Mitogen- and Stress-Activated Protein Kinase 1 Activity and Histone H3 Phosphorylation in Oncogene-Transformed Mouse Fibroblasts

The Ras–mitogen-activated protein kinase (MAPK) signaling pathway is often deregulated in cancer cells. Mutations in ras are found in about 30% of human cancers. Furthermore, defective or overexpressed cell surface receptors acting through this pathway, overexpressed ras genes and mutations in BRAF gene result in persistent activation of the Ras-MAPK pathway (1). Activation of the Ras-MAPK signaling pathway results in chromatin remodeling and expression of immediate-early genes (2, 3). One of the downstream events of this pathway is phosphorylation of histone H3 at Ser¹⁰. Mouse fibroblasts transformed with c-Ha-ras have a more decondensed chromatin structure and greater steady-state level of phosphorylated Ser¹⁰ H3 than that of parental cells (2, 4).

Phosphorylation of H3 at Ser¹⁰ occurs rapidly after stimulation of mammalian cells with agents such as epidermal growth factor (EGF), 12-O-tetradecanoylphorbol-13-acetate, or UVB (4, 5, 6). We and others (4, 7, 8) have demonstrated that the phosphorylated H3 in EGF- or 12-O-tetradecanoylphorbol-13-acetate–stimulated mouse fibroblasts is associated with immediate-early genes (c-fos, c-myc, and c-jun). Phosphorylation of the NH₂-terminal tail of H3 is thought to contribute to the disruption of immediate-early gene chromatin folding and intermolecular fiber-fiber interactions promoting the transcription of the gene. The H3 kinase responding to a stimulated Ras-MAPK pathway is mitogen- and stress-activated protein kinase-1 (MSK1). MSK1, which belongs to the AGC family of kinases and is related in structure to ribosomal p70 S6 kinase subfamily, is also activated through the p38/SAPK2 signaling pathway after activation by stressful stimuli (9).

The elevated steady-state level of phosphorylated Ser¹⁰ H3 in c-Ha-ras–transformed mouse fibroblasts may be a consequence of increased activity of the H3 kinase and/or decreased activity of the H3 phosphatase. Our previous studies demonstrated that H3 phosphatase activity, which was identified as protein phosphatase 1, was similar in Ciras-3 (c-Ha-ras–transformed 10T1 2) mouse fibroblasts and in the parental 10T1 2 cells, suggesting that the H3 kinase was responsible for the increased steady-state levels of phosphorylated H3.

In this study, we analyzed the H3 kinase activity in Ciras-3 and 10T1 2 cells and determined whether the H3 kinase activity was MSK1. Furthermore, we investigated the MSK1 protein level and the subcellular distribution of MSK1 in Ciras-3 and 10T1 2 mouse fibroblasts. Evidence is provided that the H3 kinase activity that is elevated in Ciras-3 cells is MSK1. Also, we demonstrate that the subcellular distribution and level of MSK1 protein are similar in Ciras-3 and 10T1 2 mouse fibroblasts.

The cell line Ciras-3 was derived from 10T1 2 cells by transfection with the T-24 c-Ha-ras oncogene (4). Cells were grown in plastic tissue culture plates at 37°C in a humidified atmosphere containing 5% CO₂ in α-MEM medium supplemented with 10% (v/v) fetal bovine serum, penicillin G (100 units/mL), streptomycin sulfate (100 μg/mL), and amphotericin B (250 ng/mL). The proportion of cells in the different cell cycle phases was determined by flow cytometry.

Cells were harvested and lysed in 400 μL of ice-cold Nonidet P40 buffer [150 mmol/L NaCl, 50 mmol/L Tris-HCl (pH 8.0), 0.5% Nonidet P40, 1 mmol/L phenylmethylsulfonyl fluoride, 1 mmol/L sodium orthovanadate, 1.0 mmol/L NaF, 1.0 μg/mL leupeptin, 1.0 μg/mL aprotinin, and 25 μmol/L β-glycerophosphate]. Cell extracts were subjected to centrifugation at 10,000 × g for 10 minutes at 4°C, and the supernatant was saved. Protein concentration of the supernatant was determined using the Bio-Rad Protein Assay as per manufacturer’s instructions (Bio-Rad Laboratories, Hercules, CA).

10T1 2 and Ciras-3 cells were lysed with Nonidet P40 buffer, and the insoluble material was removed by centrifugation for 10 minutes at 10,000 × g. Cell cycle–matched total cell extracts (10 μg) were incubated with 2 μg of H3-H4 tetramer fraction isolated from mature chicken erythrocytes as described previously (10), 10 mmol/L MgCl₂, 1 μmol/L microcystin-LR and ± 10 μmol/L H89 for 10 minutes at 4°C. Reactions were started with addition of 50 μmol/L ATP and 5 μCi of [γ-³²P]ATP (3,000 Ci/mmol/L) and incubation at 30°C for 30 minutes. Reactions were stopped with the addition of SDS-PAGE loading buffer and incubation on ice for 20 minutes. Furthermore, cell cycle–matched total cell extracts (500 μg) were incubated with 3.5 μg of anti-MSK1 antibody coupled to 10 μL of protein G-Sepharose at 4°C for 24 hours. The protein G beads were washed three times with 500 μL of buffer B and twice with buffer C (5). The beads were then resuspended in buffer C containing 10 mmol/L MgCl₂, 1 mmol/L phenylmethylsulfonyl fluoride, 1 mmol/L sodium orthovanadate, 1 mmol/L NaF, 1 μg/mL leupeptin, 10 μg/mL aprotinin, 25 μmol/L β-glycerophosphate, 1 μmol/L microcystin-LR, 5 μg H3-H4 tetramer, and ±10 μmol/L H89. Reactions were carried out as mentioned above. The H3-H4 tetramer and MSK1 immunoprecipitates were analyzed by SDS-15%-PAGE, and visualization and quantification of signals were analyzed by autoradiogram and phosphorimager analysis.

Fifty ng of purified active MSK1 (Upstate Biotechnology, Lake Placid, NY) was incubated with 1 μg of histone H3 (Roche Diagnostics, Mannheim, Germany) or H3-H4 fraction along with Mg²⁺/ATP solution and 5 μCi [γ-P³²]ATP at 30°C for 15 minutes. The reaction was stopped by the addition of SDS-PAGE loading buffer and incubation on ice for 10 minutes. The samples were resolved by SDS-15%-PAGE. The gel was stained with Coomassie Blue stain. Visualization of the phosphorylation signal was detected by phosphorimager analysis and autoradiography.

Cellular fractionations were carried out with a slight modification as described previously (11). In brief, 10T1 2 and Ciras-3 cells were resuspended in Tris-NaCl-Mg²⁺ buffer [100 mmol/L NaCl, 300 mmol/L sucrose, 10 mmol/L Tris-HCl (pH 7.4), 2 mmol/L MgCl₂, and 1% thiodiglycol] containing all of the inhibitors mentioned in the Nonidet P40 buffer. Lysis of cells was performed by passage through a syringe with a 22-gauge needle. The cytosol and nuclei were isolated from lysed cells by centrifugation at 6000 × g. Isolated nuclei were inspected by microscopic analyses. The nuclei were resuspended in TNM buffer, and nuclei extraction was performed by addition of Triton X-100 to a final concentration of 0.5% and incubation on ice for 5 minutes. After centrifugation at 6000 × g for 10 minutes, the supernatant, termed Triton X-100–soluble fraction, was saved. The nuclei pellet was resuspended in TNM buffer with 0.5% Triton X-100, and this fraction was termed Triton X-100–insoluble fraction.

Proteins were analyzed by SDS-(10% and 15%)-PAGE. The proteins analyzed by SDS-15%-PAGE were visualized by Coomassie Blue staining, and proteins analyzed by SDS-10%-PAGE were visualized by transfer to nitrocellulose membrane and immunochemical staining with various antibodies.

Anti-phospho-p44/p42 MAPK, anti-phospho-p38, and anti-p38 rabbit polyclonal antibodies were purchased from Cell Signaling Technology (Beverly, MA). Anti-extracellular signal-regulated kinase (ERK; sc-93-G) goat polyclonal antibody and anti-Sp3 rabbit polyclonal antibody were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-MSK1 rabbit polyclonal antibody used in immunoblotting was obtained from Sigma (St. Louis, MO). Anti-MSK1 sheep polyclonal antibody used in immunoprecipitation was purchased from Upstate Biotechnology. Anti-HDAC1 rabbit polyclonal antibody was from Affinity Bioreagents, Inc. (Golden, CO).

Throughout this study we used cell cycle–matched Ciras-3 and 10T1 2 cells. Table 1 shows that the majority of the 10T1 2 and Ciras-3 cells were in G₁ phase of the cell cycle. Expression of the c-Ha-ras oncogenes in Ciras-3 mouse fibroblasts is thought to result in the persistent stimulation of the Ras-MAPK pathway. The relative activities of the Ras-MAPK pathways in these cells were evaluated by immunoblot analyses of ERK1 and 2 and their phosphorylated isoforms. Fig. 1 shows that steady state of activated phosphorylated ERKs was greater (about 6-fold) in Ciras-3 cells. Overexpression of c-Ha-ras has also been reported to activate the p38 pathway (12). However, phosphorylated (activated) p38 was not observed in immunoblot analyses of cell extracts from cell cycle–matched parental and oncogene-transformed cells (data not shown). These data show that the steady-state activities of ERK1 and ERK2 are greater in the Ciras-3 than in the parental 10T1 2 cell line.

Because the relative activity of the H3 phosphatase, protein phosphatase 1, was similar in Ciras-3 and 10T1 2, we surmised that an increased H3 kinase activity in the Ciras-3 cells would account for the increased levels of phosphorylated Ser¹⁰ H3. In vitro H3 kinase assays with equal amounts of cell extracted protein from cell cycle–matched cells were performed. Fig. 2 shows that H3 kinase activity in Ciras-3 cell extract was greater than that from the 10T1 2 cells. We observed an increase of 3-fold in H3 kinase activity in the Ciras-3 cell extracts (average of three separate preparations). Histone H4 was also weakly labeled in this assay. As a control, we did not observe H3 kinase activity when either the histone substrate or cell extract was absent.

H89 is a potent inhibitor of MSK1 (5). To determine whether the H3 kinase activity was due to MSK1, the Ciras-3 and 10T1 2 cell extracts were preincubated with H89. Presence of the kinase inhibitor significantly reduced the H3 kinase activity in both cell extracts. The remaining H3 kinase activity may be Aurora B, the mitotic H3 kinase. We found that in contrast to MSK1, Aurora B was relatively insensitive to H89 inhibition in in vitro kinase assays (data not shown).

The results shown in Fig. 2 presented evidence that MSK1 activity was greater in Ciras-3 than in 10T1 2 cells. However, H89 is also a potent inhibitor of protein kinase A, which may also phosphorylate H3 (13, 14). To directly test whether the MSK1 H3 kinase activity was greater in Ciras-3 cells, MSK1 was immunoprecipitated from the cell extracts and assayed for H3 kinase activity. Fig. 3 shows that the MSK1 activity was greater in the immunoprecipitated Ciras-3 fraction than that from the 10T1 2 cell extract. The H3 kinase activity of MSK1 in both preparations was inhibited by H89. In addition to radiolabeling of H3, weak labeling of H4 was observed, and H89 suppressed the labeling of this histone. In control experiments when the primary antibody was not included or the immunoprecipitate was not added, H3 kinase activity was not detected. Furthermore, in control experiments, we incubated purified histone H3 with a commercial preparation of MSK1 to show that MSK1 radiolabeled purified H3 (Fig. 3, right panel).

The increased MSK1 activity observed in the Ciras-3 sample was not due to an increase in the amount of MSK1 protein immunoprecipitated from the Ciras-3 cell extract, as the stained gel shown in Fig. 3 (top panel) revealed that the MSK1 immunoprecipitates from the cell extracts had similar amounts of MSK1. In repeats of these analyses (n = 4), we observed an average 3-fold increase in the MSK1 activity in the Ciras-3 immunoprecipitates relative to that in the 10T1 2 preparations.

Constitutive activation of the Ras-MAPK pathway results in the altered expression at the protein level of signaling and cell cycle proteins (1). The previous immunoprecipitation analyses shown in Fig. 3 provided evidence that MSK1 protein levels were similar in Ciras-3 and 10T1 2 cell extracts. To explore this further, we compared the levels of MSK1 with ERKs by immunoblot analyses of cellular extracts from Ciras-3 and 10T1 2 cells. Fig. 4 shows that the protein level of MSK1 in the two cell lines were equivalent in accord with the immunoprecipitation results. Together these data demonstrate that the increased phosphorylation of H3 observed in ras-transformed mouse fibroblasts is due to an increase in the activity, but not protein level of MSK1.

MSK1 is located in the nucleus and cytoplasm. Nuclear MSK1 would be responsible for phosphorylating H3 associated with immediate-early genes, whereas MSK1 located in the cytoplasm has an interesting role in translation by phosphorylating 4E-BP1 (15). To determine whether MSK1 subcellular distribution was altered in the ras-transformed mouse fibroblasts, cell fractions were analyzed by immunoblotting. Cells were lysed in TNM buffer without any detergents to minimize loosely bound nuclear proteins from leaking out of the nuclei. The nuclei were then resuspended in TNM buffer with 0.5% Triton X-100 and incubated on ice to release loosely bound nuclear proteins (Triton S fraction). The resulting pellet contained the tightly bound nuclear proteins, which includes proteins associated with the nuclear matrix (Triton P). As a reference, we compared the distribution of MSK1 with the nuclear transcription factor Sp3 and chromatin remodeling enzyme, histone deacetylase 1 (HDAC1). Sp3 has three isoforms, the expressions of which are regulated at the level of translation by selection of different translation initiation sites on the Sp3 mRNA. A consistent observation in analyses of these cell fractions was that the relative level of the short Sp3 isoforms compared with the long isoform was greater in the Ciras-3 cells (Fig. 5). This observation suggests that the translational machinery is altered in the ras-transformed cells. Fig. 5 shows the distribution of MSK1 among the various cellular fractions was similar for Ciras-3 and 10T1 2 cells. In contrast to Sp3 and HDAC1, MSK1 was present in the cytosol fraction. However, most MSK1 was located in the nuclear fractions in Ciras-3 and 10T1 2 cells. Also dissimilar from Sp3 and HDAC1, most, if not all, MSK1 was extracted from the nucleus with 0.5% Triton X-100.

Constitutive activation of the Ras-Raf-MAP/ERK kinase-ERK pathway in c-Ha-ras-transformed mouse fibroblasts increases the steady-state level of phosphorylated Ser¹⁰ H3. Here, we show that the activity of a downstream target of this signal transduction pathway, the H3 kinase MSK1, is increased in the ras-transformed cells. Constitutive activation of the Ras-MAPK pathway did not alter the protein level of MSK1 or the subcellular distribution of the enzyme. Because the level of the H3 phosphatase is not altered in the ras-transformed cells, the net result of this constitutively activated pathway is an increase in the steady-state level of phosphorylated H3.

In our kinase assays with H3 and H4 and immunoprecipitated MSK1, we observed a low level of H4 radiolabeling. There was the possibility that immunoprecipitated MSK1 was associated with another kinase that phosphorylated H4. However, we observed that purified MSK1, when incubated with H3 and H4, would also weakly label H4, with H3 being the preferred substrate. MSK1 phosphorylates H3 at the site RKS. H4 has the sequence RIS⁴⁷, which may be the site phosphorylated albeit weakly by MSK1. Because this portion of the H4 molecule is in the histone fold, which is in the interior of the nucleosome, it is likely that H4 Ser⁴⁷ would be inaccessible to MSK1 in chromatin.

Phosphorylated H3 in cycling Ciras-3 cells and in 12-O-tetradecanoylphorbol-13-acetate– or EGF-stimulated 10T1 2 mouse fibroblasts is associated with relaxed chromatin regions that are located in specific nuclear locations (4). In chromatin immunoprecipitation assays we and others (4, 7, 8) have provided direct evidence that the induced phosphorylated Ser¹⁰ H3 is bound to the promoter and coding regions of immediate-early genes c-fos, c-myc, and c-jun. Pretreatment of the Ciras-3 and 10T1 2 cells with the potent MSK1 inhibitor H89 reduced or prevented the transiently 12-O-tetradecanoylphorbol-13-acetate–induced expression of immediate-early genes such as c-fos and urokinase plasminogen activator, suggesting that phosphorylation of H3 contributed to the induced transcription of these genes (10). Similar results were obtained with MSK1/2 knock out mouse primary embryonic fibroblasts (16). The elevated activity of MSK1 in the ras-transformed cells may be one of several deregulated chromatin modifying enzymes that through their action of remodeling chromatin lead to aberrant expression of genes in the transformed cells.

MSK1 is located primarily in the nucleus and to a lesser extent in the cytoplasm of Ciras-3 and 10T1 2 mouse fibroblasts. Most, if not all, MSK1 is extracted from the nuclei with 0.5% Triton X-100. Thus, the majority of MSK1 is loosely bound to chromatin and other nuclear substructures. This observation suggested that MSK1 was not associated with the nuclear matrix, a conclusion that was confirmed by demonstrating that nuclear matrices did not retain MSK1.¹ MSK1 subnuclear location is in contrast to other chromatin modifying enzymes, such as histone acetyltransferases (CBP and PCAF) and histone deacetylases (HDAC1 and HDAC2), which are tightly bound in the nucleus and associated with the nuclear matrix (11).

Mutations in ras are found at high frequency in different types of cancer including adenocarcinomas of the pancreas, colon, and lung. Additional mutations or overexpression of EGF receptors and HER-2/neu receptors that signal through Ras to elevate the activity of the Ras-MAPK pathway are relatively common in breast cancer. Mutations such as those in the BRAF gene and oncoproteins (e.g., c-Src) observed in melanoma and breast cancer will also activate the Ras-MAPK pathway. The wide-spread involvement of the Ras-MAPK pathway in multiple cancer sites suggests that the activation of the H3 kinase MSK1 may be a frequent alteration in neoplasia and, considering the enzymes role in chromatin remodeling, may be a worthy target for therapeutic intervention.