Long noncoding RNA HULC accelerates liver cancer by inhibiting PTEN via autophagy cooperation to miR15a

Thirty cases of liver cancer tissues used for analysis were obtained from liver cancer patients who had undergone surgery. The liver cancer tissues were fixed in formalin before embedded in paraffin. All patients were diagnosed as liver cancer according to histological examination.

Human liver cancer line Hep3B was maintained in DMEM Medium supplemented with 10% heat-inactivated fetal bovine serum (Gibco) in a humidified atmosphere of 5% CO₂ incubator at 37 °C. Plasmid pGFP-V-RS, pCMV6-A-GFP, pCMV6-XL5-PTEN, pGFP-V-RS-PTEN, pGFP-V-RS-Sirt1, pMiR-Target were purchased from Origene (Rockville, MD 20850, USA). pEZX-MT01-PTEN-3’UTR was purchased from GeneCopeia (Rockville, MD, USA).pCMV6-AC-GFP-HULC [the HULC sequence (NR_004855) was synthesized and cloned into the cloning site ofpCMV6-AC-GFP plasmid (Origene)], pcDNA3-sirt1[the Sirt1 sequence in the Flag-SIRT1 (Plasmid #1791,Addigene) was digested and subcloned into the cloning site of pcDNA3], pcDNA3-sirt mutant [the mutant Sirt1 (H363Y) sequence in the Flag-SIRT1 H363Y (Plasmid #1792, addigene) was digested and sub-cloned into the pcDNA3] were constructed by our lab.

Total RNA was purified using Trizol (Invitrogen) according to manufacturer’s instructions. cDNA was prepared by SuperScript First-Strand Synthesis System (Invitrogen). PCR analysis was performed under the special conditions. β-actin was used as an internal control.

Total RNA was isolated from cells using Trizol (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s protocol. Real-time RT-PCR-based detection of mature miR-15a and U6 snRNA was achieved with the miRNA Detection kit (including a universe primer, U6 primers, Qiagen) and miR15a specific upsteam primers (Origene, USA).

Proteins were separated on a 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto a nitrocellulose membrane. The blots were incubated with antibodyovernight at 4 °C. Following three washes, membranes were then incubated with secondary antibody overnight at 4 °C. Signals were visualized by ECL.

Cell lysates were incubated with A/G-plus agarose beads (Santa Cruz, Biotechnology, Inc. CA) together with the antibody or normal mouse or rabbit IgG for 4 h at 4 °C. Beads were subsequently washed and RNAs werethen isolated. RT-PCR was performed according to the manufacturer’s instructions.

Cells growth in vitro were detected by CCK8 assay kit according to the manufacturer instruction. Cell growth curve was based on the corresponding the normalized values of OD450.

Cells were plated on a six well plate and the six well plate was incubated at 37 °C in humidified incubator for 10 days. Cell colonies were stained with 0.5% Crystal Violet and the colonies were counted.

Four-weeks old male athymic Balb/C mice were purchased from Shi Laike Company (Shanghi, China) and maintained in the Tongji animal facilities approved by the China Association for Accreditation of Laboratory Animal Care. The athymic Balb/C mice were injected in the armpit area subcutaneously with cell suspension. The mice were observed four weeks, and then sacrificed to recover the tumors.

To identify the relationship between HULC and PTEN, miR15a in human liver cancer, we detected the expression of HULC, miR15a, PTEN in human hepatocarocinoma tissues. At the first time, we preformed In situ hybridization (ISH) and RT-PCR to detect HULC expression. The expression of HULC was significantly increased in liver cancer tissues compared to their adjacent noncancerous tissues (100%,n = 30, P < 0.01) [Fig. 1a (left and middle) & Fig. 1b]. Next, we detected the miR15a, the level of mature miR15a were significantly reduced in the liver cancer tissues compared to their adjacent noncancerous tissues (100%, n = 30, P < 0.01) (Fig. 1c). Finally, we performed immunohistochemistry staining and Western blotting to analyse PTEN expression. The expression of PTEN was significantly reduced in liver cancer tissues compared to their adjacent noncancerous tissues (100%, n = 30, P < 0.01) (Fig. 1a (right) & Fig. 1d). Significantly, there was a strong negative relevance between HULC and PTEN or miR15a. Taken Together, these observations suggest HULC was over-expressed and miR15a or PTEN was down-expressed inhuman liver cancer tissue.

To further investigate whether HULC promoted malignant growth of human liver cancer cell Hep3B, we first established two stable Hep3B cell lines transfected withpCMV6-A-GFP, pCMV6-A-GFP-HULC respectively. As shown in Fig. 2a & b, the expression of HULC was significantly increased in HULC overexpressing Hep3B on the transcriptional level. As shown in Fig. 2c, excessive HULC significantly increased the growth of liver cancer cell Hep3B compared to the control cells (P < 0.01). We further performed colony formation assay and observed a significant increase in colony formation efficiency rate in excessive HULC group compared to control (85.33 ± 14.35% vs 20.67 ± 3.83%, P = 0.00846 < 0.01) (Fig. 2d & e). Moreover, the BrdU positive rate is significantly increased in pCMV6-A-GFP-HULC group compared to the control cells (70.99 ± 16.44% versus 35.41 ± 8.03%, p = 0.0244 < 0.05) (Fig. 2f).

To explore the effect of HULC on liver cancer cells in vivo, the two stable Hep3B cell lines were injected subcutaneously into Balb/C mice. As shown in Fig. 3a, b & c, when HULC was overexpressed, the average xenograft tumor weight is significantly increased compared to the corresponding control group (0.000843 ± 0.00223 g versus 0.1996 ± 0.1676 g, P = 0.009713 < 0.01), and when HULC was overexpressed, the xenograft tumor formation rate is significantly increased in HULC overexpressing group compared to the corresponding control group (100% versus 14.29%, P < 0.01). Moreover, compared to control, xenograft tumors contained more of poorly differentiated cells in pCMV6-A-GFP-HULC group compared to control group (Fig. 2d, left). As well as the PCNA and Ki67 were significantly higher in HULC overexpressing xenograft tumors than in the control group (Fig. 2d, middle and right).

Collectively, these findings suggest that HULC accelerates malignant progression of liver cancer cells.

Given that there was a strong negative relevance between HULC and miR15a expression, we wonder whether HULC influenced on the expression and mature of miR15a.Compared to the control group, HULC inhibited the binding of METTL3 (a N6-adenosine-methyltransferase) to pri-miR15a (Fig. 4a) and decreased the methylation level (me6A) of pri-miR15a (Fig. 4b). Moreover, HULC inhibited the binding of DGCR8 and Droha to pri-miR15a (Fig. 4c & d) and the binding of exportin5 and Droha to pri-miR15a (Fig. 4e). Therefore, HULC decreased the binding of Dicer and Ago2 to mature miR15a probe (Fig. 4f) and decreased the level of pre-miR15a, pri-miR15a, and mature miR15a compared to the control group (Fig. 4g & h). Furthermore, in Hep3B cell lines with excessive miR15a, (Fig. 4i), miR15a could significantly decrease the expression of P62, Notch, JUN, mTOR and increase the expression of PTEN (Fig. 4j). Together, these observations suggest that HULC inhibits expression and mature of miR15a.

To address whether HULC was associated with autophagy, we analyse the HULC functions on autophagy in liver cancer cells. At the first time, both the CHIP and the quantitative CHIP showed that HULC overexpression enhanced RNA polII loading on the LC3 promoter region (Fig. 5a). Furthermore, both RT-PCR and real-time RT-PCR showed that HULC overexpression increased the mRNA level of LC3 (Fig. 5b). Finally, HULC overexpression increased the expression of LC3I and LC3II (a autophagy marker) compared to control (Fig. 5c). On the other hand, HULC overexpression increased the expression of Sirt1 (Fig. 5d). Furthermore, Sirt1 could decreased the acetylation of LC3, however, the mutant Sirt1 did not result in this action (Fig. 5e). Notably, excessive Sirt1 could increase the LC3 II (a LC3 I trans-located product), however, the mutant sirt1 did not result in this action (Fig. 5f). Excessive HULC could increase the LC3I and LC3 II, however, either Sirt1 knockdown or sirt1 inhibitor fully abrogated the excessive HULC action in addition to increase LC3I (Fig. 5g). Interestingly, excessive HULC could increase the LC3I and LC3 II, however, either miR15a overexpression or miR15a mimic could not alter this action (Fig. 5h & i). Importantly, HULC enhanced the interplay between LC3 and ATG3 (Fig. 6a). Moreover, HULC increased the expression of becline-1, however, both Sirt1 knockdown could and miR15a overexpression could not abrogate this HULC action (Fig. 6b). Finally, HULC increased the autophagy of Hep3B cell (Fig. 6c). Furthermore, HULC decreased the expression of PTEN, β-catenin and increased the expression of SAPK/JUNK, PKM2, CDK2, NOTCH1, C-Jun under starvation. However, in pCMV6-A-GFP-HULC plus 3-methyladenine (3-MA, autophagy inhibitor) group, the expression of PTEN, β-catenin, SAPK/JUNK, PKM2, CDK2, NOTCH1, C-Jun were significantly not altered compared to control (Fig. 6d). Taken together, HULC enhances autophagy in liver cancer cells via activating Sirt1.These data prove that HULC promotes the expression of LC3 not via Sirt1, and HULC increases the LC3II dependent on Sirt1.In particular, HULC promotes autophagy by enhancing the autophagy process of the transformation from LCI to LC3 II.

Given that HULC suppresses the expression of PTEN, miR15a and enhances the cell autophagy, and miR15a could inhibit the expression of 62, we wonder whether the effect of HULC on PTEN is associated with cell autophagy and miR15a. As shown in Fig. 7a, HULC inhibited the expression of PTEN on the translational level, but not on the transcriptional level. Moreover, HULC also inhibited the expression of recombinant PTEN on the translational level, but not on the transcriptional level (Fig. 7b). Furthermore, HULC enhanced the interplay between P62 and PTEN. However, both excessive miR15a and Sirt1 knockdown abrogated this HULC’s action (Fig. 7c). Surprisingly, excessive HULC promotes the ubiquitination of PTEN compared to control. However, both excessive miR15a and Sirt1 knockdown abrogated this HULC’s action (Fig. 7d & e). On the other hand, HULC inhibited the expression of PETN, and both excessive miR15a and Sirt1 knockdown abrogated this HULC’s action. In particular, HULC could not alter the expression of PTEN after the MG132 (ubiquitin– proteasome inhibitor) (Fig. 7f). In addition, we also found that HULC decrease the PTEN 3’-UTR luciferase activity (Fig. 7g). Taken together, these observations suggest that HULC inhibits PTEN through ubiquitin–proteasome system mediated by autophagy and P62.

To explore whether oncogenic action of HULC is associated with PTEN, we established three stable Hep3B cell lines (pCMV6-A-GFP, pCMV6 -A-GFP-HULC, pCMV6-A-GFP-HULC plus pCMV6-XL5-PTEN). Compared to control, excessive HULC inhibits the expression and phosphorylation of PTEN. Moreover, excessive HULC increases the phosphorylation of EGFR (Tyr845), AKT, mTOR, PI3K, JUN and the expression of survivin. However, excessive PTEN abrogated the HULC’s action (Fig. 8a). Furthermore, excessive HULC increased the loading of Jun on the H-Ras promoter region. However, PTEN overexpression fully abrogated this HULC’s action (Fig. 8b). Moreover, excessive HULC increased the luciferase activity of the H-Ras promoter. However, PTEN overexpression fully abrogated the HULC’s action (Fig. 8c). Ultimately, excessive HULC increased the mRNA of H-Ras and the expression of H-Ras. However, rescued PTEN fully abrogated this HULC’s action (Fig. 8d & e). Next we detected the cell proliferation and colony formation ability. As shown in Fig. 9a, excessive HULC significantly increased the growth of liver cancer cell Hep3B compared to the control cells (P < 0.01). However, HULC plus PTEN did significantly not alter the growth ability of liver cancer cells (P > 0.05). We further performed colony formation assay and observed a significant increase in colony formation efficiency rate in excessive HULC (75.68 ± 16.78% vs 31.77 ± 8.22%, P = 0.00777 < 0.01). However, HULC plus PTEN did significantly not alter the colony formation rate of liver cancer cells (36.12 ± 6.85% vs 31.77 ± 8.22%, P = 0.2446 > 0.05) (Fig. 9b). Furthermore, HULC overexpression significantly increased the BrdU positive rate compared to the control cells (60.82 ± 15.97% versus 25.59. ± 6.04%, p = 0.047 < 0.05). However, HULC plus PTEN did significantly not alter the BrdU positive rate of liver cancer cells (32.37 ± 9.68% versus 25.59. ± 6.04%, p = 0.1966 > 0.05) (Fig. 9c). Furthermore, the three stable Hep3B cell lines were injected subcutaneously into Balb/C mice. As shown in Fig. 9d, when HULC was overexpressed, the average xenograft tumor weight increased approximately5.58 folds compared to the corresponding control group (0.848 ± 0.222 g versus 0.152 ± 0.035 g, P = 0.00054 < 0.01). However, HULC plus PTEN did significantly not alter the xenograft tumor weight (0.167 ± 0.044 g versus 0.152 ± 0.035 g, P = 0.3196 > 0.05). On the other hand, when HULC was overexpressed, the xenograft tumor appearance time was significantly decreased compared to the corresponding control group (6.017 ± 1.049 days versus 10.583 ± 2.873 days, P = 0.003873 < 0.01). However, HULC plus PTEN did significantly not alter the xenograft tumor appearance time (10.133 ± 1.749 days versus 10.583 ± 2.873 days, P = 0.403494 > 0.05) (Fig. 9e). Moreover, the PCNA was significantly higher in HULC overexpressing xenograft tumors than in the control group (88.27 ± 13.49% versus 31.83 ± 6.41%, P = 0.009494 < 0.01). However, HULC plus PTEN did significantly not alter the PCNA positive rate of liver cancer cells (40.97 ± 9.34%versus 31.83 ± 6.41%, P = 0.1397 > 0.05) (Fig. 9f & g). It suggests that HULC promotes cell growth, colony formation ability and cell growth in vivo. However, PTEN overexpression abrogated this HULC’s action. Taken together, PTEN determines oncogenic function of HULC in liver cancer cells.

Given that HULC promotes autophagy by enhancing the autophagy process of the transformation from LCI to LC3 II (a autophagy marker), we try to test whether that HULC promotes the growth of hepatoma cells by targeting this process. Our results showed the cell proliferation ability in vitro was significantly increased in pCMV6-A-GFP-HULC group compared to pCMV6-A-GFP control group. However, when the 3-Methyladenine (3-MA) (a autophagy inhibitor for blocking from LC3I to LC3II specifically) was added, the cell proliferation ability did not significantly alter in pCMV6-A-GFP-HULC plus 3-MA group compared to pCMV6-A-GFP control group (Fig. 9h). Moreover, the results showed the cell colony formation ability was significantly increased in pCMV6-A-GFP-HULC group compared to pCMV6-A-GFP control group (32.6 ± 7.18% vs 78.93 ± 15.33%, P = 0.00731 < 0.01). However, when the 3-MA was added, the cell colony formation ability did not significantly alter in pCMV6-A-GFP-HULC plus 3-MA group compared to pCMV6-A-GFP control group (32.6 ± 7.18% vs 29.55 ± 4.32%, P = 0.1354 > 0.05) (Fig. 9i). These observations suggest that HULC promotes the growth of hepatoma cells by targeting the process of autophagy (the transformation LCI to LCII).