LncRNA SNHG17 interacts with LRPPRC to stabilize c-Myc protein and promote G1/S transition and cell proliferation

Oncogenic c-Myc is a master regulator of G1/S transition. Long non-coding RNAs (lncRNAs) emerge as new regulators of various cell activities. Here, we found that lncRNA SnoRNA Host Gene 17 (SNHG17) was elevated at the early G1-phase of cell cycle. Both gain- and loss-of function studies disclosed that SNHG17 increased c-Myc protein level, accelerated G1/S transition and cell proliferation, and consequently promoted tumor cell growth in vitro and in vivo. Mechanistically, the 1-150-nt of SNHG17 physically interacted with the 1035-1369-aa of leucine rich pentatricopeptide repeat containing (LRPPRC) protein, and disrupting this interaction abrogated the promoting role of SNHG17 in c-Myc expression, G1/S transition, and cell proliferation. The effect of SNHG17 in stimulating cell proliferation was attenuated by silencing c-Myc or LRPPRC. Furthermore, silencing SNHG17 or LRPPRC increased the level of ubiquitylated c-Myc and reduced the stability of c-Myc protein. Analysis of human hepatocellular carcinoma (HCC) tissues revealed that SNHG17, LRPPRC, and c-Myc were significantly upregulated in HCC, and they showed a positive correlation with each other. High level of SNHG17 or LRPPRC was associated with worse survival of HCC patients. These data suggest that SNHG17 may inhibit c-Myc ubiquitination and thus enhance c-Myc level and facilitate proliferation by interacting with LRPPRC. Our findings identify a novel SNHG17-LRPPRC-c-Myc regulatory axis and elucidate its roles in G1/S transition and tumor growth, which may provide potential targets for cancer therapy.


INTRODUCTION
The G1 to S phase (G1/S) transition is the key step that drives cell to the division cycle, and it is tightly controlled by the cyclindependent kinase (CDK)-Rb-E2F pathway [1,2]. c-Myc is a master regulator of cell proliferation via acting as a transcription factor. Upon stimulation of growth signals, c-Myc increases rapidly and then activates the transcription of CDKs, cyclins, E2F1, and/or inhibits the expression of CDK inhibitors, resulting in accelerating G1/S transition and cell proliferation [3,4].
Upregulation of c-Myc is frequently observed in a variety of tumors. Abnormal overexpression of c-Myc facilitates cancer development and progression by promoting cell proliferation, metabolic adaptation and suppressing the antitumor immune response [4][5][6][7], suggesting c-Myc as an attractive target for anti-cancer therapy. However, targeting c-Myc directly is intractable owing to its lack of a hydrophobic pocket that is compatible with a small organic molecule [8]. Nevertheless, recent studies in animal models show that selective small molecules targeting the BET bromodomains-containing proteins that regulate c-Myc activity can downregulate the c-Myc transcriptional signaling, reduce tumor burden, and prolong survival. Some of these small molecules already enter Phase I clinical trials [9,10]. Therefore, targeting c-Myc regulators may represent a potential strategy for anti-cancer therapy to inhibit c-Myc activity.
Long non-coding RNAs (lncRNAs) belong to a class of noncoding transcripts longer than 200 nucleotides [11]. They may work as signals, as scaffolds for protein-protein interactions, as molecular decoys, or guides to target elements [12]. LncRNAs have emerged as new regulators of various cell activities, such as cell proliferation, death, and differentiation [13,14]. Deregulation of lncRNAs can promote tumor development, self-renewal of cancer stem cells, and drug resistance [15,16]. Recent evidence shows that knockdown of lncRNA EPIC1 reduces the c-Myc's occupancy on the promoters of c-Myc target genes and curtails their expression [17]. In contrast, lncRNA GLCC1 and PVT1 can increase the stability of c-Myc protein [18,19]. These findings suggest that lncRNAs may play important roles in regulating c-Myc function. However, lncRNAs that regulate c-Myc and, in turn, cell cycle progression remain to be explored.
In an attempt to identify oncogenic lncRNAs that regulate G1/S transition, we screened for the lncRNAs that were differentially expressed during cell cycle progression and found that lncRNA SnoRNA Host Gene 17 (SNHG17) was upregulated at the early G1phase and in various cancer types, including hepatocellular carcinoma (HCC). It increased the stability of c-Myc protein and in turn accelerated G1/S transition and cell proliferation via physically interacting with leucine rich pentatricopeptide repeat containing (LRPPRC) protein. These findings reveal the novel roles of SNHG17 and LRPPRC in regulating c-Myc function and highlight the importance of the SNHG17-LRPPRC-c-Myc axis in G1/S transition and tumor growth, which may provide potential targets for cancer therapy.

RESULTS
SNHG17 is upregulated at the early G1-phase and increases the stability of c-Myc protein To identify oncogenic lncRNAs that may regulate G1/S transition, we adopted RNA sequencing technology to screen for molecules that were differentially expressed during cell cycle progression, using a serum starvation-stimulation assay in human skin fibroblasts (SF), a classic model for G1/S transition. The expression pattern of cell cycle-related genes indicated the successful establishment of the studied model ( Fig. 1A-B). As shown, SF cells were induced to enter a quiescent state by depriving serum for 48 h and then stimulated to enter the G1-phase and subsequent S-phase by adding serum (Fig. 1A). Next, we examined the expression profiling of genes at 0, 4, and 18 h after serum addition, representing G0, early, and late G1-phase, respectively. We found that 511 lncRNAs were differentially expressed at 4 or 18 h compared to 0 h after serum stimulation (Fig. 1C). The following criteria were then used to select the candidate lncRNAs  S1B), SNHG17 is the only one that was upregulated in all ten malignancies analyzed based on the cancer genome atlas (TCGA) data ( Fig. 1D; Supplementary Fig. S1C), and a higher SNHG17 level was associated with shorter recurrence-free survival (RFS) and overall survival (OS) of HCC patients ( Fig. 1E; Supplementary Fig.  S1D). Upregulation of SNHG17 was further confirmed in our HCC cohort (Fig. 1F). Moreover, SNHG17 expression significantly increased during the early and middle G1-phases ( Fig. 1A-B). c-Myc is a well-known proto-oncogene and early response gene at the G1-phase [3]. Notably, the level of SNHG17 began to increase right after c-Myc elevation, and lowered after c-Myc decline at the G1-phase (Fig. 1B). We therefore examined whether SNHG17 could affect c-Myc level, and found that the protein level of c-Myc was enhanced by overexpressing SNHG17 but was reduced by silencing SNHG17 ( Fig. 2A-B). Moreover, silencing SNHG17 accelerated the decline of c-Myc protein during cell cycle progression (Fig. 2C). However, neither knockdown nor overexpression of SNHG17 affected the level of c-Myc mRNA ( Supplementary Fig. S3A-B), suggesting that SNHG17 may affect c-Myc level at post-transcriptional level. Subsequent investigations, using cycloheximide to block the de novo protein synthesis, revealed that the half-life of c-Myc protein was significantly decreased by knocking down SNHG17 (Fig. 2D). Furthermore, proteasome inhibitor MG132 abolished the effect of siSNHG17 in reducing c-Myc level (Fig. 2E), and siSNHG17 increased the level of ubiquitylated c-Myc (Fig. 2F), indicating that SNHG17 may increase c-Myc level by preventing ubiquitin-proteasome-dependent degradation of c-Myc protein.

SNHG17 promotes G1/S transition and cell proliferation by increasing c-Myc level
Considering that c-Myc plays a critical role in G1/S transition, we first investigated whether SNHG17 could regulate cell cycle progression. The levels of the key regulators of G1/S transition, including cyclin D, cyclin E, cyclin A, CDK4, CDK6, CDK2, p15, p16, p21, pRb, phosphorylated pRb (ppRb), and E2F1 were examined upon SNHG17 overexpression. As shown, the protein levels of ppRb and the c-Myc downstream target genes (CDK4, CDK2, cyclin A2) were significantly increased in SNHG17-overexpressing cells (Fig. 3A, left panel; Supplementary Fig. S4), whereas the level of ppRb was decreased in siSNHG17-transfectants (Fig. 3A, right panel). Consistent with the observed molecular events, nocodazole-synchronized model revealed that the G1-population was significantly increased by silencing SNHG17 (Fig. 3B; Supplementary Fig. S5A) but was reduced by overexpressing SNHG17 (Fig. 3C; Supplementary Fig. S5B). Serum starvation-stimulation assays further showed that the entry of cells into the S-phase was obviously delayed in SNHG17-silencing transfectants ( Fig. 3D) but was accelerated in SNHG17overexpressing cells (Fig. 3E). Moreover, SNHG17 knockdown decreased the mRNA levels of genes essential for S-phase entry (Fig. 3F). Consistently, the fraction of DNA-replicating cells was reduced by siSNHG17 ( Fig. 3G; Supplementary Fig. S5C) but was enhanced by overexpressing SNHG17 (Fig. 3H; Supplementary Fig.  S5D). These results suggest that SNHG17 may promote G1/S transition and, in turn, cell proliferation.
We further evaluated whether SNHG17 regulated tumor growth. As shown, the cell number was significantly decreased by silencing SNHG17 (Fig. 4A) but was increased by overexpressing SNHG17 (Fig. 4B). Consistently, the siSNHG17-transfected hepatoma cells displayed reduced colony formation than the NCtransfectants ( Fig. 4C; Supplementary Fig. S6A), while the SNHG17overexpressing hepatoma cells formed more colonies ( Fig. 4D; Supplementary Fig. S6B). Mouse xenograft models revealed that tumors overexpressing SNHG17 grew faster than control xenografts ( Fig. 4E), suggesting that SNHG17 may function as an oncogenic lncRNA to promote tumor growth.
Next, we validated whether SNHG17 exerted its function via c-Myc, and found that overexpressing c-Myc attenuated the roles of siSNHG17 in repressing G1/S transition and cell growth (Fig. 4F), whereas inhibiting c-Myc abrogated the stimulatory effects of SNHG17 in G1/S transition and cell growth (Fig. 4G), suggesting that SNHG17 may promote G1/S transition and cell proliferation by stabilizing c-Myc protein.
SNHG17 enhances the c-Myc level and cell proliferation by directly interacting with LRPPRC To explore how SNHG17 upregulated c-Myc, RNA pulldown assays were used to identify the SNHG17-associated proteins. According to the mass spectrum assay, three candidate proteins, including LRPPRC, DHX9, and UPF1, had high scores and appropriate molecular weight (Fig. 5A). Notably, only LRPPRC, a leucine-rich PPR cassette protein with RNA-binding activity, was pulled down by SNHG17 but not its antisense RNA (Fig. 5B). Moreover, RNA immunoprecipitation assays revealed that compared with the IgG-control group, the LRPPRC-precipitated complex contained more SNHG17 but a similar amount of negative control U6 RNA ( LRPPRC has been shown to regulate the expression of genes at the transcriptional and post-transcriptional levels by binding to its target RNAs [20][21][22][23][24]. It is unknown whether LRPPRC regulates the stability of c-Myc protein or G1/S transition. We found that similar to the effects of siSNHG17, siLRPPRC had no impact on the c-Myc mRNA level ( Supplementary Fig. S9D) but observably decreased the level of c-Myc protein (Fig. 5D), and this effect was attenuated by MG132 treatment (Fig. 5E). Furthermore, siLRPPRC increased the polyubiquitination (Fig. 5F) and reduced the stability of c-Myc protein (Fig. 5G), indicating that LRPPRC may enhance c-Myc level via a post-translational regulatory mechanism. Moreover, siLRPPRC inhibited G1/S transition and cell growth, which was abrogated by overexpressing c-Myc (Fig. 5H). These data suggest that LRPPRC may promote G1/S transition and cell proliferation by increasing the stability of c-Myc protein.
Next, we explored whether SNHG17 increased c-Myc level by binding to LRPPRC. As shown, only the full-length LRPPRC and the fragments containing the 1035-1369-aa could be pulled down by SNHG17 (Fig. 6A), suggesting the 1035-1369-aa as an essential domain for LRPPRC to bind SNHG17. Further study revealed that only full-length SNHG17 and the fragments carrying the 1-150-nt region could pull down the GST-LRPPRC-1035-1369-aa fragment (Fig. 6B), indicating 1-150-nt as the core sequence (SNHG17-core) for SNHG17 to bind the 1035-1369-aa domain of LRPPRC. Notably, the roles of SNHG17 in increasing c-Myc level and promoting G1/S transition and cell growth were attenuated by deleting the 1-150nt sequence of SNHG17 (Fig. 6C, D) or by silencing LRPPRC (Fig. 6E, adenocarcinoma. E Kaplan-Meier analysis revealed a significant association between higher SNHG17 level and shorter recurrence-free survival or overall survival of HCC patients. The median SNHG17 level in all 364 HCC tissues, derived from TCGA, was chosen as the cut-off value to separate the high-SNHG17 group (n = 182) from the low-SNHG17 group (n = 182). F The level of SNHG17 was significantly increased in HCC tissues of our study cohort. The expression of SNHG17 was assessed by qPCR in 60 paired HCC (T) and adjacent non-tumor liver tissues (N). For B, error bars represent mean ± SEM from three independent experiments. P values were assessed by paired Student's t test (D, F) or log-rank test (E). *, P < 0.05; **, P < 0.01; ***, P < 0.001.
F; Supplementary Fig. S10A, B). These results suggest that SNHG17 may exert its function via LRPPRC.
We then validated the in vitro findings in human samples. As shown, the levels of SNHG17, LRPPRC, and c-Myc were significantly increased in HCC tissues (Fig. 7A, B), and they showed a positive correlation with each other (Fig. 7C). Similar to the results of SNHG17 (Fig. 1D, E), LRPPRC was upregulated in different types of cancer and high LRPPRC level was associated with worse survival of HCC patients ( Supplementary Fig. S11A, B).
Taken together, SNHG17 may inhibit c-Myc ubiquitination and thus stabilize c-Myc protein by associating with LRPPRC, which leads to upregulation of c-Myc, acceleration of G1/S transition and cell proliferation, and consequent tumor growth (Fig. 7D).

DISCUSSION
Aberrant overexpression of c-Myc oncogene, which is frequently reported in a variety of tumors, contributes to the dysregulation of G1/S transition and is a key event in tumor development [3,25,26]. The roles of lncRNAs in c-Myc regulation remain largely unknown. In this study, we identify an oncogenic lncRNA SNHG17 that is upregulated at the early G1-phase and in various cancer types, including HCC. Moreover, SNHG17 enhances the stability of c-Myc protein by directly binding to LRPPRC and, in turn, promotes G1/S transition and tumor growth.
c-Myc is an early response gene that can be activated by growth signals [27]. In normal cells, c-Myc level is tightly regulated at transcriptional and post-transcriptional levels. c-Myc protein has a . C Silencing SNHG17 accelerated the decline of c-Myc protein during cell cycle progression. NC-or siSNHG17-transfectants were serum-starved and then grown in the 15% FBS-containing medium (S + ) for the indicated hours, followed by Western blotting. SE, short exposure; LE, long exposure. D Silencing SNHG17 decreased the stability of c-Myc protein. NC-or siSNHG17-transfected cells were treated with 25 μg/mL cycloheximide for the indicated time periods before Western blotting. E The siSNHG17-induced c-Myc reduction was attenuated by proteasome inhibitor. NC-or siSNHG17-transfectants without (−) or with (+) 10 μg/mL MG132 treatment for 3 h were subjected to Western blotting. F Silencing SNHG17 increased the level of ubiquitinated c-Myc. NC-or siSNHG17-transfectants without (−) or with (+) 10 μg/mL MG132 treatment for 3 h were subjected to immunoprecipitation (IP) using anti-c-Myc antibody and then Western blotting using anti-Ubiquitin antibody (Ub). iMAX, cells exposed to Lipofectamine RNAiMAX but not RNA duplexes. NC, negative control of RNA duplex. siSNHG17-1 and siSNHG17-2, siRNAs targeting different regions of SNHG17. Different exposure times were used to achieve appropriate signalto-noise for different experiments. For A-B, error bars represent mean ± SEM from three independent experiments. P values were assessed by unpaired Student's t test. **, P < 0.01; ***, P < 0.001.
Here, we showed that SNHG17 enhanced the stability of c-Myc protein based on the following evidence. Firstly, c-Myc protein level was increased by overexpressing SNHG17 but was reduced by silencing SNHG17. However, neither knockdown nor overexpression of SNHG17 affected the level of c-Myc mRNA. Secondly, silencing SNHG17 increased the ubiquitylated c-Myc level and accelerated the decline of c-Myc protein, whereas proteasome inhibitor MG132 abolished the effect of siSNHG17 in decreasing c-Myc protein level. Thirdly, SNHG17 and c-Myc exhibited a similar expression pattern during cell cycle progression, and upregulation of SNHG17 was significantly correlated with elevation of c-Myc protein in HCC tissues. Our findings suggest that lncRNAs represent a new class of c-Myc regulators, and upregulation of SNHG17 in tumors may facilitate tumor growth via c-Myc.
As a snoRNA host gene-derived lncRNA, SNHG17 was upregulated in various cancer types, and its upregulation predicts a poor survival. Genomic amplification was reported as one cause of SNHG17 overexpression in non-small-cell lung cancer (NSCLC), and knockdown of SNHG17 inhibited the proliferation and migration and promoted the apoptosis of NSCLC cells with unknown mechanisms [35]. It was shown that helicobacter pylori infection elevated SNHG17 expression in gastric cancers, resulting in increased genome instability via a SNHG17/miR-3909/RING1/Rad51 pathway [36]. Zhang et al. showed that SNHG17 promoted cell proliferation, migration, invasion, and resistance to apoptosis by polycomb repressive complex 2-mediated epigenetic repression on cyclin-dependent kinase inhibitors, including p15 and p57 [37]. Although SNHG17 has been implicated in tumor development, its role in regulating G1/S transition via c-Myc has not been reported yet. We found that SNHG17 promoted G1/S transition and, in turn, cell proliferation and tumor growth by increasing the level of c-Myc protein. SNHG17 stabilized c-Myc protein via binding to the 1035-1369-aa domain of LRPPRC. We further characterized the 1-150-nt fragment as the essential domain for the function of SNHG17 because deletion of this region abrogated the interaction between SNHG17 and LRPPRC, and attenuated the effects of SNHG17 in enhancing c-Myc stability, G1/S transition, and cell growth.
LRPPRC is an RNA binding protein that regulates gene translation, RNA polyadenylation and transport, and maintains RNA and genome stability [20][21][22][23][24]. Recently, Chen and colleagues reported that lncRNA DANCR directly interacted with LRPPRC and guided LRPPRC protein to bind and stabilize the mRNAs of CCND1, PLAU, and IL-11, resulting in enhanced proliferation, migration, and invasion of bladder cancer cells [24]. These findings suggest that different lncRNAs may interact with LRPPRC protein to increase the mRNA or protein stability of different genes, which highlight the importance of LRPPRC in the regulation of gene expression.
The role of LRPPRC in tumor development remains inconclusive. It has been reported that LRPPRC protein levels were significantly reduced in diethylnitrosamine (DEN)-induced mouse HCC tissues, and liver-specific deletion of LRPPRC increased incidence of DENinduced HCC [23]. However, LRPPRC is upregulated in human lung adenocarcinoma [38], and knockdown of LRPPRC inhibits the growth of bladder and lung cancer cells in vitro and in vivo [24,38]. Consistently, we found significant upregulation of LRPPRC in human HCC samples. Analysis using the cancer genome atlas (TCGA) data showed that the mRNA levels of LRPPRC were significantly increased in 8 out of 10 malignancies, including HCC. And high LRPPRC expression was correlated with worse survival of HCC patients. It has been shown that the gene mutation profiles are quite different between DEN-induced HCC and human HCC [39]. The discrepancy between studies in human and mice might be attributed to the species specificity and the different pathological processes in human tumors and chemical-induced tumors in mice.
It has been reported that LRPPRC can suppress autophagy [40,41] and c-Myc protein can be degraded via autophagy pathway [42], which raises the possibility that LRPPRC may regulate the autophagic degradation of c-Myc protein. However, we found that autophagy inhibitor chloroquine could not abolish the effect of siLRPPRC in decreasing c-Myc protein level ( Supplementary Fig. S12). On the other hand, silencing LRPPRC increased the ubiquitylated c-Myc level and reduced the stability of c-Myc protein, suggesting that LRPPRC may regulate c-Myc stabilization through ubiquitin-proteasome system. However, we did not detect an association between LRPPRC and the mRNA or protein of c-Myc, suggesting that LRPPRC may act in an indirect manner, for example, through some E3 ligases to control the ubiquitination of c-Myc. Our study discloses the novel functions of LRPPRC and explores new mechanisms responsible for the tumorpromoting effect of LRPPRC.
In conclusion, we identify a SNHG17/LRPPRC/c-Myc regulatory axis and characterize its function in G1/S transition, which uncovers new mechanisms underlying cell proliferation and tumor growth, and provides potential targets for anti-cancer therapy.

Human tissues
Human HCC tissues and adjacent noncancerous liver tissues were collected from patients who undertook tumor resection at the Cancer Center, Sun Yat-sen University, P.R. China. All tissues were histologically examined. No local or systemic therapy had been conducted before surgery. After surgical resection, no other anti-cancer therapy was carried out before recurrence. Informed consent was obtained from each patient, and the protocol was approved by the Institutional Research Ethics Committee.

Cell lines and cell culture
Transformed human embryonic kidney cell line, HEK293T (ATCC, CRL-3216), and three human hepatoma cell lines, HepG2 (ATCC, HB-8065), SNU-449 (ATCC, CRL-2234), and HCCLM9 [43,44] were cultured in Dulbecco's modified Eagle's medium (DMEM) (Corning, Manassas, VA, USA) containing 10% fetal bovine serum (FBS) (Gibco, Thermo Fisher Scientific, Waltham, Fig. 3 SNHG17 promotes G1/S transition and cell proliferation. A The level of phosphorylated-pRb was increased by overexpressing SNHG17 (left panel), but was reduced by silencing SNHG17 (right panel). Cells stably expressing SNHG17 and the control cells (Ctrl) (left panel), or the siRNA-transfected cells (right panel) were subjected to Western blotting. Different exposure times were used to achieve appropriate signal-to-noise for overexpression and knockdown experiments. ppRb, S780-phosphorylated pRb; pRb, total pRb protein. B-C The fraction of cells at the G1-phase was increased by silencing SNHG17 but was reduced by overexpressing SNHG17. NC-or siSNHG17-transfectants (B) and cells stably expressing SNHG17 and the control cells (Ctrl) (C) were synchronized with nocodazole before cell cycle analysis. D-E The serumstimulated S-phase entry was inhibited by silencing SNHG17 but was promoted by overexpressing SNHG17. NC-or siSNHG17-transfectants (D) and cells stably expressing SNHG17 and the control cells (Ctrl) (E) were serum-deprived and then grown in the 15% FBS-containing medium for the indicated time periods before cell cycle analysis. The percentage of cells at the G1-phase is indicated for each sample. F Silencing SNHG17 inhibited the expression of S-phase genes. NC-or siSNHG17-transfectants were serum-deprived (S -) and then cultured in the 15% FBS-containing medium (S + ) for the indicated time periods before qPCR. G-H The fraction of DNA-replicating cells was decreased by silencing SNHG17 but was increased by overexpressing SNHG17. Cells undergoing DNA replication were detected using EdU incorporation assays. For B-C and F-H, error bars represent mean ± SEM from three independent experiments. P values were assessed by unpaired Student's t test. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, not significant. Massachusetts, USA), 100 mg/mL streptomycin and 100 U/mL penicillin in a humidified atmosphere with 5% CO 2 at 37°C. Human primary SF cells were isolated from human neonatal foreskin as described previously [45] and cultured in RPMI 1604 medium (Corning) supplemented with 10% FBS (Gibco), 100 mg/mL streptomycin, and 100 U/mL penicillin in a humidified atmosphere with 5% CO 2 at 37°C. The cell lines with stable expression of SNHG17 (SF-SNHG17, HepG2-SNHG17, SNU-449-SNHG17, HCCLM9-SNHG17), SNHG17-Δcore (HepG2-SNHG17-Δcore, SNU-449-SNHG17-Δcore), c-Myc (HepG2-c-Myc, SNU-449-c-Myc), and their matched control lines (SF-Ctrl, HepG2-Ctrl, SNU-449-Ctrl, HCCLM9-Ctrl) were established by infecting cells with lentivirus that expressed the target sequence, followed by selection under puromycin (1 μg/mL) for 6 days before used. All these sublines expressed copGFP.

RNA oligoribonucleotides
The small interfering RNAs (siRNAs) targeting the transcripts of human SNHG17 (GeneBank accession No. NR_152762.1), c-Myc (NM_002467.6) and LRPPRC (NM_133259.4) were designated as siSNHG17, sic-Myc, and siLRPPRC, respectively, and were purchased from Ribobio (Guangzhou, China). The negative control (NC) RNA duplex for siRNA was nonhomologous to any human genome sequences. A Identification of SNHG17-associated proteins by RNA pulldown and mass spectrometry assays. Biotin-labeled SNHG17 or its antisense RNA (SNHG17-AS, negative control) was incubated with lysates from HepG2 cells to pull down the SNHG17-associated proteins, followed by silver staining. A specific fragment (indicated by an arrow), pulled down by SNHG17 but not SNHG17-AS (left panel), was subjected to mass spectrometry. The candidate proteins are listed (right panel). B LRPPRC but not DHX9 or UPF1 was enriched in the protein complexes pulled down by SNHG17. Cellular proteins were pulled down with biotin-labeled SNHG17 or SNHG17-AS and then subjected to Western blotting. GAPDH, negative control. C RNA immunoprecipitation assays revealed an interaction between SNHG17 and LRPPRC in vivo. The RNAs that were precipitated by anti-LRPPRC antibody or isotype-matched control IgG were analyzed by qPCR, using primers for SNHG17 or U6. The enrichment was normalized to the IgG control. U6 was used as a negative control. D Silencing LRPPRC decreased the protein level of c-Myc. E The siLRPPRC-induced c-Myc reduction was abrogated by proteasome inhibitor. NC-or siLRPPRC-transfectants without (−) or with (+) 10 μg/mL MG132 treatment for 3 h were subjected to Western blotting. SE, short exposure; LE, long exposure. F Silencing LRPPRC increased the level of ubiquitinated c-Myc. NC-or siLRPPRC-transfectants without (−) or with (+) 10 μg/mL MG132 treatment for 3 h were subjected to immunoprecipitation (IP) using anti-c-Myc antibody and then Western blotting using anti-Ubiquitin antibody (Ub). G Silencing LRPPRC decreased the stability of c-Myc protein. NC-or siLRPPRC-transfected cells were treated with 25 μg/mL cycloheximide for the indicated time periods before Western blotting. H Ectopic expression of c-Myc antagonized the roles of siLRPPRC in decreasing DNA-replicating cells (left panel) and cell growth (right panel). Cells stably expressing c-Myc and the control cells (Ctrl) were transfected with NC or siLRPPRC and then subjected to EdU incorporation assays (left panel) and cell counting (right panel). For C and H, error bars represent mean ± SEM from three independent experiments. P values were assessed by unpaired Student's t test. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, not significant.

Rapid amplification of cDNA ends (RACE)
Total RNA of SF cells was used to amplify 5′-and 3′-ends of SNHG17. The 5′end was determined using a 5′-Full RACE Kit (D315, TaKaRa, Kyoto, Japan). 3′ RACE was performed as described previously [45]. The PCR-amplified sequences of 5′-and 3′-end fragments were determined by direct sequencing.

RNA sequencing and analysis
Total RNA was isolated from SF cells using Trizol reagent (Invitrogen). The quality of extracted RNA was assessed using the NanoDrop ND-2000 spectrophotometer. Directional RNA-seq libraries were constructed and then sequenced at the Novogene Bioinformatics Institute (Beijing, China) on an Illumina HiSeq-PE150 Platform. The obtained reads were then mapped to the human genome (hg38) database using STAR v.2.6.0, and fragments per kilobase of exon per million mapped reads (FPKM) were calculated using RSEM. Gene expression levels and differential expression were quantified with Cuffdiff. The RNA-seq data have been deposited in the Gene Expression Omnibus (GEO) database under accession number GSE166448.

Analysis of gene expression
The RNA and protein levels of gene were analyzed by real-time quantitative polymerase chain reaction (qPCR) and Western blotting, respectively.
The sequences of all oligonucleotides used for siRNA, qPCR, RACE, and vector construction are listed in Table S1.

Cell transfection and lentivirus production
Reverse transfection of RNA oligoribonucleotides was performed with Lipofectamine RNAiMAX (Invitrogen). A final concentration of 10 nM duplex was used. Plasmid DNAs were transfected using Lipofectamine 3000 (Invitrogen).
To produce lentivirus, HEK293T cells were co-transfected with lentivirus expression vector containing target sequence and packing plasmids (pMD2.G and psPAX2, Addgene MA, USA) by calcium phosphate precipitation, then refreshed with culture medium 16 h after transfection. Thirty-six hours later, the lentivirus supernatant was harvested and stored in aliquots at −80°C until use. Target cells were infected with lentiviral supernatant supplemented with 8 μg/mL polybrene (Millipore, Billerica, MA).
For the nocodazole-synchronized model, cells were treated with 50 ng/ mL nocodazole before FACS analysis. For the serum starvation-stimulation model, cells were serum-deprived for 48 h, followed by serum re-addition, and then harvested at the indicated time points.

The 5-ethynyl-2′-deoxyuridine (EdU) incorporation assay
EdU incorporation assay was conducted using Cell-Light EdU Apollo ® 567 In Vitro Kit (C10310-1, Ribobio). Briefly, 12 h after last transfection or seeding, cells were serum-deprived for 48 h and then cultured in 15% FBScontaining medium for the indicated time periods, followed by incubation in complete medium containing 50 μM EdU for additional 2 h. Cells were then fixed with 4% paraformaldehyde, and sequentially incubated with 2 mg/mL glycine for 5 min, 1 × PBS containing 0.5% Triton X-100 for 10 min, and click reaction buffer (containing Apollo ® 567) for 30 min. Nuclei were counterstained with Hoechst 33342. The fraction of DNA-replicating cells, which represented for cell proliferation rate, was expressed as the ratio of EdU-positive cells to the number of Hoechst 33342-positive cells. At least 500 cells were counted for each sample.

Colony formation assay
For the loss-of-function assay, NC-or siSNHG17-transfected cells (1000 cells for HepG2 and 300 cells for SNU-449) were placed in a 6-well plate and maintained in complete medium for 14 (HepG2) or 10 (SNU-449) days. For the gain-of-function assay, HepG2 (500 cells) and SNU-449 (150 cells) sublines stably expressing SNHG17 and their control lines were seeded in a 6-well plate for 14 (HepG2) or 12 (SNU-449) days. The colonies were fixed with methanol and then stained with 0.1% crystal violet in 20% methanol for 15 min before analysis.

Mouse xenograft models
Male NOD-Prkdcem26Cd52Il2rgem26Cd22/Nju (NCG) mice at 4 weeks of age were used. HCCLM9-SNHG17 and its control line HCCLM9-Ctrl (6.0 × 10 6 ) resuspended in 100 μL of DMEM/matrigel (1:1 volume; R&D Systems, Minneapolis, MN, USA) were subcutaneously injected into the right and left side of the posterior flank, respectively. Mice were sacrificed 20 days after implantation. Tumor growth was measured every three days, and the volume of the tumor was detected with electronic digital calipers and calculated with the formula: volume = (length × width 2 )/2. At the end of experiments, tumors were dissected, photographed, and weighed.
All mouse experiments were approved by the Institutional Animal Care and Use Committee at Sun Yat-sen University. All experiments were performed according to the Guide for the Care and Use of Laboratory Animals (NIH publications Nos. 80-23, revised 1996) and following the institutional ethical guidelines for animal experiments.

Isolation of subcellular fraction
The cytoplasmic and nuclear extracts were isolated using NE-PER Nuclear and Cytoplasmic Extraction Reagents (Pierce, Rockford, IL, USA), followed by isolation of RNAs or proteins.

RNA pulldown assay
RNA pulldown assay was performed as previously reported [45]. In brief, full-length or truncated mutants of human SNHG17 or antisense of SNHG17 were in vitro transcribed and biotin-labeled using Biotin RNA labeling mix (2147483647, Roche, Mannheim, Germany), then incubated with purified GST-fusion proteins or with HepG2 cell lysates that were precleared by yeast tRNA at RT for 1 h. After incubation with pre-washed streptavidin MagneSphere ® paramagnetic particles (Promega, Madison, USA) for 30 min at RT, the biotin-labeled RNA-protein complexes with paramagnetic particles were collected by a magnetic holder and washed with IP-lysis buffer (25 mM Tris-HCl at pH 7.4, 150 mM NaCl, 1 mM EDTA at pH 8.0, 1% NP-40, 5% glycerol and 100 U/mL RNase inhibitor) for 5 times, then digested with RNase A (Thermo Scientific, Waltham, MA, USA) in 10 mM Tris-HCl (pH 8.0) at 37°C for 15 min to release the RNA binding proteins. The retrieved proteins were separated by SDS-PAGE gel, followed by silver staining, and analyzed by mass spectrometry (Beijing Protein Innovation, Beijing, China) or Western blotting.

Immunoprecipitation (IP) assay
IP assay was carried out as previously [47] reported with modifications. Briefly, HepG2 cell lysates were pre-cleared with yeast tRNA and then incubated with the indicated antibody or isotype-matched IgG at 4°C for 4 h. Then the immunoprecipitated complexes were collected by incubation with protein A/G magnetic beads (B23201; Bimake, Houston, TX, USA) at 4°C for 2 h. After washing with IP lysis buffer for 5 times, the immunoprecipitated complexes were eluted with 1 × SDS lysis buffer and then analyzed by Western blotting.