The Melanoma-Upregulated Long Noncoding RNA SPRY4-IT1 Modulates Apoptosis and Invasion

Published OnlineFirst May 10, 2011; DOI: 10.1158/0008-5472.CAN-10-4460

Cancer Molecular and Cellular Pathobiology Research

Divya Khaitan1, Marcel E. Dinger2, Joseph Mazar1, Joanna Crawford2, Martin A. Smith2, John S. Mattick2, and Ranjan J. Perera1

Abstract The identification of cancer-associated long noncoding RNAs (lncRNAs) and the investigation of their molecular and biological functions are important to understand the molecular biology of cancer and its progression. Although the functions of lncRNAs and the mechanisms regulating their expression are largely unknown, recent studies are beginning to unravel their importance in human health and disease. Here, we report that a number of lncRNAs are differentially expressed in melanoma cell lines in comparison to melanocytes and keratinocyte controls. One of these lncRNAs, SPRY4-IT1 (GenBank accession ID AK024556), is derived from an intron of the SPRY4 gene and is predicted to contain several long hairpins in its secondary structure. RNA-FISH analysis showed that SPRY4-IT1 is predominantly localized in the cytoplasm of melanoma cells, and SPRY4-IT1 RNAi knockdown results in defects in cell growth, differentiation, and higher rates of apoptosis in melanoma cell lines. Differential expression of both SPRY4 and SPRY4-IT1 was also detected in vivo, in 30 distinct patient samples, classified as primary in situ, regional metastatic, distant metastatic, and nodal metastatic melanoma. The elevated expression of SPRY4-IT1 in melanoma cells compared to melanocytes, its accumulation in cell cytoplasm, and effects on cell dynamics, including increased rate of wound closure on SPRY4-IT1 overexpression, suggest that the higher expression of SPRY4-IT1 may have an important role in the molecular etiology of human melanoma. Cancer Res; 71(11); 3852–62. 2011 AACR.

Introduction mammalian genome expresses large numbers of long ncRNAs (lncRNA; ref. 1) that are dynamically expressed in tissue-, There is considerable interest in understanding the func- differentiation stage-, and cell type-specific patterns (3), at tion of RNA transcripts that do not code for proteins in least some of which localize into specific subcellular compart- eukaryotic cells. As evidenced by cDNA cloning projects (1) ments (3, 4). lncRNAs are also known to play important roles and genomic tiling arrays (2), more than 90% of the human during cellular development and differentiation (5–8) consis- genome undergoes transcription, but does not code for pro- tent with the view that they are under evolutionary selection teins. These transcriptional products are referred to as non- (9–11). Interestingly, lncRNAs can influence the expression of protein coding RNAs (ncRNAs). A variety of ncRNAs, such as specific target proteins at specific genomic loci (12, 13), ribosomal RNAs, transfer RNAs, and spliceosomal RNAs, are modulate the activity of protein binding partners (14, 15), essential for cell function. Similarly, large numbers of short direct chromatin-modifying complexes to their sites of action ncRNAs such as micro-RNAs (miRNAs), endogenous siRNAs, (16), and are posttranscriptionally processed to produce 0 PIWI-interacting RNAs (piRNAs), and small nucleolar RNAs numerous 5 -capped small RNAs (17, 18). Epigenetic pathways (snoRNAs) are also known to play important regulatory roles can also regulate the differential expression of lncRNAs (19). in eukaryotic cells. Recent studies have showed that the The expression of lncRNAs is misregulated in various diseases, including ischemia (20), heart disease (21), Alzheimer's disease (22), psoriasis (23), and spinocerebellar ataxia type 8 (19, 24). Authors' Affiliations: 1Sanford Burnham Medical Research Institute, Orlando, Florida; and 2Institute for Molecular Bioscience, University of This misregulation has also been shown in various types of Queensland, St. Lucia, Australia cancers, such as breast cancer (8, 25), colon cancer (26), Note: Supplementary data for this article are available at Cancer Research prostate cancer (27), hepatocellular carcinoma (28, 29), and Online (http://cancerres.aacrjournals.org/). leukemia (28). One such lncRNA, DD3 (also known as PCA3), is D. Khaitan and M.E. Dinger contributed equally to this work. listed as a prostate cancer-specific biomarker (30). Recent Corresponding Author: Ranjan J. Perera, Sanford Burnham Medical studies have revealed the contribution of ncRNAs as proto- Research Institute, 6400 Sanger Road, University of Queensland, oncogenes, for example GAGE6 (31), as tumor suppressor Orlando, FL 32827. Phone: 407-745-2084; Fax: 407-745-2001; E-mail: p15 [email protected] and John S. Mattick, Institute for Mole- genes, for example (32), in tumorigenesis, and as drivers cular Bioscience, St Lucia, QLD 4101, Australia. Phone: +61-(7)-3346- of metastatic transformation, for example HOTAIR in breast 2079; Fax: +61-(7)-3346-2101; E-mail: [email protected] cancer (33). doi: 10.1158/0008-5472.CAN-10-4460 The focus of this study was to identify and characterize 2011 American Association for Cancer Research. lncRNAs that are differentially expressed in melanoma

3852 Cancer Res; 71(11) June 1, 2011

Noncoding RNA Function SPRY4-IT1 in Human Melanoma

compared to melanocytes and normal skin. We find that a 12, and 18 nmol/L RNAi duplexes were diluted in 500 mL serum group of lncRNAs is differentially regulated in melanoma and free medium, mixed gently and 5 mL of lipofectamine RNAi- one such lncRNA, SPRY4-IT1, derived from an intron within MAX was added to each well containing the diluted RNAi the SPRY4 gene, is upregulated in melanoma cells. Knocking- molecules. This mixture was incubated for 20 minutes at room down its expression resulted in defects in cell growth, inva- temperature before the transfection. A total of 250,000 cells sion, and elevated rates of apoptosis in melanoma cells. were diluted in complete Tu growth medium (ref. 39; without antibiotics) and plated in each well. RNAi duplex—lipofecta- Materials and Methods mine RNAiMAX complexes were added to each well and mixed gently by rocking the plate. Plasmid transfection was per- RNA secondary structure prediction formed in MatTek 1.5-mm glass-bottom dishes. Five micro- The most recent versions of RNAfold (34) and RNAstructure grams of pcDNA6/SPRY4-IT1 or pcDNA6/V5-HisA (negative (35) were employed for generating RNA secondary structures. control) was diluted in 100 mL serum-free medium to which 5 Both of these programs implement an RNA partition function mL of Fugene 6 was added. Following a 20-minute incubation, algorithm, which was chosen for 2 reasons: (i) it produces a the respective transfection mixtures were added their respec- structure almost identical to the minimum free energy algo- tive dishes. In all cases, cells were incubated for 48 hours rithm with RNAfold with few proximal suboptimal structures at 37 CinaCO2 incubator and gene knockdown or over- and (ii) it is required for subsequent prediction of pseudo- expression levels were assessed by quantitative real-time PCR knots with ProbKnot (included in RNAstructure). (qRT-PCR). The evolutionary conservation of secondary structures was conducted with the consensus-based programs RNAz (36) and In vitro wound healing assay SISSIz (37) on the Enredo-Pecan-Ortheus 31-way eutherian WM1552C cells were transfected with 18 nmol/L SPRY4-IT1 mammal genome alignment from ENSEMBL. Orthologous or scrambled siRNA siRNA (Life Technologies), respectively, sequences to SPRY4-IT1 were selected and realigned with and LOX IMV1 cells were transfected with pcDNA6/SPRY4-IT1 MAFFT, using the mafft-ginsi algorithm (38). Sliding window or empty pcDNA6/V5-HisA respectively, as described earlier. ranges of 100 nt window with 25 nt slide, 150 nt window with All transfected cell samples were seeded on Mat Tek 1.5-mm 50 nt slide, and 300 nt window with 100 nt slide were tested glass-bottom dishes until 90% to 95% confluent. Cell mono- with both RNAz and SISSIz, using parameters "-d" and "-d -t -n layers were then gently scratched with a pipette tip across the 200 -p 0.02", respectively. diameter of the dish and rinsed with PBS and cell media to remove cellular debris. The surface area of the scratched siRNA to knockdown SPRY4-IT1 in melanoma cells surface was quantified after wounding and again every 20 Five different siRNA siRNAs that targeted SPRY4-IT1 RNA and a minutes for 24 hours on a Nikon BioStation IM (Nikon scrambled siRNA siRNA control were generously provided by Life Instruments, Inc.) cell incubator. The extent of wound closure Technologies. The siRNA siRNA molecules are 25 base-pair was calculated using the ratio of the surface area between the double-stranded RNA oligonucleotides with proprietary chemical remaining wound edges to the surface area of the initial modifications. The BLOCK-iT RNAi designer was used to find wound for each time point. These data were then expressed gene-specific 25 nucleotide siRNA siRNA molecules. It uses gene- as a percentage of wound closure relative to control condi- specific targets for RNAi analysis and reports up to 10 top scoring tions for each experiment. The change in surface area was siRNA siRNA targets. The freeze-dried siRNAs were dissolved in calculated using NIS Elements software (Nikon Instruments, RNase free-water and stored as aliquots at 20C. The siRNA with Inc.) and performed in triplicate. the sequence gctttctgattccaaggcctattaa yielded the highest degree of SPRY4-IT1 knockdown. RNA–FISH analysis Locked nucleic acid (LNA) modified probes for human Overexpression construct to upregulate SPRY4-IT1 in lncRNA SPRY4-IT1(5-FAM- TCCACTGGGCATATTCTAAAA- LOX-IMV1 melanoma cell line 36-FAM) and a negative/scramble control (5TYE665- Oligonucleotides for amplification of full length SPRY4-IT1 GTGTAACACGTCTATACGCCCA-3 TYE665, miRCURY-LNA (SPRY4-IT1 Forward 50-TAAGCTTGTAGAGATGGGGGTTT- detection probe, Exiqon) were used for RNA–FISH. In situ CATCCTGTTGG-30 and SPRY4-IT1 Reverse 50-ACTCGAGAAA- hybridization was performed using the RiboMap in situ hybri- GACTCCCTTTCCTTAAGCAGATTCAC-30) were designed to dization kit (Ventana Medical Systems, Inc.) on a Ventana incorporate external HindIII and XhoI sites, respectively. The machine. The cell suspension was diluted to 10,000 cells/100 melanocyte genomic DNA SPRY4-IT1 amplicon (Amplitaq mL and pipetted into clonal rings on the autoclaved glass Gold, Life Technologies) was cloned into the pCR4-TOPO slides. The following day, the clonal rings were removed; slides vector (Life Technologies), sequence verified and subcloned were washed in PBS and fixed in 4% paraformaldehyde and 5% via HindIII/XhoI digestion into the pcDNA6/V5-HisA mam- acetic acid. After acid treatment using hydrochloride-based malian expression vector (Life Technologies). RiboClear reagent (Ventana Medical Systems) for 10 min at 37C, the slides were treated with the ready-to-use protease 3 Cell culture conditions and transfection reagent. The cells were hybridized with the antisense LNA siRNA transfection was performed with lipofectamine riboprobe (40 nmol/L) using RiboHybe hybridization buffer RNAiMax (Life Technologies) in 6-well plates. A total of 6, (Ventana Medical Systems) for 2 hours at 58C after an initial

www.aacrjournals.org Cancer Res; 71(11) June 1, 2011 3853

Khaitan et al.

denaturing prehybridization step for 4 min at 80C. Next, the metastatic and nodal metastatic), and 6 normal skin samples slides were subjected to a low-stringency wash with 0.1 SSC were also analyzed using the same microarrays. Hierarchical (Ventana Medical Systems) for 4 min at 60C, and then 2 clustering was used to represent the differential lncRNA further washing steps with 1 SSC for 4 min at 60C. These expression profiles (Supplementary Fig. S1) and revealed slides were fixed in RiboFix and counterstained with 4’- that potential signatures could be used to discriminate 6’diamidino-2-phenylindole (DAPI), in an antifade reagent between normal and melanoma samples. To select candi- (Ventana). The images were acquired using a Nikon A1R dates for functional studies, we asked whether any of the VAAS laser point- and resonant-scanning confocal micro- differentially expressed lncRNAs in melanoma cell lines were scope equipped with a single photon Ar-ion laser at 60 with also differentially expressed in patient samples. As a result 4 zoom. we identified 4 candidate ncRNAs (AF085920, AK091731, AK128206, and AK024556), which were differentially Metabolic viability by MTT assay expressed in both melanoma cell lines and patients samples MTT (3-(4,5-dimethyl-2-yl)-2,5-diphenyl-2II-tetrazolium relative to melanocytes and whose differential expression we bromide) was purchased from Roche. Cells were plated in were able to validate by next generation sequencing of 96-well plates (5000 cells/100 mL/well). After 48 hours of polyadenylated RNA isolated from melanocytes and transfection, 20 mL MTT solution was added and the cells WM1552C cells (Supplementary Fig. S2 and Fig. 1A and were incubated at 37C in the dark for 4 hours. The generated B).Wenextexaminedthegenomiccontextofthe4lncRNA formazan OD was measured at 490 nm to determine the cell candidates and found that 3 of the lncRNAs (GenBank IDs viability on the Flex station (Molecular Devices; www.mole- AF085920, AK091731, and AK128206) mapped within a few culardevices.com). kilobases of the 3’UTRs of known protein-coding genes and one (GenBank ID AK024556) mapped within an intronic Invasion assays region (Supplementary Fig. S3). Because of the possibility of BD BioCoat growth factor reduced insert plates (Matrigel read-through expression from the upstream protein-coding Invasion Chamber 12 well plates) were prepared by rehydrat- genes and the complications in interpretation that this ing the BD Matrigel matrix coating in the inserts with 0.5 mL would create in subsequent studies to characterize these of serum-free complete Tu media for 2 hours at 37C. The transcripts, we excluded the 3 lncRNAs that were positioned rehydration solution was carefully removed from the inserts, downstream of 3’UTRs from further study. Therefore, given 500 mL complete Tu (2% FBS) was added to the lower wells of the intronic position of the remaining candidate, which the plate. 1 104 transfected and untransfected cells sus- incidentally also had the highest degree of upregulation in pended in 500 mL of serum-free complete Tu media was added melanoma cells compared to melanocytes (12-fold), we to the top of each insert well. Invasion assay plates were pursued this lncRNA for further study. AK024556 is a 687 incubated for 48 hours at 37C. Following incubation, the nt unspliced, polyadenylated transcript originally identified noninvading cells were removed by scrubbing the upper sur- in adipose tissue and is transcribed from the second intron face of the insert. The cells on the lower surface of the insert of the SPRY4 gene (Fig. 1A), which led to our naming the were stained with crystal violet and each trans-well membrane transcript SPRY4-IT1. This region is not conserved beyond mounted on a microscope slide for visualization and analysis. the primate genomes and there is no EST expression The slides were scanned in scanscope and the number of cells detected in mouse. Secondary structure prediction of the migrating was counted using Aperio software (www.aperio. SPRY4-IT1 sequence indicates that this ncRNA transcript com). Data are expressed as the percent invasion through the could fold into long stable hairpin structures, suggest- membrane relative to the migration through the control ing that SPRY4-IT1 may function intrinsically as an RNA membrane. molecule. See Supplementary Material (S10) for RNA extraction, A comparison of SPRY4-IT1 in kidney, blood, and breast cell microarray analysis, qRT-PCR, Northern blot analysis, and lines revealed expression to be equal to that of melanocytes or phosphatidylserine externalization. less (Supplementary Fig. S4). We then measured the expression levels of SPRY4-IT1 (Fig. 1C) as well as the SPRY4 ORF Results (Supplementary Fig. S6) in 7 additional nonpigmented melanoma cell lines (WM793B, A375, SKMEL-2, RPMI 7951, HT- Differentially regulated lncRNAs in melanoma cells 144, LOX-IMV1, and G361) by qRT-PCR and the results To identify lncRNAs involved in melanoma, we analyzed showed that the expression of both SPRY-IN1 and SPRY4 total RNA from a stage III melanoma cell line (WM1552C), was elevated in most of the melanoma cell lines relative to melanocytes, and keratinocytes by using a noncoding RNA control melanocytes. microarray (NCode human array; Life Technologies). The microarrays contain probes to target 12,784 lncRNAs and Structural prediction of SPRY4-IT1 25,409 mRNAs. In total, we identified 77 lncRNAs that were To determine whether the SPRY4-IT1 RNA contained any significantly differentially expressed (P < 0.015; fold-change particular secondary structural features, the SPRY4-IT1 geno- >2) in WM1552C relative to melanocytes. In addition to cell mic sequence was submitted to secondary structure and line profiling, 29 independent melanoma patient samples pseudoknot prediction (see Materials and Methods). Several (graded as primary in situ, regional metastatic, distant helical regions are commonly predicted by both algorithms,

3854 Cancer Res; 71(11) June 1, 2011 Cancer Research

Noncoding RNA Function SPRY4-IT1 in Human Melanoma

A 5 kb

SPRY4.1 SPRY4.2

2 _ SPRY4-IT1 Mammalian conservation –0.3 _

BC50 Melanocyte 40 tpm

0 60 WM1552C tpm 20

10 expression SPRY4-IT1

0 melanocytes) to (relative SPRY4-IT1

200 bp 0 MelanoKeratinocytecyte A375 WM1552CSK-MEL-2RPMI-7951HT144 LOX-IMV1G361 WM793B

c t t t t c t g g t 40 c t a g g a c a D g t t g g c c 20 30 a g c c c t g t t G g A G 10 g 270 g G g a G 5’ t t t G g c T 50 G g a 1 G T c 260 T t t C G G G C G A a g a g c T C C A a G C T C G G C a t c t t A G g t G G T 280 t A G c t A A C C t t c g 60 70 G T 130 t 250 A a G T C A a a c c A C 440 t a C C C C C C 120 c t t t t g g c G T g C A C c t 290 A T 450 C t c c T c T C c c T C G A T g G G T C T G 460 t a t g t C G G a G A a c c T G T C g g a c 240 C G G t a c g T G a G C 430 g 100 g G A G A C T G T A C c A G C A G A C t a 80 A T C A 230 T T C g T a t T A t t 90 A T G G T T T G T T T G 470 c T G T C C t g A G A C 420 c T A t a C 300 T G g g A 310 C t c G G c C A A c g T 140 C A G 480 c G t t c a G C G C t A C 110 T T G t C G 220 T c T A C G G t G C T A 320 t A T T A T A G 410 C A T A t G C T A A C T A t 370 T C 390 330 T A T C G A A T T C G t C C T T A 380 G G C G C G C A t G C G A 150 T G T 490 T G G C C G T A A T T t 360 G G C C C T A G C G 340 T T C C G A T T A t G A C 400 T C A T T C C 550 G G T A C C t T G C A T A C T T T 350 A G 210 A C t A A G G G T C C A G C T G G G G G T G A G t T A G 560 C C G T A C T T T C T G A T G G G A T 540 C A C T T A C 530 G A A A T T 610 A C C T T G G T T A T G 500 A G T A G G A T 570 C C A G 160 A C A A A T A C G C A T G C T 520 620 C C C G T T 600 G G C C A T A T G A C C T 200 G A 580 T A T 170 A G T A C C A A T T A G C T G T C G G T G T T G 190 C 180 G C A A A G T C G T T C A A G G A A C A A T G G G G T T C T T T T G 510 T T T T G T C T T T C T C C A A G A A C 590 T G A 680 A G C 630 C 670 G C A G C A 660 T A A

T A T 687 G C T A T 3’ G C

G T

A T 650 T T T T G C G 640 C T T A T T

Figure 1. Genomic context, expression profile in melanoma cell lines, and secondary structure of SPRY4-IT1. A, genome browser view of the SPRY4 locus showing the 2 annotated isoforms of SPRY4 (SPRY4.1 and SPRY4.2) and the position of the lncRNA SPRY4-IT1, within the second intron of SPRY4.2. Arrows indicate the direction of transcription. B, next-generation cDNA sequencing data illustrates the expression levels of AK024556 (SPRY4-IT1). The top panel depicts the transcript numbers in melanocytes and the bottom panel shows the transcript level in WM1552C. C, expression of SPRY4-IT1 in 8 melanoma cell lines, keratinocytes, and melanocytes, as determined by qRT-PCR. Error bars indicate the standard error of the mean for 3 technical replicates. Expression values are normalized to 1 in melanocytes. D, computational prediction of the secondary structure of the SPRY4-IT1 transcript, as determined by RNAfold and RNAstructure (see Materials and Methods). Blue lines indicate positions of pseudo knots. Red base-pairing indicates regions of consensus structure between the 2 algorithms. The contour of the large stem loop is highlighted in red, whereas the position of the 2 pyknons are outlined in green. including a large stem-loop from positions 220 to 321 (Fig. 1D). knots (i.e., nested helices) are predicted by ProbKnot, which The latter encompasses 1 of 2 nonrepeat associated "pyknons," provides high sensitivity and positive prediction rates (41). putative regulatory motifs that are nonrandomly distributed Consistent with the absence of expression of SPRY4-IT1 out- throughout the genome (40). In addition, 3 putative pseudo- side of primates, no compatible structures appear to be

www.aacrjournals.org Cancer Res; 71(11) June 1, 2011 3855

Khaitan et al.

significantly conserved throughout a multiple alignment of tional exon that results in translation initiation from an orthologous sequences from 31 eutherian mammals (see alternate start codon. To better understand where SPRY4 Materials and Methods). functions and the relative expression of the 2 isoforms, we used qRT-PCR to measure the expression of SPRY4.1 and Expression profiling of SPRY4 and SPRY4-IT1 in normal SPRY4.2 across 20 human tissues (Supplementary Fig. S7). human tissue The results showed that both isoforms are expressed in all SPRY4 belongs to the Sprouty (SPRY) family of genes, which tissues examined, with the highest expression found in the encode Ras/Erk inhibitor proteins. There are 4 SPRY genes lung and placenta, and the lowest in the thymus and esopha- (SPRY1, SPRY2, SPRY3, and SPRY4) in human (47, 48). Our next gus. SPRY4.1 was found to be the more abundant isoform, generation sequencing data shows that SPRY1 and SPRY3 have occurring in diverse ratios (relative to SPRY4.2) across differ- little or no expression in either melanoma or melanocytes, but ent tissues, ranging from 2.7:1 in kidney to 28:1 in thyroid. both SPRY2 and SPRY4 are highly expressed in melanoma cells Despite the differences in abundance, the expression profiles compared to melanocytes (Supplementary Fig. S5). Although of SPRY4.1 and SPRY4.2 were predominantly positively corre- the role of SPRY4 has not been examined in melanoma, SPRY4 lated (R ¼ 0.75; Pearson correlation). is down-regulated in non–small cell lung cancer (NSCLC) and Given the intronic position of SPRY4-IT1 within SPRY4,we inhibits cell growth, migration, and invasion in transfected cell next aimed to determine whether the expression of SPRY4-IT1 lines, suggesting it may function as a tumor suppressor (42). and SPRY4 were linked. Therefore, we examined the relative SPRY4 occurs in 2 alternately spliced isoforms, termed SPRY4.1 expression of SPRY4-IT1 across the same panel of 20 human and SPRY4.2 (Fig. 1A), the latter of which includes an addi- tissues (Fig. 2A). Interestingly, we found that SPRY4-IT1 was

A 0.045

0.040

0.035

0.030

0.025

0.020

(relative to RPLO ) (relative 0.015 expression SPRY4-IT1 0.010 Figure 2. Expression of SPRY4- IT1 and SPRY4 in 20 normal 0.005 human tissues. A, expression of SPRY4-IT1 relative to the 0.000 RPLO AdiposeBladderBrain CervixColon Eso Heart KidneyLiver Lung Ovary PlacentaPro Sk. muscleSm. intestineSpleenTestisThymusThyroidTrachea housekeeping control gene phagus sta in 20 normal human tissues, as te determined by qRT-PCR. Error bars indicate the standard error of B the mean for 3 technical 5 replicates. B, ratios of SPRY4-IT1 4.5 to SPRY4.1 across 20 normal human tissues. The expression of 4 SPRY4.1 was determined using RPLO 3.5 qRT-PCR using as the control gene (see Supplementary 3 Fig. S1 for SPRY4.1 and SPRY4.2 2.5 expression data).

Relative expression Relative 1.5 ) / SPRY4.1 ( SPRY4-IT1 1 0.5

0 AdiposeBladderBrain CervixColon EsophagusHeart Kidney Liver Lung Ova PlacentaProstateSk. muSm.scle intestineSpl TestisThymusThyroidTrachea ry een

3856 Cancer Res; 71(11) June 1, 2011 Cancer Research

Noncoding RNA Function SPRY4-IT1 in Human Melanoma

ABSPRY4-IT1 60 SPRY4.2 800

40 400

Figure 3. SPRY4-IT1 expression in melanoma patients. A–D, 20 measure of the relative expression 50 of SPRY4-IT1 in primary (A), nodal Relative expression (normalized to melanocytes) metastasis (B), regional Relative expression (normalized to melanocytes) 0 0 Primary Primary14 Primary 21 Prim22 Pri Noda Nodal 2Noda Nodal 5Noda Nod Nodal 11N o metastasis (C), and distant mary24 da ary 23 al 8 metastasis (D) in melanoma l 1 l 4 l 7 l 12 patient samples. The patient samples are arbitrarily numbered. CD 60 Gene expression values were determined by qRT-PCR and error 300 bars indicate the standard error of the mean for 3 technical 40 200 replicates. All expression values are normalized to 1 in melanocytes. 100 20 Relative expression (normalized to melanocytes) 0 Relative expression (normalized to melanocytes) 0 RegionalRegional 3 R 6 R R Regiona Distant Distant 1Di Dis D D e eg eg s istant 2 istan gional 27 tant 17 tant 18 ional ion al 29 t 2 l 30 9 2 3 5 6 8

more highly expressed than SPRY4.1 in several tissues, occur- expression data, confirming that SPRY4-IT1 is consistently up- ring at ratios as high as 4.5:1 in kidney (Fig. 2B). Furthermore, regulated in melanoma patient samples compared to the the range in expression for SPRY4-IT1 across the 20 different melanocyte control (Fig. 3A–D). tissues was much greater than that of SPRY4; SPRY4-IT1 varied by as much as 111-fold (placenta vs. esophagus) compared to Knockdown and localization of SPRY4-IT1 in melanoma SPRY4.1, which varied by a maximum of 10-fold (thyroid vs. cells kidney). Despite the variation in abundance and range, the To investigate the functional role of SPRY4-IT1, we used expression profile of SPRY4-IT1 was correlated with both siRNA to downregulate SPRY4-IT1 expression in melanoma SPRY4.1 (R ¼ 0.62; Pearson correlation) and SPRY4.2 (R ¼ cells. Five different siRNA molecules were tested for their 0.84; Pearson correlation). The similar expression profiles knockdown efficiency, the most efficient of which (siRNA 594) between SPRY4-IT1 and SPRY4 suggest that SPRY4-IT1 and was selected for subsequent biological studies. To determine SPRY4 may share the same transcriptional regulatory factors. the optimal concentration for knockdown, several different The 2 transcripts may either be transcribed independently concentrations of siRNA were examined in the melanoma cell from the same promoter or, alternatively, they may be tran- lines A375 and WM1552C (Fig. 4A and B). When these cells scribed as a single transcript, with SPRY4-IT1 then being were transfected with 6 nmol/L of siRNA, a 45% SPRY4-IT1 processed directly from the intron of SPRY4. In the latter silencing was observed in A375 cells, but no significant scenario, the higher abundance of SPRY4-IT1 could be changes were observed in WM1552C cells. However, 18 explained by higher stability of the lncRNA relative to the nmol/L of siRNA yielded at least 60% knockdown in both cell mRNA. lines (WM1552C and A375). These results were validated by Northern blot analysis (Fig. 4C). Although we saw a high level SPRY4-IT1 and SPRY4 expression in patient samples of SPRY4-IT1 knockdown with 30 nmol/L siRNA, we also saw In light of the complex, although generally positively cor- significant cell death (data not shown). Therefore, subsequent related, relationship between the SPRY4-IT1 and SPRY4 cell biology studies were performed with a maximum of 18 expression profiles, we next examined the expression of nmol/L siRNA. SPRY4-IT1 and SPRY4 in 25 melanoma patient samples using Given the correlated expression of SPRY4-IT1 and SPRY4,we qRT-PCR (Fig. 3A–D). We found that although the expression next aimed to determine the effect of SPRY4-IT1 knockdown of both SPRY4-IT1 and SPRY4.2 varied considerably between on SPRY4 expression in A375 cells. Using qRT-PCR, we deter- patients samples, their relative expression levels were highly mined the expression of SPRY4 following siRNA-mediated correlated (R ¼ 0.95, Pearson correlation; Supplementary knockdown of SPRY4-IT1. As a result, we found that the level Fig. S8). The qRT-PCR results also validated the microarray of SPRY4 expression was not significantly altered following

www.aacrjournals.org Cancer Res; 71(11) June 1, 2011 3857

Khaitan et al.

AB 1.6 A375 1.2 WM1552C

1.4 1.0 1.2 0.8 1.0

0.8 0.6

0.6 0.4 SPRY4-IN1 expression

SPRY4-IN1 expression 0.4 (relative to scrambled control) (relative to scrambled control) 0.2 0.2

0 0 Scrambled siRNA6 nmol/L siRNA12 nmol/L siRNA18 nmol/L siRNA Scrambled siRNA6 nmol/L siRNA12 nmol/L siRNA18 nmol/L siRNA

C MC A375 A375 A375 A375 A375 +Scr 6 nmol/L 12 nmol/L 18 nmol/L

α-SPRY4-IT1

α-U6 RNA

D Melanocyte

A375 + Scrambled siRNA A375 + 6 nmol/L siRNA A375 + 12 nmol/L siRNA A375 + 18 nmol/L siRNA

Figure 4. Knockdown and subcellular localization of SPRY4-IT1 in melanoma cells and melanocytes. A and B, expression of SPRY4-IT1 following knockdown by siRNA at different concentrations (6, 12, and 18 nmol/L) in the melanoma cell lines A375 (A) and WM1552C (B). The level of knockdown efficiency was determined by qRT-PCR. Error bars indicate the standard error of the mean for 3 technical replicates and expression values are normalized to 1 in the respective Scramble siRNA controls. C, validation of siRNA knockdown by Northern blot analysis using a probe specific to SPRY4-IT1.U6RNA was used as a load control. D, localization of SPRY4-IT1 by RNA-FISH in melanocytes and A375-transfected cells with SPRY4-IT1–targeted siRNA (or scrambled siRNA siRNA) at various concentrations. Nuclei are stained blue (DAPI) and SPRY4-IT1 is stained green. Magnification is 60 with 4.5 zoom.

3858 Cancer Res; 71(11) June 1, 2011 Cancer Research

Noncoding RNA Function SPRY4-IT1 in Human Melanoma

SPRY4-IT1 knockdown relative to the scrambled siRNA siRNA WM1552C cells transfected with SPRY4-IT1 siRNA compared control (Supplementary Fig. S9). This result confirms that the to cells transfected with scrambled siRNA siRNA. The percen- RNAi knockdown strategy did not appreciably affect the levels tage surface area between wounds was determined at 4 hours of the host SPRY4 transcript and that the phenotypic effects intervals from 0 to 24 hours. Comparison between the SPRY4- observed following knockdown of SPRY4-IT1 were driven IT1 and scrambled siRNA siRNA cells indicated a decreased directly by SPRY4-IT1. Furthermore, the result shows that mobility of the SPRY4-IT1 siRNA-transfected cells with the SPRY4-IT1 does not regulate SPRY4 expression. most pronounced effect observed at t ¼ 16 (P ¼ 0.0013; Figs. Next, we examined the expression of lncRNA SPRY4-IT1 in 5F and S10). The effect of upregulation of SPRY4-IT1 on cell A375 cell lines and melanocytes by in situ hybridization using a migration was assessed on LOX IMV1 cells, which have a locked nucleic acid (LNA) FAM-labeled probe (see Materials comparatively low level of endogenous SPRY4-IT1 expression and Methods). RNA–FISH showed that SPRY4-IT1 is localized (Fig. 1C). Comparison of the percentage surface area between as a punctate pattern in the nucleus, but the majority of the wounds of pcDNA6/SPRY4-IT1 or pcDNA6 vector only trans- signal was observed in the cell cytoplasm (Fig. 4D). Consistent fected LOX IMV1 cells, respectively, was determined at 2 hours with the abovementioned qRT-PCR results (Fig. 1C), RNA– intervals from 0 to 24 hours. The pcDNA6/SPRY4-IT1 trans- FISH also revealed that SPRY4-IT1 was highly expressed in fected cells showed an immediate increase in mobility with A375 melanoma cells compared to melanocytes. The dose- the most pronounced effect being observed at t ¼ 6 dependent reduction of RNA–FISH signal in A375 cells trans- (P < 0.0001) and wound closure occurring at 8 hours as fected with different concentrations of SPRY4-IT1–targeted compared to 20 hours for the vector transfected cells (Figs. siRNAs show that the probe was specifically detecting the 5G and S10). SPRY4-IT1 transcript. Discussion Modulation of SPRY4-IT1 expression effects metabolic viability, cell death, cell invasion, and cell migration In this study, we report the identification of melanoma- To investigate the possible role of SPRY4-IT1 on the growth specific lncRNAs and characterization of one such tran- of melanoma cells, the metabolic viability was assessed using a script, SPRY4-IT1 (GenBank accession ID AK024556) in colorimetric assay, which involves the conversion of MTT in melanoma cell lines and patient samples. SPRY4-IT1 is active mitochondria of living cells to formazan. The amount of derived from an intron of the Sprouty 4 (SPRY4) gene and formazan correlates with the number of viable cells. A375 is predicted to contain several long hairpins in its secondary melanoma cells knocked down with siRNA showed a 50% structure. Knockdown of SPRY4-IT1 expression results in decrease in metabolic viability 48 hours after transfection, defects in cell growth, decreased invasion, and increased whereas WM1552C cells showed a 30% decrease in viability rates of apoptosis in melanoma cells. A decreased capacity (Fig. 5A and B). The MTT assay shows that the downregulation for cell migration was also observed for SPRY4-IT1 knock- of SPRY4-IT1 expression decreases cell growth in melanoma down, whereas a converse increase in speed of migration cells. Next, we investigated the effects of SPRY4-IT1 knock- was noted for a melanoma cell line overexpressing SPRY4- down on apoptosis. Apoptosis was detected by labeling phos- IT1. Together with the increased expression of SPRY4-IT1 in photidylserine using FITC-conjugated Annexin V in unfixed melanoma cells and patient samples compared to melano- cells (43). The percentage of Annexin V positive–negative and cytes, we suggest that SPRY4-IT1 may play an important role PI positive–negative cells was estimated by gating the cell in melanoma pathogenesis in humans. population. A375-untreated or scrambled siRNA siRNA-trea- Although the functional significance of miRNAs is now well ted cells showed minimal annexin positive cells 48 hours after known, little is understood about the functionality of most transfection (Fig. 5C). The fraction of annexin positive cells lncRNAs. Nevertheless, the growing catalog of functionally with 6 nmol/L of siRNA was 9%. This was increased to 26% at characterized lncRNAs reveals that these transcripts are 12 nmol/L and 53% when 18 nmol/L of siRNA used for important in different physiological processes, including transfection. Interestingly, no major differences were observed embryonic stem cell differentiation (7), T-cell differentiation in propidium iodide-positive cells, indicating that the knock- (44), oligodendrogenesis (19), keratinocyte differentiation (45), down of SPRY4-IT1 induces cell death primarily through and altered expression of lncRNAs could result in cancer (46). apoptosis, not necrosis. We also examined the effect of Recently, the long intergenic RNA HOTAIR was shown to SPRY4-IT1 knockdown on the invasion of A375 melanoma regulate metastatic progression in human breast cancer. This cells (Fig. 5D and E). The results of the invasion assay shows RNA recruits Polycomb Repressive Complex 2 to specific that knockdown of SPRY4-IT1 inhibits melanoma cell invasion target genes in the genome that lead to H3K27 trimethylation by greater than 60% at 6 nmol/L of siRNA and greater than and epigenetic silencing of metastatic suppressor genes (33). 80% at 12 and 18 nmol/L. This invasion defect is significant, Interestingly, the majority of lncRNAs are transcribed close to even accounting for defects due to the loss of cell viability or within protein-coding loci, which has prompted the hypoth- (>80% invasion defect at 12 and 18 nmol/L siRNA with only a esis that lncRNAs may have cis-acting effects within these loci. 50% loss of cell viability). To investigate the independent Indeed, a number of studied lncRNAs, such as Evf2 and p15as, effects of knockdown and upregulation of SPRY4-IT1 on cell have fitted this model by influencing the expression (either migration, we utilized an in vitro scratch assay. The effect of positively or negatively) of the local protein-coding gene. SPRY4-IT1 knockdown on cell motility was assessed on Apart from intronically derived small RNAs, such as some

www.aacrjournals.org Cancer Res; 71(11) June 1, 2011 3859

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 2011 American Association for Cancer Research. 3860 hi epciecls uniiaino on eln a ae ntilct saswt ro asidctn h tnaderro h mean. the of error of point standard time the hour" indicating "0 bars the error to with compared assays was triplicate point on time Each based software. was elements healing NIS wound using of determined Quantification was cells. wounds respective the their between area surface percent software. The Aperio using counted were Matrigel the through cells invading of numbers mean Representative the experiment). of represents potential 1 bar invasive of Each the representative E, and field D (visual quadrant. assay each in presented are staining PI M1clstasetdwt pcDNA6/ either with with transfected transfected cells cells WM1552C of IMV1 motility cell of (P Comparison asterisks assay. healing by wound indicated is siRNA siRNA or siRNA siRNA acrRs 11)Jn ,2011 1, June 71(11) Res; Cancer mean of analyses either triplicate ( with represents asterisk transfected bar an cells Each by of assay. represented Viability MTT is cells. by control melanoma determined WM1552C was or siRNA A375 siRNA of viability reduces knockdown 5. Figure al. et Khaitan Downloaded from ouainof Modulation SPRY4-IT1 FG DE C A SPRY4-IT1

% Distance across the wound Published OnlineFirstMay10,2011;DOI:10.1158/0008-5472.CAN-10-4460 100 20 40 60 80 – 0 agtdsRAwr tie o nei /Iadaaye yfo yoer.Tepretg fclspstv o nei and V Annexin for positive cells of percentage The cytometry. flow by analyzed and V/PI Annexin for stained were siRNA targeted

cancerres.aacrjournals.org Absorbance OD 490 nm 0.2 0.4 0.6 0.8 1.0 01 41 82 22 26 24 22 20 18 16 14 12 10 8 6 4 2 0 0 % Cell invasion 100 Do onsfo ebae.Sgiiatyrdcdivso in invasion reduced Significantly membranes. 3 from counts of SD 50 25 75 Propidium iodide 0 xrsinefcsmtblcvaiiy eldah elivso,adcl irto nmlnm el.AadB, and A cells. melanoma in migration cell and invasion, cell death, cell viability, metabolic effects expression 01010 01010 01010 0101000 100 10 1 1000 100 10 1 1000 100 10 1 1000 100 10 1 SPRY4-IT1 .%00 .%14 .%37 .%13.7% 2.6% 3.7% 1.7% 1.4% 1.9% 0.0% 2.4%

A375

P cabe iN mlLsRA1 mlLsRA18nmol/LsiRNA 12nmol/LsiRNA 6nmol/LsiRNA Scrambled siRNA Control

< < Scramble .0) ,kokonof knockdown C, 0.008).

.4) n ,temgainptnilo el ihaltered with cells of potential migration the G, and F 0.049). Scrambled Scrambled

Time (hr) WM1552C +scrambled siRNA d WM1552C +SPRY4-IT1 siRNA

repypDA/5Hs etr()wsfloe,satn t0hu n otnigtruht 4hours. 24 to through continuing and hour 0 at starting followed, was (G) vector pcDNA6/V5-HisA empty or 6 * nmol/L siR nmol/L

0.1% siRN nmol/L 6

2 nmol/L siRNA nmol/L 2 *

12 nmol/L nmol/L 12

18 nmol/L siRNA nmol/L 18 A *

nue ppoi.Clstetdfr4 or ihete scrambled either with hours 48 for treated Cells apoptosis. induced % Distance across the wound 100

20 40 60 80 Absorbance OD 490 nm 0 0.5 1.0 1.5 2.5 2.0 Scrambled siRNA 2no/ iN 18nmol/LsiRNA 12 nmol/LsiRNA 0 iN ncdw nA7 el a xmndb arglinvasion Matrigel by examined was cells A375 in knockdown siRNA 101214161820222426 8 6 4 2 0 SPRY4-IT1 WM1552C

26.0% Control

D hr h infcn ifrnefo h scrambled the from difference significant the where SD, SPRY4-IT1 Scr –

agtdsRAo cabe iN iN F n LOX and (F) siRNA siRNA scrambled or siRNA targeted amb

Time (hr) LOX IMV1+ VO LOX IMV1+SPRY4-IT1 led led

SPRY4-IT1

6 nmol/LsiRNA 6 nmol/ 6 *

SPRY4-IT1 – siRNA L

agtdsRAa oprdt scrambled to compared as siRNA targeted 12 nmol/L si nmol/L 12 * 53.2%

xrsinwseaie ya by examined was expression

A RN

18 nmol/L si nmol/L 18 *

– RNA agtdsRAo scrambled or siRNA targeted acrResearch Cancer SPRY4-IT1 Published OnlineFirst May 10, 2011; DOI: 10.1158/0008-5472.CAN-10-4460

Noncoding RNA Function SPRY4-IT1 in Human Melanoma

snoRNAs and miRNAs, the role of other intronic ncRNAs is Previous examinations of lncRNAs have shown that many unknown. are localized to specific subcellular compartments (3, 4). The Previous large-scale analyses examining the relative expres- subcellular location of RNA can provide important insight sion of protein-coding genes and associated intronic lncRNAs into its potential function. For example, the lncRNA MEN e/b have revealed that, although there is a general trend that the is localized to nuclear paraspeckles, which is a nuclear expression of the host gene and the intronic ncRNA are domain thought to be involved in various aspects of RNA positively correlated, the relationship is seldom exclusive (3, processing. Using fluorescent in situ hybridization, we found 7). As an illustrative example, the expression of Odz3 and its that SPRY4-IT1 was localized to the cytoplasm. Although the associated intronic lncRNA are localized to the hippocampus majority of functionally characterized lncRNAs are nuclear of the adult mouse brain, but although Odz3 is expressed only localized, it has been previously noted that many lncRNAs in the CA1 subfield, the ncRNA is expressed throughout all are transported to the cytoplasm (50). Within the cytoplasm, hippocampal subfields (3). Such an expression profile suggests it is possible that the lncRNAs are functioning as protein- that intronic lncRNAs are not simply coexpressed with the scaffolds or as signaling molecules. Determination of any host mRNA, but rather that there is some independence in proteins interacting with SPRY4-IT1 will be important in their regulatory control mechanisms. This resonates with the elucidating its role and put additional context to its cellular comparative expression profiles of SPRY4-IT1 and the host localization. gene SPRY4, whose expression is correlated across most In summary, we have shown that SPRY4-IT1 expression is tissues, but is highly divergent in some others, such as kidney substantially increased in patient melanoma cell samples where it is expressed at 4.5-fold higher levels than SPRY4. compared to melanocytes. The elevated expression of Nevertheless, the generally similar expression profiles, which SPRY4-IT1 in melanoma cells, its accumulation in cell cyto- are suggestive of shared regulatory mechanisms, is consistent plasm, and its effects on cell dynamics suggest that the mis- with the notion that the function of SPRY4-IT1 is likely to be expression of SPRY4-IT1 may have an important role in related to the biological pathway of SPRY4. melanoma development, and could be an early biomarker SPRY4, a member of the Ras/Erk inhibitor encoding Sprouty and a key regulator for melanoma pathogenesis in human. family of genes, is an inhibitor of the receptor-transduced mitogen-activated protein kinase (MAPK) signaling pathway. Disclosure of Potential Conflicts of Interest It functions upstream of RAS activation and impairs the for- mation of active GTP-RAS (49). Previous studies have reported No potential conflicts of interest were disclosed. SPRY4 mRNA expression to be decreased in non– NSCLC cell lines, with SPRY4 shRNA knockdown showing increased Acknowledgments cell growth. Conversely, stable transfection of SPRY4 in NSCLC We thank Dr. James Goydos (UMDNJ-Robert Wood Johnson Medical School) cell lines lead to reduced cell growth, migration, and invasion. for clinical samples, Dr. Ravichandran (Life Technologies) for siRNA and Together, these results indicate a putative tumor suppressor Dr. Nathan Salamonis (UCSF) for preliminary data analysis. role for SPRY4 (42). The intronic positioning of SPRY4-IT1 within SPRY4, coupled with the largely shared expression profiles, Grant Support raises the hypothesis that SPRY4-IT1 may also be involved in the MAPK-signaling pathway, which would be consistent with This work was supported by an Australian Research Council/University of Queensland co-sponsored Federation Fellowship (FF0561986; J.S. Mattick), a the defects in cell growth, migration, differentiation, and National Health and Medical Research Council of Australia (NHMRC) Career increased rates of apoptosis in melanoma cells following its Development Award (CDA631542; M.E. Dinger), and a Queensland Government in silico Department of Employment, Economic Development and Innovation Smart knockdown. Examination of pathway maps indicate Futures Fellowship (M.E. Dinger). that SPRY4-IT1 interacts with Raf1, B-Raf, MEK1/2, TESK1, The costs of publication of this article were defrayed in part by the payment MARKK, and MARK2, further supporting the notion that of page charges. This article must therefore be hereby marked advertisement in SPRY4-IT1 may effect MAPK signaling. Further investigation accordance with 18 U.S.C. Section 1734 solely to indicate this fact. SPRY4-IT1 is currently underway to determine whether or not is Received December 9, 2010; revised March 3, 2011; accepted March 24, 2011; indeed involved in the MAPK signaling pathway. published OnlineFirst May 10, 2011.

References 1. Carninci P, Kasukawa T, Katayama S, Gough J, Frith MC, Maeda N, up-regulated upon muscle differentiation and are essential com- et al. The transcriptional landscape of the mammalian genome. ponents of paraspeckles. Genome Res 2009;19:347–59. Science 2005;309:1559–63. 5. Amaral PP, Mattick JS. Noncoding RNA in development. Mamm 2. Cheng J, Kapranov P, Drenkow J, Dike S, Brubaker S, Patel S, et al. Genome 2008;19:454–92. Transcriptional maps of 10 human chromosomes at 5-nucleotide 6. Blackshaw S, Harpavat S, Trimarchi J, Cai L, Huang H, Kuo WP, et al. resolution. Science 2005;308:1149–54. Genomic analysis of mouse retinal development. PLoS Biol 2004;2: 3. Mercer TR, Dinger ME, Sunkin SM, Mehler MF, Mattick JS. Specific E247. expression of long noncoding RNAs in the mouse brain. Proc Natl 7. Dinger ME, Amaral PP, Mercer TR, Pang KC, Bruce SJ, Gardiner Acad Sci U S A 2008;105:716–21. BB, et al. Long noncoding RNAsinmouseembryonicstem 4. Sunwoo H, Dinger ME, Wilusz JE, Amaral PP, Mattick JS, Spector cell pluripotency and differentiation. Genome Res 2008;18: DL. MEN varepsilon/beta nuclear-retained non-coding RNAs are 1433–45.

www.aacrjournals.org Cancer Res; 71(11) June 1, 2011 3861

Khaitan et al.

8. Rinn JL, Kertesz M, Wang JK, Squazzo SL, Xu X, Brugmann SA, et al. 30. de Kok JB, Verhaegh GW, Roelofs RW, Hessels D, Kiemeney LA, Functional demarcation of active and silent chromatin domains in Aalders TW, et al. DD3(PCA3), a very sensitive and specific marker to human HOX loci by noncoding RNAs. Cell 2007;129: 1311–23. detect prostate tumors. Cancer Res 2002;62:2695–8. 9. Pollard KS, Salama SR, King B, Kern AD, Dreszer T, Katzman S, et al. 31. Li L, Feng T, Lian Y, Zhang G, Garen A, Song X. Role of human Forces shaping the fastest evolving regions in the human genome. noncoding RNAs in the control of tumorigenesis. Proc Natl Acad Sci U PLoS Genet 2006;2:e168. S A 2009;106:12956–61. 10. Pheasant M, Mattick JS. Raising the estimate of functional human 32. Yu W, Gius D, Onyango P, Muldoon-Jacobs K, Karp J, Feinberg AP, sequences. Genome Res 2007;17:1245–53. et al. Epigenetic silencing of tumour suppressor gene p15 by its 11. Guttman M, Amit I, Garber M, French C, Lin MF, Feldser D, et al. antisense RNA. Nature 2008;451:202–6. Chromatin signature reveals over a thousand highly conserved large 33. Gupta RA, Shah N, Wang KC, Kim J, Horlings HM, Wong DJ, et al. non-coding RNAs in mammals. Nature 2009;458:223–7. Long non-coding RNA HOTAIR reprograms chromatin state to pro- 12. Zhao J, Sun BK, Erwin JA, Song JJ, Lee JT. Polycomb proteins mote cancer metastasis. Nature 2010;464:1071–6. targeted by a short repeat RNA to the mouse X chromosome. Science 34. Hofacker IL. RNA secondary structure analysis using the Vienna RNA 2008;322:750–6. package. Curr Protoc Bioinformat 2004;Chapter 12:Unit 12 2. 13. Prasanth KV, Spector DL. Eukaryotic regulatory RNAs: an answer to 35. Reuter JS, Mathews DH. RNAstructure: software for RNA second- the ‘genome complexity’ conundrum. Genes Dev 2007;21:11–42. ary structure prediction and analysis. BMC Bioinformat 2010; 14. Espinoza CA, Allen TA, Hieb AR, Kugel JF, Goodrich JA. B2 RNA binds 11:129. directly to RNA polymerase II to repress transcript synthesis. Nat 36. Gruber AR, Findeiss S, Washietl S, Hofacker IL, Stadler PF. Rnaz 2.0: Struct Mol Biol 2004;11:822–9. improved noncoding RNA detection. Pac Symp Biocomput 15. Mariner PD, Walters RD, Espinoza CA, Drullinger LF, Wagner SD, 2010;15:69–79. Kugel JF, et al. Human Alu RNA is a modular transacting repressor of 37. Gesell T, Washietl S. Dinucleotide controlled null models for com- mRNA transcription during heat shock. Mol Cell 2008;29:499–509. parative RNA gene prediction. BMC Bioinformat 2008;9:248. 16. Mattick JS, Amaral PP, Dinger ME, Mercer TR, Mehler MF. RNA 38. Katoh K, Asimenos G, Toh H. Multiple alignment of DNA sequences regulation of epigenetic processes. Bioessays 2009;31:51–9. with MAFFT. Methods Mol Biol 2009;537:39–64. 17. Fejes-Toth K, Sotirova V, Sachidanandam R, Assaf G, Hannon GJ, 39. Mazar J, DeYoung K, Khaitan D, Meister E, Almodovar A, Goydos J, Kapranov P, et al. Post-transcriptional processing generates a diver- et al. The regulation of miRNA-211 expression and its role in mela- sity of 5’-modified long and short RNAs. Nature 2009;457:1028–32. noma cell invasiveness. PLoS One 2010;5:e13779. 18. Mercer TR, Dinger ME, Bracken CP, Kolle G, Szubert JM, Korbie DJ, 40. Tsirigos A, Rigoutsos I. Human and mouse introns are linked to et al. Regulated post-transcriptional RNA cleavage diversifies the the same processes and functions through each genome's eukaryotic transcriptome. Genome Res 2010;20:1639–50. most frequent non-conserved motifs. Nucleic Acids Res 2008; 19. Mercer TR, Qureshi IA, Gokhan S, Dinger ME, Li G, Mattick JS, et al. 36:3484–93. Long noncoding RNAs in neuronal-glial fate specification and oligo- 41. Bellaousov S, Mathews DH. ProbKnot: fast prediction of RNA sec- dendrocyte lineage maturation. BMC Neurosci 2010;11:14. ondary structure including pseudoknots. RNA 2010;16:1870–80. 20. Zhao J, Sinclair J, Houghton J, Bolton E, Bradley A, Lever A. Cyto- 42. Tennis MA, Van Scoyk MM, Freeman SV, Vandervest KM, Nemenoff megalovirus beta2.7 RNA transcript protects endothelial cells against RA, Winn RA. Sprouty-4 inhibits transformed cell growth, migration apoptosis during ischemia/reperfusion injury. J Heart Lung Transplant and invasion, and epithelial-mesenchymal transition, and is regulated 2010;29:342–5. by Wnt7A through PPARgamma in non-small cell lung cancer. Mol 21. Ishii N, Ozaki K, Sato H, Mizuno H, Saito S, Takahashi A, et al. Cancer Res 2010;8:833–43. Identification of a novel non-coding RNA, MIAT, that confers risk of 43. Vermes I, Haanen C, Steffens-Nakken H, Reutelingsperger C. A novel myocardial infarction. J Hum Genet 2006;51:1087–99. assay for apoptosis. Flow cytometric detection of phosphatidylserine 22. Faghihi MA, Modarresi F, Khalil AM, Wood DE, Sahagan BG, Morgan expression on early apoptotic cells using fluorescein labelled Annexin TE, et al. Expression of a noncoding RNA is elevated in Alzheimer's V. J Immunol Methods 1995;184:39–51. disease and drives rapid feed-forward regulation of beta-secretase. 44. Pang KC, Dinger ME, Mercer TR, Malquori L, Grimmond SM, Chen W, Nat Med 2008;14:723–30. et al. Genome-wide identification of long noncoding RNAs in CD8þ T 23. Sonkoly E, Bata-Csorgo Z, Pivarcsi A, Polyanka H, Kenderessy- cells. J Immunol 2009;182:7738–48. Szabo A, Molnar G, et al. Identification and characterization of a 45. Mazar J, Sinha S, Dinger ME, Mattick JS, Perera RJ. Protein-coding novel, psoriasis susceptibility-related noncoding RNA gene, PRINS. and non-coding gene expression analysis in differentiating human J Biol Chem 2005;280:24159–67. keratinocytes using a three-dimensional epidermal equivalent. Mol 24. Daughters RS, Tuttle DL, Gao W, Ikeda Y, Moseley ML, Ebner TJ, et al. Genet Genom 2010;284:1–9. RNA gain-of-function in spinocerebellar ataxia type 8. PLoS Genet 46. Hall PA, Russell SH. New perspectives on neoplasia and the RNA 2009;5:e1000600. world. Hematol Oncol 2005;23:49–53. 25. Woo CJ, Kingston RE. HOTAIR lifts noncoding RNAs to new levels. 47. FongCW,ChuaMS,McKieAB,LingSH,MasonV,LiR,etal. Cell 2007;129:1257–9. Sprouty 2, an inhibitor of mitogen-activated protein kinase signal- 26. Pibouin L, Villaudy J, Ferbus D, Muleris M, Prosperi MT, Remvikos Y, ing, is down-regulated in hepatocellular carcinoma. Cancer Res et al. Cloning of the mRNA of overexpression in colon carcinoma-1: a 2006;66:2048–58. sequence overexpressed in a subset of colon carcinomas. Cancer 48. Lo TL, Fong CW, Yusoff P, McKie AB, Chua MS, Leung HY, et al. Genet Cytogenet 2002;133:55–60. Sprouty and cancer: the first terms report. Cancer Lett 2006;242:141– 27. Fu X, Ravindranath L, Tran N, Petrovics G, Srivastava S. Regulation of 50. apoptosis by a prostate-specific and prostate cancer-associated 49. Leeksma OC, Van Achterberg TA, Tsumura Y, Toshima J, Eldering E, noncoding gene, PCGEM1. DNA Cell Biol 2006;25:135–41. Kroes WG, et al. Human sprouty 4, a new ras antagonist on 5q31, 28. Calin GA, Liu CG, Ferracin M, Hyslop T, Spizzo R, Sevignani C, et al. interacts with the dual specificity kinase TESK1. Eur J Biochem Ultraconserved regions encoding ncRNAs are altered in human leu- 2002;269:2546–56. kemias and carcinomas. Cancer Cell 2007;12:215–29. 50. Barthelson RA, Lambert GM, Vanier C, Lynch RM, Galbraith DW. 29. Lin R, Maeda S, Liu C, Karin M, Edgington TS. A large noncoding RNA Comparison of the contributions of the nuclear and cytoplasmic is a marker for murine hepatocellular carcinomas and a spectrum of compartments to global gene expression in human cells. BMC Genom human carcinomas. Oncogene 2007;26:851–8. 2007;8:340.

3862 Cancer Res; 71(11) June 1, 2011 Cancer Research

The Melanoma−Upregulated Long Noncoding RNA SPRY4-IT1 Modulates Apoptosis and Invasion

Divya Khaitan, Marcel E. Dinger, Joseph Mazar, et al.

Cancer Res 2011;71:3852-3862. Published OnlineFirst May 10, 2011.

Updated version Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-10-4460

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2011/05/06/0008-5472.CAN-10-4460.DC1

Cited articles This article cites 49 articles, 16 of which you can access for free at: http://cancerres.aacrjournals.org/content/71/11/3852.full#ref-list-1

Citing articles This article has been cited by 20 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/71/11/3852.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/71/11/3852. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.