6/STAT3 Signaling Pathway - Wiley Online Library

A MERICAN A SSOCIATION FOR T HE STUDY OF LIVER D I S E ASES

HEPATOLOGY, VOL. 67, NO. 1, 2018

Long Noncoding RNA TSLNC8 Is a Tumor Suppressor That Inactivates the Interleukin-6/STAT3 Signaling Pathway Jiwei Zhang,1* Zhe Li,1* Longzi Liu,2* Qifeng Wang,1 Shengli Li,1 Di Chen,1 Zhixiang Hu,1 Tao Yu,3 Jie Ding,1 Jinjun Li,3 Ming Yao,3 Shenglin Huang,1 Yingjun Zhao,1 and Xianghuo He

1,4

Long noncoding RNAs can serve as oncogenes or tumor suppressors in human cancer; however, their biological functions and underlying mechanism in hepatocarcinogenesis are largely unknown. Here, we report a novel tumor suppressor long noncoding RNA on chromosome 8p12 (termed TSLNC8) that is frequently deleted and down-regulated in hepatocellular carcinoma (HCC) tissues. The loss of TSLNC8 is highly associated with the malignant features of HCC and serves as a prognostic indicator for HCC patients. TSLNC8 significantly suppresses the proliferation and metastasis of HCC cells in vitro and in vivo. TSLNC8 exerts its tumor suppressive activity by competitively interacting with transketolase and signal transducer and activator of transcription 3 (STAT3) and modulating the STAT3-Tyr705 and STAT3Ser727 phosphorylation levels and STAT3 transcriptional activity, thus resulting in inactivation of the interleukin-6– STAT3 signaling pathway in HCC cells. Conclusion: TSLNC8 is a promising prognostic predictor for patients with HCC, and the TSLNC8–transketolase–STAT3 axis is a potential therapeutic target for HCC treatment. (HEPATOLOGY 2018;67:171-187).

L

ong noncoding RNAs (lncRNAs) are more than 200 bp in length with no or weak protein coding abilities.(1) Studies have reported that lncRNAs regulate various biological processes, including cell differentiation, immune response, apoptosis, and development.(2) The molecular functions of lncRNAs include preventing RNA and protein from binding to intended targets, acting as host genes for microRNAs, and serving as molecular scaffolds to guide proteins to their direct chromosomal targets.(3) Increasing evidence indicates that lncRNAs play essential roles in cancer development and progression.(4) Hepatocellular carcinoma (HCC) is one of the most common and malignant tumors in adults, accounting for

one-fifth of the incidence of malignant tumors in China.(5) HCC is also the most common cancer in Chinese men under the age of 60 years, and Chinese HCC mortality rates are the highest worldwide.(6) Despite continuous improvements in medical technology over the past decades, such as surgical resection, liver transplantation, radiation therapy, and chemotherapy, the 5-year survival rate of HCC is 10% of the total samples (Supporting Table S1). We analyzed the RNA-sequencing data of 50 paired HCC tissues and adjacent nontumor tissues of The Cancer Genome Atlas. There are four lncRNAs—TSLNC8, RP11-148O21.2, AC133633.2, and CTD-3247F14.2—which were significantly down-regulated in HCC tissues, and subsequent polit experiments proved that only TSLNC8 was significantly deleted and down-regulated in HCC (Supporting Table S2), suggesting that TSLNC8 is the important tumor suppressor in chromosome 8p. The TSLNC8 gene has two annotated transcripts in the National Center for Biotechnology Information database (https://www.ncbi.nlm.nih.gov/; Supporting Fig. S1A), with limited protein-coding potential (https://www.ncbi.nlm.nih.gov/orffinder/, the open reading frame finder; Supporting Fig. S1B). The dominant TSLNC8 transcript of 1,404 bp was identified by the 50 and 30 rapid amplification of complementary DNA ends assays and northern blot assays in Huh-7 and SMMC-7721 cells (Supporting Fig. S1C,D). TSLNC8 was expressed in HCC cell lines to varying

ZHANG, LI, ET AL.

degrees (Supporting Fig. S1E), with about 75% distribution in the nucleus (Supporting Fig. S1F). Next, we confirmed that TSLNC8 was significantly deleted in HCC tissues compared with their paired nontumor tissues (Fig. 1A), and the TSLNC8 locus was lost in 68.0% (49/72) of the HCC tissues (Fig. 1B). Moreover, the expression of TSLNC8 RNA was significantly down-regulated in 65% (78/120) of the 120 HCC cases (Fig. 1C,D). Importantly, the TSLNC8 RNA level was correlated with several clinicopathological features of HCC patients (Supporting Table S3) and inversely correlated with the numbers of tumor nodules or the existence of cancer embolus, and lower TSLNC8 RNA levels exhibited poorer differentiation stage (Fig. 1E). Furthermore, patients with lower TSLNC8 RNA levels exhibited poor overall survival (Fig. 1F). These results suggested that loss of TSLNC8 at 8p12 was a critical event in hepatic carcinogenesis and could be used as a prognostic indicator for HCC patients in the clinic.

TSLNC8 SUPPRESSES HCC CELL PROLIFERATION AND TUMORIGENICITY IN VITRO AND IN VIVO To further determine the effects of TSLNC8 on HCC cell proliferation and tumorigenicity, stable cell lines (pWPXL-TSLNC8) were established through a lentiviral infection in HCC cells (Supporting Fig. S2A). TSLNC8 overexpression significantly inhibited the colony formation ability of HCC cells (Fig. 2A; Supporting Fig. S2B); in addition, the silencing of TSLNC8 by specific small interfering RNAs (siRNAs) significantly increased HCC cell colony formation (Fig. 2B; Supporting Fig. S2C,D). Consistently, cell proliferation assays showed that TSLNC8 overexpression decreased the proliferation rates of HCC cells (Fig. 2C), whereas the TSLNC8 siRNAs significantly accelerated HCC cell proliferation (Fig. 2D). To examine the effect of TSLNC8 on the tumorigenicity of HCC cells in vivo, pWPXL-TSLNC8 cells and control cells, derived from SMMC-7721 cells, were subcutaneously injected into nude mice. With no obvious effects on mouse weight (Supporting Fig. S2E), the volume and the weight of the tumors were dramatically lower in the pWPXL-TSLNC8 group than in the control group (Fig. 2E; Supporting Fig. S2F), indicating that TSLNC8 overexpression remarkably inhibited the tumorigenic ability of HCC cells in vivo.

173

ZHANG, LI, ET AL.

HEPATOLOGY, January 2018

FIG. 1. TSLNC8 is decreased in HCC and clinically relevant to patient prognosis. (A) DNA copy number analysis of TSLNC8 in 72 matched HCC/nontumor tissues. Log2 ratio dot plots were generated from data obtained from genomic quantitative real-time PCR analyses. The y axis shows the log2 ratio of paired tumor/nontumor DNA levels, and the x axis represents the scatter distribution of the 72 samples. (B) The pie chart shows the proportions of samples in the copy number loss (green), gain (red), and normal (yellow) categories. (C) The RNA levels of TSLNC8 in 120 matched HCC/nontumor tissues. Log2 ratio dot plots were generated from data obtained from quantitative real-time PCR analyses. The y axis shows the log2 ratio of paired tumor/nontumor RNA levels, and the x axis represents the scatter distribution of the 120 samples. (D) The pie chart shows the proportions of samples in the down-regulation (green), up-regulation (red), and no change (yellow) categories. (E) Relative TSLNC8 RNA expression in 120 HCC tumor samples, with tumor numbers (1 or >1), with or without cancer embolus, with tumor–node–metastasis stage (I-II or III-IV). (F) Kaplan–Meier plots of HCC patients stratified by TSLNC8 RNA levels (log-rank test). Values are expressed as median with interquartile range. Statistical analysis was performed using the Student t test in (A,C) and the Wilcoxon test (D,E).

TSLNC8 INHIBITS HCC CELL INVASION AND METASTASIS IN VITRO AND IN VIVO The inverse relationship between TSLNC8 RNA level and the existence of cancer embolus prompted us to explore whether TSLNC8 could affect HCC cell invasion and metastasis. Transwell assays with or without Matrigel showed that TSLNC8 overexpression repressed HCC cell migration and invasion (Fig. 3A; Supporting Fig. S3A). In contrast, TSLNC8 knockdown by siRNA significantly enhanced the migration

174

and invasion of HCC cells (Fig. 3B; Supporting Fig. S3B,C). Moreover, TSLNC8 had no impact on HCC cell proliferation in the same situation of Transwell assays (Supporting Fig. S3D,E), suggesting that the inhibition of migration and invasion by TSLNC8 was not due to the suppressed cell proliferation. Next, pWPXL-TSLNC8 cells and control cells, derived from SMMC-7721 cells, were transplanted into the livers of nude mice. After the transplanted cells grew in situ for 8 weeks, the mice were sacrificed and the nodules examined. TSLNC8 remarkably inhibited the metastatic ability of SMMC-7721 cells in vivo

HEPATOLOGY, Vol. 67, No. 1, 2018

ZHANG, LI, ET AL.

FIG. 2. TSLNC8 inhibits HCC cell proliferation in vitro and in vivo. (A) Colony formation assays for SMMC-7721 and SNU-449 cells infected with the lentivirus expressing TSLNC8 or the control. (B) Colony formation assay for Huh-7 and SNU-449 cells transfected with the TSLNC8 siRNAs or the mock controls. (C) Cell proliferation assays for SMMC-7721 and SNU-449 cells infected with the lentivirus expressing TSLNC8 or the control, using Cell Counting Kit-8 assay. (D) Cell proliferation assays for Huh-7 and SNU-449 cells transfected with TSLNC8 siRNAs or mock controls. (E) The tumor growth curve and the tumor weight of SMMC7721 cells infected with the lentivirus expressing TSLNC8 or the control were subcutaneously injected into nude mice. Values are expressed as mean 6 SEM, n 5 3 (A-D) or n 5 12 (E). **P < 0.01 and ***P < 0.001. Abbreviations: NC, normal control; OD, optical density.

(Supporting Fig. S3F): 90% (9/10) of control mice had intrahepatic metastatic nodules, whereas only 30% of mice in the pWPXL-TSLNC8 group did. Additionally, 70% (7/10) of control mice had distal lung metastases, whereas 30% of mice in the pWPXL-TSLNC8 group did. Regarding distal intestine metastasis, 80% (8/10) of control mice and only 20% (2/10) of mice in the pWPXL-TSLNC8 group had metastatic nodules

(Fig. 3C). Moreover, the numbers of metastatic nodules in the liver, lungs, and intestinal region were significantly decreased in the pWPXL-TSLNC8 group (Fig. 3D-F); however, no obvious effect on mouse weight was observed (Supporting Fig. S3G). Taken together, these findings demonstrated that TSLNC8 inhibited HCC cell invasion and metastasis in vitro and in vivo.

175

ZHANG, LI, ET AL.


FIG. 3. TSLNC8 inhibits HCC cell invasion and metastasis in vitro and in vivo. (A) Transwell migration and invasion assays for SMCC-7721 and SNU-449 cells infected with the lentivirus expressing TSLNC8 or the control, respectively. (B) Transwell migration and invasion assays for Huh-7 and SNU-449 cells transfected with the TSLNC8 siRNAs or mock controls, respectively. (C) In vivo metastatic events in SMCC-7721 cells infected with the lentivirus expressing TSLNC8 or the control. The cells were orthotopically injected into the livers of nude mice. Sixty days later, the mice were sacrificed, and the livers, lungs, and intestines were subjected to immunohistochemical staining. The P values were determined using the v2 test. (D-F) Hematoxylin and eosin staining of sections with metastatic nodules in the liver, lung, and intestine. Metastatic nodules were counted and analyzed using the Student t test. Values are expressed as mean 6 SEM, n 5 3 (A, B) or n 5 10 (C-F). Abbreviation: NC, normal control.

176


TSLNC8 MODULATES MULTIPLE SIGNALING PATHWAYS INVOLVED IN HCC CELL PROLIFERATION AND METASTASIS To obtain better understanding of the inhibitory roles of TSLNC8 on HCC cell proliferation and metastasis, RNA-sequencing analysis was performed to analyze the gene expression profile affected by TSLNC8 knockdown. Hierarchical clustering revealed a total of 377 up-regulated genes and 737 downregulated genes in SMMC-7721 cells transfected with two independent TSLNC8 siRNAs (Fig. 4A; Supporting Table S4). Gene set enrichment analysis showed that TSLNC8 knockdown primarily affected the signaling pathway gene sets, including the insulin signaling pathway, the chemokine signaling pathway, and the Janus kinase (JAK)–/STAT signaling pathway, resulting in remarkable changes in the biological processes and cellular components of the cell cycle, axon guidance, focal adhesion, and cytokine–cytokine interactions (Fig. 4B; Supporting Fig. S4). Functional annotation revealed the genes in the top-scoring processes, such as the JAK–STAT3 pathway, cell cycle process, and focal adhesion process (Fig. 4C). And the mRNA and protein levels of selected genes related to tumor growth and metastasis were indeed regulated by TSLNC8 in the HCC cells (Fig. 4D,E), suggesting that TSLNC8 might exert its biological activities on HCC cell growth, invasion, and metastasis through modulating these signaling pathways in HCC cells.

TSLNC8 PHYSICALLY INTERACTS WITH STAT3 AND INHIBITS ITS TRANSCRIPTIONAL ACTIVITY To explore the molecular mechanism by which TSLNC8 exerted its tumor suppressive function on HCC cells, we performed biotin-labeled RNA pulldown accompanied by mass spectrometric assays to identify TSLNC8-interacting proteins in HCC cells. The results in three independent experiments repeatedly showed a specific protein band at approximately 90 kDa in the TSLNC8 pull-down samples, and 27 potential interacting proteins were obtained based on a confidence score >100 in mass spectrometric assays (Supporting Fig. S5A and Table S5). Notably, STAT3, the primary signaling molecule in the JAK– STAT3 pathway, was confirmed as a specific binding

ZHANG, LI, ET AL.

protein for TSLNC8 (Supporting Fig. S5B). This interaction was further confirmed through RNA immunoprecipitation (RIP) assays (Supporting Fig. S5C). Moreover, a series of deletion mapping analyses revealed that the truncated TSLNC8 fragment 480831 nucleotides was responsible for the interaction of TSLNC8 with STAT3 (Fig. 5A), which is the right arm of TSLNC8 (http://www.lncipedia.org/; Supporting Fig. S5D). Next, RIP and pull-down assays revealed that the DNA-binding domain of STAT3 was required for the interaction (Fig. 5B; Supporting Fig. S5E). Moreover, the activity of the luciferase reporter with the STAT3 response element was diminished in TSLNC8-overexpressing HCC cells, whereas the activity of the reporter was augmented after TSLNC8 knockdown (Fig. 5C). The mRNA levels of the downstream genes in the JAK–STAT3 pathway were also affected by TSLNC8 knockdown or overexpression (Fig. 5D). Furthermore, the TSLNC8 fragments with STAT3-binding ability exhibited significant inhibitory effects on HCC cell proliferation and migration (Fig. 5E; Supporting Fig. S5F). Notably, there was a predicted STAT3 recognized site at the 748-758 nucleotides of TSLNC8 (http://jaspar. genereg.net/; core sequence, 50 -GTGCAAGGAAA30 ), just in the region of TSLNC8 for the interaction with STAT3. A mutant TSLNC8, nominated as TSLNC8-M, was generated with the 748-758 nucleotide sequence substituted by an unrelated sequence (50 GAATTGGATCC-30 ). Functional assays showed that TSLNC8 could significantly suppress the proliferation and migration of SMMC-7721 cells, whereas TSLNC8-M did not have a similar function (Fig. 5F; Supporting Fig. S5G), indicating that the binding of STAT3 with the recognized site at the 748-758 nucleotides is important for the biological function of TSLNC8. Taken together, these findings indicate that TSLNC8 can physically interact with STAT3 and inhibit its transcriptional activity, leading to the suppression of HCC cell proliferation and migration.

TSLNC8 REGULATES STAT3Tyr705 AND STAT3-Ser727 PHOSPHORYLATION LEVELS IN HCC CELLS STAT3 requires phosphorylated modification for its homodimerization, nuclear translocation, and transcriptional activity.(21) Thus, we examined whether TSLNC8 had an effect on STAT3 phosphorylation in HCC cells. TSLNC8 knockdown significantly

177

ZHANG, LI, ET AL.


FIG. 4

178


enhanced STAT3-Tyr705 (Y705) phosphorylation but suppressed STAT3-Ser727 (S727) phosphorylation and total STAT3 protein level, whereas TSLNC8 overexpression inhibited STAT3-Y705 phosphorylation but increased STAT-S727 phosphorylation and total STAT3 protein level (Fig. 6A). Moreover, TSLNC8 knockdown in Huh7 cells significantly promoted the nuclear distribution of phosphorylated STAT3-Y705 but reduced the amount of phosphorylated STAT3-S727 and total STAT3 in the nuclear components (Fig. 6B, left). On the other hand, TSLNC8 overexpression in SMMC-7721 cells caused an obvious reduction of phosphorylated STAT3-Y705 in the nuclear components, accompanied by the nuclear accumulation of phosphorylated STAT3-S727 and total STAT3 (Fig. 6B, right). These results suggested that TSLNC8 could affect the phosphorylation status and the nuclear distribution of STAT3, thus regulating the biological functions of this protein. Moreover, the truncated TSLNC8 fragments with STAT3-binding potency could also regulate the phosphorylation status of STAT3 in HCC cells (Fig. 6C). Interestingly, the effects of TSLNC8 on the phosphorylation of STAT3-Y705, STAT3-S727, and total STAT3 protein level are diverse in HCC cells. Although the oncogenic effects of STAT3-Y705 phosphorylation have been confirmed in various cancer types,(22) the role of STAT3-S727 phosphorylation is not well defined (i.e., S727 phosphorylation may either enhance or inhibit STAT3 transcriptional activity).(23) To determine the function of STAT3 phosphorylation, an Asp (D) was introduced to mimic the constitutively active state of the phosphorylated site. The Y705D mutant significantly enhanced HCC cell proliferation and migration, whereas the S727D mutant remarkably inhibited the proliferation and migration of HCC cells (Supporting Fig. S6A,B). Furthermore, introduction of TSLNC8 significantly inhibited the proliferation and migration of HCC cells, and importantly, reintroduction of STAT3-Y705D rescued the TSLNC8-induced decrease in HCC cell proliferation

ZHANG, LI, ET AL.

and migration (Fig. 6D; Supporting Fig. S6C). On the other hand, inhibition of endogenous TSLNC8 by specific siRNAs promoted the proliferation and migration of HCC cells, while restoration of STAT3S727D remarkably abolished the promoting effects of TSLNC8 knockdown on HCC cell proliferation and migration (Fig. 6E; Supporting Fig. S6D). Collectively, these findings suggested that the decrease in Y705 phosphorylation and the increase in S727 phosphorylation may contribute to the tumor suppressive effects of TSLNC8 on HCC cell proliferation, invasion, and metastasis.

TSLNC8 REGULATES STAT3 PHOSPHORYLATION BY INTERACTING WITH TKT IN HCC CELLS Next, we explored the underlying mechanism through which TSLNC8 inhibited STAT3-Y705 phosphorylation and enhanced STAT3-S727 phosphorylation in HCC cells. A previous report showed that TKT, a key enzyme in the pentose phosphate pathway, might regulate STAT3-Y705 and STAT3S727 phosphorylation in prostate cancer cells.(24) Notably, our mass spectrometric analysis identified TKT as a candidate interacting protein with TSLNC8, suggesting that TSLNC8 may regulate STAT3 phosphorylation by interacting with TKT in HCC cells. To test this hypothesis, we first examined whether TKT could regulate STAT3 phosphorylation and activation in HCC cells. As expected, enhanced Y705 phosphorylation and inhibited S727 phosphorylation were observed in TKT-overexpressing cells, whereas an opposing status was detected in TKTknockdown cells (Fig. 7A). Importantly, introduction of TKT rescued the TSLNC8-induced decrease of Y705 phosphorylation, whereas the enhancement of Y705 phosphorylation by the TSLNC8 siRNA was suppressed by TKT knockdown (Fig. 7B). Additionally, TKT overexpression

FIG. 4. TSLNC8 causes deregulation of the genes involved in cell proliferation and metastasis in HCC cells. (A) Gene expression profiles from SMMC-7721 cells transfected with the TSLNC8 siRNAs or the mock control. Genes are shaded with green, black, or red in the heat map to indicate low, intermediate, or high expression, respectively. (B) Functional annotation clustering of genes regulated by TSLNC8 in SMMC-7721 cells is shown. Significantly enriched groups nominated by the gene ontology term are ranked based on the group enrichment scores. Red indicates signaling pathway; green, biological process; and blue, cellular component. (C) Gene expression levels of the subsets of genes involved in the JAK–STAT pathway, cell cycle process, and focal adhesion components. Red indicates high expression; black, intermediate expression; and green, low expression. (D) Quantitative real-time PCR for selected genes from the ranked pathways in SMMC-7721 cells transfected with TSLNC8 siRNAs. Values are expressed as mean 6 SEM, n 5 3. (E) Western blotting analysis for selected genes from the ranked pathways in SMMC-7721 cells transfected with TSLNC8 siRNAs. b-actin served as the control. Abbreviations: FDR, false discovery rate; NC, normal control; NES, normalized enrichment score.

179

ZHANG, LI, ET AL.


FIG. 5. TSLNC8 interacts with STAT3 and inhibits its transcriptional activity. (A) Deletion mapping of the STAT3-binding domain in TSLNC8. Above, the in vitro transcribed full-length TSLNC8 and deletion fragments showing the correct sizes. Below, immunoblotting analysis of STAT3 in protein samples pulled down by the different TSLNC8 constructs. (B) Deletion mapping of the TSLNC8-binding domain in STAT3. Left, diagrams of full-length STAT3 and the domain-truncated fragments. Right, quantitative real-time PCR detection of TSLNC8 retrieved by full-length or domain-truncated STAT3-Flag using a Flag antibody. RIP assays were performed using SMMC-7721 cells transfected with the indicated vectors. (C) Luciferase assays for SMMC-7721 cells infected with the lentivirus expressing TSLNC8 or transfected with TSLNC8 siRNAs. (D) The mRNA levels of STAT3 downstream genes in SMMC-7721 cells infected with the lentivirus expressing TSLNC8 or transfected with TSLNC8 siRNAs. (E) Cell proliferation assays and Transwell migration assays for SMMC-7721 cells infected with the lentivirus expressing full-length TSLNC8, the deletion fragments, or the control. (F) Cell proliferation assays and Transwell migration assays for SMMC-7721 cells infected with the lentivirus expressing full-length TSLNC8 or the TSLNC8-M (748-758 nucleotide sequence substituted by unrelated sequence). b-Actin served as the control in (C) and (D). Values are expressed as mean 6 SEM, n 5 3 (C-F). *P < 0.05, **P < 0.01, and ***P < 0.001. Abbreviations: CC, coiled-coil domain; C-end, C-terminus; DB, DNA-binding domain; IB, immunoblot; Int, protein interaction domain; NC, normal control; OD, optical density.

180


ZHANG, LI, ET AL.

FIG. 6. TSLNC8 regulates STAT3-Tyr705 and STAT3-Ser727 phosphorylation level in HCC cells. (A) Immunoblotting analysis for the phosphorylation status of STAT3-Y705 and STAT3-S727 in SMMC-7721 cells infected with the lentivirus expressing TSLNC8 or in Huh-7 cells transfected with TSLNC8 siRNA. (B) Immunoblotting analysis for the distribution of the phosphorylated STAT3-Y705 and STAT3-S727 in SMMC-7721 cells infected with the lentivirus expressing TSLNC8 or in Huh-7 cells transfected with TSLNC8 siRNA. (C) Immunoblotting analysis for the phosphorylation status of STAT3-Y705 and STAT3-S727 in SMMC7721 cells infected with the lentivirus expressing full-length TSLNC8, deletion fragments, or the control. (D) Cell proliferation assays and Transwell migration assays for SMMC-7721 cells infected with the lentivirus expressing wild-type STAT3, the STAT3-Y705D mutant, or the control. (E) Cell proliferation assays and Transwell migration assays for Huh-7 cells infected with a lentivirus expressing wild-type STAT3, the STAT3-S727D mutant, or the control. b-Actin served as the control in (A) and (C). Glyceraldehyde 3phosphate dehydrogenase and H3 served as the controls in (B). Values are expressed as mean 6 SEM, n 5 3 (D, E). *P < 0.05, **P < 0.01, and ***P < 0.001. Abbreviations: GAPDH, glyceraldehyde 3-phosphate dehydrogenase; NC, normal control; OD, optical density.

promoted the proliferation and migration of HCC cells (Supporting Fig. S7A-C), whereas TKT knockdown suppressed HCC cell proliferation and migration (Supporting Fig. S7D-F). Moreover, coimmunoprecipitation assays showed that TKT primarily bound the STAT3 C-terminal domain (Supporting Fig. S8A), in which the Y705 and S727 sites are located. These data confirmed the effect of TKT on STAT3 phosphorylation in HCC cells. The luciferase reporter assay also revealed that TKT could enhance the activity of STAT3 response element in HCC cells (Fig. 7C). Moreover, the restoration of TKT remarkably abolished the inhibitory effects of TSLNC8 on cell proliferation, migration, and

invasion, whereas TKT siRNA erased the promoting effect of TSLNC8 knockdown on cell proliferation, migration, and invasion (Fig. 7D; Supporting Fig. S8B,C). These results indicated that TSLNC8 might regulate cell growth and metastasis by modulating the STAT3 phosphorylation status through TKT. Next, RNA pull-down and RIP assays confirmed that TSLNC8 was physically associated with TKT in HCC cells (Supporting Fig. S8D,E). Interestingly, TKT also interacted with the right arm of TSLNC8 containing the crucial 480-831 nucleotide region (Supporting Fig. S8F), which overlapped with the 480-831 nucleotide region recognized by STAT3, suggesting that TSLNC8 might execute its function by

181

ZHANG, LI, ET AL.


FIG. 7. TSLNC8 regulates STAT3 phosphorylation by interacting with TKT in HCC cells. (A) Immunoblotting analysis for the phosphorylation status of STAT3-Y705 and STAT3-S727 in Huh-7 cells transfected with TKT siRNA or in SMMC-7721 cells infected with the lentivirus expressing TKT. (B) Immunoblotting analysis for the phosphorylation status of STAT3-Y705 and STAT3-S727 in cells with the indicated treatments. (C,D) Effect of additional TKT on the luciferase assays, cell proliferation assays, Transwell migration assays, and invasion assays in cells with the indicated treatments. (E) Identification of the interaction between STAT3 and TKT in SMMC-7721 cells infected with the lentivirus expressing TSLNC8 or in Huh-7 cells transfected with TSLNC8 siRNA. (F) Identification of the TSLNC8–TKT complex after incubation of the biotinylated TSLNC8 probe with protein extracts from SMMC-7721 cells transfected with the STAT3 siRNA or TKT siRNA. b-Actin served as the control (A-F). Values are expressed as mean 6 SEM, n 5 3 (C,D). *P < 0.05, **P < 0.01, and ***P < 0.001. Abbreviations: IB, immunoblot; IgG, immunoglobulin G; IP, immunoprecipitation; NC, normal control; OD, optical density.

preventing TKT and STAT3 from binding to each other. Then, the interaction between STAT3 and TKT was examined by coimmunoprecipitation assays, in the presence and absence of TSLNC8. TSLNC8 knockdown could enhance the interaction between

182

TKT and STAT3 (Fig. 7E, left). When TSLNC8 was overexpressed, there was less TKT that interacted with STAT3, as well as less STAT3 that bound to TKT (Fig. 7E, right), indicating that TSLNC8 exhibits a negative regulatory effect on the interaction


between TKT and STAT3. Subsequent RNA pulldown assays showed that STAT3 knockdown strengthened the assembly of TSLNC8 with TKT, and TKT knockdown enhanced the interaction between TSLNC8 with TKT (Fig. 7F), without obvious effects on the mRNA levels of TSLNC8 in the processes (Supporting Fig. S8G,H), suggesting competitive binding of TKT or STAT3 to TSLNC8. Collectively, these results demonstrated that TSLNC8 regulated STAT3 phosphorylation and exerted its tumor suppressive function through competitively binding to TKT or STAT3, thereby preventing the interaction between these two proteins in HCC cells.

TSLNC8 HARBORS TUMOR SUPPRESSOR PROPERTIES THROUGH INACTIVATING THE IL-6–STAT3 PATHWAY IN HCC CELLS IL-6 is the most important STAT3 activator and promotes tumor cell proliferation, survival, invasion, and angiogenesis.(25) In most HCC cell lines, IL-6 enhances STAT3-Y705 phosphorylation and inhibits STAT3-S727 phosphorylation (Supporting Fig. S9A). Because TSLNC8 can regulate STAT3 phosphorylation and activity, we next evaluated its effects on the IL-6–STAT3 signaling pathway in HCC cells. Although IL-6 did not affect TSLNC8 RNA levels and TKT mRNA levels (Supporting Fig. S9B), it induced Y705 phosphorylation; and this induction was significantly repressed by TSLNC8 overexpression, whereas the inhibitory effect of IL-6 on S727 phosphorylation was restored by TSLNC8 (Fig. 8A; Supporting Fig. S9C). Not surprisingly, the enhancement of the luciferase activity of STAT3 response elements by IL-6 was also inhibited by TSLNC8 introduction and facilitated by TSLNC8 knockdown (Fig. 8B). Consistently, the IL-6-induced downstream gene expression in the JAK–STAT3 signaling pathway was suppressed by TSLNC8 overexpression but fostered by introduction of TKT (Fig. 8C). These findings confirmed that TSLNC8 inhibited IL-6-mediated STAT3 signaling in HCC cells. Furthermore, rescue experiments indicated that TKT could significantly restore the TSLNC8-induced decrease of Y705 phosphorylation and the increase of S727 phosphorylation in the condition of IL-6 treatment (also seen in Fig. 8A). Consequently, similar restoration effects of TKT

ZHANG, LI, ET AL.

against TSLNC8 were observed in the luciferase assays evaluating the STAT3 response element and the expression of genes downstream of the JAK–STAT3 pathway (Supporting Fig. S9D). Taken together, TSLNC8 could modulate IL-6–STAT3 signaling pathways, which involved a complementary role of TKT in HCC cells. To determine the pathological significance of TKT, STAT3-Y705 phosphorylation, and STAT3-S727 phosphorylation, we analyzed their levels through immunohistochemical staining of 334 HCC/nontumor samples (Supporting Table S6). The median follow-up period of these patients was 79.0 months (range, 4.0-109.0; standard deviation, 36.60). The overall survival rates at 1, 3, 5, and 7 years posthepatectomy were 93.4%, 72.3%, 60.5%, and 54.3%, respectively. The survival analysis showed no significant differences in the patients with high or low TKT expression (Supporting Fig. S10A,B), in the patients with high or low levels of Y705 phosphorylation (Supporting Fig. S10C,D) and in the patients with high or low levels of S727 phosphorylation (Supporting Fig. S10E,F). Notably, patients with a simultaneously high TKT level and high Y705 phosphorylation showed significantly diminished survival. Consistently, the results showed that patients with simultaneously low TKT level and high S727 phosphorylation showed significantly optimal survival, and patients with simultaneously low Y705 phosphorylation and high S727 phosphorylation showed a significantly better survival (Fig. 8D). The multivariate analysis (Supporting Table S7) revealed that the tumor TKT, Y705 phosphorylation, and S727 phosphorylation status in combinations of two were independent prognostic factors for postoperative survival.

Discussion Deletions in the short arm of chromosome 8 are the most frequent genetic events in a variety of tumors(2629) ; for instance, the 8p deletion contributes to HCC metastasis and patient mortality.(10,30,31) Several established or putative tumor suppressor genes are located on 8p, including CSMD1 (8p23),(8) DLC1 (8p22),(9) ELP3 (8p21),(10) HTPAP (8p12),(11) LZTS1 (8p21),(12) SH2D4A (8p22),(15) and TNFRSF10C (8p21).(16) In the present study, we identified the lncRNA TSLNC8, which is located at 8p12, as a tumor suppressor in HCC. The TSLNC8 locus was frequently deleted in HCC and markedly suppresses

183

ZHANG, LI, ET AL.


FIG. 8

184


the tumorigenicity and metastasis of HCC cells. Importantly, TSLNC8 expression was inversely correlated with tumor nodules, cancer embolus, and differentiation stage. Patients with lower TSLNC8 expression also exhibited poorer overall survival, indicating that TSLNC8 expression could serve as a promising prognostic indicator for HCC patients. TSLNC8 controls the expression of target genes involved in HCC cell proliferation, invasion, and metastasis by disrupting the IL-6–STAT3 signaling pathway. Particularly, TSLNC8 can directly interact with STAT3 and regulate its phosphorylation and transcriptional activity. As a promising molecular target for the treatment of various cancers, STAT3 is activated in many cancer types.(32,33) Several lncRNAs regulate the phosphorylation status of STAT3 and alter its nuclear import–export dynamics.(34,35) The lnc-DC binds directly to STAT3 and promotes STAT3 phosphorylation at Y705 by preventing STAT3 from binding to phosphatase small heterodimer partner 1.(34) OLA1P2 binds to STAT3, inhibits the phosphorylation of Y705, and blocks phosphorylated STAT3 homodimer formation.(35) Here, we showed that TSLNC8 binds to the DNAbinding domain of STAT3 but not the C-terminal domain containing the Y705 and S727 sites, which binds with lnc-DC or OLA1P2, and that the STAT3 recognized site at the 748-758 nucleotides of TSLNC8 is important for the biological function of this lncRNA. In some cancers, STAT3 is primarily activated through phosphorylation at Y705 and/or S727.(21) The oncogenic effect of Y705 has been confirmed in various cancer types.(22) However, the role of S727 phosphorylation is not well defined, and this modification may either enhance or inhibit the DNA binding ability of Y705-phosphorylated STAT3.(23) In the present study, the oncogenic effect of Y705 phosphorylation and the suppressive effect of S727 phosphorylation were observed in HCC cells, which

ZHANG, LI, ET AL.

supported the interdependence of STAT3 serine and tyrosine phosphorylation in different cancer types. The effects of TSLNC8 on the protein levels of phosphorylated STAT3-Y705, phosphorylated STAT3-S727, and total STAT3 protein level are diverse in HCC cells. Intriguingly, the TSLNC8-induced effects on the protein levels and nuclear distribution of phosphorylated STAT3-Y705 exhibited an independent manner with those on total STAT3 protein. These results suggested that TSLNC8 could affect the phosphorylation status and the nuclear distribution of STAT3, thus leading to the distinguished biological function of STAT3 in HCC cells. The decrease in Y705 phosphorylation and the increase in S727 phosphorylation may contribute to the tumor suppressive effects of TSLNC8 on HCC cell proliferation, invasion, and metastasis. The oncogenic roles of IL-6–STAT3 signaling have been demonstrated in HCC and other solid tumors.(25) Several tumor suppressor genes at 8p, such as SH2D4A and SORBS3, have been associated with the repression of IL-6 signaling in HCC.(15) However, the lncRNAs with these genomic changes in the crosslink between tumor cells and the tumor microenvironment remain unclear. In the present study, we demonstrated that TSLNC8, in the frequently deleted region 8p12, was associated with the repression of IL-6 signaling in HCC cells, thereby representing a distinct layer of regulatory circuitry through lncRNAs. TKT is a ubiquitous enzyme that catalyzes the reversible transfer of two-carbon ketol units between ketose and aldose phosphates, which govern the carbon flow through the nonoxidative branch of the pentose phosphate pathway.(36) Up-regulation of TKT and reactive oxygen species–mediated extracellular signal– regulated kinase activation by sugiol inhibited STAT3 activity in prostate carcinoma cells.(24) Although these data suggest that TKT may regulate STAT3 activity, to the best of our knowledge the mechanism

FIG. 8. TSLNC8 harbors tumor suppressor properties through dampening of the IL-6–STAT3 signaling pathway. (A) Immunoblotting analysis for the phosphorylation status of STAT3-Y705 and STAT3-S727 in cells with the indicated treatments. (B) Luciferase assays for cells with the indicated treatments. (C) The mRNA levels of STAT3 downstream genes in cells with the indicated treatments. (D) Combined influence of TKT, phosphorylated STAT3-Y705, or phosphorylated STAT3-S727 dimorphisms on the risk of HCC death. Left, patients with simultaneously high Y705 phosphorylation and TKT expression showed significantly poor survival rates. Middle, patients with simultaneously high S727 phosphorylation expression and low TKT expression showed significantly optimal survival. Right, patients with simultaneously high S727 phosphorylation and low Y705 phosphorylation showed significantly better survival. Cumulative survival rates were determined using the Kaplan-Meier method (log-rank test). (E) The working model shows that TSLNC8 loci on chromosome 8p12 are deleted in HCC, resulting in a decrease of TSLNC8 expression. TSLNC8 interacts with TKT and STAT3, thus modulating STAT3 phoshorylation and inhibiting its transcriptional activity. TSLNC8 exerts its tumor suppressive activities through inactivation of the IL-6–STAT3 signaling pathway in HCC. b-Actin served as the control (A-C); values are expressed as mean 6 SEM, n 5 3 (B,C). *P < 0.05, **P < 0.01, and ***P < 0.001. Abbreviations: ISRE, interferon-stimulated response element; NC, normal control.

185

ZHANG, LI, ET AL.

underlying this process has not been demonstrated. We showed that TKT interacted with the C-terminal domain of STAT3 in HCC cells, resulting in the increase of Y705 phosphorylation and the decrease of S727 phosphorylation. Furthermore, TKT, Y705 phosphorylation, or S727 phosphorylation status in combinations were independent prognostic factors for HCC. Both TKT and STAT3 competitively bind to the right arm of TSLNC8, thereby preventing the interaction of these two proteins. These results established a molecular link between TKT, STAT3 phosphorylation, and lncRNA TSLNC8. In conclusion, we characterized the lncRNA TSLNC8 as a tumor suppressor on chromosome 8p12. TSLNC8 can physically interact with TKT and STAT3 and inhibit STAT3 phosphorylation and transcriptional activity in HCC cells. TSLNC8 exerts its tumor suppressive activity through inactivation of the IL-6–STAT3 signaling pathway in HCC (Fig. 8E). TSLNC8 may serve as a prognostic predictor for patients with HCC, and the TSLNC8–TKT–STAT3 axis is a potential therapeutic target for HCC treatment. Acknowledgment: We are most grateful for Dr. T. Didier’s gifts of the pWPXL, psPAX2, and pMD2.G plasmids.


10)

11)

12)

13)

14)

15)

16)

17)

18)

REFERENCES 1) Necsulea A, Soumillon M, Warnefors M, Liechti A, Daish T, Zeller U, et al. The evolution of lncRNA repertoires and expression patterns in tetrapods. Nature 2014;505:635-640. 2) Mercer TR, Dinger ME, Mattick JS. Long non-coding RNAs: insights into functions. Nat Rev Genet 2009;10:155-159. 3) Sahu A, Singhal U, Chinnaiyan AM. Long noncoding RNAs in cancer: from function to translation. Trends Cancer 2015;1:93109. 4) Schmitt AM, Chang HY. Long noncoding RNAs in cancer pathways. Cancer Cell 2016;29:452-463. 5) Chen W, Zheng R, Baade PD, Zhang S, Zeng H, Bray F, et al. Cancer statistics in China, 2015. CA Cancer J Clin 2016;66: 115-132. 6) Rapisarda V, Loreto C, Malaguarnera M, Ardiri A, Proiti M, Rigano G, et al. Hepatocellular carcinoma and the risk of occupational exposure. World J Hepatol 2016;8:573-590. 7) Goh GB, Chang PE, Tan CK. Changing epidemiology of hepatocellular carcinoma in Asia. Best Pract Res Clin Gastroenterol 2015;29:919-928. 8) Midorikawa Y, Yamamoto S, Tsuji S, Kamimura N, Ishikawa S, Igarashi H, et al. Allelic imbalances and homozygous deletion on 8p23.2 for stepwise progression of hepatocarcinogenesis. HEPATOLOGY 2009;49:513-522. 9) Zimonjic DB, Popescu NC. Role of DLC1 tumor suppressor gene and MYC oncogene in pathogenesis of human

186

19)

20)

21)

22) 23) 24)

25)

hepatocellular carcinoma: potential prospects for combined targeted therapeutics (review). Int J Oncol 2012;41:393-406. Roessler S, Long EL, Budhu A, Chen Y, Zhao X, Ji J, et al. Integrative genomic identification of genes on 8p associated with hepatocellular carcinoma progression and patient survival. Gastroenterology 2012;142:957-966. Ren N, Wu JC, Dong QZ, Sun HJ, Jia HL, Li GC, et al. Association of specific genotypes in metastatic suppressor HTPAP with tumor metastasis and clinical prognosis in hepatocellular carcinoma. Cancer Res 2011;71:3278-3286. He Y, Liu X. The tumor-suppressor gene LZTS1 suppresses hepatocellular carcinoma proliferation by impairing PI3K/Akt pathway. Biomed Pharmacother 2015;76:141-146. Di Benedetto M, Pineau P, Nouet S, Berhouet S, Seitz I, Louis S, et al. Mutation analysis of the 8p22 candidate tumor suppressor gene ATIP/MTUS1 in hepatocellular carcinoma. Mol Cell Endocrinol 2006;252:207-215. Huang J, Zheng DL, Qin FS, Cheng N, Chen H, Wan BB, et al. Genetic and epigenetic silencing of SCARA5 may contribute to human hepatocellular carcinoma by activating FAK signaling. J Clin Invest 2010;120:223-241. Ploeger C, Waldburger N, Fraas A, Goeppert B, Pusch S, Breuhahn K, et al. Chromosome 8p tumor suppressor genes SH2D4A and SORBS3 cooperate to inhibit interleukin-6 signaling in hepatocellular carcinoma. HEPATOLOGY 2016;64:828-842. Ganten TM, Haas TL, Sykora J, Stahl H, Sprick MR, Fas SC, et al. Enhanced caspase-8 recruitment to and activation at the DISC is critical for sensitisation of human hepatocellular carcinoma cells to TRAIL-induced apoptosis by chemotherapeutic drugs. Cell Death Differ 2004;11(Suppl. 1):S86-S96. Jia D, Wei L, Guo W, Zha R, Bao M, Chen Z, et al. Genomewide copy number analyses identified novel cancer genes in hepatocellular carcinoma. HEPATOLOGY 2011;54:1227-1236. Xu D, Yang F, Yuan JH, Zhang L, Bi HS, Zhou CC, et al. Long noncoding RNAs associated with liver regeneration 1 accelerates hepatocyte proliferation during liver regeneration by activating Wnt/beta-catenin signaling. HEPATOLOGY 2013;58: 739-751. Huang JF, Guo YJ, Zhao CX, Yuan SX, Wang Y, Tang GN, et al. Hepatitis B virus X protein (HBx)–related long noncoding RNA (lncRNA) down-regulated expression by HBx (Dreh) inhibits hepatocellular carcinoma metastasis by targeting the intermediate filament protein vimentin. HEPATOLOGY 2013;57: 1882-1892. Pan YF, Qin T, Feng L, Yu ZJ. Expression profile of altered long non-coding RNAs in patients with HBV-associated hepatocellular carcinoma. J Huazhong Univ Sci Technolog Med Sci 2013;33:96-101. Wen Z, Zhong Z, Darnell JE Jr. Maximal activation of transcription by Stat1 and Stat3 requires both tyrosine and serine phosphorylation. Cell 1995;82:241-250. Bowman T, Garcia R, Turkson J, Jove R. STATs in oncogenesis. Oncogene 2000;19:2474-2488. Sansone P, Bromberg J. Targeting the interleukin-6/Jak/stat pathway in human malignancies. J Clin Oncol 2012;30:1005-1014. Jung SN, Shin DS, Kim HN, Jeon YJ, Yun J, Lee YJ, et al. Sugiol inhibits STAT3 activity via regulation of transketolase and ROS-mediated ERK activation in DU145 prostate carcinoma cells. Biochem Pharmacol 2015;97:38-50. Bromberg J, Wang TC. Inflammation and cancer: IL-6 and STAT3 complete the link. Cancer Cell 2009;15:79-80.


26) Lai J, Flanagan J, Phillips WA, Chenevix-Trench G, Arnold J. Analysis of the candidate 8p21 tumour suppressor, BNIP3L, in breast and ovarian cancer. Br J Cancer 2003;88:270-276. 27) Seitz S, Rohde K, Bender E, Nothnagel A, Kolble K, Schlag PM, et al. Strong indication for a breast cancer susceptibility gene on chromosome 8p12-p22: linkage analysis in German breast cancer families. Oncogene 1997;14:741-743. 28) Stoehr R, Wissmann C, Suzuki H, Knuechel R, Krieg RC, Klopocki E, et al. Deletions of chromosome 8p and loss of sFRP1 expression are progression markers of papillary bladder cancer. Lab Invest 2004;84:465-478. 29) Tanaka K, Kikuchi-Yanoshita R, Muraoka M, Konishi M, Oshimura M, Miyaki M. Suppression of tumorigenicity and invasiveness of colon carcinoma cells by introduction of normal chromosome 8p12-pter. Oncogene 1996;12:405-410. 30) Li SP, Wang HY, Li JQ, Zhang CQ, Feng QS, Huang P, et al. Genome-wide analyses on loss of heterozygosity in hepatocellular carcinoma in southern China. J Hepatol 2001;34:840849. 31) Qin LX, Tang ZY, Sham JS, Ma ZC, Ye SL, Zhou XD, et al. The association of chromosome 8p deletion and tumor metastasis in human hepatocellular carcinoma. Cancer Res 1999;59:56625665.

ZHANG, LI, ET AL.

32) Bromberg JF, Wrzeszczynska MH, Devgan G, Zhao Y, Pestell RG, Albanese C, et al. Stat3 as an oncogene. Cell 1999;98:295-303. 33) Chai EZ, Shanmugam MK, Arfuso F, Dharmarajan A, Wang C, Kumar AP, et al. Targeting transcription factor STAT3 for cancer prevention and therapy. Pharmacol Ther 2016;162:86-97. 34) Wang P, Xue Y, Han Y, Lin L, Wu C, Xu S, et al. The STAT3-binding long noncoding RNA lnc-DC controls human dendritic cell differentiation. Science 2014;344:310-313. 35) Guo H, Liu J, Ben Q, Qu Y, Li M, Wang Y, et al. The aspirininduced long non-coding RNA OLA1P2 blocks phosphorylated STAT3 homodimer formation. Genome Biol 2016;17:24. 36) Schenk G, Duggleby RG, Nixon PF. Properties and functions of the thiamin diphosphate dependent enzyme transketolase. Int J Biochem Cell Biol 1998;30:1297-1318.

Author names in bold designate shared co-first authorship.

Supporting Information Additional Supporting Information may be found at onlinelibrary.wiley.com/doi/10.1002/hep.29405/suppinfo.

187