The ablation of EZH2 uncovers its crucial role in rhabdomyosarcoma formation

RD cells fail to complete the differentiation program and show high levels of EZH2

In order to confirm the data showed in previous studies, we analyzed expressions of MyoD and two muscle-specific genes, myogenin and myosin heavy chain (Myh), in normal myoblasts (C2C12) and in RMS cells (RD) in both proliferative and differentiation conditions. Protein expression was examined by immunoblot. As shown in Figure 1A, MyoD dependent-genes myogenin and Myh exhibit significantly lower protein levels in RD, according to previous studies. Interestingly, expression of MyoD, the principal activator of myogenesis, shows very low levels if compared with normal myoblasts. In agreement with this result, after 96 h in differentiation medium (DM), RD cells fail cell cycle arrest and do not form myotubes (Fig. 1B).

Recent studies showed that expression of EZH2, the catalytic subunit of PRC2 complex, frequently elevated in several tumors,²⁵ is overexpressed in RMS cell lines and samples.³³ In order to confirm these data, we analyzed gene and protein expression in RD after myogenesis induction.

According to previous studies, protein levels of EZH2 were considerably higher in proliferating RMS cells, and its expression persist in differentiation conditions; contrariwise, normal myoblasts express low levels of EZH2 in all stages of differentiation process (Fig. 1A). Moreover, transcript analysis showed lack of transcriptional repression of EZH2 in RD cells, after differentiation induction, compared with C2C12 (Fig. 1C).

Previous studies demonstrated that EZH2 promotes a displacement of MyoD from its target genes in proliferating myoblasts up to muscle differentiation triggering. We hypothesized a similar mechanism in RMS, where a deregulated PRC2 complex would prevent MyoD binding at muscle specific genes. Nevertheless, this result indicates that the inhibition of myogenesis triggering is accounted for by a downregulation in MyoD expression.

EZH2 binds MyoD, myogenin and Myh promoter in RMS cells

The PcG proteins represent a global gene silencing system that plays a crucial role in organisms development by contributing to the maintenance of a stemness potential.³⁴ Furthermore, EZH2 levels are abnormally elevated in cancer tissues vs. corresponding normal ones.²⁵

In order to clarify the role of EZH2 in RMS formation, we investigated whether EZH2 is able to bind muscle-specific promoters in RD cells. Chromatin immunoprecipitation (ChIP) was performed using anti-EZH2 antibody. Enrichment of Myogenin and Myh promoter regions in immunoprecipitated RD DNA was evaluated by real-time PCR. EZH2 was detected in both promoter regions analyzed (Fig. 2A and B).

Furthermore, ChIP assay with anti-EZH2 antibody was performed in order to investigate a putative role of PRC2 in the inhibition of MyoD expression. Surprisingly, our results demonstrated that EZH2 is able to bind MyoD promoter, differently from current literature (Fig. 2C).

EZH2 inhibits the expression of genes involved in skeletal muscle differentiation in RMS

In order to evaluate the functional role of EZH2 in RMS cells, a stable EZH2 knockdown RD cell line was generated using a vector-based shRNA. EZH2 knockdown RD cells showed a specific transcript and protein reduction of 80% in MyoD compared with normal and shRNA scrambled-infected cells (Fig. 3A and B).

Therefore, all cell lines were grown in growth medium (GM) or induced to differentiate, replacing GM with differentiating medium (DM) for various time periods. As shown in Figure 3C, after 96 h in DM, EZH2 knockdown cells showed a partial recovery of the differentiated phenotype compared with wild type, and some myotube can be detected. In order to assess whether the reactivation of the myogenic program overlaps with a resumption of the expression of differentiation markers, transcripts and proteins levels were assessed every 24 h after induction of differentiation in both wild type and EZH2 knockdown cells. As expected, differentiation induction in the absence of EZH2 resulted in a significant increase of muscle specific factors (Fig. 4A and B). In order to confirm the role of EZH2 in downregulating muscle-specific genes, and, in particular, MyoD, wild type and EZH2 knockdown cells were transiently transfected with MyoD promoter luciferase reporter and induced to differentiate for 72h. EZH2 knockdown cells showed a greater MyoD promoter activation following differentiation promotion, confirming the data previously obtained (Fig. 4C). The partial reactivation of muscle-specific genes, in particular MyoD, strongly supports the role of EZH2 in myogenesis control and, if deregulated, in RMS formation.

Downregulation of EZH2 restores MyoD binding and complex activation to muscle-specific promoters

As discussed in the introduction, in normal myoblasts EZH2 is able to silence muscle-specific promoters. In order to investigate whether silencing of EZH2 would allow a restoration of MyoD binding to its target genes, we performed a ChIP assay using anti-MyoD antibody. As previously explained, cells were growth in GM and induced to differentiate in DM for 72 h. Enrichment of myogenin and Myh promoter regions in immunoprecipitated DNA was evaluated by real-time PCR. As expected, MyoD is unable to bind muscle-specific genes in RMS cells, whereas its binding is restored in EZH2 knockdown RD cells (Fig. 5A).

ChIP assay was performed evenly with anti-H3K27me3 antibody that confirmed a decrease of try-methylation of H3K27 in EZH2 knockdown cells, demonstrating the inhibition of EZH2 activity in knockdown cells (Fig. 5B).

MyoD is able to recruit numerous factors in order to activate transcription of its target genes, such as acetyltransferases p300 and PCAF, and pTEFIIb, which promotes transcriptional activation. In particular, pTEFIIb complex, by phosphorylation of RNApolII CTD, converts the pre-initiation form of RNApolII into the active form. Furthermore, recent studies demonstrated that UTX mediates demethylation of H3K27me3 at muscle-specific genes during myogenesis. The removal of the UTX-dependent-repressive H3K27me3 mark requires RNA PolII elongation form.³⁵

Therefore, we analyzed the phosphorylation of RNApolII CTD in RD wild type and EZH2 knockdown after 72 h of myogenesis induction. ChIP assays were performed using specific antibodies for phosphorylated forms of RNAPolII at Serine 2 and Serine 5. As control, chromatin immunoprecipitation using total RNApolII antibody was added.

As shown in Figure 6A and B, Ser2 and Ser5 phosphorylation of RNA Pol II at Myogenin and Myh promoters were significantly increased in the absence of EZH2. In contrast, RNA Pol II occupancy at the promoter regions analyzed did not show significant differences.

These data demonstrated the crucial role of EZH2 in muscle-specific gene inhibition and in RMS progression.

Cell culture and differentiation

Myoblasts (C2C12) were grown in DMEM supplemented with 20% FBS 1% L-glutamine and antibiotics (growth medium, GM). RMC cells (RD) were grown in DMEM supplemented with 10% FBS, 1% L-glutamine and antibiotics. The differentiation was induced by serum withdrawal in presence of 2% horse serum (differentiation medium, DM). All cell lines were obtained from ATCC. Pellets were collected every 24h for 96h.

Lentivirus production and infection

Knockdown cells were generated as described in www.broadinstitute.org/genome_bio/trc/publicProtocols.html. 
Briefly 293FT-packaging cells (Invitrogen) were seed at 1.5 × 10⁵ cells/ml (6 ml per plate) in low-antibiotic growth media (DMEM + 10% FBS) in 6 cm tissue culture plates.

Packaging cells were transfected with 3μg of Hairpin-pLKO.1 vector, shRNA for EZH2 (SHCLNG-NM_004456- Sigma-Aldrich) or scrambled shRNA (Addgene plasmid 17920), 0.9 μg of packaging plasmid psPAX2 (Addgene plasmid 12260), 0.1 μg of envelope plasmid pMD2G, (Addgene plasmid 12259) using 24 μl of FuGene HD (Promega). The media containing virus were filtered with 0.45 μm filter. Polyclonal populations of EZH2-knockdown and scrambled RD cells were generated by infection with 1 MOI (multiplicity of infectious) of shRNA lentiviral particles. After 3 d, the cells were selected with 2 μg /ml of puromycin for 1 week.

Plasmids, transient transfections and luciferase assay

The plasmid MyoD promoter-luc reporter was constructed as follow. RD cells DNA was extracted by DNeasy blood and tissue kit (Qiagen) following the manufacture’s protocol. PCR was performed with specific primers for MyoD promoter: 
For: 5′ -GGGTGGCTAGCAGGACTTTGCTGCCTCTGCCA-3′; Rev: 5′- GCAAGATCTCGATATAGCGGATGGCGTTGC-3′. 
The amplified product was cloned in NheI and BglII site of the pGL3 basic vector (Promega). The correct sequence of the construct was confirmed by sequencing.

Transient transfections were performed using FuGene HD (Promega). 2 μg of total DNA diluted in 100 μl Opti-MEM (CellGro) was incubated with 8 μl of FuGene HD for 20 min to let the transfection complex form. The transfection complex was added in a ratio 1:16 to the volume of the incubation medium of the cells (6 μl of transfection complex was added to 100 μl of complete medium in 96 wells plate).

Dual luciferase reporter assay (Promega) was used to measure the firefly luciferase and renilla luciferase activities within the transfected cells. Luciferase assay was conducted with MyoD promoter-luc reporter and RL-TK renilla. Both RD and EZH2-knockdown transfected cells were cultured in DM for 72 h. Luciferase activity was normalized to TK-directed Renilla expression, in a ratio 1:20 respect to firefly luciferase vector, to control the variability in transfection efficiencies. The assay was performed with Sirius Luminometer (Berthold detection systems).

The results are expressed in arbitrary units relative to the activity of the basic luciferase vector in RD wild type.

Total RNA extraction, cDNA synthesis and real-time PCR

Total RNA was extracted using High Pure RNA Isolation Kit (Roche). Retrotranscription was performed with 1 μg of RNA and random primers, using High Capacity cDNA Reverse Transcription Kit (Applied Biosystems), following the manufacture’s protocol. Real-time PCR was performed using FastStart Universal SYBR Green Master (ROX) (Roche) and primers 250 nM. Accumulation of fluorescent products was monitored using an Applied Biosystem 7,300 system (Applied Biosystems). Each data point was obtained from at least three independent experiments. Transcripts for glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were used as a reference. To ensure specific PCR amplification, every real-time PCR run was followed by a dissociation phase analysis (denaturation curve) and by gel electrophoresis. ΔΔCT method was used to calculate relative changes in gene expression; primer efficiency was calculated for every target using 5 × 10-fold serial dilutions of PCR products.

The primers used in Real-time PCR are listed below: EZH2: For: 5′-GCGGGACGAAGAATAATCATGG -3′, Rev: 5′-CCAAAATTTTCTGACGATTGGAAC-3′; GAPDH: For: 5′-GAAGGTGAAGGTCGGAGT-3′, Rev: 5′-CATGGGTGGAATCATATTGGA-3′; MyoD: For: 5′-GACGGCATGATGGACTACAGC-3′, Rev: 5′-GGAGATGCGCTCCACGATGC-3′; Myogenin: For: 5′-CCTGCTCAGCTCCCTCAACC-3′, Rev: 5′-AGGGTCAGCCGTGAGCAGATG-3′; MYH: For: 5′-GGAGGAGGCTGATGAACAAGC-3′, Rev: 5′ CTGGCTCACTCTTCACTCTCG-3′.

Immunoblotting

Cells were lysed in lysis buffer (20 mM TRIS-HCl pH 8; 137 mM NaCl; 10% glycerol 1% Nonidet P-40; 2 mM EDTA; Protease Inhibitor Cocktails).

The protein concentration was determined by Bradford assay (Biorad), following the manufacturer’s instructions and by using BSA as a standard.

The protein extract (50 μg) was resolved in 8% -10% SDS/PAA gel and transferred to a nitrocellulose membrane. The protein levels were detected with the followed antibody anti-MyoD, anti-Myogenin, anti-Myh (Santa Cruz Biotechnology), anti EZH2 (Invitrogen). Equal loading was controlled with anti-GAPDH and anti-Hsp70 (Santa Cruz Biotechnology).

Anti-mouse, rabbit (1:10000) peroxidase conjugated (Pierce) and ECL detection system (PerkinElmer) were used for detection.

Chromatin immunoprecipitation assay

Chromatin immunoprecipitation (XChIP) assays were performed using the ChIP-IT Express Enzymatic kit (Active Motif) according to the protocol provided by manufacturer. Briefly, cells were inducted to differentiate for 72 h in 15 cm plates and fixed with 1% formaldehyde for 5 min at room temperature. The nuclei of cells were enzymatically treated for 10 min at 37°C in digestion buffer. Ten μg of sheared chromatins were used for immunoprecipitation with normal rabbit/mouse IgG (negative controls) or anti-EZH2 (Millipore), anti-H3K27me3 (Active Motif), anti-MyoD, anti-RNApolII, (Santa Cruz Biotechnology), anti-phospho RNA polII (S2) and anti- phospho RNA polII (S5) (Bethyl) at 4°C with rotation overnight. DNA from ChIP was purified and analyzed by real-time PCR.

Real-time PCR was performed using the Applied Biosystem 7300 System with a FastStart Universal SYBR Green Master (ROX), (Roche). Each data point was obtained from three independent experiments. Input DNA (1/5 of total chromatin used for the IP reactions) was used as reference and the background signal as calibrator. Reported data represent real-time PCR amplification values obtained with the formula:
₂ ^{[ ( Ct background − Ct Input background) − ( Ct Ab − Ct Input Ab ) ]} 
where Ct is the threshold cycle, background is the normal rabbit IgG immunoprecipitated sample, and Ab is the EZH2, H3K27me3, MyoD, RNA polII, phospho RNA polymerases immmunoprecipitated sample.

To ensure specific PCR amplification, every real-time PCR run was followed by a dissociation phase analysis (denaturation curve).

The primers used in real-time PCR are listed below:
MyoD for: GTTCCTATTGGCCTCGGACTC, rev: CTGTCCGGCCTGATTTGTGGT; Myogenin for: GAGAGGGGAATGTGTCCTCC, rev: CTTCTTTCTCCCGCATGGCC; MYH: for: CTCGTCAGAAGCCGATTCTAC, rev: CCAACTTCTCACCTGAGAGTC;