Phosphorylation of Histone H3 by Protein Kinase C Signaling Plays a Critical Role in the Regulation of the Developmentally Important TBX2 Gene^*

Cell Culture and Reagents

WI-38 (ATCC CCL-75) normal human embryonic lung fibroblasts, their in vitro transformed counterparts WI-38 CT-1 (23) and SVWI-38 cell lines were maintained in Dulbecco's-modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum (FCS), 100 units/ml penicillin, and 100 μg/ml streptomycin. Cells were maintained at 37 °C in an atmosphere of 5% CO₂. For TPA treatment, we used 100 nm TPA (Sigma). The PKC inhibitor GF109203 (20 μm), MEK-1 inhibitor PD98095 (25 μm), p38 inhibitor SB203580 (10 μm), and JNK inhibitor SP600125 (20 μm) (Calbiochem) were added 1 h prior to treatment with TPA.

Plasmids, Transfections, and Reporter Assays

The human TBX2 promoter luciferase reporter construct was generated by inserting the SacI(-1606)-HindIII(+32) fragment of the human TBX2 gene into the appropriately cleaved luciferase reporter vector pGL3-basic (Promega). The pRL-TK vector (Promega) was used as an internal control reporter to test for transfection efficiency. The 4xAP-1-luc construct, pCMV-c-Jun and pCMV-JunB expression vectors were kindly provided by Dr. Michael Birrer (National Institutes of Health). Cells were plated at 1.5 × 10⁵ cells/ml in 6-well plates 1 day before transfection. Non-liposomal-mediated gene transfer was performed using FuGENE®6 (Roche Applied Science) according to manufacturer's instructions using 1 μg of DNA consisting of the reporter construct, expression vectors, and the internal control vector. Thirty hours after transfection, cells were analyzed for luciferase activity using the Dual-Luciferase® Reporter Assay (Promega) following manufacturer's instructions and quantified with a Luminoskan Ascent Luminometer (Thermo LabSystems).

Small Interfering RNA (siRNA)

Suppression of MSK1 cellular expression was achieved using siRNA that specifically targets MSK1 mRNA. WI-38 cells were transfected with 5 μm anti-MSK1 siRNA or control (non-silencing) siRNA (Qiagen) for 72 h, using HiPerfect reagent (Qiagen) according to the manufacturer's instructions. Cells were treated with TPA 8 h prior to harvesting protein for Western blotting.

Western Blotting

Western blot assays were carried out as described previously (24). Proteins were resolved on 8–12% SDS-polyacrylamide gels, as required and transferred to Hybond ECL (Amersham Biosciences). The membranes were probed with appropriate primary antibodies and detected using peroxidase-conjugated anti-mouse or anti-rabbit antibodies (1:5000) and visualized by ECL (Pierce). The primary antibodies used were: mouse monoclonal anti-Tbx2 62-2 antibody (1:2000), rabbit polyclonal anti-phospho-H3 (1:2000, Upstate Biotechnology), anti-phospho-p38 (1:1000), anti-p38 (1:5000), anti-phospho- MSK1 (1:1000), anti-phospho-ERK1/2 (1:1000), anti-phospho-JNK (1:1000) (Cell Signaling Technology Inc., Beverly, MA), goat polyclonal anti-MSK1 (1:1000), rabbit polyclonal anti-c-Jun, JunB, ERK2, Sp1 (1:500), and mouse monoclonal α-tubulin (1:500) from Santa Cruz Biotechnology. Densitometric analysis of Western blots was done with a Syngene scanner and GeneTool and GeneSnap software (Syngene).

Real-time RT-PCR

Total RNA was extracted from cultured cells and quantitative RT-PCR assays were carried out using the LightCycler as previously described (24). Primers used for RT-PCR were human TBX2: (forward) 5′-ATGGGCATGGGTCACCTACT-3′; (reverse) 5′-GGTGTAGGGGTATTTTAAGA-3′; glyceraldehyde-3-phosphate dehydrogenase (GAPDH): forward 5′-GAAGGCTGGGGCTCATTT-3′; reverse 5′-CAGGAGGCATTGCTGATGAT-3′ 3 PCR conditions were as follows: 95 °C for 10 min; and 45 cycles of 95 °C for 45 s (denaturation), 59 °C for 20 s (annealing), and 72 °C for 20 s (extension). The PCR products of TBX2 were cloned into the pGEM-T easy vector (Promega) and verified by sequencing.

DNA Affinity Immunoblotting (DAI)

Nuclear extracts were prepared from WI-38 cells as previously described (25). For DAI assays, biotinylated DNA probes were generated by PCR using primer pairs that were synthesized and 5′-end-labeled with biotin (Forward: 5′-TGGCCTGAGCTGTCAAAAC-3′; Reverse: 5′-GCGCGACTGGTTAGATCTTG-3′), and immobilized on Dynabeads Streptavidin (Dynal Invitrogen) according to the manufacturer's instructions. For each reaction, 40 μg of nuclear extract was incubated with 1 μg of DNA probes in binding buffer (20 mm Tris-HCl (pH 7.6), 50 mm NaCl, 1 mm MgCl_2, 0.2 mm EDTA, 0.5 mm dithiothreitol, 5% glycerol, and 2 μg of poly(dI-dC)) in a final volume of 400 μl. The beads were extensively washed with binding buffer and then boiled in 25 μl of 2× loading buffer before SDS-polyacrylamide gel electrophoresis followed by immunoblotting.

Chromatin Immunoprecipitation Assays (ChIP)

ChIP assays were carried out as previously described (13). Briefly, WI-38 cells (60–70% confluence) were serum-starved 24 h prior to treatment with TPA or vehicle. Cells were fixed in 1% formaldehyde after 8-h treatment. The chromatin was extracted, sonicated, and immunoprecipitated using antibodies against Sp1 (Santa Cruz Biotechnology) or phospho-histone H3 (Upstate Biotechnology). DNA precipitation was analyzed by PCR using human TBX2-specific primer pairs (Forward: 5′-TGGCCTGAGCTGTCAAAAC-3′; Reverse: 5′-GCGCGACTGGTTAGATCTTG-3′).

Induction of TBX2 Gene Expression by TPA in Human Fibroblasts

We have previously shown that TBX2 is expressed at high levels in the normal human lung fibroblast cell line, WI-38, but that it is expressed at very low levels in WI-38-transformed cell lines, CT-1 and SVWI-38 (24). To identify signaling pathways that regulate TBX2 expression, WI-38 cells were treated with TPA and TBX2 levels were examined by Western blot analysis. Our results showed that TPA significantly induced TBX2 expression in a time-dependent manner with levels increasing from 2 h and peaking at 8 h (Fig. 1A). Interestingly, TBX2 expression was strongly induced in the SVWI-38 and CT-1 cells treated with TPA for 8 h (Fig. 1B) suggesting that the TPA signaling pathway plays an important role in the regulation of TBX2 expression.

To determine whether the above induction of TBX2 expression by TPA is due to an increase in transcription, cells were treated with or without TPA for a period spanning 1 h to 24 h and TBX2 mRNA levels assessed by real-time PCR. Our results showed that TBX2 mRNA levels were up-regulated from 2 h with levels peaking at 8 h (Fig. 1C) in WI-38 cells. Similarly, TBX2 mRNA levels increased when SVWI-38 and CT-1 cells were treated with TPA for 8 h (Fig. 1D). TBX2 mRNA levels therefore closely correlated with levels of TBX2 protein, which suggested that TPA induced TBX2 expression at a transcriptional level.

Induction of TBX2 Gene Expression by TPA in WI-38 Cells Is PKC-dependent but MAPK-independent

It is well established that in response to TPA, PKCs can activate the Ras/Raf/MEK/ERK MAPK pathway. Thus to investigate the pathway(s) by which TPA up-regulates TBX2 gene expression in WI-38 cells, we first examined the effect of TPA on the phosphorylation status of MAPKs by Western blot analyses. Fig. 2A shows that while TPA treatment had no effect on JNK phosphorylation, it led to an increase in phosphorylation of ERK1/2 up to 8 h and to a marginal increase in p38 phosphorylation at the 4-h time point.

To assess the involvement of the above signaling molecules in TPA-induced expression of TBX2, we next pretreated WI-38 cells with inhibitors to PKC, ERK1/2, p38, and JNK before TPA treatment and then analyzed TBX2 expression by Western blotting. As expected, the PKC inhibitor (GF109203X) repressed TPA-induced TBX2 expression confirming that PKC is an important mediator of TBX2 regulation by the TPA pathway (Fig. 2B). It is important to note, that inhibiting PKC also resulted in reduced basal TBX2 expression, which implies that PKC may in fact be essential for TBX2 expression in WI-38 cells. Moreover, as shown in Fig. 2B, at 8 h of TPA treatment both ERK1/2 (PD98059) and p38 (SB203580) inhibitors had no effect on the induction of TBX2 expression but the JNK inhibitor (SP600125) blocked TPA-induced TBX2 expression. Similar results were also obtained at 4 h of TPA treatment (data not shown) which implicated the JNK pathway in the TPA-mediated induction of TBX2.

To determine the blocking efficiency and specificity of the above inhibitors, their effect on phosphorylation of their respective MAPKs were examined. Our results showed that PD98059 and SB203580 can effectively reduce phosphorylation of ERK1/2 and p38, respectively (Fig. 2C). Surprisingly, while TPA-mediated TBX2 expression was repressed by SP600125 (see Fig. 2B), TPA had no effect on JNK phosphorylation (Fig. 2, C and D). These results suggested that the effect of the SP600125 inhibitor on TBX2 expression was not due to the JNK pathway.

TPA and AP-1 Does Not Activate a TBX2 Gene Reporter Construct

It is well documented that TPA induces the expression of numerous genes through up-regulation of AP-1, and we therefore examined the expression of c-Jun and JunB in WI-38, SVWI-38, and CT-1 cells treated with or without TPA by Western blot analyses. Results showed that TPA increased both c-Jun and JunB protein levels in a time-dependent manner, which correlated with an up-regulation of TBX2 in all three cell lines tested (Fig. 3, A and B and see Fig. 1, A and B).

To determine whether AP-1 can indeed regulate the TBX2 promoter we searched the 3000 bp 5′-flanking regulatory sequence of the human TBX2 gene, which was obtained from the human genome sequence data base, for potential cis-acting elements using the TFSEARCH program. A putative AP-1 site was found at −1069 relative to the transcriptional start site (data not shown), and we thus performed luciferase assays using a reporter construct driven by a 1638-bp sequence of the TBX2 gene containing this putative site (26). TPA failed to stimulate transcriptional activity of the human TBX2 promoter in SVWI-38 and CT-1 cells (Fig. 3C), and overexpression of c-Jun or JunB did not significantly increase TBX2 promoter activity in HT1080 cells (Fig. 3D). As can been seen in Fig. 3, C and D, a reporter plasmid containing four AP-1 DNA binding sites fused to luciferase (4×AP-1-luciferase), which served as a positive control, was responsive to TPA, c-Jun, and JunB. These findings led us to hypothesize that either distal elements not present in the 1638 bp TBX2 regulatory fragment may be mediating the effects of TPA or that a chromatin remodeling mechanism may be involved in transcriptional regulation of TBX2 gene expression by TPA.

Phosphorylation of Histone H3-Ser-10 Is Required for Binding of Sp1 to the TBX2 Gene in WI-38 Cells

Several lines of evidence suggest that translational modification of histone H3 leads to the activation of genes in a Sp1 site-dependent manner (24, 27–29). We have previously demonstrated that the basal transcription of the human TBX2 gene is dependent on a Sp1 site in its proximal promoter (26). This led us to ask the question as to whether TPA induces TBX2 expression through phosphorylating histone H3 and consequently recruiting Sp1 to the Sp1 site identified above. Western blot analyses were therefore performed to examine the effect of TPA on the phosphorylation status of histone H3 in WI-38 cells. Our results show that TPA treatment resulted in an increase in the levels of phosphorylated histone H3 at Ser-10 for up to 8 h (Fig. 4A), which was blocked by the PKC (GF109203X) and JNK (SP600125) inhibitors (Fig. 4B). Phosphorylation of histone H3-Ser-10 by TPA, therefore, correlated well with the regulation of TBX2 expression by TPA.

We next examined whether TPA induced histone H3-Ser-10 phosphorylation at the TBX2 promoter using ChIP assays. Primers spanning the proximal promoter of TBX2 (Fig. 4C) were used for PCR amplification of immunoprecipitated DNA fragments and quantitative analysis was performed by real-time PCR. Compared with untreated cells, a ∼2-fold increase in histone H3-Ser-10 phosphorylation was obtained at the TBX2 promoter in WI-38 cells treated with TPA for 8 h (Fig. 4, C and D). To explore the possibility that TPA enhanced Sp1 binding to the TBX2 promoter, the same primers shown in Fig. 4C were used to amplify DNA fragments immunoprecipitated with an antibody specific to Sp1. A significant increase of Sp1 was observed at the TBX2 promoter in WI-38 cells treated with TPA (Fig. 4, C and D). Our results clearly show that TPA increased both histone H3-Ser-10 phosphorylation and Sp1 recruitment to a Sp1 site in the TBX2 proximal promoter, which correlated with the induction of TBX2 expression. These findings suggested that TPA-activated transcription of TBX2 through a mechanism involving the phosphorylation of histone H3 and consequent recruitment of Sp1 to the TBX2 proximal promoter.

PKC-dependent Phosphorylation of Histone H3-Ser-10 Correlated with Activation of MSK1

Whereas several kinases, including MSK1/2 (30, 31), ribosomal S6 kinase 2 (RSK2) (32), PKA (33), PKC (34), and IκB kinase-α (IKK-α) (35) have been proposed to phosphorylate histone H3 in response to various stimuli, current evidence suggests that MSK1 and 2 are the primary kinases involved (22). We, therefore, investigated whether MSK1 was activated in WI-38 cells treated with TPA by reprobing the Western blots shown in Fig. 4, A and B with an antibody to phosphorylated MSK1. An increase in MSK1 phosphorylation was evident from 1 h and persisted over a period of 24 h of TPA treatment (Fig. 4A) and as seen for histone H3-Ser-10, phosphorylation was abolished by the PKC (GF109203X) and JNK (SP600125) inhibitors. The result obtained with the JNK inhibitor is worth commenting on at this stage because while results shown in Fig. 2, A, C, and D indicated that TPA did not significantly activate JNK in our system, the JNK inhibitor, SP600125, blocked TPA-induced expression of TBX2 (Fig. 2B), and the phosphorylation of histone H3-Ser-10 and MSK1 (Fig. 4B). Interestingly, SP600125 has recently been identified to act as an inhibitor of histone H3-Ser-10 phosphorylation through an unknown mechanism (36). These findings suggested that SP600125 may repress phosphorylation of histone H3-Ser-10 through blocking MSK1 activation. Furthermore, these results raised the possibility that the TPA-induced TBX2 expression resulted from the phosphorylation of histone H3 by MSK1. To test this possibility, we knocked down MSK1 in WI-38 cells treated with or without TPA and examined the effect on phosphorylated histone H3-Ser-10 and TBX2 levels. Western blot analyses show that MSK1-specific siRNA results in substantial reduction in MSK1 levels since it was undetectable (Fig. 4E). As expected, cells transfected with control siRNA had increased levels of MSK1, phosphorylated histone H3-Ser-10, and TBX2. In contrast, these changes were abrogated in cells transfected with siMSK1. Our results thus confirm that in the absence of MSK1, TPA has no effect on levels of both phosphorylated histone H3-Ser-10 and TBX2. In summary, the above findings revealed a potential functional link between PKC and MSK1 activation during TPA-mediated phosphorylation of histone H3-Ser-10 and TBX2 expression.

Effect of TPA on Sp1 Expression and DNA Binding Activity

We have clearly shown here that TPA-induced phosphorylation of histone H3-Ser-10 was associated with recruitment of Sp1 to the human TBX2 gene. However, several lines of evidence have also shown that TPA-mediated expression of certain genes is associated with increased levels of total and/or phosphorylated Sp1 (37). To test whether the TPA-mediated increase in TBX2 gene expression is also due to the latter mechanism involving increased levels of Sp1 we next examined the expression pattern of Sp1 in WI-38 cells treated with or without TPA. Western blotting, however, revealed that TPA treatment had no significant effect on total (Fig. 5A) and nuclear (Fig. 5B) Sp1 protein levels at time points tested. In addition, Sp1 protein levels remained unchanged when cells were pretreated with the PKC, JNK, p38, and ERK inhibitors (Fig. 5C).

We finally explored the possibility that TPA treatment may enhance the affinity of Sp1 to bind to the proximal TBX2 promoter. Results from DNA affinity immunoblotting experiments showed that while the Sp1 site in the proximal TBX2 promoter was specifically bound by Sp1 (Fig. 5D, left), TPA treatment did not enhance this binding (Fig. 5D, right). Interestingly, an unlabeled probe was less able to compete with the labeled probe in nuclear extracts from TPA-treated cells, which suggests that TPA treatment may increase the stability of Sp1 binding (Fig. 5D, right).