RETRACTED ARTICLE: Ras-ERK1/2 signaling contributes to the development of colorectal cancer via regulating H3K9ac

Human colorectal cancer cell line SW48 purchased from American Type Culture Collection (Catalogue number: CCL-231™, ATCC, Manassas, VA, USA) was cultured in Dulbecco’s Modified Eagle’s Medium (DMEM, Gibco, Grand Island, NY, USA) supplemented with 10% heat-inactivated fetal bovine serum (FBS, Gibco). The cells were maintained at 37 °C in a humidified atmosphere with 5% CO₂.

MG132 (≥95% HPLC), an inhibitor of proteasome, was purchased from Sigma-Aldrich (St. Louis, MO, USA). SW48 cells were treated by 25 μM MG132 for 0–12 h. SCH772984, SB203580, LY294002 and SP600125, the specific inhibitors for ERK1/2, MAPK, PI3K and JNK pathways were all purchased from Selleck Chemicals (Houston, TX). These inhibitors were used with concentrations of 4 nM, 5 μM, 0.5 μM and 40 nM for treating cells for 12 h.

The coding regions of human wild-type K-Ras and H3 were amplified by PCR and were subcloned into pEGFP-N1 plasmid (Clontech Inc., Palo Alto, CA). The amplified PCAF and HDAC2 were subcloned into pcDNA6.0/HA-tag vector (Invitrogen, Carlsbad, CA, USA). The pEGFP-K-Ras^G12V/T35S construct was mutated using site-directed mutagenesis. The pEGFP-H3K9Q construct was constructed using the TaKaRa MutanBEST Kit (#D401, TaKaRa, Dalian, China). Two different sequences of siRNAs specific for HDAC2 (si-HDAC2–1 and si-HDAC2–2) were purchased from GenePharma (Shanghai, China). A non-targeting sequence was used as a negative control (si-con). MDM2-MU for expression of MDM2^C464A was constructed and recombination into pIB/V5-His Vector (Invitrogen).

SW48 cells were seeded in 6-well plates with a density of 1 × 10⁵ cells/well. When 50% confluence was researched, the cells were transfected with plasmids or siRNAs by using lipofectamine 3000 (Invitrogen). At 48 h of transfection, the culture medium was replaced by the complete medium to stop transfection. Transfection efficiency was confirmed by using Western blot and/or RT-qPCR.

The transfected SW48 cells were collected by using trypsin (Sigma-Aldrich) and were seeded in 96-well plates with a density of 5 × 10³ cells/well. After 48 h of incubation at 37 °C, 20 μL of MTT solution (Sigma-Aldrich) with a final concentration of 5 mg/mL was added into each well and the plates were incubated at 37 °C for another 4 h. Then, the culture medium was removed and 100 μL dimethyl sulfoxide (DMSO, Sigma-Aldrich) was added. Following 10 min of shaking in an ELISA reader (Bio-Rad Laboratories, Hercules, CA, USA), the absorbance of each well was recorded at 570 nm.

The transfected SW48 cells were collected and seeded in the upper side of 24-well transwell chamber with 8-μm pore filters (Costar, Boston, MA). The cells were maintained at serum-free culture medium. The lower side of the chamber was filled with 600 μL complete culture medium. After 12 h of incubation at 37 °C, the cells migrated into the lower side were fixed with methanol and stained with 0.5% crystal violet (Beyotime, Shanghai, China). The absorbance of cells that had been washed with acetic acid was measured at 570 nm.

Total RNAs were extracted from the transfected cells by using the TRIzol reagent (Invitrogen). Five micrograms of total RNA was subjected to reverse transcription using the Transcriptor First Strand cDNA Synthesis Kit (Roche, Basel, Switzerland). FastStart Universal SYBR Green Master (Roche) was used in qPCR and each qPCR was carried out in triplicate for a total of 20 μL reaction mixtures on ABI PRISM 7500 Real-time PCR System (Applied Biosystems, Foster City, CA, USA). GAPDH served as an internal control. Data were analyzed according to the classic 2^−ΔΔCt method.

Low melting agarose (Thermo Scientific®, Rockford, IL, USA) with concentration of 0.5% was placed in 6-well plates, and the plates were incubated at 4 °C for 30 min. The transfected SW48 cells were seeded in 6-well with a density of 600 cells/well, and were cultured in DMEM containing 0.33% agarose at 37 °C for 2 weeks. The number of the colonies was counted microscopically.

Total protein was extracted from the transfected SW48 cells by using 1% Triton X-100 (Invitrogen) and 1 mM PMSF (pH 7.4) (Solarbio, Beijing, China) over ice for 30 min. The protein concentration was determined by BCA protein assay kit (Novagen, Madison, WI, USA). Equal amount of protein samples were subjected to SDS-PAGE and proteins were transferred onto PVDF membranes (Millipore, Bedford, MA, USA). The membranes were blocked in 130 mM NaCl, 2.5 mM KCl, 10 mM Na₂HPO₄, 1.5 mM KH₂PO₄, 0.1% Tween-20 and 5% BSA (pH 7.4) for 1 h at room temperature. Then the membranes were probed by primary antibodies overnight at 4 °C. Anti-p-ERK1/2 (MA5–15173) and anti-PCAF (MA5–11186) were purchased from Invitrogen; anti-ERK1/2 (ab17942), anti-H3K9ac (ab4441), anti-H3 (ab12079), anti-HA (ab1424), anti-GFP (ab6556), anti-actin (ab8227), anti-His (ab197049), and anti-MDM2 (ab38618) were all purchased from Abcam (Cambridge, MA). Followed by incubation with the secondary antibody for 1 h at room temperature, the positive signal was detected by using enhanced chemiluminescence (ECL) and analyzed with ImageJ 1.49 (National Institutes of Health, Bethesda, MD, USA).

The transfected SW48 cells in 6-well plates were cultured in serum-deprived medium for 12 h to synchronize cells to G0-phase. Then, the cells were harvested by trypsinisation, washed twice with PBS and fixed in 70% ethanol at 4 °C overnight. The cells were re-suspended in the solution containing 0.2 mg/mL PI (Sigma-Aldrich), 0.1% Triton X-100 (Invitrogen), and 20 μg/mL RNase A (Roche) for 30 min at room temperature in the dark. The percentage of cells in the G0/G1, S and G2/M phases of the cell cycle were analyzed by flow cytometry (FACS Calibur, Becton Dickson, San Jose, CA, USA) and ModFit software (Verity Software House, Topsham, ME, USA).

The transfected SW48 cells (3 × 10⁶ cells per sample) were incubated in 1% formaldehyde for 10 min at room temperature, and the cells were collected and lysed in 200 μL SDS Lysis Buffer (Beyotime). After ultrasonication, the DNA was sheered to an average length of 200–800 bp. The samples were centrifuged at 10,000 g at 4 °C for 10 min, and the supernatant was probed by anti-H3K9ac (ab4441, Abcam) and anti-PCAF (MA5–11186, Invitrogen) at 4 °C overnight. The sample treated by anti-IgG (ab190475, Abcam) was used as a blank control. After incubation, 60 μL ProteinA Agarose/SalmonSperm DNA (Thermo Scientific®, Rockford, IL, USA) was added, and the samples were incubated at 4 °C for 2 h. The beads were washed sequentially for 10 min in low-salt wash buffer, high-salt wash buffer, LiCl wash buffer and TE buffer, as previously described [12]. Lastly, the beads were washed in 100 μL 10% SDS, 100 μL 1 M NaHCO₃, and 800 μL ddH₂O. 20 μL 5 M NaCl was added, and the crosslinks were reversed for 6 h at 65 °C. RT-qPCR was performed to analyze the amount of immunoprecipitated DNA and input DNA.

To examine whether H3K9ac can be modulated by Ras-ERK1/2 pathway, pEGFP-K-Ras^G12V/T35S was construct and transfected into SW48 cells. Figure 1a showed that phosphorylation levels of ERK1/2 were remarkably up-regulated in Ras^G12V/T35S group as compared to pEGFP group (transfected with an empty plasmid), indicating ERK1/2 pathway was activated by K-Ras mutated at G12 V and T35S. Then, the expression changes of H3K9ac were measured by performing Western blot analysis. Results in Fig. 1b and c displayed that, K-Ras mutated at G12 V and T35S significantly down-regulated H3K9ac expression (P < 0.01), but has no effects on H3 expression. These data suggested that Ras-ERK1/2 activation repressed the acetylation of H3 at lysine 9. In order to reveal whether H3K9 acetylation is specifically mediated by ERK1/2, four inhibitors specific for ERK1/2, MAPK, PI3K and JNK pathways were used, and the expression of H3K9ac was reassessed. As a result, we found that only the inhibitor of ERK1/2 (SCH772984) could recover H3K9ac expression following K-Ras mutation at G12 V and T35S (Fig. 1d and e). No such effects were observed in cells treated with the inhibitors specific for MAPK, PI3K and JNK, i.e., SB203580, LY294002 and SP600125. These findings suggested that H3K9 acetylation was specifically mediated by ERK1/2, rather than MAPK, PI3K and JNK.

Next, pEGFP-H3K9Q plasmid was constructed to mimic the acetylated H3K9, and the plasmid was co-transfected with pEGFP-K-Ras^G12V/T35S into SW48 cells. Transfection efficiency tested by western blotting revealed that H3K9ac expression was remarkably up-regulated by transfection with pEGFP-H3K9Q, in the presence and absence of pEGFP-K-Ras^G12V/T35S (Fig. 2a). MTT assay result showed that, co-transfection of cells with pEGFP-K-Ras^G12V/T35S and pEGFP-H3 significantly increased OD-value, compared to the transfection of pEGFP-H3 alone (P < 0.001, Fig. 2b). Of note, co-transfection of cells with pEGFP-K-Ras^G12V/T35S and pEGFP-H3K9Q attenuated Ras^G12V/T35S-induced evaluation of OD-value (P < 0.001). Same trends were observed in Fig. 2c and d, colony number and OD-value in migration assay were both significantly increased in Ras+H3 group compared to GFP + H3 group (P < 0.001). And they were both significantly decreased in Ras+H3K9Q group, as compared to Ras+H3 group (P < 0.01). Taken together, recovering the acetylation of H3K9 attenuated the promoting effects of Ras-ERK1/2 on tumor cells growth and migration.

Next, the involvement of H3K9ac in the transcription of Ras downstream genes was addressed. RT-qPCR data in Fig. 3a showed that the mRNA levels of CYR61 (P < 0.01), IGFBP3 (P < 0.01) and WNT16B (P < 0.05) were significantly up-regulated, while the mRNA levels of NT5E (P < 0.001), GDF15 (P < 0.01), and CDC14A (P < 0.01) were significantly down-regulated in Ras+H3 group, as compared to GFP + H3 group. However, the alteration of these mRNAs induced in Ras+H3 group were abolished in Ras+H3K9Q group. ChIP assay results in Fig. 3b showed that H3K9ac level was reduced at the promoters of these genes (P < 0.05, P < 0.01 or P < 0.001) following the activation of Ras-ERK1/2. Based on these data, we speculated that Ras-ERK1/2 mediated the transcription of its downstream genes also via regulating H3K9 acetylation.

Two different sequences of siRNAs specific for HDAC2 (si-HDAC2–1 and si-HDAC2–2) were transfected into SW48 cells. The mRNA levels of HDAC2 were remarkably silenced by siRNA transfection (Fig. 4a). Additionally, protein levels of H3K9ac were obviously up-regulated by siRNA transfection, even in K-Ras^G12V/T35S-expressing cells (Fig. 4b), suggesting H3K9 acetylation could be recovered by HDAC2 silence. Subsequently, cell viability, migration, and cell cycle progression were tested to see the effect of HDAC2 silence on SW48 cells growth and migration. Figure 4c-e demonstrated that both si-HDAC2–1 and si-HDAC2–2 remarkably attenuated the effects of Ras-ERK1/2 activation on SW48 cells viability (P < 0.01), migration (P < 0.001) and S-phase arrest. And also, the transcription of CYR61 (P < 0.01), IGFBP3 (P < 0.05), WNT16B (P < 0.05), NT5E (P < 0.001), GDF15 (P < 0.001), and CDC14A (P < 0.01) altered by Ras-ERK1/2 activation were attenuated by si-HDAC2–1 plus si-HDAC2–2 (Fig. 4f). These data suggested that silence of HDAC2 recovered the acetylation of H3K9, which in turn inhibited the growth and migratory capacities of SW48 cells.

In order to reveal how Ras-ERK1/2 repressed H3K9 acetylation, we focused on investigating PCAF, a reported upstream gene of H3K9ac [13]. Figure 5a displayed that mRNA levels of PCAF and HDAC2 were both unaffected by K-Ras^G12V/T35S. Figure 5b results indicated that the protein level of PCAF was down-regulated by K-Ras^G12V/T35S, but the protein level of HDAC2 was unaffected. These results implied that Ras-ERK1/2 post-transcriptionally down-regulated PCAF expression. This was also confirmed in Fig. 5c, that both PCAF and H3K9ac protein expression was repressed in Ras^G12V/T35S group. To further confirmed whether PCAF involved in the transcription of Ras downstream genes, ChIP was performed. Results from Fig. 5d showed that all of these genes that exhibited reduced H3K9ac following Ras-ERK1/2 activation also exhibited significant reduction in PCAF binding, suggesting PCAF was responsible for Ras-ERK1/2-repressed H3K9 acetylation.

Finally, SW48 cells were treated with MG132 (a proteasome inhibitor) to confirm whether PCAF regulated H3K9ac post-transcriptionally. Figure 5e showed that MG132 remarkably reversed the reduction of PCAF in K-Ras^G12V/T35S-expressing cells. Results in Fig. 5f showed that H3K9ac levels were down-regulated by Ras^G12V/T35S after 48 h of transfection in absence of MG132. However, treating cells with 25 μM MG132 gradually recovered the expression of H3K9ac (Fig. 5g). Thus, it is possible that Ras-ERK1/2 pathway repressed H3K9 acetylation through regulating PCAF.

It has been reported that the E3 ubiquitin ligase MDM2 could bind to acetylases, such as p300/CBP or PCAF [14]. Thus, we explored whether MDM2 was implicated in PCAF degradation in K-Ras^G12V/T35S-expressing cells. Figure 6a and b showed that PCAF expression was gradually repressed with MDM2 expression. However, in MDM2-mutant (MDM2-MU) transfected cells, no such down-regulations were observed in PCAF expression (Fig. 6c and d), indicating MDM2 was responsible for PCAF degradation.

Next, we established a link between H3K9ac expression and MDM2 activity to reveal whether MDM2-mediated PCAF degradation was required to modulate Ras-ERK1/2-repressed H3K9 acetylation. Figure 6e showed that, MDM2 was up-regulated, while H3K9ac was down-regulated in K-Ras^G12V/T35S-expressing cells. Thereafter, the expression of MDM2 was repressed in K-Ras^G12V/T35S-expressing cells by siRNA transfection (Fig. 6f). As a result, we found that silence of MDM2 resulted in an up-regulation of H3K9ac in K-Ras^G12V/T35S-expressing cells (Fig. 6g). Collectively, these data implied that Ras-ERK1/2 regulates H3K9ac via MDM2-mediated PCAF degradation.