Members of the hSWI/SNF chromatin remodeling complex associate with and are phosphorylated by protein kinase B/Akt

Abstract

In an adenosine triphosphate (ATP)-dependent process, the hSWI/SNF chromatin remodeling complex functions to alter chromatin structure, thereby regulating transcription factor access to DNA. In addition to interactions with transcription factors and recognition of acetylated histone residues, the chromatin remodeling activity of hSWI/SNF has also been shown to respond to a variety of cell signaling pathways. Our results demonstrate a novel interaction between the serine/threonine kinase Akt and members of the hSWI/SNF chromatin remodeling complex. Activation of Akt in HeLa cells resulted in its association with hSWI/SNF subunits: INI1, BAF155 and BAF170, as well as actin. BAF155 became preferentially recognized by an antibody that detects phosphorylated Akt substrates upon activation of Akt, suggesting that BAF155 may be an in vivo target for phosphorylation by Akt. Glutathione-S-transferase (GST) pulldown experiments demonstrated that INI1 and BAF155 were both capable of directly interacting with Akt. Finally, in vitro kinase assays provided additional evidence that BAF155 and potentially INI1 are substrates for Akt phosphorylation. These data provide the first evidence that Akt signaling may modulate function of the hSWI/SNF complex.

Introduction

The hSWI/SNF complex is an important contributor to chromatin remodeling, involved in the activation and repression of target gene transcription. Each complex contains four core subunits (BRG1/hBrm, BAF155, BAF170 and INI1) and a variable number of additional subunits (Phelan et al., 1999). BRG1 and hBrm are the catalytic ATPase subunits and each complex contains one of the two. The human homologs of the yeast SWI3 protein, BAF155 and BAF170, are 62% identical to each other at the protein level (Wang et al., 1996). Loss of the subunit INI1, also known as BAF47, hSNF5 and SMARCB1 (SWI/SNF-related matrix associated actin dependent regulator of chromatin B1) is a recurrent genetic characteristic of malignant rhabdoid tumor (MRT), a rare and aggressive pediatric cancer (Versteege et al., 1998; Biegel et al., 1999). Reintroduction of INI1 into MRT cell lines has been shown to cause cell cycle arrest at the G0–G1 stage and flat-cell morphology (Betz et al., 2002; Versteege et al., 2002; Zhang et al., 2002; Reincke et al., 2003). Mouse knockout experiments have also demonstrated that while homozygous loss of INI1 results in early embryonic lethality, heterozygous loss predisposes mice to tumors consistent with MRT (Klochendler-Yeivin et al., 2000; Roberts et al., 2000; Guidi et al., 2001). Finally, actin and actin-related proteins have been found in stoichiometric amounts in hSWI/SNF complexes and are thought to facilitate association with the nuclear matrix (Zhao et al., 1998; Rando et al., 2002).

The chromatin remodeling ability of hSWI/SNF has been shown to respond to various signaling pathways, including response to phosphoinositols (Zhao et al., 1998; Rando et al., 2002), the p38 pathway during skeletal myogenesis (Simone et al., 2004), interferons (Agalioti et al., 2000; Huang et al., 2002; Liu et al., 2002; Cui et al., 2004) and nuclear hormone receptors (Fryer and Archer, 1998; DiRenzo et al., 2000; Belandia et al., 2002; Métivier et al., 2003; Link et al., 2005). Phosphatidyl inositol 4,5-bisphosphate (PIP2) has been shown to help increase the association of the SWI/SNF complex with the nuclear matrix (Zhao et al., 1998; Rando et al., 2002). Phosphatidyl inositol 4,5-bisphosphate is also involved in the activation of the serine/threonine kinase Akt, by recruiting it to the plasma membrane, where it is activated via phosphorylation. In addition, INI1 was recently shown to interact with the catalytic subunit of protein phosphatase 1 (Wu et al., 2002), which has a multitude of substrates, including Akt (Xu et al., 2003). These observations led us to question whether a potential relationship also existed between Akt and the hSWI/SNF complex.

Akt, also referred to as protein kinase B (PKB), plays an important role in signaling pathways related to cell survival. It is a downstream effector of phosphoinositide-3-kinase (PI3K) and its dysregulation has been demonstrated in many types of cancer. Activated Akt protects cells from apoptosis by phosphorylating substrates such as proapoptotic enzymes, caspase-9 and BAD (BCL2-antagonist of cell death) resulting in their inactivation, as well as phosphorylating forkhead-related transcription factors, preventing the transcription of proapoptotic genes (Datta et al., 1997; Franke, 1999; Franke et al., 2003 and references therein).

Here we have used a combination of experimental approaches to examine potential interactions between Akt and two of the core hSWI/SNF subunits: INI1 and BAF155. Results demonstrate that active Akt is able to associate with INI1 and BAF155 both in vivo and in vitro. Also, BAF155, at least, appears to be a target of Akt phosphorylation. These results suggest that signaling through Akt may modulate chromatin remodeling in mammalian cells.

Results

Akt and SWI/SNF components interact in cells

To determine whether hSWI/SNF subunits associate with Akt in cells, a series of immunoprecipitation (IP) experiments was performed. In preliminary experiments, a HeLa cell line stably expressing Flag-tagged INI1 at levels approximating endogenous INI1 expression was established (called Flag (FL)-INI1 cells). To modulate Akt activation state, cells were grown continuously in the presence of serum (randomly cycling) or were serum-starved (SS), and were fractionated into nuclear and cytoplasmic fractions. Western blot analysis using an antibody that detects Akt phosphorylated on S473 demonstrated that Akt was unphosphorylated, and therefore inactive, under SS conditions (Figure 1a). Both the parental HeLa and the FL-INI1 cells had phospho-Akt in both the nuclear and cytoplasmic compartments when grown in the presence of serum. In this experiment, levels of phosphorylated Akt in the FL-INI1 randomly cycling cells appeared to be lower than in the parental HeLa cell line; however, this was not a consistently observed phenomenon as minor variations were seen between experiments. Western blot analysis also revealed that, as expected, BAF155 and the majority of FL-INI1 were localized in the nucleus, although some FL-INI1 was present in the cytoplasm. In addition, serum starvation had no obvious impact on the expression levels of either protein. Subsequently, anti-Flag M2 resin was used to immunoprecipitate Flag-tagged INI1 from the randomly cycling and SS FL-INI1 cell lysates using the parental, randomly cycling HeLa cell lysates as a negative control. The immunoprecipitates were examined for the presence of FL-INI1, BAF155, Akt and actin (Figure 1b). BAF155 co-immunoprecipitated with FL-INI1 under both conditions. Although the two proteins were expected to co-immunoprecipitate given that they are both known members of the hSWI/SNF complex, this result further indicated that the complex is stable in SS cells. Akt was also present in the FL-INI1 immunoprecipitates; Akt association with FL-INI1 was greater in extracts from randomly cycling cells. This result demonstrated a potentially novel interaction between Akt and the hSWI/SNF complex, and also indicated that Akt may need to be active for this interaction to occur. In addition, actin also preferentially associated with FL-INI1 in extracts prepared from randomly cycling cells, suggesting that under SS conditions a number of regulators of the hSWI/SNF complex may disengage or that the association of actin and Akt is in some way co-dependent.

Figure 1
figure1

Modulation of Akt phosphorylation affects complex formation. (a) Serum starvation reduces Akt activity in HeLa cells. HeLa (HL) and FL-INI1 cells were grown continuously in the presence of serum (RC for randomly cycling), or the FL-INI1 cells were serum-starved (SS) and fractionated into nuclear (N) and cytoplasmic (C) fractions. Duplicate aliquots containing 20 μg of protein from each fraction were separated by 10% sodium dodecyl sulfate (SDS)–PAGE and Western blots were performed on one membrane probed consecutively with anti-Flag, anti-BAF155 and anti-tubulin antibodies, with washes in between. The other membrane was probed with anti-phospho-Akt. The tubulin signal indicated that the randomly cycling cells contained some cytoplasmic contamination in the nuclear fraction, but the amount is not enough to account for the presence of phospho-Akt in the nuclear extract. (b) INI1 interacts with Akt and actin preferentially in randomly cycling cells and interacts with BAF155 despite serum starvation. In all, 300 μg of whole-cell extract from HeLa RC and FL-INI1 RC and SS cells was immunoprecipitated using an anti-Flag M2 affinity gel. One-third of the immunoprecipitated material and 30 μg of control lysates were fractionated by 10% SDS–PAGE and transferred to polyvinylidene difluoride (PVDF) membrane. The membrane was cut into strips and the top portion probed with anti-BAF155 antibody. The bottom portion was probed consecutively with anti-Flag, anti-Akt1/2 and anti-actin, with washes in between. The bracket indicates the location of the visible heavy chain in immunoprecipitation (IP) reactions. Similar results were obtained in three independent experiments.

To confirm these results, interactions were also examined using anti-Akt resin to immunoprecipitate complexes. The model system of inducible Akt activation was again used and immunoprecipitates from randomly cycling or SS FL-INI1 cells were analysed using a panel of Western blots (Figure 2, lanes 4 and 5). Akt was immunoprecipitated under both conditions; however, in accordance with our previous results, Akt interactions with the hSWI/SNF subunits were only detected when Akt was active.

Figure 2
figure2

Only active Akt interacts with INI1 and BAF155 and requires the presence of INI1 for interactions. In all, 100 μg of nuclear extract from RC and serum-starved (SS) FL-INI1 cells and 300 μg of nuclear extract from the malignant rhabdoid tumor (MRT) cell line TTC549 was immunoprecipitated with an agarose-conjugated anti-Akt1 affinity resin. 10% sodium dodecyl sulfate (SDS)–PAGE was used to separate immunoprecipitation (IP) reactions and 10 μg control nuclear lysates. Western blotting was then performed by cutting the membrane into strips. The top portion was probed with anti-BAF155 and anti-BAF170 antibodies, with washes in between. The bottom portion was probed consecutively with anti-Akt1/2, anti-Flag, anti-INI1 antibodies and anti-actin, with washes in between. The bracket indicates the location of the visible heavy chain in IP reactions. The asterisk designates the Flag-tagged INI1. Similar results were obtained in at least two independent experiments.

Akt was also immunoprecipitated from a MRT cell line to determine if it is able to interact with BAF155, BAF170 and actin in the absence of INI1. These cells contain high levels of total and phosphorylated Akt compared to HeLa cells (Figure 2 and data not shown). No interaction between Akt and hSWI/SNF subunits could be detected, even using three times the amount of MRT lysate for the IP than is sufficient for FL-INI1 cells (Figure 2, lane 6). This suggested that an interaction between Akt and hSWI/SNF subunits in cells depends on the presence of INI1.

BAF155 is a substrate for Akt phosphorylation

The presence of an interaction between Akt and members of the hSWI/SNF complex raised the possibility that hSWI/SNF subunits were substrates for phosphorylation by Akt. Therefore, the sequences of hSWI/SNF component proteins were examined using the NetPhosK1.0 server for potential cellular kinase recognition sites (Blom et al., 2004). Although many potential sites for recognition by various kinases were present on the subunits, no Akt consensus sequences were found on BRG1, hBrm, BAF60, BAF57, or on INI1. However, BAF170 contained one Akt consensus site and three Akt consensus sites were present on BAF155 (Figure 4c). To investigate whether Akt can phosphorylate BAF155 in vivo, the status of BAF155 phosphorylation was examined by immunoprecipitating it from randomly cycling and SS FL-INI1 cells. Immunoprecipitates were probed with an antibody that detects phosphorylated Akt substrates. The results (Figure 3) demonstrate that the Akt substrate antibody recognized a strong signal for BAF155 in randomly cycling cells (lane 1), whereas the signal in SS cells was greatly reduced (lane 3). Probing the immunoprecipitates with an anti-Flag antibody revealed a BAF155–INI1 interaction under both conditions, confirming the results of Figure 1b. Similar experiments have been attempted to examine INI1 phosphorylation; however, they were inconclusive because FL-INI1 and the heavy chain of the immunoprecipitating antibody are poorly resolved by sodium dodecyl sulfate (SDS)–PAGE.

Figure 4
figure4

Akt and hSWI/SNF subunits interact directly in vitro. (a) INI1 and BAF155 interact directly with Akt. Active, purified, full-length human recombinant Akt1 (Upstate) was incubated with glutathione beads containing glutathione-S-transferase alone (GST), truncations of GST-BAF155 at residues 202 and 334, full-length GST-INI1 (FL), and truncations of GST-INI1 at residues 131 and 246. Bound Akt was fractionated by 10% sodium dodecyl sulfate (SDS)–PAGE and visualized by Western blotting using anti-Akt1/2 and anti-phospho (ser473) Akt antibodies. Similar results were obtained in at least two independent experiments. (b) BAF155 interacts directly with INI1. Full-length 35S-labeled INI1 was incubated with glutathione beads containing GST alone (GST) or truncations of GST-BAF155 at residues 202 and 334. Bound INI1 was fractionated by 10% SDS–PAGE and the gel was dried and visualized on a PhosphorImager. Similar results were obtained in two independent experiments. ‘IN’ is the input lane and represents 100% of the protein added to each reaction for Akt and INI1. (c) Schematic of BAF155 and INI1 proteins. Full-length BAF155 is represented and GST-BAF155 truncations are depicted above, numbers indicate amino-acid residues. Potential Akt consensus sites are marked and the sequences are shown below with putative phosphorylated residues underlined. Full-length INI1 is also represented with GST-INI1 truncations depicted above. For INI1, Rpt 1: repeat 1; Rpt 2: repeat 2; NES: nuclear export signal.

Figure 3
figure3

BAF155 is phosphorylated by Akt in randomly cycling cells. In total, 100 μg of nuclear extract from FL-INI1 randomly cycling (RC) or serum-starved (SS) cells was immunoprecipitated using anti-BAF155 antibody coupled to protein A beads. The RC extract was also immunoprecipitated using preimmune serum coupled to protein A beads (IgG). The immunoprecipitated material was divided in half and duplicate aliquots were separated by 8% sodium dodecyl sulfate (SDS)–PAGE and Western blotted. One membrane was cut into strips and the top half probed with anti-BAF155 antibodies and the bottom half probed with anti-Flag. The other membrane was probed with anti-phospho-(ser/thr) Akt substrate antibody. Similar results were obtained in two independent experiments.

Direct interactions between BAF155 and Akt, BAF155 and INI1, and INI1 and Akt

Our results indicated that cellular complexes exist containing Akt, INI1 and BAF155; however, these experiments could not determine if these interactions were mediated by direct protein–protein contacts. To examine this, a series of GST pulldown experiments was conducted. Glutathione-S-transferase fusions containing full-length INI1 or truncated versions containing stop codons at residues 131 and 246 were expressed in bacteria. INI1 contains two highly conserved imperfect repeat domains (Rpt): Rpt1 spans amino acids (aa) 186–245 and Rpt2 spans aa 259–319 (Morozov et al., 1998; Cheng et al., 1999). The truncation mutants expressed the N-terminus of INI1 alone (aa 1–131) and the N-terminus plus Rpt 1 (aa 1–246) (Figure 4c). Two N-terminal fragments of BAF155 were also cloned as GST fusions, encompassing amino acid residues 1–334 or 1–202, which would contain, or omit, the three Akt consensus sites, respectively (Figure 4c). These GST fusion proteins were bound to glutathione beads and tested for interactions with active, purified, recombinant Akt (Figure 4a). Both truncations of BAF155 appeared to directly bind to Akt, with the shorter one consistently demonstrating a slightly stronger interaction (lanes 3 and 4). All three INI1 constructs interacted with Akt as well, suggesting that Akt interacted with the N-terminus of INI1. Interestingly, our results repeatedly showed a slightly shifted-down signal for full-length INI1 (lane 9). One possibility was that it was also interacting with residual unphosphorylated Akt in the sample. To test this, the pulldown was repeated and first analysed using an antibody to phosphorylated Akt. Full-length INI1 did appear to interact with the phosphorylated Akt, as did the truncated proteins (lanes 12–14, top penel). On reprobing the membrane with an antibody to total Akt, the Akt input band broadened, suggesting that some of the input Akt was unphosphorylated (lane 10, lower panel). Of the pulldown reactions, only the band in the full-length INI1 lane appeared to broaden, with a slightly lower signal becoming visible (lane 14, lower panel). This suggests that full-length INI1 is able to interact with both phosphorylated and unphosphorylated Akt in vitro. Using 35S-labeled full-length INI1 as a target, our results also demonstrate that both BAF155 truncations are capable of directly interacting with INI1 (Figure 4b). This observation corroborates previous data in the literature in which BAF155 and BAF170 co-immunoprecipitated with INI1 after mixed baculovirus expression experiments (Phelan et al., 1999), and, in addition, demonstrates that the N-terminus of BAF155 is sufficient for such an interaction.

BAF155 and INI1 are phosphorylated by Akt in vitro

In vitro kinase assays were used to investigate whether the active, purified, recombinant Akt could phosphorylate either BAF155 or INI1. Purified GST-tagged INI1 and BAF155 proteins were used as substrates in these assays. Coomassie blue staining was used to visualize the eluted proteins and ensure that comparable amounts were used in the kinase assay (Figure 5a). Purified Akt was incubated separately with the proteins in the presence of adenosine triphosphate (ATP) and reactions were run on a polyacrylamide gel. The results (Figure 5b) demonstrated that the truncation of BAF155 ending at residue 334 was phosphorylated in the reaction, whereas the truncation ending at 202 was not. These results are consistent with the presence of the three Akt consensus sites only in the longer fragment. The results for INI1 were less clear – the N-terminal fragment ending at residue 131 was consistently phosphorylated, whereas the longer fragments were only marginally labeled in some experiments, at best.

Figure 5
figure5

Akt labels BAF155 and INI1 in vitro. (a) Coomassie blue staining of glutathione-S-transferase (GST) fusion proteins. Approximately 0.4 μg purified, soluble GST-fusion proteins were fractionated by 10% sodium dodecyl sulfate (SDS)–PAGE and stained with Coomassie blue. (b) Akt kinase assay. Proteins as in part (a) were incubated with Akt or buffer alone (input) for 10 min at 30°C; reactions were terminated by boiling in Laemmli sample buffer and fractionated by 10% SDS–PAGE. Western blots were performed using the anti-phospho-(ser/thr) Akt substrate antibody. For BAF155 the film was exposed for 20 s, for INI1 the film was exposed for 45 s. Similar results were obtained in three independent experiments.

Discussion

This is the first study to demonstrate interactions between Akt and members of the hSWI/SNF chromatin remodeling complex and to further demonstrate that Akt is capable of phosphorylating hSWI/SNF subunits, potentially affecting their function. Under conditions in which Akt is active, it was found associated with the hSWI/SNF subunits INI1, BAF155, BAF170 and actin. Glutathione-S-transferase pulldown experiments demonstrated direct interactions between Akt and BAF155 and between Akt and INI1. Furthermore, a direct interaction was seen between BAF155 and INI1, extending previous work that showed that a complex could be formed between INI1, BAF155 and BAF170 after baculovirus expression (Phelan et al., 1999). The relevance of an interaction between Akt and hSWI/SNF subunits in vivo was demonstrated by the fact that BAF155 immunoprecipitated from cells was preferentially recognized by the Akt substrate antibody under conditions where Akt was active. Finally, we found that purified, active Akt was able to label a fragment of BAF155 containing Akt consensus sites in in vitro kinase assays. The kinase assay results for INI1 were somewhat perplexing in that the shortest N-terminal fragment of INI1 was labeled, but a longer fragment and full-length INI1 were not. One explanation for this result is that Akt may modify a region of the protein at the N-terminus that is masked in the longer proteins, at least in vitro.

Our co-IP experiments also revealed that Akt was unable to interact with hSWI/SNF subunits in MRT cells, which lack INI1, suggesting that INI1 may be necessary for Akt to interact with the hSWI/SNF complex. One possibility is that INI1 may play a role in recruiting Akt to the complex. Although INI1 has been described as predominantly a nuclear protein, it contains a masked nuclear export signal (Craig et al., 2002) (Figure 4c). A role for INI1 in the cytoplasm has not been elucidated, but it may perhaps involve Akt. Another possibility is that INI1 may act as a scaffold for other hSWI/SNF subunits, helping them to stabilize and achieve a conformation that allows interaction with Akt to occur. Interestingly, our laboratory has repeatedly seen that reintroduction of INI1 in MRT cell lines leads to increased levels of BAF155, as detected by Western blot analysis (data not shown). Also, although Doan et al. (2004) concluded that INI1 was not necessary for the expression or assembly of hSWI/SNF subunits, their gel filtration analysis of the hSWI/SNF complex revealed several differences between a cell line with endogenous INI1 (HeLa) and two MRT cell lines. As noted by the authors, BAF170 and BAF155 appeared as doublets in the MRT cell lines, compared to single bands in HeLa fractions, and they hypothesized that this may be due to differential modification or proteolysis. Also, levels of hBrm were lower in the INI1-deficient cells. Another study saw a significant decrease in protein levels of BRG1 and hBrm in HeLa cells when INI1 levels were reduced using small interfering RNA (Cui et al., 2004). Chen and Archer (2005) demonstrated that protein–protein interactions between BAF155 and BAF57 are essential for the stability of BAF57 and hypothesized that, without such an interaction, BAF57 may be degraded by the proteasome. Our results have shown that purified Akt can interact with and phosphorylate a BAF155 peptide in vitro; however, Akt may be unable to co-immunoprecipitate BAF155 in the MRT cell line because BAF155 is less stable or has an aberrant conformation in cells in the absence of INI1.

The hSWI/SNF complex is known to be localized on the promoters and important in the transcriptional regulation of many genes important for cell cycle control. Chromatin immunoprecipitation experiments have demonstrated the presence of hSWI/SNF on the cyclin D promoter, where it repressed transcription (Zhang et al., 2002). Chromatin immunoprecipitation also demonstrated that hSWI/SNF members occupy the p16INK4a and p21WAF1/CIP1 promoters in vivo, where their presence is correlated with increased transcription (Lee et al., 2002; Kadam and Emerson, 2003; Kang et al., 2004; Oruetxebarria et al., 2004; Chai et al., 2005). The hSWI/SNF complex also regulates a diverse group of other mammalian target genes. Chromatin immunoprecipitation experiments have demonstrated the presence of hSWI/SNF components on the promoters of IFNγ- and IFNα-inducible genes, as well as on the IFNβ promoter (Agalioti et al., 2000; Huang et al., 2002; Liu et al., 2002; Pattenden et al., 2002; Cui et al., 2004). hSWI/SNF is also present on the CD44 and E-cadherin promoters (Banine et al., 2005; Gresh et al., 2005), on the promoters for myogenin and muscle creatine kinase (Simone et al., 2004), and on steroid receptor targets (Fryer and Archer, 1998; DiRenzo et al., 2000; Belandia et al., 2002; Métivier et al., 2003; Link et al., 2005). Despite the growing knowledge of promoters regulated by hSWI/SNF, it is not well understood how the complex becomes localized and activated at various promoters; however, there is evidence for the involvement of signaling pathways. For example, the p38 pathway is activated during skeletal myogenesis and it has been shown that the hSWI/SNF subunit, BAF60, can be phosphorylated by p38αβ in vitro. Furthermore, pharmacologic inhibition of p38 prevents hSWI/SNF recruitment to muscle-specific elements (Simone et al., 2004). Our experiments have additionally identified the PI3K/Akt pathway as a novel signaling pathway that may also target the hSWI/SNF complex.

The interaction between members of the hSWI/SNF complex, which often function as tumor suppressors, and Akt, a proto-oncogene, begs the question of what is the end point of such interactions. Traditionally, Akt has been thought of as a molecule important in cell proliferation and survival, and it is aberrantly activated by various means in many different cancers. However, recent experiments have illustrated the importance of the Akt pathway in cellular differentiation and senescence, as well. For example, it has been shown that insulin-like growth factor I promotes muscle cell survival through activation of Akt and subsequent upregulation of p21WAF1/CIP1, which forces withdrawal of the cells from the cell cycle to allow them to differentiate (Lawlor and Rotwein, 2000). Additionally, regulation of endothelial cell sprouting to form capillary structures is mediated by the proteoglycan decorin in an Akt-dependent fashion (Schönherr et al., 2001). Finally, senescent endothelial cells have been shown to have higher levels of phospho-Akt compared to young endothelial cells. Akt was shown to contribute to the replicative senescence of these cells by increasing the transcriptional activity of p53 and thereby upregulating p21WAF1/CIP1 (Miyauchi et al., 2004). As mentioned above, hSWI/SNF components have been localized to the promoters of the p21WAF1/CIP1 and p16INK4a genes, both of which are important in differentiation and/or senescence. We propose that in MRT cells the loss of hSWI/SNF function may prevent signaling pathways from being able to induce these molecules, thereby allowing the cells to escape appropriate regulation of differentiation or senescence. Future studies to investigate the effects of Akt activation on the recruitment of hSWI/SNF to various cancer-related target genes should help to resolve this issue.

Materials and methods

Cells, cell culture and treatments

HeLa cells were grown in minimum essential medium (MEM) (Gibco, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (FBS) (Gibco). HeLa cells stably expressing a Flag-tagged INI1 fusion protein (FL-INI1 cells) were grown in MEM supplemented with 10% FBS and 0.3 μg puromycin per ml. Cloning of the Flag-INI1 construct was done as described previously by Reincke et al. (2003). Transfections were carried out using Lipofectin (Invitrogen, Carlsbad, CA, USA) per the manufacturer's instructions. Stably transfected clones were obtained after antibiotic selection. In order to reduce Akt phosphorylation, cells were grown in serum-free MEM for 20–24 h (SS). TTC549 cells were grown in RPMI (Gibco) supplemented with 10% FBS and were a generous gift from Dr Timothy Triche (Los Angeles Children's Hospital, Los Angeles, CA, USA).

Cellular fractionation and Western blot analysis

Cells were fractionated into nuclear and cytoplasmic extracts using a NE-PER kit per the manufacturer's instructions (Pierce, Rockford, IL, USA). Whole-cell extracts were made using cell lysis buffer (20 mM Tris, pH 7.5, 150 mM NaCl, 1 mM ethylenediamine tetraacetate, 1 mM ethyleneglycol-bis(β-aminoether)-N,N′-tetra-acetic acid, 1% Triton-X-100, 2.5 mM sodium pyrophosphate, 1 mM β-glycerolphosphate), plus 0.5 mM phenylmethylsulfonylfluoride (PMSF) and 1 mM sodium orthovanadate. Cells were rinsed in cold phosphate-buffered saline (PBS) containing 1 mM sodium orthovanadate, incubated for 5 min on ice in cell lysis buffer, scraped off plates, sonicated and centrifuged. Extracts were fractionated on 8% or 10% SDS–polyacrylamide gels and electroblotted onto Immobilon polyvinylidene difluoride (PVDF) membranes (Millipore, Billerica, MA, USA). Western blots were performed using the following antibodies: anti-Akt 1,2 (H-136, Santa Cruz, Santa Cruz, CA, USA), anti-phospho Akt (S473, Cell Signaling #9271, Beverly, MA, USA), anti-phospho-(ser/thr) Akt substrate, which preferentially recognizes proteins containing phosphor-ser/thr preceded by Lys or Arg at posititions −5 and −3 (Cell Signaling #9611), anti-β-actin (A5441, Sigma, St Louis, MO, USA), anti-BAF155 (H-76, Santa Cruz), anti-BAF170 (H-116, Santa Cruz), anti-Flag M2 (F3165, Sigma), anti-INI1 (gift from Dr Willliam Schubach, University of Washington, Seattle, WA, USA), and anti-α-tubulin (T5168, Sigma). Secondary antibodies were horseradish peroxidase-conjugated and were detected using chemiluminescence (enhanced chemiluminescence; Amersham Biosciences, Piscataway, NJ) and exposure of blots to X-ray film.

Immunoprecipitations

Anti-Flag immunoprecipitations

Whole-cell extracts were diluted with 150 mM IP buffer (150 mM KCl, 50 mM Tris, pH 7.5, 5 mM MgCl2, 0.1% IGEPAL (Sigma)), plus 0.5 mM PMSF and 1 mM sodium orthovanadate, in a final volume of approximately 300 μl, and immunoprecipitated using 30 μl of anti-Flag M2 affinity gel (Sigma) by rolling at 4°C for 2.5 h. Immunoprecipitates were washed three times in Tris-buffered saline (20 mM Tris base, 140 mM NaCl, pH 7.6) plus 1 mM sodium orthovanadate and proteins were boiled off the beads using Laemmli sample buffer (0.0625 M Tris, pH 6.8, 2% SDS, 10% glycerol, 0.001% bromophenol blue, 0.7 M β-mercaptoethanol).

Anti-Akt immunoprecipitations

Nuclear extracts were diluted in IP buffer lacking KCl in a final volume of 300–400 μl and immunoprecipitated with 20 μl of an agarose-conjugated anti-Akt1 affinity resin (Santa Cruz) by rolling overnight at 4°C. Immunoprecipitates were washed four times with IP buffer containing 100 mM KCl and proteins were boiled off the beads in Laemmli sample buffer.

Anti-BAF155 immunoprecipitations

An anti-BAF155 affinity resin was generated by incubating 20 μl of a 50% slurry of protein A beads (Sigma), 170 μl PBS (140 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4, pH 7.4) and 10 μl anti-BAF155 antibody for 1 h rolling at 4°C. Beads were washed two times with PBS and used in IP reactions. Control resin was preimmune serum coupled to protein A beads (IgG). Nuclear extracts were diluted in PBS and immunoprecipitated using the anti-BAF155 antibody coupled to protein A beads. Immunoprecipitates were washed three times with PBS and proteins were boiled off the beads using Laemmli sample buffer.

Construction and expression of glutathione-S-transferase fusion proteins

Full-length INI1 cDNA (GenBank accession # NM_003073) was generated and cloned into the pcDNA3.1(+) vector (Invitrogen) as described by Reincke et al. (2003). The open reading frame was subcloned into the pGEX2TK vector (Amersham Biosciences). The Stratagene Quick Change mutagenesis protocol (La Jolla, CA, USA) was used to introduce stop codons (bold) into GST-INI1 at residues 131 (sequence: 5′-IndexTermAGG AAC AGC CAG TAG GTA CCC ACC CTG C-3′) and 246 (sequence: 5′-IndexTermTC AGA CAG CAG ATC TAG TCC TAC CCC ACG G-3′).

Wild-type BAF155 fragments were generated for cloning by reverse transcribing 1 μg of total RNA from HeLa cells in a 20 μl reaction containing 25 mM MgCl2, PCR buffer II (Roche, Indianapolis, IN, USA), 10 mM deoxynucleoside triphosphates, 2 μg RNase inhibitor, 40 U reverse transcriptase (Invitrogen), 0.025 mM random hexamers and 0.025 mM oligo d(T)16 primers. Primers were used to amplify the N-terminus of BAF155 from the ATG start codon to amino acid 334 by PCR (sequences: 5′-IndexTermAAA AGG ATC CCA TAT GGC CGC AGC GGC GGG C-3′ and 5′-IndexTermAAA AGA ATT CTT ACG GAG GGG GAG GCG AAG GCG AAT GTT TCC TCT TTC GAG-3′). PCR reactions contained 0.33 μ M of forward primer, 0.27 μ M of reverse primer, PCR buffer II (Roche), 5 μl dimethylsulfoxide, 2.5 U of AmpliTaq Gold (Roche) and 20 μl of the reverse transcriptase reaction in a total volume of 100 μl. Amplification was performed using the following cycling parameters: 10 min at 94°C; 40 cycles of 94, 65 and 72°C for 30 s each; and a single final extension at 72°C for 7 min. The product was digested with BamHI and EcoRI and cloned into pGEX2TK. To create a BAF155 construct that lacked the Akt consensus sites, the above construct was digested with HindIII and EcoRI, the ends filled in, and self-ligated to create a GST construct containing residues 1–202 of BAF155. All pGEX2TK constructs were verified by sequence analysis and transformed into BL21(DE3)pLysS bacteria.

Bacteria were grown to late log phase and induced for 2 h with 0.3–0.4 mM isopropyl-β-D-thiogalactopyranoside. They were centrifuged and frozen for 1 h before lysis in PBS, containing 0.5 mM PMSF. After sonication and centrifugation, total proteins were bound to glutathione sepharose beads (Amersham Biosciences) by rocking lysates and beads on ice for 1 h. Beads were washed twice with PBS and relative protein levels were determined by Coomassie blue staining of aliquots released from the beads and separated on a 10% SDS–polyacrylamide gel. Beads were then adjusted to approximately equal protein concentrations by adding empty glutathione beads as necessary and verified by again viewing with Coomassie blue staining after fractionation on a 10% SDS–polyacrylamide gel. To elute the proteins, washed beads were incubated with 40 mM reduced glutathione in 200 mM Tris, pH 7.5, 100 mM NaCl, and incubated rolling at 4°C for 1 h to overnight. Eluted proteins were dialysed against 100 volumes of dialysis buffer (20 mM N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid, 50 mM NaCl, 10% glycerol) prior to use in kinase assays.

Akt kinase assays

Kinase assays were performed using 40 ng of active, purified, full-length human recombinant Akt1 (Upstate, Charlottesville, VA, USA). Purified, soluble GST-fusion proteins were incubated with Akt in 1 × Akt kinase buffer (25 mM Tris, pH 7.5, 5 mM β-glycerophosphate, 2 mM dithiothreitol, 10 mM MgCl2, 1 mM sodium orthovanadate) containing 200 μ M ATP. Reactions were incubated for 10 min at 30°C, terminated by boiling in Laemmli sample buffer, and fractionated on a 10% SDS–polyacrylamide gel, followed by Western blot analysis using the Akt substrate antibody.

Glutathione-S-transferase pulldown assays

In all, 40 ng of active, purified, full-length human recombinant Akt1 (Upstate) was incubated with protein-bound glutathione beads in 100 mM IP buffer, containing 0.5 mM PMSF and 1 mM sodium orthovanadate, by rolling overnight at 4°C in a total volume of 200 μl. Beads were washed three times with 100 mM IP buffer, containing 0.5 mM PMSF. Proteins were boiled off the beads in Laemmli sample buffer, fractionated on 10% SDS–polyacrylamide gels and Western blotted.

Target protein INI1 was produced and radiolabeled with 35S-methionine using the TnT-coupled reticulocyte lysate system (Promega, Madison, WI, USA). Pulldowns were performed as above. Samples were fractionated on a 10% SDS–polyacrylamide gel that was dried, exposed to a PhosphorImager intensifying screen and analysed using the Storm 860 imaging system (Molecular Dynamics, Inc., Sunnyvale, CA, USA).

Accession codes

Accessions

GenBank/EMBL/DDBJ

Abbreviations

hSWI/SNF:

hSWI (switch)/SNF (sucrose non-fermenting)

BRG1/hBrm:

Brahma-related gene 1/human Brahma

BAF:

BRG1-associated factor

INI1:

integrase interactor 1

SMARCB1:

SWI/SNF-related matrix-associated actin-dependent regulator of chromatin B1

References

  1. Agalioti T, Lomvardas S, Parekh B, Yie J, Maniatis T, Thanos D . (2000). Cell 103: 667–678.

  2. Banine F, Bartlett C, Gunawardena R, Muchardt C, Yaniv M, Knudsen ES et al. (2005). Cancer Res 65: 3542–3547.

  3. Belandia B, Orford RL, Hurst HC, Parker MG . (2002). EMBO J 21: 4094–4103.

  4. Betz BL, Strobeck MW, Reisman DN, Knudsen ES, Weissman BE . (2002). Oncogene 21: 5193–5203.

  5. Biegel J, Zhou J, Rorke L, Stenstrom C, Wainwright L, Fogelgren B . (1999). Cancer Res 59: 74–79.

  6. Blom N, Sicheritz-Pontén T, Gupta R, Gammeltoft S, Brunak S . (2004). Proteomics 4: 1633–1649.

  7. Chai J, Charboneau AL, Betz BL, Weissman BE . (2005). Cancer Res 65: 10192–10198.

  8. Chen J, Archer TK . (2005). Mol Cell Biol 25: 9016–9027.

  9. Cheng SW, Davies KP, Yung E, Beltran RJ, Yu J, Kalpana GV . (1999). Nat Genet 22: 102–105.

  10. Craig E, Zhang Z-K, Davies KP, Kalpana G . (2002). EMBO J 21: 31–42.

  11. Cui K, Tailor P, Liu H, Chen X, Ozato K, Zhao K . (2004). Mol Cell Biol 24: 4476–4486.

  12. Datta SR, Dudek H, Tao X, Masters S, Fu H, Gotoh Y et al. (1997). Cell 91: 231–241.

  13. DiRenzo J, Shang Y, Phelan M, Sif S, Myers M, Kingston R et al. (2000). Mol Cell Biol 20: 7541–7549.

  14. Doan DN, Veal TM, Yan Z, Wang W, Jones SN, Imbalzano AN . (2004). Oncogene 23: 3462–3473.

  15. Franke TF . (1999). Neural Notes V: 3–7.

  16. Franke TF, Hornik CP, Segev L, Shostak GA, Sugimoto C . (2003). Oncogene 22: 8983–8998.

  17. Fryer CJ, Archer TK . (1998). Nature 393: 88–91.

  18. Gresh L, Bourachot B, Reimann A, Guigas B, Fiette L, Garbay S et al. (2005). EMBO J 24: 3313–3324.

  19. Guidi CJ, Sands AT, Zambrowica BP, Turner TK, Demers DA, Webster W et al. (2001). Mol Cell Biol 21: 3598–3603.

  20. Huang M, Qian F, Hu Y, Ang C, Li Z, Wen Z . (2002). Nat Cell Biol 4: 774–781.

  21. Kadam S, Emerson BM . (2003). Mol Cell 11: 377–389.

  22. Kang H, Cui K, Zhao K . (2004). Mol Cell Biol 24: 1188–1199.

  23. Klochendler-Yeivin A, Fiette L, Barra J, Muchardt C, Babinet C, Yaniv M . (2000). EMBO Rep 1: 500–506.

  24. Lawlor MA, Rotwein P . (2000). Mol Cell Biol 20: 8983–8995.

  25. Lee D, Kim JW, Seo T, Hwang SG, Choi EJ, Choe J . (2002). J Biol Chem 277: 22330–22337.

  26. Link KA, Burd CJ, Williams E, Marshall T, Rosson G, Henry E et al. (2005). Mol Cell Biol 25: 2200–2215.

  27. Liu H, Kang H, Liu R, Chen X, Zhao K . (2002). Mol Cell Biol 22: 6471–6479.

  28. Métivier R, Penot B, Hübner MR, Reid G, Brand H, Kos M et al. (2003). Cell 115: 751–763.

  29. Miyauchi H, Minamino T, Tateno K, Kunieda T, Toko H, Komuro I . (2004). EMBO J 23: 212–220.

  30. Morozov A, Yung E, Kalpana G . (1998). Proc Natl Acad Sci USA 95: 1120–1125.

  31. Oruetxebarria I, Venturini F, Kekarainen T, Houweling A, Zuijderduijn LM, Mohd-Sarip A et al. (2004). J Biol Chem 279: 3807–3816.

  32. Pattenden SG, Klose R, Karaskov E, Bremner R . (2002). EMBO J 21: 1978–1986.

  33. Phelan M, Sif S, Narlikar G, Kingston RE . (1999). Mol Cell 3: 247–253.

  34. Rando OJ, Zhao K, Janmey P, Crabtree GR . (2002). Proc Natl Acad Sci USA 99: 2824–2829.

  35. Reincke BS, Rosson GB, Oswald BW, Wright CF . (2003). J Cell Physiol 194: 303–313.

  36. Roberts CWM, Galusha SA, McMenamin ME, Fletcher CDM, Orkin SH . (2000). Proc Natl Acad Sci USA 97: 13796–13800.

  37. Schönherr E, Levkau B, Schaefer L, Kresse H, Walsh K . (2001). J Biol Chem 276: 40687–40692.

  38. Simone C, Forcales SV, Hill DA, Imbalzano AN, Latella L, Puri PL . (2004). Nat Genet 36: 738–743.

  39. Versteege I, Medjkane S, Rouillard D, Delattre O . (2002). Oncogene 21: 6403–6412.

  40. Versteege I, Seveenet N, Lange J, Rousseau-Merch M-F, Ambros P, Handgretinger R et al. (1998). Nature 394: 203–206.

  41. Wang W, Cote J, Xue Y, Zhou S, Khavari P, Biggar S et al. (1996). EMBO J 15: 5370–5382.

  42. Wu DY, Tkachuck DC, Roberson RS, Schubach WH . (2002). J Biol Chem 277: 27706–27715.

  43. Xu W, Yuan X, Jung YJ, Yang Y, Basso A, Rosen N et al. (2003). Cancer Res 63: 7777–7784.

  44. Zhang Z-K, Davies KP, Allen J, Ahu L, Pestell RG, Zagzag D et al. (2002). Mol Cell Biol 22: 5975–5988.

  45. Zhao K, Wang W, Rando OJ, Xue Y, Swiderek K, Kuo A et al. (1998). Cell 95: 625–636.

Download references

Acknowledgements

We thank Drs Yusuf Hannun (MUSC) and Dennis Watson (MUSC) for critical review of this manuscript and Drs David Kurtz (MUSC), Robin Muise-Helmericks (MUSC), Dennis Watson and Daohong Zhou (MUSC) for helpful discussions and ideas. We additionally thank Dr Robin Muise-Helmericks for reagents and assistance with techniques, as well as Dr William Schubach (University of Washington) for the anti-INI1 antibody, and Dr Timothy Triche (Los Angeles Children's Hospital) for cell lines. Funding for this work was provided by the MUSC Department of Pathology and Laboratory Medicine (CFW, KSJF), Department of Defense grant #N00014-96-1298 (CFW, KSJF), a Hollings Cancer Center Abney Foundation Scholarship (KSJF), and National Institutes of Health grants #T32 DK007431-21 (WJM) and #T35 HL007769-14 (JSR). We also thank the MUSC Nucleic Acids Analysis and Molecular Biology Core Facilities.

Author information

Correspondence to C F Wright.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Foster, K., McCrary, W., Ross, J. et al. Members of the hSWI/SNF chromatin remodeling complex associate with and are phosphorylated by protein kinase B/Akt. Oncogene 25, 4605–4612 (2006). https://doi.org/10.1038/sj.onc.1209496

Download citation

Keywords

  • hSWI/SNF
  • INI1
  • BAF155
  • Akt

Further reading