay All cell types are human. K562 is a hematopoietic stem cell–like line. THP1 For Northern analysis, total RNA from K562 was isolated with TRIzol is an immature monocytic leukemia line. MonoMac6 cells are a more reagent (Invitrogen). A 674-bp 32P-labeled sense or antisense RNA probe highly differentiated monocytic cell line (59). Each was maintained in was generated by in vitro transcription with the MAXIscript kit (Invi- RPMI 1640 with 10% cosmic calf serum (Fisher Scientific, Pittsburgh, trogen). Hybridization was performed using QuickHyb (Agilent, La Jolla, PA). Primary monocytes were obtained from normal human donors and CA), following the manufacturer9s instructions. purified by elutriation at the Penn Center for AIDS Research and then The modified run-on assay was performed, as described (64). Nuclei ∼ further purified by adherence. They were 90% pure, as defined by CD14 were isolated, and transcription was allowed to proceed in the presence of expression. Transfection of cells was performed by electroporation with biotin-16-UTP (Roche, Indianapolis, IN). The nascent transcripts were the Amaxa Cell Line Lonza Nucleofector Kit (Amaxa Biosystems, collected on avidin magnetic beads (Dynal, Invitrogen) and AMV reverse Gaithersburg, MD). HPLC-purified LPS and PMA were from Sigma-Aldrich transcriptase used to generate cDNA. The proximal cDNA was quantitated (St. Louis, MO). The SMART vector 2.0 Lentiviral LRRFIP1 short hairpin by quantitative RT-PCR (qRT-PCR) using custom primers with actin pri- RNA (shRNA) or nontargeting negative control viral particles were pur- mers (Applied Biosystems) for normalization. chased from Dharmacon (Chicago, IL). The target sequences of three Three different-sized ncRNAs (674, 377, and 150) were produced 9 LRRFIP1-shRNA used in the studies were as follows: 5 -GGUUAU- by different primers (Supplemental Table I) and then ligated into the 9 9 9 9 CACCCAGAUUAGA-3 ,5-AAUGGAGAGACUUCCGACA-3 , and 5 - pGEM-T Easy vector. The sense and antisense ncRNAs were generated 9 GUAGGGAUCACAACGAAGA-3 . Three shRNAs were evaluated, and by in vitro transcription with T7 or SP6 polymerase using MaxiScript the most effective was used for the experiments. Transduction of cells (Invitrogen). The 300-bp actin RNA derives from the 39 end of ACTB, Downloaded from and stable cell line development were done following the manufacturer’s and the GFP RNA derives from 735 bp of GFP cDNA. The RNA- 9 protocol. The EZH2 small interfering RNA (siRNA) sequence was 5 - binding assay was modified from that described (65). In brief, different- 9 9 GGAUGGUACUUUCAUUGAATT-3 , and the SUZ12 sequence was 5 - sized ncRNAs were incubated with purified protein in gel shift buffer GGAUAGAUGUUUCUAUCAATT-39 (Ambion Intvitrogen, Grand Island, (10 mM MOPS [pH 7.0], 50 mM KCl, 5 mM MgCl2, 1 mM DTT, 10% NY). The ncRNA overexpression system used the PMK CMV vector with glycerol and heparin) at 4˚C for 30 min. The binding reactions were the ncRNA ligated downstream of the CMV promoter. loaded onto 0.5% agarose gel (prepared by TB buffer) and run in TB buffer

RNA extraction, quantitative real-time PCR, and NanoString (45 mM Tris, 45 mM boric acid). http://www.jimmunol.org/ nCounter assay Chromatin and nuclear RNA immunoprecipitations RNA was prepared, DNase treated, and reverse transcribed using the Ad- vantage RT for PCR Kit (Clontech, Mountainview, CA). Nuclear RNA was Chromatin immunoprecipitation (ChIP) assays were carried out as previ- extracted with the PARIS Kit (Ambion, Grand Island, NY). Primer–probe ously described (66, 67) and used the Abs for SUZ12 (Abcam, Cambridge, combinations for all targets are listed in Supplemental Table I. Custom MA), LRRFIP1 (Sigma-Aldrich), H3K27me3, and EZH2 (Millipore, primers and labeled probes were synthesized by IDT (Coralville, IA). Billerica, MA). A negative control Ab (anti-GST; Abcam) was always Spliced TNF mRNA was detected by proprietary gene-specific primers included but is omitted from some of the figures for simplicity. Duplicates or triplicates were analyzed from each experiment. ChIP data were nor- from Applied Biosystems on a TaqMan SDS 7900HT. Primers to amplify 3 2 18S rRNA were included in each amplification and served as the internal malized to input according to the following formula: 2 (10% input Ct standard for normalization. sample Ct). by guest on October 1, 2021 Nuclear RNA immunoprecipitation (RNA-IP) was prepared as described NanoString nCounter analysis was performed using NanoString custom- 3 7 synthesized probes (NanoString, Seattle, WA). Total RNA, nuclear RNA, or by Rinn et al. (68), with a few modifications. Briefly, 4 10 cells were lysate was directly hybridized with gene-specific color-coded probes, and harvested and resuspended in 2 ml PBS, 2 ml nuclear isolation buffer (1.28 M data collection was carried out in the nCounter Digital Analyzer, as de- sucrose, 40 mM Tris-HCl [pH 7.5], 20 mM MgCl2, 4% Triton X-100), and 6 ml water on ice for 20 min. The nuclei were obtained by douncing scribed by the manufacturer. Transcript numbers for each gene were nor- 3 malized to the mean of housekeeping genes. In addition, six positive-control with pestle B. Nuclei were pelleted by centrifugation at 2,500 g for 15 and eight negative-control probes were added to each reaction. Normali- min and resuspended in 1 ml receptor interacting protein (RIP) buffer (150 mM KCl, 25 mM Tris [pH 7.4], 5 mM EDTA, 0.5 mM DTT, 0.5% IGE- zation then used the top 100 geometric mean. All the reaction counts were m m m m within the linear dynamic range of the standard curve. PAL, 9 g/ml leupeptin, 9 g/ml pepstatin, 10 g/ml chymostatin, 3 g/ The 59 and 39 RACE were performed with the FirstChoice RLM-RACE ml aprotinin, 1 mM PMSF, 100 U/ml RNasin). Nuclear membranes and Kit (Ambion), according to the manufacturer’s instructions. Total RNA debris were pelleted by centrifugation at 13,000 rpm for 10 min. Ab was was extracted from K562 and treated with DNase I. The PCR products added to supernatant and incubated for 2 h at 4˚C. Protein A beads were were purified and cloned into a pGEM-T vector (Promega) for sequencing. added and incubated for 1 h at 4˚C with gentle rotation. The beads were pelleted at 2,500 rpm for 30 s, resuspended in 500 ml RIP buffer, and Subcellular fractionation, m7G-cap analysis, and poly A+ RNA repeated for a total of three RIP washes, followed by one wash in PBS. The purification beads were resuspended in 1 ml TRIzol. Coprecipitated RNAs were isolated and reverse transcribed, as above. Real-time PCR was performed Subcellular fractions were prepared as described, with a few modi- with TNF1–4 primers and normalized to 18S rRNA. fications (60, 61). A total of 5–10 3 106 K562 cells was lysed in RSB-100 buffer (100 mM Tris-HCl [pH 7.4], 100 mM NaCl, 2.5 mM Protein analysis MgCl2,50mg/ml digitonin, 100 U/ml RNasin, 13 phosphatase inhibitor mixture [Sigma-Aldrich]), followed by centrifugation at 2000 3 g for The TNF ELISAs used the BD Biosciences BD OptEIA TNF ELISA kit. A 8 min. The supernatant was collected as the cytosolic fraction. The nu- BIA3000 instrument (Biacore, Piscataway, NJ) was used to detect binding clear pellet was then resuspended in RSB-100 T (0.5% Triton X-100 in to the TNF promoter RNA or DNA sequences. Biotinylated probes centered RSB-100). After centrifugation at 2000 3 g for 8 min, the supernatant on 2308, consisting of 25-bp dsRNA or DNA or DNA:RNA and a 6-bp was collected as the nuclear fraction. The resulting chromatin pellet spacer, were captured to the CM5 sensor chip via streptavidin. The pro- was resuspended in RSB-100T and sonicated. The soluble DNA-bound teins were suspended in running buffer (10 mM HEPES [pH 7.4], 150 mM RNA fraction was collected after centrifugation at 4000 3 g for 15 NaCl, 0.005% Tween 20). The experiments were performed at 25˚C with min. RNA was extracted with TRIzol (Invitrogen, Grand Island, NY) a flow rate of 30 ml/min. Kinetic constants were obtained by fitting the and treated with RNase-free DNase I (Qiagen, Valencia, CA). Ab for data to a simple one-to-one binding model using BIA EVALUATION 3.1 the m7G-cap was from Synaptic Systems (Goettingen, Germany), and software. we performed the immunoprecipitation, according to the manufac- The coimmunprecipitations used EZH2 and SUZ12 Abs. Protein A turer’s instructions, with 30 mg total K562 RNA and protein G beads beads were used for collection (Invitrogen). Protein was quantitated using (GeneScript, Piscataway, NJ) (62, 63). The RNA-IP and the non–RNA- a Bradford assay, equally loaded on 4–12% NuPAGE gel (Invitrogen Life IP were reverse transcribed and analyzed by quantitative PCR. PolyA+ Technologies, Carlsbad, CA), and blotted with LRRFIP1 Ab (BD Bio- RNA was purified by oligo(dT)-cellulose (Sigma-Aldrich) with 150 mg sciences, Franklin Lakes, NJ). The Journal of Immunology 3

Results very low levels of TNF message. HeLa cells are fully repressed Identification of TNF upstream ncRNAs for TNF expression. Jurkat, THP1, and primary monocytes do not produce TNF protein at baseline, but, in each case, the cells can be We previously identified a transcriptional repressor of TNF, induced to express significantly more TNF protein and mRNA after LRRFIP1, that was a known RNA-binding protein (35, 69). To stimulation. All probes detected ncRNA signal in five different cell understand potential mechanisms, UCSC genome browser was types, confirming the presence of ncRNAs (Fig. 1D, 1E). used to explore the existence of ncRNAs upstream of the TNF We further confirmed the presence of the ncRNAs by RACE, TSS. GM12878 and K562 tracks both showed low-abundance which showed heterogeneous lengths of sense and antisense ncRNAs upstream of the TNF TSS. We also explored our own ncRNAs. The shortest product was ∼200 bp, and the longest prod- RNA-seq data for evidence of transcription upstream of the TNF uct was .900 bp (Supplemental Fig. 1B). To better define the sizes TSS and found low levels of transcripts identified, similar to those of the RNAs, we used directional primers and RT-PCR (shown seen in the UCSC tracks (Supplemental Fig. 1A). We examined on the map in Fig. 1). This showed strong bands of 600 bp and the coding potential of these transcripts through in silico transla- smaller in both sense- and antisense-primed cDNAs (Fig. 2A). An tion and found a very limited potential for peptide production. 800-bp band was seen faintly in the sense-primed sample. Northern To directly identify ncRNAs, we initially designed qRT-PCR blot confirmation was attempted, but the signal was dispersed at amplimers across this region, as follows: TNF1, 2, 3, 4, and 5 ∼700 bp, and no clear band was visualized (data not shown). There- (Fig. 1A). RNA species were detected in the lymphotoxin alpha fore, the ncRNAs are bidirectional and heterogeneous in length, (LTA)–TNF intergenic region in K562 cells after reverse transcrip- withapeaklengthof∼600–700 bp. tion of RNA with directional primers. RNase treatment abrogated the The ncRNA structure signal, demonstrating that the amplification was truly RNA. The 39 Downloaded from end of the LTA gene lies only 1.24 Kb upstream of the TNF TSS, To examine the structure of the ncRNAs, we used an Ab for the 59 and RNA species were found throughout this region (Fig. 1B, 1C). m7G-cap structure to determine whether the ncRNAs are capped We wanted to confirm our detection of the ncRNAs without the and oligo(dT) cellulose to capture polyadenylated RNAs. These potential confounder of skewing due to PCR amplification, and we two strategies determined that the ncRNAs were largely capped therefore used a digital method called nCounter. Seven sense or and polyadenylated (Fig. 2B, 2C). Controls for the oligo(dT) and

antisense probes were designed across the intergenic region, as Ab capture demonstrated the appropriate modiﬁcation patterns for http://www.jimmunol.org/ follows: S 1–7 and AS A–G (Fig. 1A). We compared the ncRNAs 18S and IL-1b mRNAs (data not shown). Different species of in several cell types of differing competence for TNF expression. ncRNAs localize to different cellular compartments and are as- The K562 cell line does not make TNF protein, although it produces sociated with different posttranscriptional modiﬁcations. To lo- by guest on October 1, 2021

FIGURE 1. ncRNAs identified upstream of TNF. RT-PCR and direct detection were used to characterize the ncRNAs upstream of TNF. (A) A map demonstrates the location of the primers and probes to detect ncRNAs. Nanostring nCounter probes for digital detection are labeled S1-S7 (for sense targeting) and AS-A-G (for antisense targeting). TNF1–5 represent the qRT-PCR amplimers. The 674-bp ncRNA used for overexpression studies is shown near the bottom of the map. The 410 and 430 phosphorothioate oligonucleotides used for knockdown of sense and antisense ncRNAs, respectively, are shown. The LRRFIP1 binding site at 2308 previously identified is indicated with an oval (37). The arrows indicate the direction of the oligonucleotide sequences. (B) Sense ncRNAs and (C) antisense ncRNAs were identified by qRT-PCR from total RNA with or without RNase A treatment. Directional reverse-transcription primers at exon 1 (for sense) and at 2991 (for antisense) were used. The chart legend applies to both sense and antisense graphs. All the differences between K562 RNA and RNase A-treated K562 RNA and the No-RT K562 RNA (No RT) were significant (p , 0.05, n = 3). Error bars denote SE. NanoString nCounter technology was used to quantify the (D) sense or (E) antisense ncRNAs in five different types of cells. 4 NONCODING RNA REGULATION OF TNF EXPRESSION

FIGURE 2. ncRNA structure and localization. The biochemical structure of the ncRNAs was defined. (A) Detection of ncRNA transcripts by RT-PCR. Downloaded from RNA from K562 cells was reverse transcribed using either no primer (2), sense (S)-, or antisense (AS)-specific primers. For the PCR amplification, the reverse primer was TNF3R, and the locations of the forward primers are shown on the bottom of the gel. The primers TNF3, 4, 5, 2974, and 2606 refer to positions as shown in Fig. 1A. The strong bands at the bottom of the gel for the 2974F and TNF5F amplifications represent primer dimers and not authentic amplification. This gel is representative of three experiments. (B) Poly A+ RNA was purified by oligo(dT)-cellulose, and the RNA extracted from supernatant was used as the negative control. Poly A+ selection enriched for ncRNA recovery detected by qRT-PCR. n =4.(C) qRT-PCR was performed for the RNA immunoprecipitated by M7G, IgG, or mock immunoprecipitated. The M7G-cap–precipitated material enriched the ncRNA. n =4.(D) RNA was http://www.jimmunol.org/ extracted from K562 cytoplasm, nucleoplasm, and chromatin, and qRT-PCR was performed after reverse transcription with random primers. The ncRNAs localized primarily to the chromatin fraction. n = 3. Error bars in panels denote SE. *p , 0.05, **p , 0.01. calize the ncRNAs, we fractionated K562 cells into cytoplasm, cells, we treated the cells with low-dose PMA to induce low levels nucleoplasm, and chromatin. The ncRNAs were detected by qRT- of TNF mRNA. We found that transfection of ncRNAs did suc- PCR and were found almost exclusively in the chromatin fraction cessfully target chromatin (Fig. 4C). Transfection of antisense (Fig. 2D). Thus, the ncRNAs are chromatin associated, capped, ncRNA led to significantly decreased TNF mRNA levels with

and polyadenylated. transfection of sense ncRNA having a more modest effect (Fig. 4D). by guest on October 1, 2021 We additionally examined the effects in THP1 cells, a monocyte Function of the ncRNAs line competent for expression. We did not observe any effect of Knowing that the ncRNAs were chromatin associated, we hypoth- knockdown of the ncRNAs (data not shown), probably because the esized that the peaks of RNA at the 39 end of LTA and the 59 end ncRNA levels are already so low in THP1 cells (Fig. 3D). Over- of TNF represented “spill” from adjacent LTA and TNF mRNAs expression of the ncRNAs was associated with repression of TNF (Fig. 1). We therefore purified nuclear RNA to eliminate most mRNA both in resting and stimulated THP1 cells (Fig. 4E). To mRNAs and were still able to detect signal across the region using demonstrate the specificity of the effect, we examined IFN regu- nCounter, without the peaks near the ends of LTA and TNF (Fig. 3A, latory factor 1 (IRF1), a transcription factor induced by LPS. 3B). We noted that the repressed K562 nuclear RNA preparations Transfection of the ncRNAs had no effect on IRF1 transcript had higher levels of the ncRNAs than the competent THP1 nuclear abundance in THP1 cells (Fig. 4F). Furthermore, the effect of RNA preparations (Fig. 3A, 3B). When THP1 cells were stimulated the ncRNAs was demonstrable at the level of protein production with LPS, there was a rapid increase in TNF message (Fig. 3C), as (Fig. 4G). Therefore, knockdown of the ncRNAs derepressed expected. There was a concomitant decline of sense and antisense expression of TNF mRNA, and overexpression led to repression ncRNAs in the 2300regionofTNFpromoter(Fig.3D). of TNF mRNA. These studies suggested that these ncRNAs are To determine whether these ncRNAs function in the regulation mechanistically involved in repression of TNF. of TNF expression, we designed a set of phosphorothioate oligonucleotides to knock down these ncRNAs. They had variable efficacy Chromatin characteristics at the site of the ncRNAs derepressing TNF mRNA and variable efficacy at knocking down the To better understand the chromatin environment at the site of the ncRNAs (Supplemental Fig. 2). We selected a pair of oligonucleo- ncRNAs, we first examined histone modifications. One epigenetic tides (410, 430) that successfully targeted the ncRNA, derepressed mark of particular interest is trimethylation of lysine 27 on histone TNF expression and were near the site where we had previously 3H (H3K27me3). EZH2 and SUZ12 are two important polycomb identified histone marks of repression (Fig. 1A). In the repressed protein family members that have been implicated in ncRNA- K562 cell type, the 410 and 430 oligonucleotides effectively knocked mediated repression (70–72). EZH2 methylates H3K27, and SUZ12 down the ncRNAs and derepressed the TNF message (Fig. 4A, 4B). regulates methyltransferase activity at both H3K9 and H3K27. We Although the level of TNF mRNA is quite low in these cells, the used ChIP assays to investigate chromatin marks of repression as depletion of the ncRNAs clearly led to increased spliced message. well as the presence of EZH2 and SUZ12. We also examined a To further investigate the role of the ncRNAs, we made ncRNA known repressor of TNF, LRRFIP1, which was of interest because overexpression constructs of 674 bp in both the sense and antisense of its reported RNA-binding function. As expected, H3K27me3, orientation (Fig. 1A). Because K562 cells are already repressed EZH2, SUZ12, and LRRFIP1 were all present on the TNF pro- The Journal of Immunology 5

FIGURE 3. ncRNA levels are inversely related to TNF mRNA expression. Nano- String nCounter analysis indicated that there were more nuclear (A) sense and (B) antisense ncRNAs in K562 than THP1 cells. n =2.(C) TNF mRNA levels greatly increased after LPS stimulation of THP1 cells. n =3.(D) LPS stimulation led to decreased ncRNAs in THP1 cells. n =2. Cells were treated with 1 mg/ml LPS for 30 or 60 min, and NanoString technology was used to measure nuclear ncRNAs and total TNF message. Error bars in panels denote SE. *p , 0.05, **p , 0.01, ***p , 0.001. (E) The purity of the cell fractions for the nuclear RNA analyses was conﬁrmed by Western blot. Heat shock protein 90 was used as a marker for cytoplasm, and histone 3H was used as a marker for nuclear material. Downloaded from

moter in repressed K562 cells but not competent THP1 cells itation was performed, demonstrating that EZH2 and SUZ12 both (Fig. 5). They appeared to occupy a similar genomic space with interacted with LRRFIP1 (Fig. 5E). a peak of ampliﬁcation centered in the TNF3 amplimer site. http://www.jimmunol.org/ These data suggested that a complex of proteins resided on the ncRNAs interact with LRRFIP1 and are required for LRRFIP1 repressed promoter and that the repressed state was associated binding to chromatin with H3K27me3. To determine whether LRRFIP1 might directly LRRFIP1 has been previously described as a RNA-binding protein interact with EZH2 and SUZ12, noncross-linked coimmunoprecip- (69), and we hypothesized that it might interact with chromatin via

FIGURE 4. The ncRNA regulation of TNF mRNA expression. Levels of ncRNAs were mod- by guest on October 1, 2021 ulated to examine the effect. (A) K562 cells were transfected with the indicated 20-bp phosphorothioate oligonucleotides that knock down either the sense (410) or the antisense (430) transcripts. The ncRNAs were measured directly using qRT- PCR with random primers, demonstrating appropriate knockdown. n =5.(B) Both sense and antisense ncRNA knockdown increased TNF mRNA in K562 cells. n =5.(C) K562 cells were pre- treated with 10 ng/ml PMA for 3 d to induce low levels of TNF expression and then transfected with the 674 ncRNA sense (S), antisense (AS) overexpression constructs or the empty vector alone (PMK). These exogenously transcribed ncRNAs successfully targeted chromatin. n =3.(D) ncRNA overexpression decreased the TNF mRNA level in low-dose PMA-treated K562 cells. n =3.(E) The effect of ncRNA overexpression was examined in THP1 cells. Both basal and LPS-induced TNF expression was diminished in the ncRNA-trans- ected cells compared with vector (PMK) alone. n =4.(F) As a speciﬁcity control, we measured IRF1 transcript abundance from the same cultures. No effect was seen. n =4.(G) THP1 cells were transfected with the overexpression constructs or the control vector. Cells were stimulated with LPS, and supernatants were collected. A TNF ELISA was used to quantitate the TNF protein production (n = 2). Error bars in panels denote SE. *p , 0.05, **p , 0.01, ***p , 0.001. Asterisks refer to the difference between the 410 or 430 and LacZ. 6 NONCODING RNA REGULATION OF TNF EXPRESSION

FIGURE 5. Chromatin characteristics upstream of TNF. Chromatin marks were compared in K562 and THP1 cell lines. ChIP assays showed that (A) H3K27me3, (B) EZH2, and (C) Suz12 were increased in K562 cells compared with THP1 cells. n =5. (D) LRRFIP1 was also found to bind to the same region of TNF promoter in K562 cells. n = 5. Error bars in all panels denote SE. *Indicates the p value of K562 compared with THP1 cells for the same Ab. *p , 0.05, **p , 0.01. (E) Coimmunoprecipi- tation analysis of the interaction of LRRFIP1 with SUZ12 and EZH2 in K562 cells. The lysates were precipitated with anti-IgG, EZH2, or SUZ12 and analyzed by Western blot with Downloaded from LRRFIP1 Abs. This is one representative Western blot from three experiments. http://www.jimmunol.org/ a RNA-mediated process. LRRFIP1 exists as two major RNA LFFRIP1 equally (Supplemental Fig. 3F), but there did not seem splice variants, which in turn lead to two very different proteins to be a strict sequence requirement as sense and antisense were (Supplemental Fig. 3A). We made a full-length exon 2 start equally capable of binding. MFold (73) was used to predict the LRRFIP1 construct that expressed the 160-kDa LRRFIP1 protein, secondary structure of the 674 antisense ncRNA and predicted and we made a truncated LRRFIP1 construct that was predicted to a highly folded arrangement that would lead to extensive dsRNA lack the entire DNA-binding region and half of the RNA-binding structures even when transcribed as a single-stranded species

domain (Supplemental Fig. 3A). Using a biosensor approach with (Supplemental Fig. 3H). Thus, this ncRNA would have the struc- by guest on October 1, 2021 32-bp oligonucleotides centered on the LRRFIP1 binding site at tural requirements that have been identified as ideal for LRRFIP1 2308 or an off-target sequence, we examined binding to different binding. nucleic acid structures. We found that the full-length LRRFIP1 Therefore, LRRFIP1 was found to bind dsRNA . ssRNA by bound to dsRNA, ssRNA, DNA:RNA, dsDNA, and ssDNA, but both biosensor and RNA-binding analyses. To investigate LRRFIP1 it had higher affinity for dsRNA, ssRNA, and DNA:RNA compared binding to ncRNA in vivo, we performed a RNA-IP using K562 with dsDNA (Table I). It recognized a dsRNA off-target structure nuclei. ncRNAs were detected by qRT-PCR from LRRFIP1- (actin) similarly. In contrast, the truncated LRRFIP1 protein and immunoprecipitated nuclear RNA but not with H3K9me3 or GST did not exhibit any binding over background (data not shown). GST-negative control Abs (Fig. 6B). Globin mRNA was also not To further confirm direct RNA binding, a RNA-binding assay recognized, thereby demonstrating specificity. These analyses was performed (Fig. 6A). A RNA species of 674 bp was produced demonstrated LRRFIP1 binding to RNA by in vitro and in vivo. by in vitro transcription (Fig. 1A) and was incubated with several We wished to determine whether LRRFIP1 binding to chromatin different proteins, including LRRFIP1. The free RNA appeared as was dependent on the ncRNAs. Transfection of both the reverse a low smear, whereas the LRRFIP1-RNA complex appeared as oligonucleotide 410 (targeting sense ncRNA) and forward oligo- a shifted, higher band, confirming direct RNA binding. There was nucleotide 430 (targeting antisense ncRNA) led to diminished no binding to tRNA. We further examined the structural require- LRRFIP1 binding to the upstream ncRNAs in K562 nuclei, with ments of RNA binding of LRRFIP1 (Supplemental Fig. 3). Longer antisense targeting having slightly more of an effect. Mock- RNAs bound LRRFIP1 better than shorter species. dsRNA of all transfected cells and LacZ oligonucleotide-transfected cells were sizes bound LRRFIP1 better than ssRNA. Not all RNAs bound comparable, and immunoprecipitation with a GST Ab demon-

Table I. Biosensor analysis of LRRFIP1 binding

Ka (1/MS) Kd (1/S) KD (M) p Value (Compared with dsRNA) dsRNA 1.62E+04 (64.09E+3) 4.36E-04 (61.3E-4) 3.97E-08 (62.73E-08) dsDNA 1.25E+04 (69.19E+3) 2.39E-03 (69.1E-4) 2.46E-07 (68.90E-08) ,0.0005 DNA-RNA 2.01E+04 (61.10E+4) 1.07E-03 (63.60E-4) 5.74E-08 (61.82E-08) =0.245 ssDNA 4.80E+04 (63.60E+3) 4.44E-03 (61.78E-3) 1.85E-07 (61.21E-07) ,0.05 ssRNA 1.31E+04 (68.05E+3) 9.18E-04 (65.2E-4) 7.00E-08 (65.81E-08) =0.1881 Off-target dsRNA 1.91E+04 9.72E-04 5.08E-08 Average of ﬁve experiments for TNF oligonucleotides. The Journal of Immunology 7 Downloaded from

FIGURE 6. LRRFIP1 interactions. LRRFIP1 is a RNA-binding protein implicated in TNF repression. (A) LRRFIP1 bound ncRNAs in vitro. The 674 ncRNA species or yeast tRNA were incubated with puriﬁed full-length LRRFIP1 (F-LRRFIP1) or truncated LRRFIP1 protein (T-LRRFIP1). GST and CLCA1 protein were used as negative controls. The free RNA is indicated as the (2) lane. Only full-length LRRFIP1 bound to the 674 ncRNA. LRRFIP1 bound tRNAs very poorly. This is representative of three experiments. (B) LRRFIP1 bound ncRNAs in vivo. A RNA IP was performed with LRRFIP1, http://www.jimmunol.org/ H3K9me3, or GST Abs using K562 nuclei. The recovered RNA was reverse transcribed and ampliﬁed with the indicated upstream primer/probe pairs. LRRFIP1 immunoprecipitation brought down the ncRNAs but not globin. n =3.(C) K562 cells were transfected with the indicated 20-bp phosphorothioate oligonucleotides that knock down either the sense or the antisense transcripts. Immunoprecipitation of LRRFIP1 recovered less ncRNA in the knockdown cells. n =3.(D) LRRFIP1 binding to the TNF promoter was diminished in K562 after transfection of the 430 oligo (which knocked down antisense ncRNA) or the 410-transfected cells (which knocked down sense ncRNA). A GST ChIP is included as a negative control. n = 5. Error bars in panels denote SE. *p , 0.05. All asterisks refer to the difference between the 410 or 430 and LacZ.

strated background levels of signal (Fig. 6C). These data dem- LRRFIP1 regulation of TNF mRNA by guest on October 1, 2021 onstrated that the 410 and 430 oligonucleotides interfere with Our data suggested that the ncRNAs, especially antisense ncRNAs, LRRFIP1 binding to the chromatin-associated ncRNA, presum- are involved in LRRFIP1 binding to the promoter as a transcrip- ably by depleting the ncRNAs. Having demonstrated that the 410 tional repressor. To confirm the function of LRRFIP1, we created and 430 oligonucleotides diminished ncRNA interactions with stably transfected LRRFIP1 knockdown cells with short hairpin LRRFIP1, we investigated whether that was sufficient to disrupt RNAs (shRNAs) in K562 and THP1 cells (Fig. 9A). We found that LRRFIP1 binding to chromatin. Transfection of the two phos- TNF mRNA was significantly increased in both the K562 and phorothioate oligonucleotides led to decreased binding of LRRFIP1 THP1 LRRFIP1 knockdown cells (Fig. 9B, 9C), and TNF protein to chromatin in a ChIP assay (Fig. 6D). Therefore, the ncRNAs are was also increased in the LRRFIP1 knockdown THP1 cells after critical for LRRFIP1 targeting. stimulation (Fig. 9D), consistent with its role as a repressor. We Overexpression of ncRNAs can drive repressive chromatin used PMA as an acute stimulus in THP1 cells and also found knockdown of LRRFIP1 was associated with increased tran- We had demonstrated that overexpression of the ncRNA re- scription as defined by a modified run-on assay (Fig. 9E). Other pressed TNF message levels (Fig. 4). We hypothesized that overex- genes, both induced or not by PMA, did not exhibit increased pression of the ncRNA could also drive marks of repression. The transcription on the LRRFIP1 knockdown cells. We also per- same overexpression constructs were used as before. We were formed ChIP assays to define the effects of LRRFIP1 knockdown restricted to using the TNF2 probe due to the high level of exon chromatin marks of repressors; however, no changes were ogenous DNA after transfection. In this assay, we found increased observed in H3K27me3, EZH2, and SUZ12 in both LRRFIP1 levels of H3K27me3, EZH2, and SUZ12 in antisense 674–trans- knockdown K562 and THP1 cells (data not shown). Therefore, fected cells compared with vector-transfected cells (PMK) (Fig. 7). LRRFIP1 binding to chromatin is dependent on EZH2 and SUZ12, For the 674 sense transfection, only EZH2 binding was found to be but EZH2 and SUZ12 binding is not dependent on LRRFIP1. significantly increased. Overexpression of the ncRNA was suf- However, we found that knockdown of LRRFIP1 resulted in di- ficient to induce marks of repression, but the role of LRRFIP1 minished ncRNA abundance, especially around the LRRFIP bind- was still not clear. To identify reciprocal relationships between ing site (Fig. 9F). This suggests that the stability and function of LRRFIP1, EZH2, and SUZ12, we knocked down EZH2 and SUZ12 ncRNA might depend on LRRFIP1 binding. (Fig. 8). Knocking down either EZH2 or SUZ12 compromised binding of EZH2, SUZ12, and LRRFIP1 to chromatin (Fig. 8B– D), demonstrating that the protein–protein interactions seen by Discussion coimmunoprecipitation are functionally significant and impact Our data have defined a novel transcriptional regulatory pathway recruitment to chromatin. for the TNF gene. We first observed that the intergenic ncRNAs 8 NONCODING RNA REGULATION OF TNF EXPRESSION

FIGURE 7. Overexpression of ncRNAs induced repressor recruitment to the TNF promoter. The effect of ncRNAs on chromatin conformation was examined. K562 cells were treated with 10 ng/ml PMA for 3 d to induce low levels of message and then transfected with the 674 ncRNA sense or antisense overexpression constructs or the vector (PMK). CHIP assays were used to investigate (A) H3K27me3, (B) EZH2, and (C) SUZ12 binding at the TNF promoter. Overexpression of the ncRNA led to increased marks of repression. n = 4. Error bars in panels denote SE. *p , 0.05. were tightly associated with chromatin and were of diverse lengths. of LRRFIP1 to the chromatin. Localization was also dependent on K562 cells, repressed for TNF expression, had more abundant EZH2 and SUZ12. ncRNAs than THP1 cells, which are competent for expression LRRFIP1 (also known as TRIP, GCF2, and FLAP) was orig- of TNF protein. We used both knockdown and overexpression inally identified as a GC-rich–binding protein that repressed epi- strategies and demonstrated that the more abundant the ncRNAs, dermal growth factor receptor expression and platelet-derived Downloaded from thelessTNFwasexpressed.Theeffectsweresmall,suggesting growth factor expression (74–76). It has also been described as that this mechanism may represent a fine-tuning strategy for a transcriptional repressor in other settings, a tumor suppressor, transcriptional regulation. Alternatively, the knockdown strat- a platelet regulator, an early responder to foreign nucleic acids, egy and the overexpression strategy may not have led to as and a b-catenin cofactor (69, 77–85). It exists as two major effective targeting as the normal endogenous pathway, although isoforms, a long 160-kDa isoform transcribed starting from we did show that the overexpressed ncRNA localized to chro- exon 2 that includes RNA- and DNA-binding motifs and a shorter http://www.jimmunol.org/ matin. We believe the phenomenon is biologically relevant be- 120-kDa isoform transcribed from exon 1 that includes part cause effects on protein production were significant. We then of the RNA-binding motif but not the DNA-binding motif. The pursued a strategy to define the mechanism of the ncRNA effect. diverse functional descriptions may be in part due to isoform- We hypothesized that LRRFIP1, a known RNA-binding protein, specific effects. There is a homolog designated as LRRFIP2 which we had demonstrated actedasarepressorofTNF,might that shares sequence homology only at the 59 end (83). LRRFIP1 interact directly with the ncRNAs (35, 69). Our data demon- has no other homologs and no other defined motifs other than strated that not only did LRRFIP1 interact with the ncRNAs potential phosphorylation sites and a nuclear localization do- in vitro, but the interaction in vivo was required for localization main (86). by guest on October 1, 2021

FIGURE 8. Knockdown of SUZ12 and EZH2 impairs LRRFIP1 binding. Functional interactions of LRRFIP1, EZH2, and SUZ12 were examined. (A)A Western blot was used to determine the SUZ12 and EZH2 protein levels in the K562 cells transfected with siRNA-control, siRNA-EZH2, and siRNA- SUZ12. Knockdowns were effective. This is representative of three experiments. (B–D) ChIP assays demonstrated that knockdown of SUZ12 or EZH2 affected their own binding and that of LRRFIP1 at the TNF promoter. n = 3. Error bars in panels denote SE. *p , 0.05. The Journal of Immunology 9 Downloaded from

FIGURE 9. LRRFIP1 regulation of TNF mRNA. LRRFIP1 was knocked down to examine the effect on transcription. ( ) THP1 or K562 cells were http://www.jimmunol.org/ transfected with LRRFIP1 shRNA (sh-LRRFIP1) or nontarget control shRNA (NT), and puromycin was used to establish the stable LRRFIP1 knockdown cell lines. The Western blot demonstrates effective knockdown of LRRFIP1. (B) LRRFIP1 knockdown increased resting TNF mRNA level in K562 cells and (C) THP1cells. n =3.(D) TNF protein was markedly increased in PMA-stimulated LRRFIP1 knockdown THP1 cells compared with the control cells. A total of 10 ng/ml PMA was used to treat the cells for 6 h. n =3.(E) Nascent transcripts were measured after 100 ng/ml PMA treatment for 0, 30, and 60 min in THP1 cells. LRRFIP1 knockdown cells expressed higher levels of TNF message, but this effect was limited to TNF. NT, nontarget shRNA; sh-L, LRRFIP1 knockdown. CCL20 is PMA inducible; MMP1 and plasminogen are not PMA inducible. n =3.(F) LRRFIP1 knockdown altered the ncRNA abundance at the LRRFIP1 binding region. n = 2. NanoString nCounter technology was used to detect the ncRNAs. Error bars in panels denote SE. *p , 0.05, **p , 0.01, ***p , 0.001. by guest on October 1, 2021

In our study, we demonstrated that LRRFIP1 interacts with the cis or trans localize regulatory proteins and provide locus speci- ncRNAs in a sequence-independent fashion. The sense and anti- ficity (48, 93, 94). Our data clearly support the binding of sense sequences both bound LRRFIP1 and off-target dsRNA, LRRFIP1 to the RNA and via the RNA to the chromatin, con- supporting a model in which the structure is more important than sistent with a model in which the RNA provides locus specificity the sequence. dsRNA bound more effectively than ssRNA, and in cis. Our data are also consistent with a scaffolding function with longer species were also favored. These qualities are both found in EZH2, SUZ12, and LRRFIP1 assembling on a RNA tether to the the ncRNA upstream of TNF, and we speculate that the other region. These data suggest a model in which repressed cells ex- chromatin sites of LRRFIP1 binding share similar structural press ncRNAs in a region where polycomb proteins bind. Once qualities. We noted some asymmetry in the effects of the ncRNAs, bound, LRRFIP1 is recruited to the site where it participates in the with knockdown of antisense ncRNA slightly more effective at repressive complex. It is anchored to the site by the ncRNA. This altering LRRFIP1 binding. Overexpression of antisense RNA also model represents a hybrid of scenarios 2 and 3 above. seemed to have a more robust effect on chromatin marks of re- This study identified a novel transcriptional regulatory pathway. pression and TNF mRNA abundance. It is not known whether the We presume that this pathway is not unique to TNF, although our functions of the sense and antisense are different, and further studies did not extend beyond the one gene. Instead, we pursued studies will better define the structural requirement of the ncRNA a detailed analysis of one gene as an archetype. TNF is a relevant for LRRFIP1 binding. subject because of its importance in human disease states. Dis- There are now numerous examples of ncRNAs regulating tran- secting the mechanisms underlying the regulatory functions of scription, and their functional complexity, pervasive nature, and ncRNAs is just beginning. This study adds to that fundamental structural characteristics are just beginning to be understood. TSS, knowledge by examining a gene with highly dynamic expression enhancers, and CpG islands all produce short transcripts of un- and by identifying the role of LRRFIP1 in ncRNA function. certain function. The ncRNAs described in this work were found to be strongly associated with chromatin, a population of RNAs Acknowledgments initially described in 1975 (87). Several models have been posited We gratefully acknowledge the support of the Protein Core Facility at Child- to explain the role of ncRNAs in transcriptional regulation, as ren’s Hospital of Philadelphia. follows. 1) The ncRNA can serve as a decoy that titrates DNA- binding proteins away from genomic target sequences (88, 89). 2) The ncRNA can serve as a structural scaffold to facilitate inter- Disclosures actions of multiple proteins (90–92). 3) Guide RNAs operating in The authors have no financial conflicts of interest. 10 NONCODING RNA REGULATION OF TNF EXPRESSION

References 26. Barthel, R., A. V. Tsytsykova, A. K. Barczak, E. Y. Tsai, C. C. Dascher, M. B. Brenner, and A. E. Goldfeld. 2003. Regulation of tumor necrosis factor 1. Beutler, B., and T. Brown. 1991. A CAT reporter construct allows ultrasensitive alpha gene expression by mycobacteria involves the assembly of a unique estimation of TNF synthesis, and suggests that the TNF gene has been silenced enhanceosome dependent on the coactivator proteins CBP/p300. Mol. Cell. Biol. in non-macrophage cell lines. J. Clin. Invest. 87: 1336–1344. 23: 526–533. 2. Cerami, A., and B. Beutler. 1986. Cachectin: the dark side of tumor necrosis 27. Tsai, E. Y., J. V. Falvo, A. V. Tsytsykova, A. K. Barczak, A. M. Reimold, factor. Cold Spring Harb. Symp. Quant. Biol. 51: 625–629. L. H. Glimcher, M. J. Fenton, D. C. Gordon, I. F. Dunn, and A. E. Goldfeld. 3. Cerami, A., and B. Beutler. 1988. The role of cachectin/TNF in endotoxic shock 2000. A lipopolysaccharide-specific enhancer complex involving Ets, Elk-1, and cachexia. Immunol. Today 9: 28–31. Sp1, and CREB binding protein and p300 is recruited to the tumor necrosis 4. Gardam, M. A., E. C. Keystone, R. Menzies, S. Manners, E. Skamene, R. Long, factor alpha promoter in vivo. Mol. Cell. Biol. 20: 6084–6094. and D. C. Vinh. 2003. Anti-tumour necrosis factor agents and tuberculosis risk: 28. Kuprash, D. V., I. A. Udalova, R. L. Turetskaya, D. Kwiatkowski, N. R. Rice, mechanisms of action and clinical management. Lancet Infect. Dis. 3: 148–155. and S. A. Nedospasov. 1999. Similarities and differences between human and 5. Jacobs, M., D. Togbe, C. Fremond, A. Samarina, N. Allie, T. Botha, D. Carlos, murine TNF promoters in their response to lipopolysaccharide. J. Immunol. 162: S. K. Parida, S. Grivennikov, S. Nedospasov, et al. 2007. Tumor necrosis factor is 4045–4052. critical to control tuberculosis infection. Microbes Infect. 9: 623-628. 29. Lai, W. S., E. Carballo, J. R. Strum, E. A. Kennington, R. S. Phillips, and 6. Roach, D. R., A. G. Bean, C. Demangel, M. P. France, H. Briscoe, and P. J. Blackshear. 1999. Evidence that tristetraprolin binds to AU-rich elements W. J. Britton. 2002. TNF regulates chemokine induction essential for cell re- and promotes the deadenylation and destabilization of tumor necrosis factor cruitment, granuloma formation, and clearance of mycobacterial infection. J. alpha mRNA. Mol. Cell. Biol. 19: 4311–4323. Immunol. 168: 4620–4627. 30. Lo´pez de Silanes, I., S. Galbań, J. L. Martindale, X. Yang, K. Mazan-Mamczarz, 7. Bean, A. G., D. R. Roach, H. Briscoe, M. P. France, H. Korner, J. D. Sedgwick, F. E. Indig, G. Falco, M. Zhan, and M. Gorospe. 2005. Identification and and W. J. Britton. 1999. Structural deficiencies in granuloma formation in TNF gene-targeted mice underlie the heightened susceptibility to aerosol Mycobac- functional outcome of mRNAs associated with RNA-binding protein TIA-1. terium tuberculosis infection, which is not compensated for by lymphotoxin. J. Mol. Cell. Biol. 25: 9520–9531. Immunol. 162: 3504–3511. 31. Ronkina, N., M. B. Menon, J. Schwermann, C. Tiedje, E. Hitti, A. Kotlyarov, and 8. Marino, M. W., A. Dunn, D. Grail, M. Inglese, Y. Noguchi, E. Richards, M. Gaestel. 2010. MAPKAP kinases MK2 and MK3 in inflammation: complex A. Jungbluth, H. Wada, M. Moore, B. Williamson, et al. 1997. Characterization of regulation of TNF biosynthesis via expression and phosphorylation of tristetraprolin. Biochem. Pharmacol. 80: 1915–1920. tumor necrosis factor-deficient mice. Proc. Natl. Acad. Sci. USA 94: 8093–8098. Downloaded from 9. Butler, D. M., A. M. Malfait, L. J. Mason, P. J. Warden, G. Kollias, R. N. Maini, 32. Loflin, P., C. Y. Chen, and A. B. Shyu. 1999. Unraveling a cytoplasmic role for M. Feldmann, and F. M. Brennan. 1997. DBA/1 mice expressing the human hnRNP D in the in vivo mRNA destabilization directed by the AU-rich element. TNF-alpha transgene develop a severe, erosive arthritis: characterization of the Genes Dev. 13: 1884–1897. cytokine cascade and cellular composition. J. Immunol. 159: 2867–2876. 33. Min, H., C. W. Turck, J. M. Nikolic, and D. L. Black. 1997. A new regulatory 10. Carballo, E., W. S. Lai, and P. J. Blackshear. 1998. Feedback inhibition of protein, KSRP, mediates exon inclusion through an intronic splicing enhancer. macrophage tumor necrosis factor-alpha production by tristetraprolin. Science Genes Dev. 11: 1023–1036. 281: 1001–1005. 34. Zhang, W., B. J. Wagner, K. Ehrenman, A. W. Schaefer, C. T. DeMaria, 11. Dal Canto, R. A., M. K. Shaw, G. P. Nolan, L. Steinman, and C. G. Fathman. D. Crater, K. DeHaven, L. Long, and G. Brewer. 1993. Purification, character- http://www.jimmunol.org/ 1999. Local delivery of TNF by retrovirus-transduced T lymphocytes exacer- ization, and cDNA cloning of an AU-rich element RNA-binding protein, AUF1. bates experimental autoimmune encephalomyelitis. Clin. Immunol. 90: 10–14. Mol. Cell. Biol. 13: 7652–7665. 12. Douni, E., K. Akassoglou, L. Alexopoulou, S. Georgopoulos, S. Haralambous, 35. Suriano, A. R., A. N. Sanford, N. Kim, M. Oh, S. Kennedy, M. J. Henderson, S. Hill, G. Kassiotis, D. Kontoyiannis, M. Pasparakis, D. Plows, et al. 1995- K. Dietzmann, and K. E. Sullivan. 2005. GCF2/LRRFIP1 represses tumor ne- 1996. Transgenic and knockout analyses of the role of TNF in immune regu- crosis factor alpha expression. Mol. Cell. Biol. 25: 9073–9081. lation and disease pathogenesis. J. Inflamm. 47: 27–38. 36. Sullivan, K. E., A. Suriano, K. Dietzmann, J. Lin, D. Goldman, and M. A. Petri. 13. Kontoyiannis, D., M. Pasparakis, T. T. Pizarro, F. Cominelli, and G. Kollias. 2007. The TNFalpha locus is altered in monocytes from patients with systemic 1999. Impaired on/off regulation of TNF biosynthesis in mice lacking TNF AU- lupus erythematosus. Clin. Immunol. 123: 74–81. rich elements: implications for joint and gut-associated immunopathologies. 37. Cabili, M. N., C. Trapnell, L. Goff, M. Koziol, B. Tazon-Vega, A. Regev, and Immunity 10: 387–398. J. L. Rinn. 2011. Integrative annotation of human large intergenic noncoding 14. Willuweit, A., G. Sass, A. Scho¨neberg, U. Eisel, G. Tiegs, and M. Clauss. 2001. RNAs reveals global properties and specific subclasses. Genes Dev. 25: 1915– Chronic inflammation and protection from acute hepatitis in transgenic mice 1927. by guest on October 1, 2021 expressing TNF in endothelial cells. J. Immunol. 167: 3944–3952. 38. Ponting, C. P. 2008. The functional repertoires of metazoan genomes. Nat. Rev. 15. Armstrong, L., S. I. Godinho, K. M. Uppington, H. A. Whittington, and Genet. 9: 689–698. A. B. Millar. 2006. Contribution of TNF-alpha converting enzyme and protein- 39. Guttman, M., I. Amit, M. Garber, C. French, M. F. Lin, D. Feldser, M. Huarte, ase-3 to TNF-alpha processing in human alveolar macrophages. Am. J. Respir. O. Zuk, B. W. Carey, J. P. Cassady, et al. 2009. Chromatin signature reveals over Cell Mol. Biol. 34: 219–225. a thousand highly conserved large non-coding RNAs in mammals. Nature 458: 16. Buxade´,M.,J.L.Parra,S.Rousseau,N.Shpiro,R.Marquez,N.Morrice,J.Bain, 223–227. E. Espel, and C. G. Proud. 2005. The Mnks are novel components in the control of TNF 40. Carninci, P., T. Kasukawa, S. Katayama, J. Gough, M. C. Frith, N. Maeda, alpha biosynthesis and phosphorylate and regulate hnRNP A1. Immunity 23: 177–189. R. Oyama, T. Ravasi, B. Lenhard, C. Wells, et al; FANTOM Consortium; 17. Yang, Y., J. F. Chang, J. R. Parnes, and C. G. Fathman. 1998. T cell receptor RIKEN Genome Exploration Research Group and Genome Science Group (TCR) engagement leads to activation-induced splicing of tumor necrosis factor (Genome Network Project Core Group). 2005. The transcriptional landscape of (TNF) nuclear pre-mRNA. J. Exp. Med. 188: 247–254. the mammalian genome. Science 309: 1559–1563. 18. Osman, F., N. Jarrous, Y. Ben-Asouli, and R. Kaempfer. 1999. A cis-acting el- 41. Xu, Z., W. Wei, J. Gagneur, S. Clauder-Munster,€ M. Smolik, W. Huber, and ement in the 39-untranslated region of human TNF-alpha mRNA renders splicing L. M. Steinmetz. 2011. Antisense expression increases gene expression vari- dependent on the activation of protein kinase PKR. Genes Dev. 13: 3280–3293. ability and locus interdependency. Mol. Syst. Biol. 7: 468. 19. Sullivan, K. E., A. B. Reddy, K. Dietzmann, A. R. Suriano, V. P. Kocieda, 42. Ogawa, Y., B. K. Sun, and J. T. Lee. 2008. Intersection of the RNA interference M. Stewart, and M. Bhatia. 2007. Epigenetic regulation of tumor necrosis factor and X-inactivation pathways. Science 320: 1336–1341. alpha. Mol. Cell. Biol. 27: 5147–5160. 43. Splinter, E., E. de Wit, E. P. Nora, P. Klous, H. J. van de Werken, Y. Zhu, 20. Mahlknecht, U., J. Will, A. Varin, D. Hoelzer, and G. Herbein. 2004. Histone L. J. Kaaij, W. van Ijcken, J. Gribnau, E. Heard, and W. de Laat. 2011. The deacetylase 3, a class I histone deacetylase, suppresses MAPK11-mediated ac- inactive X chromosome adopts a unique three-dimensional conformation that is tivating transcription factor-2 activation and represses TNF gene expression. J. dependent on Xist RNA. Genes Dev. 25: 1371–1383. Immunol. 173: 3979–3990. 44. Carpenter, S., D. Aiello, M. K. Atianand, E. P. Ricci, P. Gandhi, L. L. Hall, 21. Wessels, I., E. Rosenkranz, M. Ventura Ferreira, S. Neuss, M. Zenke, L. Rink, M. Byron, B. Monks, M. Henry-Bezy, J. B. Lawrence, et al. 2013. A long and P. Uciechowski. 2013. Activation of IL-1b and TNFa genes is mediated noncoding RNA mediates both activation and repression of immune response by the establishment of permissive chromatin structures during monopoiesis. genes. Science 341: 789–792. Immunobiology 218: 860–868. 45. Guttman, M., J. Donaghey, B. W. Carey, M. Garber, J. K. Grenier, G. Munson, 22. Heinz, S., C. Benner, N. Spann, E. Bertolino, Y. C. Lin, P. Laslo, J. X. Cheng, G. Young, A. B. Lucas, R. Ach, L. Bruhn, et al. 2011. lincRNAs act in the C. Murre, H. Singh, and C. K. Glass. 2010. Simple combinations of lineage- circuitry controlling pluripotency and differentiation. Nature 477: 295–300. determining transcription factors prime cis-regulatory elements required for 46. Peng, X., L. Gralinski, M. T. Ferris, M. B. Frieman, M. J. Thomas, S. Proll, macrophage and B cell identities. Mol. Cell 38: 576–589. M. J. Korth, J. R. Tisoncik, M. Heise, S. Luo, et al. 2011. Integrative deep sequenc- 23. Pham, T. H., S. Langmann, L. Schwarzfischer, C. El Chartouni, M. Lichtinger, ing of the mouse lung transcriptome reveals differential expression of diverse classes M. Klug, S. W. Krause, and M. Rehli. 2007. CCAAT enhancer-binding protein of small RNAs in response to respiratory virus infection. MBio 2: pii: e00198-11. beta regulates constitutive gene expression during late stages of monocyte to 47. Kretz, M., Z. Siprashvili, C. Chu, D. E. Webster, A. Zehnder, K. Qu, C. S. Lee, macrophage differentiation. J. Biol. Chem. 282: 21924–21933. R. J. Flockhart, A. F. Groff, J. Chow, et al. 2013. Control of somatic tissue 24. Pham, T. H., C. Benner, M. Lichtinger, L. Schwarzfischer, Y. Hu, R. Andreesen, differentiation by the long non-coding RNA TINCR. Nature 493: 231–235. W. Chen, and M. Rehli. 2012. Dynamic epigenetic enhancer signatures reveal 48. Chu, C., K. Qu, F. L. Zhong, S. E. Artandi, and H. Y. Chang. 2011. Genomic key transcription factors associated with monocytic differentiation states. Blood maps of long noncoding RNA occupancy reveal principles of RNA-chromatin 119: e161–e171. interactions. Mol. Cell 44: 667–678. 25. Pham, T. H., J. Minderjahn, C. Schmidl, H. Hoffmeister, S. Schmidhofer, 49.Khalil,A.M.,M.Guttman,M.Huarte,M.Garber,A.Raj,D.RiveaMorales, W. Chen, G. La¨ngst, C. Benner, and M. Rehli. 2013. Mechanisms of in vivo K. Thomas, A. Presser, B. E. Bernstein, A. van Oudenaarden, et al. 2009. Many binding site selection of the hematopoietic master transcription factor PU.1. human large intergenic noncoding RNAs associate with chromatin-modifying com- Nucleic Acids Res. 41: 6391–6402. plexes and affect gene expression. Proc. Natl. Acad. Sci. USA 106: 11667–11672. The Journal of Immunology 11

50. Redrup, L., M. R. Branco, E. R. Perdeaux, C. Krueger, A. Lewis, F. Santos, 73. Zuker, M. 2003. Mfold web server for nucleic acid folding and hybridization T. Nagano, B. S. Cobb, P. Fraser, and W. Reik. 2009. The long noncoding RNA prediction. Nucleic Acids Res. 31: 3406–3415. Kcnq1ot1 organises a lineage-speciﬁc nuclear domain for epigenetic gene si- 74. Santiago, F. S., and L. M. Khachigian. 2004. Ets-1 stimulates platelet-derived lencing. Development 136: 525–530. growth factor A-chain gene transcription and vascular smooth muscle cell 51. Kanduri, C. 2008. Functional insights into long antisense noncoding RNA growth via cooperative interactions with Sp1. Circ. Res. 95: 479–487. Kcnq1ot1 mediated bidirectional silencing. RNA Biol. 5: 208–211. 75. Khachigian, L. M., F. S. Santiago, L. A. Rafty, O. L. Chan, G. J. Delbridge, 52. Martianov, I., A. Ramadass, A. Serra Barros, N. Chow, and A. Akoulitchev. A. Bobik, T. Collins, and A. C. Johnson. 1999. GC factor 2 represses platelet- 2007. Repression of the human dihydrofolate reductase gene by a non-coding derived growth factor A-chain gene transcription and is itself induced by arterial interfering transcript. Nature 445: 666–670. injury. Circ. Res. 84: 1258–1267. 53. Suzuki, K., T. Juelich, H. Lim, T. Ishida, T. Watanebe, D. A. Cooper, S. Rao, and 76. Reed, A. L., H. Yamazaki, J. D. Kaufman, Y. Rubinstein, B. Murphy, and A. D. Kelleher. 2008. Closed chromatin architecture is induced by an RNA A. C. Johnson. 1998. Molecular cloning and characterization of a transcription duplex targeting the HIV-1 promoter region. J. Biol. Chem. 283: 23353–23363. regulator with homology to GC-binding factor. J. Biol. Chem. 273: 21594– 54. Ting, A. H., K. E. Schuebel, J. G. Herman, and S. B. Baylin. 2005. Short double- 21602. stranded RNA induces transcriptional gene silencing in human cancer cells in the 77. Soler, G., S. Nusbaum, B. Varet, E. A. Macintyre, M. Vekemans, S. P. Romana, absence of DNA methylation. Nat. Genet. 37: 906–910. and I. Radford-Weiss. 2009. LRRFIP1, a new FGFR1 partner gene associated 55. Weinberg, M. S., L. M. Villeneuve, A. Ehsani, M. Amarzguioui, L. Aagaard, with 8p11 myeloproliferative syndrome. Leukemia 23: 1359-1361. Z. X. Chen, A. D. Riggs, J. J. Rossi, and K. V. Morris. 2006. The antisense strand 78. Lee, Y. H., and M. R. Stallcup. 2006. Interplay of Fli-I and FLAP1 for regulation of small interfering RNAs directs histone methylation and transcriptional gene of beta-catenin dependent transcription. Nucleic Acids Res. 34: 5052–5059. silencing in human cells. RNA 12: 256-262. 79. Ohtsuka, H., M. Oikawa, K. Ariake, T. Rikiyama, F. Motoi, Y. Katayose, 56. Janowski, B. A., K. E. Huffman, J. C. Schwartz, R. Ram, R. Nordsell, M. Unno, and A. C. Johnson. 2011. GC-binding factor 2 interacts with dishev- D. S. Shames, J. D. Minna, and D. R. Corey. 2006. Involvement of AGO1 and elled and regulates Wnt signaling pathways in human carcinoma cell lines. Int. AGO2 in mammalian transcriptional silencing. Nat. Struct. Mol. Biol. 13: 787–792. J. Cancer 129: 1599-1610. 57. Han, J., D. Kim, and K. V. Morris. 2007. Promoter-associated RNA is required 80. Goodall, A. H., P. Burns, I. Salles, I. C. Macaulay, C. I. Jones, D. Ardissino, for RNA-directed transcriptional gene silencing in human cells. Proc. Natl. B. de Bono, S. L. Bray, H. Deckmyn, F. Dudbridge, et al; Bloodomics Con- Acad. Sci. USA 104: 12422–12427. sortium. 2010. Transcription proﬁling in human platelets reveals LRRFIP1 as 58. Kim, D. H., L. M. Villeneuve, K. V. Morris, and J. J. Rossi. 2006. Argonaute-1 a novel protein regulating platelet function. Blood 116: 4646–4656. directs siRNA-mediated transcriptional gene silencing in human cells. Nat. 81. Yang, P., H. An, X. Liu, M. Wen, Y. Zheng, Y. Rui, and X. Cao. 2010. The Downloaded from Struct. Mol. Biol. 13: 793–797. cytosolic nucleic acid sensor LRRFIP1 mediates the production of type I in- 59. Lee, J. Y., and K. E. Sullivan. 2001. Gamma interferon and lipopolysaccharide terferon via a beta-catenin-dependent pathway. Nat. Immunol. 11: 487–494. interact at the level of transcription to induce tumor necrosis factor alpha ex- 82. Li, Y., W. Li, Y. Yang, Y. Lu, C. He, G. Hu, H. Liu, J. Chen, J. He, and H. Yu. pression. Infect. Immun. 69: 2847–2852. 2009. MicroRNA-21 targets LRRFIP1 and contributes to VM-26 resistance in 60. Wang, X., S. Arai, X. Song, D. Reichart, K. Du, G. Pascual, P. Tempst, glioblastoma multiforme. Brain Res. 1286: 13–18. M. G. Rosenfeld, C. K. Glass, and R. Kurokawa. 2008. Induced ncRNAs allosterically 83. Dai, P., S. Y. Jeong, Y. Yu, T. Leng, W. Wu, L. Xie, and X. Chen. 2009. Mod- modify RNA-binding proteins in cis to inhibit transcription. Nature 454: 126–130. ulation of TLR signaling by multiple MyD88-interacting partners including

61. Mapendano, C. K., S. Lykke-Andersen, J. Kjems, E. Bertrand, and T. H. Jensen. leucine-rich repeat Fli-I-interacting proteins. J. Immunol. 182: 3450–3460. http://www.jimmunol.org/ 2010. Crosstalk between mRNA 39 end processing and transcription initiation. 84. Rikiyama, T., J. Curtis, M. Oikawa, D. B. Zimonjic, N. Popescu, B. A. Murphy, Mol. Cell 40: 410–422. M. A. Wilson, and A. C. Johnson. 2003. GCF2: expression and molecular 62. Seidl, C. I., S. H. Stricker, and D. P. Barlow. 2006. The imprinted Air ncRNA is analysis of repression. Biochim. Biophys. Acta 1629: 15–25. an atypical RNAPII transcript that evades splicing and escapes nuclear export. 85. Sjo¨blom,T.,S.Jones,L.D.Wood,D.W.Parsons,J.Lin,T.D.Barber, EMBO J. 25: 3565–3575. D. Mandelker, R. J. Leary, J. Ptak, N. Silliman, et al. 2006. The consensus 63. Bochnig, P., R. Reuter, P. Bringmann, and R. Luhrmann. 1987. A monoclonal coding sequences of human breast and colorectal cancers. Science 314: 268– antibody against 2,2,7-trimethylguanosine that reacts with intact, class U, small 274. nuclear ribonucleoproteins as well as with 7-methylguanosine-capped RNAs. 86. Arakawa, R., A. Bagashev, L. Song, K. Maurer, and K. E. Sullivan. 2010. Eur. J. Biochem. 168: 461-467. Characterization of LRRFIP1. Biochem. Cell Biol. 88: 899–906. 64. Patrone, G., F. Puppo, R. Cusano, M. Scaranari, I. Ceccherini, A. Puliti, and 87. Paul, J., and J. D. Duerksen. 1975. Chromatin-associated RNA content of het- R. Ravazzolo. 2000. Nuclear run-on assay using biotin labeling, magnetic bead erochromatin and euchromatin. Mol. Cell. Biochem. 9: 9–16.

capture and analysis by fluorescence-based RT-PCR. Biotechniques 29: 1012– 88. Kino, T., D. E. Hurt, T. Ichijo, N. Nader, and G. P. Chrousos. 2010. Noncoding by guest on October 1, 2021 1014, 1016–1017. RNA gas5 is a growth arrest- and starvation-associated repressor of the gluco- 65. Bendak, K., F. E. Loughlin, V. Cheung, M. R. O’Connell, M. Crossley, and corticoid receptor. Sci. Signal. 3: ra8. J. P. Mackay. 2012. A rapid method for assessing the RNA-binding potential of 89. Hung, T., Y. Wang, M. F. Lin, A. K. Koegel, Y. Kotake, G. D. Grant, a protein. Nucleic Acids Res. 40: e105. H. M. Horlings, N. Shah, C. Umbricht, P. Wang, et al. 2011. Extensive and 66. Lee, J. Y., N. A. Kim, A. Sanford, and K. E. Sullivan. 2003. Histone acetylation coordinated transcription of noncoding RNAs within cell-cycle promoters. Nat. and chromatin conformation are regulated separately at the TNF-alpha promoter Genet. 43: 621–629. in monocytes and macrophages. J. Leukoc. Biol. 73: 862–871. 90. Tsai, M. C., O. Manor, Y. Wan, N. Mosammaparast, J. K. Wang, F. Lan, Y. Shi, 67. Garrett, S., K. Dietzmann-Maurer, L. Song, and K. E. Sullivan. 2008. Polarization E. Segal, and H. Y. Chang. 2010. Long noncoding RNA as modular scaffold of of primary human monocytes by IFN-gamma induces chromatin changes and histone modification complexes. Science 329: 689–693. recruits RNA Pol II to the TNF-alpha promoter. J. Immunol. 180: 5257–5266. 91. Pandey, R. R., T. Mondal, F. Mohammad, S. Enroth, L. Redrup, J. Komorowski, 68. Rinn, J. L., M. Kertesz, J. K. Wang, S. L. Squazzo, X. Xu, S. A. Brugmann, T. Nagano, D. Mancini-Dinardo, and C. Kanduri. 2008. Kcnq1ot1 antisense L. H. Goodnough, J. A. Helms, P. J. Farnham, E. Segal, and H. Y. Chang. 2007. noncoding RNA mediates lineage-specific transcriptional silencing through Functional demarcation of active and silent chromatin domains in human HOX chromatin-level regulation. Mol. Cell 32: 232–246. loci by noncoding RNAs. Cell 129: 1311–1323. 92. Kotake, Y., T. Nakagawa, K. Kitagawa, S. Suzuki, N. Liu, M. Kitagawa, and 69. Wilson, S. A., E. C. Brown, A. J. Kingsman, and S. M. Kingsman. 1998. TRIP: Y. Xiong. 2011. Long non-coding RNA ANRIL is required for the PRC2 re- a novel double stranded RNA binding protein which interacts with the leucine cruitment to and silencing of p15(INK4B) tumor suppressor gene. Oncogene 30: rich repeat of flightless I. Nucleic Acids Res. 26: 3460–3467. 1956–1962. 70. Cao, R., and Y. Zhang. 2004. The functions of E(Z)/EZH2-mediated methylation 93. Huarte, M., M. Guttman, D. Feldser, M. Garber, M. J. Koziol, D. Kenzelmann- of lysine 27 in histone H3. Curr. Opin. Genet. Dev. 14: 155–164. Broz, A. M. Khalil, O. Zuk, I. Amit, M. Rabani, et al. 2010. A large intergenic 71. Cao, R., and Y. Zhang. 2004. SUZ12 is required for both the histone methyl- noncoding RNA induced by p53 mediates global gene repression in the p53 transferase activity and the silencing function of the EED-EZH2 complex. Mol. response. Cell 142: 409–419. Cell 15: 57–67. 94. Simon, M. D., C. I. Wang, P. V. Kharchenko, J. A. West, B. A. Chapman, 72. Vire´, E., C. Brenner, R. Deplus, L. Blanchon, M. Fraga, C. Didelot, L. Morey, A. A. Alekseyenko, M. L. Borowsky, M. I. Kuroda, and R. E. Kingston. 2011. A. Van Eynde, D. Bernard, J. M. Vanderwinden, et al. 2006. The Polycomb The genomic binding sites of a noncoding RNA. Proc. Natl. Acad. Sci. USA 108: group protein EZH2 directly controls DNA methylation. Nature 439: 871–874. 20497–20502.