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

There is considerable interest in understanding the function of RNA transcripts that do not code for proteins in eukaryotic cells. As evidenced by cDNA cloning projects (1) and genomic tiling arrays (2), more than 90% of the human genome undergoes transcription, but does not code for proteins. These transcriptional products are referred to as nonprotein coding RNAs (ncRNAs). A variety of ncRNAs, such as ribosomal RNAs, transfer RNAs, and spliceosomal RNAs, are essential for cell function. Similarly, large numbers of short ncRNAs such as micro-RNAs (miRNAs), endogenous siRNAs, PIWI-interacting RNAs (piRNAs), and small nucleolar RNAs (snoRNAs) are also known to play important regulatory roles in eukaryotic cells. Recent studies have showed that the mammalian genome expresses large numbers of long ncRNAs (lncRNA; ref. 1) that are dynamically expressed in tissue-, differentiation stage-, and cell type-specific patterns (3), at least some of which localize into specific subcellular compartments (3, 4). lncRNAs are also known to play important roles during cellular development and differentiation (5–8) consistent with the view that they are under evolutionary selection (9–11). Interestingly, lncRNAs can influence the expression of specific target proteins at specific genomic loci (12, 13), modulate the activity of protein binding partners (14, 15), direct chromatin-modifying complexes to their sites of action (16), and are posttranscriptionally processed to produce numerous 5′-capped small RNAs (17, 18). Epigenetic pathways can also regulate the differential expression of lncRNAs (19). 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). This misregulation has also been shown in various types of cancers, such as breast cancer (8, 25), colon cancer (26), prostate cancer (27), hepatocellular carcinoma (28, 29), and leukemia (28). One such lncRNA, DD3 (also known as PCA3), is listed as a prostate cancer-specific biomarker (30). Recent studies have revealed the contribution of ncRNAs as proto-oncogenes, for example GAGE6 (31), as tumor suppressor genes, for example p15 (32), in tumorigenesis, and as drivers of metastatic transformation, for example HOTAIR in breast cancer (33).

The focus of this study was to identify and characterize lncRNAs that are differentially expressed in melanoma compared to melanocytes and normal skin. We find that a group of lncRNAs is differentially regulated in melanoma and one such lncRNA, SPRY4-IT1, derived from an intron within the SPRY4 gene, is upregulated in melanoma cells. Knocking-down its expression resulted in defects in cell growth, invasion, and elevated rates of apoptosis in melanoma cells.

The most recent versions of RNAfold (34) and RNAstructure (35) were employed for generating RNA secondary structures. Both of these programs implement an RNA partition function algorithm, which was chosen for 2 reasons: (i) it produces a structure almost identical to the minimum free energy algorithm with RNAfold with few proximal suboptimal structures and (ii) it is required for subsequent prediction of pseudoknots with ProbKnot (included in RNAstructure).

The evolutionary conservation of secondary structures was conducted with the consensus-based programs RNAz (36) and SISSIz (37) on the Enredo-Pecan-Ortheus 31-way eutherian mammal genome alignment from ENSEMBL. Orthologous sequences to SPRY4-IT1 were selected and realigned with MAFFT, using the mafft-ginsi algorithm (38). Sliding window ranges of 100 nt window with 25 nt slide, 150 nt window with 50 nt slide, and 300 nt window with 100 nt slide were tested with both RNAz and SISSIz, using parameters “-d” and “-d -t -n 200 -p 0.02”, respectively.

Five different siRNA siRNAs that targeted SPRY4-IT1 RNA and a scrambled siRNA siRNA control were generously provided by Life Technologies. The siRNA siRNA molecules are 25 base-pair double-stranded RNA oligonucleotides with proprietary chemical modifications. The BLOCK-iT RNAi designer was used to find gene-specific 25 nucleotide siRNA siRNA molecules. It uses gene-specific targets for RNAi analysis and reports up to 10 top scoring siRNA siRNA targets. The freeze-dried siRNAs were dissolved in RNase free-water and stored as aliquots at −20°C. The siRNA with the sequence gctttctgattccaaggcctattaa yielded the highest degree of SPRY4-IT1 knockdown.

Oligonucleotides for amplification of full length SPRY4-IT1 (SPRY4-IT1 Forward 5′-TAAGCTTGTAGAGATGGGGGTTTCATCCTGTTGG-3′ and SPRY4-IT1 Reverse 5′-ACTCGAGAAAGACTCCCTTTCCTTAAGCAGATTCAC-3′) were designed to incorporate external HindIII and XhoI sites, respectively. The melanocyte genomic DNA SPRY4-IT1 amplicon (Amplitaq Gold, Life Technologies) was cloned into the pCR4-TOPO vector (Life Technologies), sequence verified and subcloned via HindIII/XhoI digestion into the pcDNA6/V5-HisA mammalian expression vector (Life Technologies).

siRNA transfection was performed with lipofectamine RNAiMax (Life Technologies) in 6-well plates. A total of 6, 12, and 18 nmol/L RNAi duplexes were diluted in 500 μL serum free medium, mixed gently and 5 μL of lipofectamine RNAiMAX was added to each well containing the diluted RNAi molecules. This mixture was incubated for 20 minutes at room temperature before the transfection. A total of 250,000 cells were diluted in complete Tu growth medium (ref. 39; without antibiotics) and plated in each well. RNAi duplex—lipofectamine RNAiMAX complexes were added to each well and mixed gently by rocking the plate. Plasmid transfection was performed in MatTek 1.5-mm glass-bottom dishes. Five micrograms of pcDNA6/SPRY4-IT1 or pcDNA6/V5-HisA (negative control) was diluted in 100 μL serum-free medium to which 5 μL of Fugene 6 was added. Following a 20-minute incubation, the respective transfection mixtures were added their respective dishes. In all cases, cells were incubated for 48 hours at 37°C in a CO₂ incubator and gene knockdown or overexpression levels were assessed by quantitative real-time PCR (qRT-PCR).

WM1552C cells were transfected with 18 nmol/L SPRY4-IT1 or scrambled siRNA siRNA (Life Technologies), respectively, and LOX IMV1 cells were transfected with pcDNA6/SPRY4-IT1 or empty pcDNA6/V5-HisA respectively, as described earlier. All transfected cell samples were seeded on Mat Tek 1.5-mm glass-bottom dishes until 90% to 95% confluent. Cell monolayers were then gently scratched with a pipette tip across the diameter of the dish and rinsed with PBS and cell media to remove cellular debris. The surface area of the scratched surface was quantified after wounding and again every 20 minutes for 24 hours on a Nikon BioStation IM (Nikon Instruments, Inc.) cell incubator. The extent of wound closure was calculated using the ratio of the surface area between the remaining wound edges to the surface area of the initial wound for each time point. These data were then expressed as a percentage of wound closure relative to control conditions for each experiment. The change in surface area was calculated using NIS Elements software (Nikon Instruments, Inc.) and performed in triplicate.

Locked nucleic acid (LNA) modified probes for human lncRNA SPRY4-IT1(5-FAM- TCCACTGGGCATATTCTAAAA-36-FAM) and a negative/scramble control (5TYE665-GTGTAACACGTCTATACGCCCA-3 TYE665, miRCURY-LNA detection probe, Exiqon) were used for RNA–FISH. In situ hybridization was performed using the RiboMap in situ hybridization kit (Ventana Medical Systems, Inc.) on a Ventana machine. The cell suspension was diluted to 10,000 cells/100 μL and pipetted into clonal rings on the autoclaved glass slides. The following day, the clonal rings were removed; slides were washed in PBS and fixed in 4% paraformaldehyde and 5% acetic acid. After acid treatment using hydrochloride-based RiboClear reagent (Ventana Medical Systems) for 10 min at 37°C, the slides were treated with the ready-to-use protease 3 reagent. The cells were hybridized with the antisense LNA riboprobe (40 nmol/L) using RiboHybe hybridization buffer (Ventana Medical Systems) for 2 hours at 58°C after an initial denaturing prehybridization step for 4 min at 80°C. Next, the slides were subjected to a low-stringency wash with 0.1× SSC (Ventana Medical Systems) for 4 min at 60°C, and then 2 further washing steps with 1× SSC for 4 min at 60°C. These slides were fixed in RiboFix and counterstained with 4′-6′diamidino-2-phenylindole (DAPI), in an antifade reagent (Ventana). The images were acquired using a Nikon A1R VAAS laser point- and resonant-scanning confocal microscope equipped with a single photon Ar-ion laser at 60× with 4× zoom.

MTT (3-(4,5-dimethyl-2-yl)-2,5-diphenyl-2II-tetrazolium bromide) was purchased from Roche. Cells were plated in 96-well plates (5000 cells/100 μL/well). After 48 hours of transfection, 20 μL MTT solution was added and the cells were incubated at 37°C in the dark for 4 hours. The generated formazan OD was measured at 490 nm to determine the cell viability on the Flex station (Molecular Devices; www.moleculardevices.com).

BD BioCoat growth factor reduced insert plates (Matrigel Invasion Chamber 12 well plates) were prepared by rehydrating the BD Matrigel matrix coating in the inserts with 0.5 mL of serum-free complete Tu media for 2 hours at 37°C. The rehydration solution was carefully removed from the inserts, 500 μL complete Tu (2% FBS) was added to the lower wells of the plate. 1 × 10⁴ transfected and untransfected cells suspended in 500 μL of serum-free complete Tu media was added to the top of each insert well. Invasion assay plates were incubated for 48 hours at 37°C. Following incubation, the noninvading cells were removed by scrubbing the upper surface of the insert. The cells on the lower surface of the insert were stained with crystal violet and each trans-well membrane mounted on a microscope slide for visualization and analysis. The slides were scanned in scanscope and the number of cells migrating was counted using Aperio software (www.aperio.com). Data are expressed as the percent invasion through the membrane relative to the migration through the control membrane.

See Supplementary Material (S10) for RNA extraction, microarray analysis, qRT-PCR, Northern blot analysis, and phosphatidylserine externalization.

To identify lncRNAs involved in melanoma, we analyzed total RNA from a stage III melanoma cell line (WM1552C), melanocytes, and keratinocytes by using a noncoding RNA microarray (NCode human array; Life Technologies). The microarrays contain probes to target 12,784 lncRNAs and 25,409 mRNAs. In total, we identified 77 lncRNAs that were significantly differentially expressed (P < 0.015; fold-change >2) in WM1552C relative to melanocytes. In addition to cell line profiling, 29 independent melanoma patient samples (graded as primary in situ, regional metastatic, distant metastatic and nodal metastatic), and 6 normal skin samples were also analyzed using the same microarrays. Hierarchical clustering was used to represent the differential lncRNA expression profiles (Supplementary Fig. S1) and revealed that potential signatures could be used to discriminate between normal and melanoma samples. To select candidates for functional studies, we asked whether any of the differentially expressed lncRNAs in melanoma cell lines were also differentially expressed in patient samples. As a result we identified 4 candidate ncRNAs (AF085920, AK091731, AK128206, and AK024556), which were differentially expressed in both melanoma cell lines and patients samples relative to melanocytes and whose differential expression we were able to validate by next generation sequencing of polyadenylated RNA isolated from melanocytes and WM1552C cells (Supplementary Fig. S2 and Fig. 1A and B). We next examined the genomic context of the 4 lncRNA candidates and found that 3 of the lncRNAs (GenBank IDs AF085920, AK091731, and AK128206) mapped within a few kilobases of the 3′UTRs of known protein-coding genes and one (GenBank ID AK024556) mapped within an intronic region (Supplementary Fig. S3). Because of the possibility of read-through expression from the upstream protein-coding genes and the complications in interpretation that this would create in subsequent studies to characterize these transcripts, we excluded the 3 lncRNAs that were positioned downstream of 3′UTRs from further study. Therefore, given the intronic position of the remaining candidate, which incidentally also had the highest degree of upregulation in melanoma cells compared to melanocytes (12-fold), we pursued this lncRNA for further study. AK024556 is a 687 nt unspliced, polyadenylated transcript originally identified in adipose tissue and is transcribed from the second intron of the SPRY4 gene (Fig. 1A), which led to our naming the transcript SPRY4-IT1. This region is not conserved beyond the primate genomes and there is no EST expression detected in mouse. Secondary structure prediction of the SPRY4-IT1 sequence indicates that this ncRNA transcript could fold into long stable hairpin structures, suggesting that SPRY4-IT1 may function intrinsically as an RNA molecule.

A comparison of SPRY4-IT1 in kidney, blood, and breast cell lines revealed expression to be equal to that of melanocytes or less (Supplementary Fig. S4). We then measured the expression levels of SPRY4-IT1 (Fig. 1C) as well as the SPRY4 ORF (Supplementary Fig. S6) in 7 additional nonpigmented melanoma cell lines (WM793B, A375, SKMEL-2, RPMI 7951, HT-144, LOX-IMV1, and G361) by qRT-PCR and the results showed that the expression of both SPRY-IN1 and SPRY4 was elevated in most of the melanoma cell lines relative to control melanocytes.

To determine whether the SPRY4-IT1 RNA contained any particular secondary structural features, the SPRY4-IT1 genomic sequence was submitted to secondary structure and pseudoknot prediction (see Materials and Methods). Several helical regions are commonly predicted by both algorithms, including a large stem-loop from positions 220 to 321 (Fig. 1D). The latter encompasses 1 of 2 nonrepeat associated “pyknons,” putative regulatory motifs that are nonrandomly distributed throughout the genome (40). In addition, 3 putative pseudoknots (i.e., nested helices) are predicted by ProbKnot, which provides high sensitivity and positive prediction rates (41). Consistent with the absence of expression of SPRY4-IT1 outside of primates, no compatible structures appear to be significantly conserved throughout a multiple alignment of orthologous sequences from 31 eutherian mammals (see Materials and Methods).

SPRY4 belongs to the Sprouty (SPRY) family of genes, which encode Ras/Erk inhibitor proteins. There are 4 SPRY genes (SPRY1, SPRY2, SPRY3, and SPRY4) in human (47, 48). Our next generation sequencing data shows that SPRY1 and SPRY3 have little or no expression in either melanoma or melanocytes, but both SPRY2 and SPRY4 are highly expressed in melanoma cells compared to melanocytes (Supplementary Fig. S5). Although the role of SPRY4 has not been examined in melanoma, SPRY4 is down-regulated in non–small cell lung cancer (NSCLC) and inhibits cell growth, migration, and invasion in transfected cell lines, suggesting it may function as a tumor suppressor (42). SPRY4 occurs in 2 alternately spliced isoforms, termed SPRY4.1 and SPRY4.2 (Fig. 1A), the latter of which includes an additional exon that results in translation initiation from an alternate start codon. To better understand where SPRY4 functions and the relative expression of the 2 isoforms, we used qRT-PCR to measure the expression of SPRY4.1 and SPRY4.2 across 20 human tissues (Supplementary Fig. S7). The results showed that both isoforms are expressed in all tissues examined, with the highest expression found in the lung and placenta, and the lowest in the thymus and esophagus. SPRY4.1 was found to be the more abundant isoform, occurring in diverse ratios (relative to SPRY4.2) across different tissues, ranging from 2.7:1 in kidney to 28:1 in thyroid. Despite the differences in abundance, the expression profiles of SPRY4.1 and SPRY4.2 were predominantly positively correlated (R = 0.75; Pearson correlation).

Given the intronic position of SPRY4-IT1 within SPRY4, we next aimed to determine whether the expression of SPRY4-IT1 and SPRY4 were linked. Therefore, we examined the relative expression of SPRY4-IT1 across the same panel of 20 human tissues (Fig. 2A). Interestingly, we found that SPRY4-IT1 was more highly expressed than SPRY4.1 in several tissues, occurring at ratios as high as 4.5:1 in kidney (Fig. 2B). Furthermore, the range in expression for SPRY4-IT1 across the 20 different tissues was much greater than that of SPRY4; SPRY4-IT1 varied by as much as 111-fold (placenta vs. esophagus) compared to SPRY4.1, which varied by a maximum of ∼10-fold (thyroid vs. kidney). Despite the variation in abundance and range, the expression profile of SPRY4-IT1 was correlated with both SPRY4.1 (R = 0.62; Pearson correlation) and SPRY4.2 (R = 0.84; Pearson correlation). The similar expression profiles between SPRY4-IT1 and SPRY4 suggest that SPRY4-IT1 and SPRY4 may share the same transcriptional regulatory factors. The 2 transcripts may either be transcribed independently from the same promoter or, alternatively, they may be transcribed as a single transcript, with SPRY4-IT1 then being processed directly from the intron of SPRY4. In the latter scenario, the higher abundance of SPRY4-IT1 could be explained by higher stability of the lncRNA relative to the mRNA.

In light of the complex, although generally positively correlated, relationship between the SPRY4-IT1 and SPRY4 expression profiles, we next examined the expression of SPRY4-IT1 and SPRY4 in 25 melanoma patient samples using qRT-PCR (Fig. 3A–D). We found that although the expression of both SPRY4-IT1 and SPRY4.2 varied considerably between patients samples, their relative expression levels were highly correlated (R = 0.95, Pearson correlation; Supplementary Fig. S8). The qRT-PCR results also validated the microarray expression data, confirming that SPRY4-IT1 is consistently up-regulated in melanoma patient samples compared to the melanocyte control (Fig. 3A–D).

To investigate the functional role of SPRY4-IT1, we used siRNA to downregulate SPRY4-IT1 expression in melanoma cells. Five different siRNA molecules were tested for their knockdown efficiency, the most efficient of which (siRNA 594) was selected for subsequent biological studies. To determine the optimal concentration for knockdown, several different concentrations of siRNA were examined in the melanoma cell lines A375 and WM1552C (Fig. 4A and B). When these cells were transfected with 6 nmol/L of siRNA, a 45% SPRY4-IT1 silencing was observed in A375 cells, but no significant changes were observed in WM1552C cells. However, 18 nmol/L of siRNA yielded at least 60% knockdown in both cell lines (WM1552C and A375). These results were validated by Northern blot analysis (Fig. 4C). Although we saw a high level of SPRY4-IT1 knockdown with 30 nmol/L siRNA, we also saw significant cell death (data not shown). Therefore, subsequent cell biology studies were performed with a maximum of 18 nmol/L siRNA.

Given the correlated expression of SPRY4-IT1 and SPRY4, we next aimed to determine the effect of SPRY4-IT1 knockdown on SPRY4 expression in A375 cells. Using qRT-PCR, we determined the expression of SPRY4 following siRNA-mediated knockdown of SPRY4-IT1. As a result, we found that the level of SPRY4 expression was not significantly altered following SPRY4-IT1 knockdown relative to the scrambled siRNA siRNA control (Supplementary Fig. S9). This result confirms that the RNAi knockdown strategy did not appreciably affect the levels of the host SPRY4 transcript and that the phenotypic effects observed following knockdown of SPRY4-IT1 were driven directly by SPRY4-IT1. Furthermore, the result shows that SPRY4-IT1 does not regulate SPRY4 expression.

Next, we examined the expression of lncRNA SPRY4-IT1 in A375 cell lines and melanocytes by in situ hybridization using a locked nucleic acid (LNA) FAM-labeled probe (see Materials and Methods). RNA–FISH showed that SPRY4-IT1 is localized as a punctate pattern in the nucleus, but the majority of the signal was observed in the cell cytoplasm (Fig. 4D). Consistent with the abovementioned qRT-PCR results (Fig. 1C), RNA–FISH also revealed that SPRY4-IT1 was highly expressed in A375 melanoma cells compared to melanocytes. The dose-dependent reduction of RNA–FISH signal in A375 cells transfected with different concentrations of SPRY4-IT1–targeted siRNAs show that the probe was specifically detecting the SPRY4-IT1 transcript.

To investigate the possible role of SPRY4-IT1 on the growth of melanoma cells, the metabolic viability was assessed using a colorimetric assay, which involves the conversion of MTT in active mitochondria of living cells to formazan. The amount of formazan correlates with the number of viable cells. A375 melanoma cells knocked down with siRNA showed a 50% decrease in metabolic viability 48 hours after transfection, whereas WM1552C cells showed a 30% decrease in viability (Fig. 5A and B). The MTT assay shows that the downregulation of SPRY4-IT1 expression decreases cell growth in melanoma cells. Next, we investigated the effects of SPRY4-IT1 knockdown on apoptosis. Apoptosis was detected by labeling phosphotidylserine using FITC-conjugated Annexin V in unfixed cells (43). The percentage of Annexin V positive–negative and PI positive–negative cells was estimated by gating the cell population. A375-untreated or scrambled siRNA siRNA-treated cells showed minimal annexin positive cells 48 hours after transfection (Fig. 5C). The fraction of annexin positive cells with 6 nmol/L of siRNA was 9%. This was increased to 26% at 12 nmol/L and 53% when 18 nmol/L of siRNA used for transfection. Interestingly, no major differences were observed in propidium iodide-positive cells, indicating that the knockdown of SPRY4-IT1 induces cell death primarily through apoptosis, not necrosis. We also examined the effect of SPRY4-IT1 knockdown on the invasion of A375 melanoma cells (Fig. 5D and E). The results of the invasion assay shows that knockdown of SPRY4-IT1 inhibits melanoma cell invasion by greater than 60% at 6 nmol/L of siRNA and greater than 80% at 12 and 18 nmol/L. This invasion defect is significant, even accounting for defects due to the loss of cell viability (>80% invasion defect at 12 and 18 nmol/L siRNA with only a 50% loss of cell viability). To investigate the independent effects of knockdown and upregulation of SPRY4-IT1 on cell migration, we utilized an in vitro scratch assay. The effect of SPRY4-IT1 knockdown on cell motility was assessed on WM1552C cells transfected with SPRY4-IT1 siRNA compared to cells transfected with scrambled siRNA siRNA. The percentage surface area between wounds was determined at 4 hours intervals from 0 to 24 hours. Comparison between the SPRY4-IT1 and scrambled siRNA siRNA cells indicated a decreased mobility of the SPRY4-IT1 siRNA-transfected cells with the most pronounced effect observed at t = 16 (P = 0.0013; Figs. 5F and S10). The effect of upregulation of SPRY4-IT1 on cell migration was assessed on LOX IMV1 cells, which have a comparatively low level of endogenous SPRY4-IT1 expression (Fig. 1C). Comparison of the percentage surface area between wounds of pcDNA6/SPRY4-IT1 or pcDNA6 vector only transfected LOX IMV1 cells, respectively, was determined at 2 hours intervals from 0 to 24 hours. The pcDNA6/SPRY4-IT1 transfected cells showed an immediate increase in mobility with the most pronounced effect being observed at t = 6 (P < 0.0001) and wound closure occurring at 8 hours as compared to 20 hours for the vector transfected cells (Figs. 5G and S10).

In this study, we report the identification of melanoma-specific lncRNAs and characterization of one such transcript, SPRY4-IT1 (GenBank accession ID AK024556) in melanoma cell lines and patient samples. SPRY4-IT1 is derived from an intron of the Sprouty 4 (SPRY4) gene and is predicted to contain several long hairpins in its secondary structure. Knockdown of SPRY4-IT1 expression results in defects in cell growth, decreased invasion, and increased rates of apoptosis in melanoma cells. A decreased capacity for cell migration was also observed for SPRY4-IT1 knockdown, whereas a converse increase in speed of migration was noted for a melanoma cell line overexpressing SPRY4-IT1. Together with the increased expression of SPRY4-IT1 in melanoma cells and patient samples compared to melanocytes, we suggest that SPRY4-IT1 may play an important role in melanoma pathogenesis in humans.

Although the functional significance of miRNAs is now well known, little is understood about the functionality of most lncRNAs. Nevertheless, the growing catalog of functionally characterized lncRNAs reveals that these transcripts are important in different physiological processes, including embryonic stem cell differentiation (7), T-cell differentiation (44), oligodendrogenesis (19), keratinocyte differentiation (45), and altered expression of lncRNAs could result in cancer (46). Recently, the long intergenic RNA HOTAIR was shown to regulate metastatic progression in human breast cancer. This RNA recruits Polycomb Repressive Complex 2 to specific target genes in the genome that lead to H3K27 trimethylation and epigenetic silencing of metastatic suppressor genes (33). Interestingly, the majority of lncRNAs are transcribed close to or within protein-coding loci, which has prompted the hypothesis that lncRNAs may have cis-acting effects within these loci. Indeed, a number of studied lncRNAs, such as Evf2 and p15as, have fitted this model by influencing the expression (either positively or negatively) of the local protein-coding gene. Apart from intronically derived small RNAs, such as some snoRNAs and miRNAs, the role of other intronic ncRNAs is unknown.

Previous large-scale analyses examining the relative expression of protein-coding genes and associated intronic lncRNAs have revealed that, although there is a general trend that the expression of the host gene and the intronic ncRNA are positively correlated, the relationship is seldom exclusive (3, 7). As an illustrative example, the expression of Odz3 and its associated intronic lncRNA are localized to the hippocampus of the adult mouse brain, but although Odz3 is expressed only in the CA1 subfield, the ncRNA is expressed throughout all hippocampal subfields (3). Such an expression profile suggests that intronic lncRNAs are not simply coexpressed with the host mRNA, but rather that there is some independence in their regulatory control mechanisms. This resonates with the comparative expression profiles of SPRY4-IT1 and the host gene SPRY4, whose expression is correlated across most tissues, but is highly divergent in some others, such as kidney where it is expressed at 4.5-fold higher levels than SPRY4. Nevertheless, the generally similar expression profiles, which are suggestive of shared regulatory mechanisms, is consistent with the notion that the function of SPRY4-IT1 is likely to be related to the biological pathway of SPRY4.

SPRY4, a member of the Ras/Erk inhibitor encoding Sprouty family of genes, is an inhibitor of the receptor-transduced mitogen-activated protein kinase (MAPK) signaling pathway. It functions upstream of RAS activation and impairs the formation of active GTP-RAS (49). Previous studies have reported SPRY4 mRNA expression to be decreased in non-NSCLC cell lines, with SPRY4 shRNA knockdown showing increased cell growth. Conversely, stable transfection of SPRY4 in NSCLC cell lines lead to reduced cell growth, migration, and invasion. Together, these results indicate a putative tumor suppressor role for SPRY4 (42). The intronic positioning of SPRY4-IT1 within SPRY4, coupled with the largely shared expression profiles, raises the hypothesis that SPRY4-IT1 may also be involved in the MAPK-signaling pathway, which would be consistent with the defects in cell growth, migration, differentiation, and increased rates of apoptosis in melanoma cells following its knockdown. Examination of in silico pathway maps indicate that SPRY4-IT1 interacts with Raf1, B-Raf, MEK1/2, TESK1, MARKK, and MARK2, further supporting the notion that SPRY4-IT1 may effect MAPK signaling. Further investigation is currently underway to determine whether or not SPRY4-IT1 is indeed involved in the MAPK signaling pathway.

Previous examinations of lncRNAs have shown that many are localized to specific subcellular compartments (3, 4). The subcellular location of RNA can provide important insight into its potential function. For example, the lncRNA MEN ϵ/β is localized to nuclear paraspeckles, which is a nuclear domain thought to be involved in various aspects of RNA processing. Using fluorescent in situ hybridization, we found that SPRY4-IT1 was localized to the cytoplasm. Although the majority of functionally characterized lncRNAs are nuclear localized, it has been previously noted that many lncRNAs are transported to the cytoplasm (50). Within the cytoplasm, it is possible that the lncRNAs are functioning as protein-scaffolds or as signaling molecules. Determination of any proteins interacting with SPRY4-IT1 will be important in elucidating its role and put additional context to its cellular localization.

In summary, we have shown that SPRY4-IT1 expression is substantially increased in patient melanoma cell samples compared to melanocytes. The elevated expression of SPRY4-IT1 in melanoma cells, its accumulation in cell cytoplasm, and its effects on cell dynamics suggest that the mis-expression of SPRY4-IT1 may have an important role in melanoma development, and could be an early biomarker and a key regulator for melanoma pathogenesis in human.

No potential conflicts of interest were disclosed.

We thank Dr. James Goydos (UMDNJ-Robert Wood Johnson Medical School) for clinical samples, Dr. Ravichandran (Life Technologies) for siRNA and Dr. Nathan Salamonis (UCSF) for preliminary data analysis.

This work was supported by an Australian Research Council/University of Queensland co-sponsored Federation Fellowship (FF0561986; J.S. Mattick), a National Health and Medical Research Council of Australia (NHMRC) Career Development Award (CDA631542; M.E. Dinger), and a Queensland Government Department of Employment, Economic Development and Innovation Smart Futures Fellowship (M.E. Dinger).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.