HistoneH3 demethylase JMJD2A promotes growth of liver cancer cells through up-regulating miR372

Changes in histone lysine methylation status have been observed during cancer formation. JMJD2A protein is a demethylase that is overexpressed in several tumors. Herein, our results demonstrate that JMJD2A accelerates malignant progression of liver cancer cells in vitro and in vivo. Mechanistically, JMJD2A promoted the expression and mature of pre-miR372 epigenetically. Notably, miR372 blocks the editing of 13th exon-introns-14th exon and forms a novel transcript(JMJD2AΔ) of JMJD2A. In particular, JMJD2A inhibited P21(WAF1/Cip1) expression by decreasing H3K9me3 dependent on JMJD2AΔ. Thereby, JMJD2A could enhance Pim1 transcription by suppressing P21(WAF1/Cip1). Furthermore, through increasing the expression of Pim1, JMJD2A could facilitate the interaction among pRB, CDK2 and CyclinE which prompts the transcription and translation of oncogenic C-myc. Strikingly, JMJD2A may trigger the demethylation of Pim1. On the other hand, Pim1 knockdown and P21(WAF1/Cip1) overexpression fully abrogated the oncogenic function of JMJD2A. Our observations suggest that JMJD2A promotes liver cancer cell cycle progress through JMJD2A-miR372-JMJD2AΔ-P21WAF1/Cip1-Pim1-pRB-CDK2-CyclinE-C-myc axis. This study elucidates a novel mechanism for JMJD2A in liver cancer cells and suggests that JMJD2A can be used as a novel therapeutic targets of liver cancer.


INTRODUCTION
It is well known that Histone modifications control chromatin function. JMJD2A belongs to a demethylase for histone H3 on lysine 9/36. JMJD2A is overexpressed in cancer [1] and inhibits repair of DNA damage by reducing homologous recombination (HR) repair [2,3]. Strikingly, JMJD2A is negatively regulated by SIRT2 in cancer [4] and is enhanced via Hypoxia to provide unmethylated H3K9 residues [5]. Of significance, JMJD2A level is negatively associated with miR-34a and JMJD2A inhibits the output of miR-34a [6].
Moreover, a miR372/let-7 axis regulates human somatic cell and cancer cell fates [12,13]. Studies indicates that miR372 inhibits the Ras homolog gene family member C (RhoC) [14] and p62 in human cancer cells [15]. However, the tumor suppressing roles of miR372 was also found in some cancer cells, possibly via the down-regulation of CDK2 and CCNA1 [16].
PIM kinases (a family of Ser/Thr kinases) promotes proliferation and progression of cancer cell [17]. Our previous study showed that upregulation of Pim1 promotes hepatocarcinogenesis through PKM2 and CUDR [18]. Moreover, the Pim1 kinase promotes migration and adhesion of cancer cell [19]. Importantly, Pim1 promotes chemoresistance in cancer [20].
In this study, we indicate that JMJD2A accelerates malignant progression of liver cancer cells. Moreover, JMJD2A inhibited P21(WAF1/Cip1) and enhanced Pim1 through JMJD2AΔ dependent on miR372. This study elucidates a novel mechanism for JMJD2A in liver cancer cells and suggests that JMJD2A can be used as a novel therapeutic targets for liver cancer.

RT-PCR
Total RNA was purified using Trizol (Invitrogen) according to manufacturer's instructions. cDNA was prepared by using oligonucleotide (dT) 15 , random primers, and a SuperScript First-Strand Synthesis System (Invitrogen). PCR analysis was performed under the specical conditions. β-actin was used as an internal control.

MicroRNA detection
Total RNA was isolated from cultured cells using Trizol (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol. Real-time RT-PCR-based detection of mature miR372 and U6 snRNA was achieved with the miRNA Detection kit (including a universe primer, U6 primers, Qiagen) and miR372 specific upsteam primers (Origene, USA). qRT-PCR was performed with a StepOnePlus real-time PCR system (Applied Biosystems), a SuperScript First-Strand Synthesis System (Invitrogen, Carlsbad, CA, USA) and Power SYBR Green PCR Master Mix (Applied Biosystems) in accordance with the manufacturers' protocols.
Northern-western blotting for miRNA 20 µg total RNA (add 20 µl DEPC H2O, 20 µl formamide to RNA sample and then heat RNA at 65°C for 10 min following chill on ice for 1 min) was separated on 12% polyacrylamide/8M urea gel (Amersham Pharmacia, Uppsala, Sweden) in a Protean II apparatus (BioRad, Hercules, CA, USA) (gel preheated to 55°C by electric current and water bath circulation before loading sample). 21nt RNAs are near the middle position between dye of bromophenol blue and xylene FF in such PAGE gel. Soak Hybond-N+ membrane (Amersham) in ddH2O for a few seconds and in transfer buffer (0.5X TBE) for 15 minutes, and soak 2 pieces of whatman paper in 0.5XTBE. Set up transfer as such From Bottom (anode) to top : whatman, membrane, gel, whatman, cathode plate. Separated RNA in gel was electro-blotted onto Hybond-N+ membrane (Amersham) by Trans-Blot SD electrophoretic transfer cell (BioRad) at 400mA for 1 hour. After UV cross-linking and air drying, blotted membrane was prehybridized with Ultrahyb-oligo hybridization buffer (Ambion) at 37~42 ºC for 60 min, hybridized with Biotin-labeled antisense probe of the target miR372 and incubated at 40°C for overnight. The membrane was washed 2~4 times at 40°C with 2× SSC and 0.5% SDS for 15 min and then western blotting with anti-Biotin.

Western blotting
The logarithmically growing cells were washed twice with ice-cold phosphate-buffered saline (PBS, Hyclone) and lysed in a RIPA lysis buffer. Cells lysates were centrifuged at 12,000g for 20 minutes at 4°C after sonication on ice, and the supernatant were separated. After being boiled for 5-10 minutes in the presence of 2-mercaptoethanol, samples containing cellulars proteins were separated on a 10% sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto a nitrocellulose membrane, blocked in 10% dry milk-TBST (20 mM Tris-HCl [PH 7.6], 127 mM NaCl, 0.1% Tween 20) for 1 h at 37°C. Following three washes in Tris-HCl pH 7.5 with 0.1% Tween 20, the blots were incubated with 0.2 µg/ml of antibody(appropriate dilution) overnight at 4°C. Following three washes, membranes were then incubated with secondary antibody for 60 min at 37°C or 4°C overnight in TBST. Signals were visualized by ODYSSEY infrared imaging system.

Chromatin immunoprecipitation (CHIP) assay
Cells were cross-linked with 1% (v/v) formaldehyde (Sigma) for 10 min at room temperature and stopped with 125 mm glycine for 5 min. Crossed-linked cells were washed with phosphate-buffered saline, resuspended in lysis buffer, and sonicated for 8-10 min in a SONICS VibraCell to generate DNA fragments with an average size of 500 bp. Chromatin extracts were diluted 5-fold with dilution buffer, pre-cleared with Protein-A/G-Sepharose beads, and immunoprecipitated with antibodies on Protein-A/G-Sepharose beads. After washing, elution and de-cross-linking, the ChIP DNA was detected by PCR.

Super-EMSA(Gel-shift)
Cells were washed and scraped in ice-cold PBS to prepare nuclei for electrophoretic gel mobility shift assay with the use of the gel shift assay system modified according to the manufacturer's instructions (Promega).

Chromosome conformation capture (3C) -chromatin immunoprecipitation (ChIP) assay(ChIP-3C/ChIP-Loop assays)
Antibody-specific immunoprecipitated chromatin was obtained as described above for ChIP assays. Chromatin still bound to the antibody-Protein-A/G-Sepharose beads were resuspended in 500 μl of 1.2× restriction enzyme buffer at 37°C for 1 h. 7.5 μl of 20% SDS was added, the mixture was incubated for 1 h, followed by addition of 50 μl of 20% Triton X-100, incubated for an additional 1 h. Samples were then incubated with 400 units of selected restriction enzyme at 37°C overnight. After digestion, 40 μl of 20% SDS was added to the digested chromatin, and the mixture was incubated at 65°C for 10 min. 6.125 ml of 1.15× ligation buffer and 375 μl of 20% Triton X-100 was added, the mixture was incubated at 37°C for 1 h, and then 2000 units of T4 DNA ligase was added at 16°C for a 4-h incubation. Samples were then de-cross-linked at 65°C overnight followed by phenol-chloroform extraction and ethanol precipitation. After purification, the ChIP-3C material was detected for long range interaction with specific primers.

Cells proliferation CCK8 assay
Cells were synchronized in G0 phase by serum deprivation, then released from growth arrest by reexposure to serum, and grown in complete medium according to the manufacturer's instructions (Boshide, Wuhan, China).

Colony-formation efficiency assay
5 × 10 2 cells were plated on a 10 cm dish, the 10 ml DMEM containing 10%FBS was added into each 10 cm dish of the three replicate. Then these dishes were incubated at 37°C in humidified incubator for 10 days. Cell colonies on the dishes were stained with 1 ml of 0. 5% Crystal Violet for more than 1 hour and the colonies were counted.

Xenograft transplantation in vivo
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 Hep3B suspension of 1 × 10 8 cells in 100 μl of phosphate buffered saline. The mice were observed four weeks, and then sacrificed to recover the tumors. The wet weight of each tumor was determined for each mouse. A portion of each tumor was fixed in 4% paraformaldehyde and embedded in paraffin for histological hematoxylin-eosin(HE) staining.

JMJD2A accelerates growth of liver cancer cells
To investigate whether JMJD2A promoted malignant growth of human liver cancer cell line Hep3B, we first screened two stable Hep3B cell lines transfectd with pCMV6-AC-GFP(GFP ctrl), pCMV6-AC-GFP-JMJD2A(JMJD2A) respectively. As shown in Figure 1Aa-1Ad, the expression of JMJD2A mRNA or protein was significantly increased in JMJD2A overexpressing Hep3B. As shown in Figure  1B, excessive JMJD2A significantly increased the growth Oncotarget 49096 www.impactjournals.com/oncotarget of liver cancer cell Hep3B compared to the control group (P < 0.01). Furthermore, we performed colony formation assay and observed a significant increase in colony formation efficiency rate in excessive JMJD2A compared to control(100 ± 0%% versus 33.07 ± 13.98%, P = 0.00711 < 0.01) ( Figure 1C). Moreover, JMJD2A overexpression significantly increased the BrdU positive rate compared to the control cells (33.25 ± 5.39% versus 78.91 ± 8.97%, p = 0.01477 < 0.05) ( Figure 1D). To explore the effect of JMJD2A on liver cancer cells in vivo, the two stable Hep3B were injected subcutaneously into Balb/C mice. As shown in Figure 2A, 2B, when JMJD2A was overexpressed, the xenograft tumor weight increased approximately 16.419 folds compared to the corresponding control group (0.006728 ± 0.01585 grams versus 0.11047 ± 0.142461 grams, P = 0.0228 < 0.05). Moreover, compared to control, xenograft tumors contained more of poorly differentiated cells in JMJD2A overexpression group ( Figure 2C). Taken together, these findings demonstrate that JMJD2A accelerates malignant progression of liver cancer cells.

JMJD2A enhances miR372 expression epigenetically
Given our previous study showed JMJD2A is positively associated with miR372 in human liver cancer tissues, we consider whether JMJD2A enhances miR372 expression. As shown in Figure 3A, JMJD2A was overexpressed in Hep3B cell line transfected with pCMV6-AC-GFP-JMJD2A. In the JMJD2A overexpressed Hep3B cell lines, the JMJD2A inhibited the interplay between H3K36me3 and miR372 promoter ( Figure 3B) and the interplay between DNMT1 and miR372 promoter ( Figure 3C). JMJD2A inhibited the methylation of miR372 promoter region (Figure 3Da and 3Db).Moreover, JMJD2A enhanced the CRE element luciferase activity ( Figure 3Ea) and the loading of CREB on the miR372 promoter region ( Figure 3Eb). Furthermore, JMJD2A enhanced the the loading of P300 and RNApolII on the miR372 promoter region ( Figure 3E, 3G). Intriguingly, JMJD2A promoted the formation of CTCF mediated promoter-enhancer DNA loop of miR372 and triggered CREB,P300,RNApolII into the DNA loop ( Figure 3H). Ultimately, pri-miR372,pre-miR372 and mature miR372 were significantly increased in JMJD2A overexpressing Hep3B compared to control group ( Figure 3I, 3J). Together, these observations suggest that JMJD2A promoted the expression and mature of miR372 epigenetically.

miR372 triggers a novel transcript (JMJD2AΔ) of JMJD2A
Given JMJD2A exists different transcripts, we wonder whether miR372 is involved in JMJD2A exon-intron editing. As shown in Figure 4A, a 39KD JMJD2A transcript, JMJD2AΔ, was significantly increased in JMJD2A or miR372 overexpressing Hep3B cell line. However, the JMJD2AΔ was disappeared in JMJD2A overexpressing plus miR372 knockdown Hep3B cell line. Moreover, JMJD2A decreased the loading of DNMT1 and increased the loading of TET1 on the left flank of 14th exon and on the right flank of 13th exon. However, miR372 inhibitor abrogated this JMJD2A action ( Figure 4B). Therefore, JMJD2A decreased the DNA methylation on the left flank of 14th exon and on the on the right flank of 13th exon. However, miR372 inhibitor abrogated this JMJD2A action by MSP PCR experiment (Figure 4Ca, 4Cb) or Dot blot assay (Figure 4Da, 4Db). Furthermore, JMJD2A decreased intron DNA looping within the 13th exon-14th exon by CTCF-3C-CHIP and decreased the CTCF,AgoI, HP1α in the DNA looping.However, miR372 inhibitor abrogated this JMJD2A action by CHIP ( Figure  4E). In assay in vitro, we found that JMJD2A decreased the interplay between CTCF, AgoI, HP1α and the probe of right flank of 13th exon. However, miR372 inhibitor abrogated this JMJD2A action by DNA pulldown ( Figure  4F). And excessive JMJD2A decreased the interplay between CTCF, AgoI, HP1α and the probe of left flank of 14th exon. However, miR372 inhibitor abrogated this JMJD2A action by CHIP ( Figure 4G). Informatics analysis suggests that JMJD2A contains 23 exons(1-23 exon), while JMJD2AΔ contains only contain 7 exons z(15-22) (Figure 4Ha). Therefore, we infer it is associated with exon-intron editing, e.g. 13th exon-introns-14th exon (Figure 4Hb). Taken together, miR372 blocks the editing of 13th exon-introns-14th exon of JMJD2A and a novel transcript( JMJD2AΔ) was formed in liver cancer cells.

JMJD2A promotes cell cycle progression via Pim1-ppRB1-CDK2-CycinE-C-myc pathway
Given that JMJD2A increased expression of Pim1 and decreased the methylation of Pim1 by inhibiting P21(WAF1/Cip1), we consider whether excessive JMJD2A may destroy the regulation and control of cell cycle proteins. As shown in Figure 7A, the interplay between CDK2 and CyclinE, CDK2 and ppRB1 was enhanced and the interplay between CDK2 and pRB1 was reduced in JMJD2A overexpressing Hep3B. However, Pim1 depletion fully abrogated the action of excessive JMJD2A that altered the interplay between CDK2 and CyclinE, CDK2 and ppRB1 and between CDK2 and pRB1 in Hep3B (Figure 7Ba, 7Bb). Moreover, the CDK2,CyclinE, and ppRB1 loading to C-myc promoter region was significantly increased in excessive JMJD2A group compared to control ( Figure 7C). Furthermore, the C-myc promoter luciferase activity was significantly increased in Hep3B transfected with pCMV6-AC-GFP-JMJD2A compared to control, however, Pim1 knockdown abrogated this action of JMJD2A ( Figure 7D). Ultimately, C-myc was significantly reduced in excessive JMJD2A group compared to control. However, the expression of C-myc were significantly not altered in Hep3B transfected with pCMV6-AC-GFP-JMJD2A plus pGFP-V-RS-Pim1 compared to control ( Figure 7E).Taken together, these observations suggest JMJD2A promotes cell cycle progression via Pim1-ppRB1-CDK2-CycinE-C-myc pathway.

DISCUSSION
Dysregulation of histone lysine methyltransferases is associated with carcinogenesis. Our studies indicated the effects of demethylase JMJD2A during  Each value was presented as mean ± standard error of the mean (SEM).**P < 0.01. (b) The appearance time of each tumor was determined for each mouse. Each value was presented as mean ± standard error of the mean (SEM).**P < 0.01. (c)A portion of each tumor was fixed in 4% paraformaldehyde and embedded in paraffin for PCNA staining (DAB stainning, original magnification×100).
To date, accumulating evidence indicates that JMJD2A results in a critical role in carcinogenesis and Oncotarget 49107 www.impactjournals.com/oncotarget cancer metastasis. Excessive JMJD2A produces strong oncogenic functions [31]. JMJD2A is involved in the regulation of cancer cell cycle [32]. In particular, JMJD2A induces site-specific copy gain and amplification in cancer cancer [33,34]. Herein, the functions of JMJD2A promotion of liver cancer stem cell growth is explained by results from two parallel sets of experiments: (1) JMJD2A is overexpressed and associated with miR372, P21(WAF1/ Cip1), Pim1 in human liver cancer tissues; (2) JMJD2A accelerates malignant progression of liver cancer cells in vitro and in vivo.
Of significance, our observations demonstrated that miR372 influences on JMJD2A functions in liver cancer cells.The increase of miR372 may partly decide the JMJD2A-medicated promotion of liver cancer cell growth. Our findings in this study provide novel evidence for an active role of miR372 in this action. This assertion is based on several observations: (1) JMJD2A enhances miR372 expression epigenetically. (2) miR372 prompts s a novel transcript( JMJD2AΔ) of JMJD2A, a 39KD JMJD2A transcript (3) miR372 blocks the editing of 13th exon-introns-14th exon of JMJD2A.
Strikingly, our data suggest that miR372 influence on the editing of JMJD2A, involving in CTCF,DNMT1,HP1α. In particular, a novel transcript, JMJD2AΔ was formed. Increase of JMJD2AΔ may partly contribute to JMJD2A-medicated promotion of liver cancer cell growth. This assertion is identified by several observations:(1)JMJD2A inhibited P21(WAF1/Cip1) and enhanced Pim1 in liver cancer cell cycle progress through JMJD2AΔ dependent on miR372.(2) JMJD2A promoted cell growth, colony formation ability and cell growth in vivo. However, both JMJD2A plus P21WAF1/Cip1 and JMJD2A plus Pim1RNAi abrogated the JMJD2A action. Pim1 and P21(WAF1/Cip1) determine JMJD2A oncogenic function. These findings are noteworthy, given that JMJD2AΔ is a and functions as a key oncogene mediatng various biological processes including cell proliferation, differentiation.
Studies show that CTCF has been found as a regulation factor of gene splicing [35,36].In particular, CTCF, the Argonaute protein AGO1, and the heterochromatin protein 1 (HP1) have been indicated in the process of gene splicing [37].Moreover, HP1 is associated with gene's alternative splicing [38].
Furthermore, our present results show that JMJD2A oncogenic action is associated with Cyclin-dependent kinase 2 (CDK2) and RB activity. It is well known that CDK2 regulates and controls cell cycle progression through several signaling pathway,e.g. EGF-ELK4/c-Fos pathway [49]. Of significance, specific cyclinE/ CDK2 inhibitors can block tumor progression [50]. Futhermore, defects in pRb tumor suppressor may lead to tumorigenesis [51].

In conclusions
The present study indicates a novel evidence for JMJD2A in tumorigenesis by upregulating miR372 in liver cancer cells, which may have potential therapeutic significance. Alteration of the expression of JMJD2A may also mediate changes to affect gene expression and contribute to hepatocarcinogenesis. JMJD2A knockdown in combination with blocking miR372 might represent a promising treatment strategy targeting tumors with overactivated JMJD2A.