| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
1 INSERM U509, Laboratoire de Pathologie Moléculaire des Cancers, and 2 Unité de Bioinformatique, Institut Curie, Paris, France
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
hSNF5/INI1 encodes a subunit of the SWI-SNF chromatin remodeling complex, which is one of several evolutionarily conserved multiprotein complexes that disrupt nucleosomes in vitro in an ATP-dependent manner.
In addition to the cancer-related mutations of hSNF5/INI1, mounting evidence suggests a role of SWI-SNF complexes in oncogenesis. Indeed, BRG1 and hBRM interact and collaborate with the retinoblastoma (RB) tumor suppressor and histone deacetylase (HDAC) to repress E2F1 activity (8) . This functional link between E2F1 and SWI-SNF is further supported by genetic screens in Drosophila that identified members of SWI-SNF complexes as negative regulators of E2F activity (9) . Inactivating mutations in BRG1 have also been identified in several human tumor cell lines or tumors, and the heterozygous inactivation of BRG1 is associated with a tumor-prone phenotype in mice (10, 11, 12) .
Recently, we and others have shown that hSNF5/INI1 is implicated in cell growth control by inhibiting S-phase entry and by promoting a G1 arrest in hSNF5/INI1-deficient cell lines (13, 14, 15, 16, 17) . This arrest can be reversed by coexpression of adenoviral E1a but not by RB binding-deficient E1a mutants, strongly suggesting that the presence of a functional RB is required. However, the ability of RB to arrest hSNF5/INI1-deficient rhabdoid cells indicates that hSNF5/INI1, in contrast to BRG1, is not mandatory to RB function. Regarding the G1 arrest induced by hSNF5/INI1, other groups have provided results suggesting that hSNF5/INI1 may participate in the repression of the cyclin D1 promoter or in the induction of INK4a gene expression (16 , 17) .
In this study, we have taken advantage of the construction of two different conditional systems for hSNF5/INI1 activity in hSNF5/INI1-deficient rhabdoid cells, one based on the tetracycline-inducible system and another based on the fusion between hSNF5/INI1 and the estrogen receptor (ER) hormone binding domain, to perform time course experiments of the gene expression profile and analysis of the cell phenotype after restoration of the hSNF5/INI1 protein.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Plasmid Construction.
To construct the pBI-HA-INI1 vector, the HA-INI1 coding sequence (20)
was PCR amplified using forward primer 5'-GCTAGCTGCTACTGAAGATCTCCCAAGCTCCACCATGGGG-3' and reverse primer 5'-CATGACATCGGATGCTAGCTAGCGATGGGCTGGTTACCAGGCC-3' containing BglII and NheI restriction sites, respectively. The PCR product was digested by both restriction enzymes and cloned into BglII/NheI linearized tet-responsive reporter plasmid pBI-EGFP vector (21)
. The hormone binding domain of ER (22)
was cloned in the XhoI site 3' to HA-INI1, creating the pCDNA-HA-INI1-ER vector. Subsequently, the HA-INI1-ER cDNA was cloned in place of luciferase in the XbaI/NheI sites of the pRLTK vector (Promega, Madison, WI). The BamHI/SpeI fragment containing the HA-INI1-ER and globin intron sequences was then cloned by PCR in the BglII/NheI linearized pBI-EGFP vector. Constructs were verified by sequencing.
Generation of Inducible Rhabdoid Cell Lines.
MON-tTa-expressing clones were obtained by transfection using the Effectene method (Qiagen Gmbh) of the human rhabdoid MON cell lines with the tTA-IRES-NEO vector (21)
. The MON-tTA25 clone, which yielded an optimal induction of luciferase on transient transfection with the pUHC 13.3 reporter plasmid, was then transfected with the PBI-EGFP/HA-INI1 and PBI-EGFP/HA-INI1-ER constructs. Inducible clones were sought by reverse transcription-PCR and Western blot.
Immunoblotting, Immunoprecipitation, and Immunofluorescence.
The 738 polyclonal antibody was obtained by immunizing rabbits with the RNTGDADQWCPLLET peptide. Antibodies against ß-actin (ac-74), BRG1 (2SN-2E12-AS), BAF155/170 (sc 9746/sc 9744) were obtained from Sigma, Euromedex (Montreal, Canada), and Santa Cruz Biotechnology (Santa Cruz, CA), respectively. Western blotting was conducted as described previously (15)
. For immunoprecipitations, precleared nuclear cell extracts were incubated with protein G-coupled with anti-hemagglutinin [HA (12CA5; Roche Molecular Biochemicals, Indianapolis, IN)] or anti-FLAG antibodies (F3165; Sigma). Protein-antibody complexes were subsequently washed in buffer containing 150 mM NaCl and 0.2% NP40 and analyzed by immunoblotting. Immunofluorescence experiments were conducted with tetramethylrhodamine isothiocyanate-conjugated phalloidin (Sigma) or a paxillin-specific monoclonal antibody (clone 349; BD Biosciences) on 3% formaldehyde-fixed and 0.2% Triton X-100-permeabilized cells.
Affinity Precipitation of Cellular GTP-Rho or Measuring Rho-GTP Levels.
Glutathione S-transferase-Rhotekin pull-down assay was performed using the EZ-Detected Rho Activation Kit (Pierce Biotechnology), according to the manufacturer’s instructions. Briefly, lysates from induced or uninduced I2A cells were incubated with glutathione S-transferase-Rhotekin and gluthatione-coated beads to capture GTP-bound proteins. Bound proteins were washed several times and eluted with sample buffer before analysis by immunoblot using an anti-Rho antibody.
Microarray Analysis.
Total RNAs were extracted using the TRIzol reagent (Gibco-BRL) and quality controlled by migration on Agilent Bioanalyser 2100 (Waldbronn, Germany). The RNA samples were labeled and hybridized on Affymetrix Genechip human U133A (22K) as recommended by the manufacturer. The arrays were scanned using a Hewlett Packard confocal laser scanner, and data were processed with the MICROARRAY SUITE (Mas 5.0; Affymetrix) and dChip softwares (23)
. We used Rosetta’s algorithm to calculate the baseline signal resulting from the calculation of mean of the three uninduced samples, taking into account the SE of the measurements. The fold change was calculated for each of the seven time points as the ratio of signals from induced or uninduced samples. To identify differentially expressed genes, probe sets were filtered to include only genes responding to the following criteria: (a) at least one value scored "present" and a signal difference higher than 60 as compared with the control; and (b) for early time points (28–96 h), the analysis included genes with a fold change higher than 2 for at least two time points. For the late time point (10 days), a 3-fold variation was required, given the high number of genes demonstrating variation at this time. A total of 482 unique genes met all conditions. We used the logarithm of fold change to cluster genes with a similar pattern of expression by applying the k-means clustering algorithm using Euclidian distances (Supplementary Data 2).
Quantitative Reverse Transcription-PCR.
Total RNAs treated with DNase I (Invitrogen) were reverse-transcribed with oligo-random hexamers using the GeneAmpRNA PCR kit (Applied Biosystems) and amplified with specific primers according to the manufacturer’s instructions. Glyceraldehyde-3-phosphate dehydrogenase was used for calibration (primer sequences are indicated in Supplementary Data 1).
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). HA-INI1 was detectable 12 h after tetracycline depletion, and its expression reached a maximum at 30 h. The expression level was then stable over time and similar to that of the endogenous protein in non-rhabdoid cell lines (Fig. 1C)
. An efficient integration of HA-INI1 in the SWI-SNF complex was documented by immunoprecipitation experiments, which showed that in extracts from induced cells, subunits of the complex, including BRG1, BAF155 and BAF170, were brought down by an anti-HA antibody, but not by a control anti-FLAG antibody (Fig. 1D)
. Therefore, this HA-INI1 inducible tet-off system allows a physiological level of expression of the protein, which is properly integrated in the SWI-SNF complex.
|
. Interestingly, the repression of HA-INI1 expression, through reintroduction of tetracycline in the medium, resulted in the reversion of the cell cycle arrest (Fig. 2B)
. These data showed that most of the HA-INI1-expressing cells arrested in G1 phase still remained competent to reenter the cell cycle. Accordingly, HA-INI1 expression did not lead to significant apoptosis as monitored by annexinV staining (Fig. 2C)
and sub-G1 analysis or to senescence as evaluated by ß-galactosidase staining (data not shown).
|
; Supplementary Data 2). The number of genes demonstrating induction was higher (332; 69%) than that of repressed genes (150; 31%). Induced genes could be divided into seven groups (clusters A–G). Early-induced genes were grouped in clusters A (2 genes), B (11 genes), C (71 genes), and D (91 genes) according to their magnitudes of induction. As expected, HA-INI1 was induced very strongly and rapidly and belonged to group A. Genes with intermediate and late induction were classified in clusters F (63 genes) and G (70 genes), respectively. Genes demonstrating a transient induction were clustered in group E (24 genes). Finally, groups H (100 genes) and I (50 genes) included repressed genes at early and late times, respectively.
|
Identification of Direct Targets.
To identify direct transcriptional targets, we developed a conditional hSNF5/INI1-ER fusion protein comprising hSNF5/INI1 and the ER ligand-binding domain (22)
. This fusion protein demonstrated a slight residual activity in the absence of 4-hydroxytamoxifen, which precluded the selection of clones with stable expression. To take into account this parameter, we constructed clone ER4A with a tetracycline-based inducible expression of HA-INI1-ER (Fig. 4A)
. On tetracycline withdrawal, a slight cell cycle effect was observed, consistent with the weak basal activity of the HA-INI1-ER protein. The addition of 4-hydroxytamoxifen resulted in a complete cell cycle arrest, indicating that 4-hydroxytamoxifen restored full activity of the fusion protein (Fig. 4B)
.
|
Functional Groups.
Differentially expressed genes were classified in functional clusters according to their known or inferred functions using gene ontology. The classification of genes shown in Table 1
and 2
was based on this analysis, which identified highly significant groups of genes (P < 0.001). The first one corresponds to genes encoding cell cycle regulators including components of the prereplication complex (MCM2, MCM5, MCM10, MCM4, RF-C, CDT1, CDC6, and CDC7) and a number of E2F targets (Table 1
; Ref. 24
). Within this group, which mainly contains late down-regulated genes, it is worth mentioning the DNA replication licensing factor CDT1, which is essential for the initiation of DNA replication and demonstrates a very early down-regulation. The second main functional group contains numerous genes that participate in cytoskeleton organization and cell-matrix or cell-cell interactions (Table 2
; Refs. 25, 26, 27
).
|
|
. Interestingly, most genes from this group are regulated early after HA-INI1 induction and are therefore susceptible to mediate the late effects of HA-INI1 expression. Finally, in light of the essential role of hSNF5/INI1 orthologs during early steps of development, it is worth mentioning the early variation of a set of genes involved in developmental processes (Table 2
; Refs. 4, 5, 6
, 28
, and 29
).
hSNF5/INI1 Regulates Actin Cytoskeleton Organization.
The induction of hSNF5/INI1 expression was associated with dramatic modification of cell shape, including a rounding of the cell body and a strong decrease of the cell size (Fig. 5A)
. The well-organized actin stress fiber network of uninduced cells, detected by phalloidin staining, was completely modified on HA-INI1 induction with loss of stress fibers and concomitant appearance of a strong cortical actin staining (Fig. 5B)
. The modifications of cell shape and actin organization were also associated with a strong tendency of cells to detach from the dish as documented by time-lapse videomicroscopy (data not shown), suggesting a decreased adhesion of cells to the extracellular matrix. In support of this hypothesis, a complete disappearance of paxillin-stained focal adhesions was observed (Fig. 5C)
. This reorganization occurred with constant amounts of ß-actin and paxillin, as indicated by immunoblots showing similar levels of these proteins in induced and uninduced cells (data not shown).
|
, together with dramatic modifications of the actin cytoskeleton. The critical role of the activity of members of the Rho family of small GTPases in the regulation of actin stress fibers and focal adhesion formations (25)
prompted us to analyze Rho activity in HA-INI1 induced and uninduced cells. We used the glutathione S-transferase-Rhotekin Rho binding domain to specifically pull down the active, GTP-bound Rho proteins. Fig. 6
shows that whereas the overall level of Rho protein is similar in induced and uninduced cells, the GTP-bound fraction is strongly decreased after HA-INI1 expression.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Time course gene expression profiling indicates that this arrest is associated with the down-regulation of a high number of genes involved in cell cycle progression. This includes different MCM proteins (CDC6 and CDT1, which are essential members of the DNA replication complex). Numerous E2F target genes are retrieved among down-regulated genes (24) . This supports the hypothesis of hSNF5/INI1, together with BRG1, being involved in RB-mediated corepression of E2F (8) . Whereas most of these genes demonstrate a relatively late down-regulation, concomitant with the cell cycle arrest, the decreased expression of the DNA replication licensing factor CDT1 occurs much earlier, which may suggest a causal role in the inhibition of DNA replication.
We identify four genes, which constitute excellent candidates for being direct transcriptional targets of hSNF5/INI1. These genes exhibit strong and early increases after hSNF5/INI1 induction without requirement for de novo protein synthesis. ATP1B1 encodes the ß1 subunit of the Na/K-ATPase, an ATP-dependent sodium and potassium transporter across the plasma membrane, and pleiotrophin is a cytokine with diverse functions in growth, apoptosis, angiogenesis, and cell differentiation (30) . HESR1 is a bHLH transcription factor related to the hairy/enhancer of split (HES) family of downstream targets of the Notch signaling pathway involved in cell fate commitment in a broad range of developmental processes (31) . Interestingly, recent results have shown that BRM and hSNF5/INI1 are recruited to the HES1 and HES5 promoters (32) . Taken together, these data strongly suggest that hSNF5/INI1 is involved in transcriptional regulation of targets of the Notch pathway. Finally, FRIZZLED 7 is a receptor of the Wnt signaling pathway. Previous experiments have indicated that BRG1 cooperates with ß-catenin to activate TCF target genes (33) . Our work suggests that SWI-SNF could also contribute to the activation of Wnt signaling through increased expression of the FRIZZLED 7 receptor. Hence, through the direct induction of ATP1B1, pleiotrophin, HESR1, and FRIZZLED 7, hSNF5/INII1 may directly regulate diverse cellular processes including ionic homeostasis, cytokine signaling, and the Notch and Wnt pathway. One may speculate that this direct involvement in various pathways may account for the extremely detrimental effect of hSNF5/INI1 loss of function during development and adult life.
The reexpression of hSNF5/INI1 protein in hSNF5/INI1-deficient cells induces dramatic modifications of the cell shape associated with a complete reorganization of the actin cytoskeleton. These cellular phenotypic changes occur concomitantly with the up-regulation of genes involved in the polymerization and organization of the actin cytoskeleton. These genes include Arp2, N-WASP, WAVE 3, cortactin, Rho-GAP4, and RhoE. Arp2 is a core component of the Arp2/3 complex, which is involved in the nucleation of actin filaments and regulated by N-WASP, WAVE 3, and cortactin (34) . Previous reports have shown that WASP and WAVE family proteins play a critical role in the organization of cortical actin filaments (35) . When bound to GTP, Rho GTPase proteins increase the stability of actin-based structures such as stress fibers and focal adhesions. They therefore constitute major regulators of the actin cytoskeleton. The present microarray experiments suggest that hSNF5/INI1 expression could result in the inhibition of this function through induction of Rho-GAP4, which inhibits Rho functions by increasing its GDP-bound form, and RhoE, which antagonizes some effects of the RhoA protein and inhibits the assembly of actin stress fibers and the formation of focal adhesions in fibroblasts and Madin-Darby canine kidney cells (36) . The analysis of the activity of Rho protein in I2A rhabdoid cells reveals that the induction of hSNF5/INI1 leads to a dramatic decrease of the amount of active GTP-Rho proteins. These data indicate that the modulation of Rho activity may constitute an essential step of the cytoskeleton modifications induced by hSNF5/INI1. Interestingly, recent results also implicate BRG1 in the modulation of the actin cytoskeleton. In SW13 BRG1-deficient cells, restoring BRG1 activity results in enhanced formation of actin stress fibers (37) . The opposite effects of BRG1 and hSNF5/INI1, two members of the SWI-SNF complex, on the organization of the actin stress fiber network strongly suggest that the effect of SWI-SNF on cytoskeleton organization may be highly dependent on the cellular context. Strikingly, highly tumorigenic MON cells demonstrate a well-organized network of stress fibers and focal adhesions that are altered on reversion of the transformed phenotype associated with hSNF5/INI1 reexpression. This pattern is somehow paradoxical, given the disruption of actin filaments and decrease of focal adhesions commonly observed after transformation by various oncogenes (38) . However, because much of our knowledge of cytoskeleton modifications during cell transformation is derived from experiments performed with fibroblasts, it is likely that this apparent discrepancy reflects cell type specificity (39, 40, 41) . Importantly, we observe that Rho activity is decreased when hSNF5/INI1 is induced, an observation consistent with the previously documented role of Rho in various transformation processes (39) .
Altogether, this study correlates temporal changes in the gene expression profile with phenotypic modifications, including cell cycle arrest and remodeling of actin cytoskeleton. These phenomena are particularly tightly regulated during development and differentiation and altered in oncogenic processes. This work also allows the identification of putative direct targets, which should prove critical to decipher the transcriptional mechanisms of action of hSNF5/INI1.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org). I. Versteege is currently at the Institut de Recherche Servier, Division Cancérologie P04 France, 125 Chemin de Ronde, 78290 Croissy sur Seine, France.
Requests for reprints: Olivier Delattre, INSERM U509, Laboratoire de Pathologie Moléculaire des Cancers, Institut Curie, 26 rue d’Ulm, 75248 Paris Cedex 05, France. Phone: 33-1-42-34-66-81; Fax: 33-1-42-34-66-30; E-mail: olivier.delattre{at}curie.fr
Received 9/24/03. Revised 1/28/04. Accepted 2/24/04.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  X. Mao, X. Nie, F. Cao, and J. Chen Functional analysis of ScSwi1 and CaSwi1 in invasive and pseudohyphal growth of Saccharomyces cerevisiae Acta Biochim Biophys Sin, July 1, 2009; 41(7): 594 - 602. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A.-M. Lepagnol-Bestel, A. Zvara, G. Maussion, F. Quignon, B. Ngimbous, N. Ramoz, S. Imbeaud, Y. Loe-Mie, K. Benihoud, N. Agier, et al. DYRK1A interacts with the REST/NRSF-SWI/SNF chromatin remodelling complex to deregulate gene clusters involved in the neuronal phenotypic traits of Down syndrome Hum. Mol. Genet., April 15, 2009; 18(8): 1405 - 1414. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Caramel, F. Quignon, and O. Delattre RhoA-Dependent Regulation of Cell Migration by the Tumor Suppressor hSNF5/INI1 Cancer Res., August 1, 2008; 68(15): 6154 - 6161. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Morozov, S. J. Lee, Z.-K. Zhang, V. Cimica, D. Zagzag, and G. V. Kalpana INI1 Induces Interferon Signaling and Spindle Checkpoint in Rhabdoid Tumors Clin. Cancer Res., August 15, 2007; 13(16): 4721 - 4730. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. B. Zraly, F. A. Middleton, and A. K. Dingwall Hormone-response Genes Are Direct in Vivo Regulatory Targets of Brahma (SWI/SNF) Complex Function J. Biol. Chem., November 17, 2006; 281(46): 35305 - 35315. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A S-Y Leong and T Y-M Leong Newer developments in immunohistology. J. Clin. Pathol., November 1, 2006; 59(11): 1117 - 1126. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. G. Nagl Jr., D. R. Zweitzig, B. Thimmapaya, G. R. Beck Jr., and E. Moran The c-myc Gene Is a Direct Target of Mammalian SWI/SNF-Related Complexes during Differentiation-Associated Cell Cycle Arrest Cancer Res., February 1, 2006; 66(3): 1289 - 1293. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. R. Ramirez-Carrozzi, A. A. Nazarian, C. C. Li, S. L. Gore, R. Sridharan, A. N. Imbalzano, and S. T. Smale Selective and antagonistic functions of SWI/SNF and Mi-2beta nucleosome remodeling complexes during an inflammatory response Genes & Dev., February 1, 2006; 20(3): 282 - 296. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Muller, Y. Oma, L. Vallar, E. Friederich, O. Poch, and B. Winsor Sequence and Comparative Genomic Analysis of Actin-related Proteins Mol. Biol. Cell, December 1, 2005; 16(12): 5736 - 5748. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. G. Nagl Jr., A. Patsialou, D. S. Haines, P. B. Dallas, G. R. Beck Jr., and E. Moran The p270 (ARID1A/SMARCF1) Subunit of Mammalian SWI/SNF-Related Complexes Is Essential for Normal Cell Cycle Arrest Cancer Res., October 15, 2005; 65(20): 9236 - 9244. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Modena, E. Lualdi, F. Facchinetti, L. Galli, M. R. Teixeira, S. Pilotti, and G. Sozzi SMARCB1/INI1 Tumor Suppressor Gene Is Frequently Inactivated in Epithelioid Sarcomas Cancer Res., May 15, 2005; 65(10): 4012 - 4019. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. A. Hill, S. Chiosea, S. Jamaluddin, K. Roy, A. H. Fischer, D. D. Boyd, J. A. Nickerson, and A. N. Imbalzano Inducible changes in cell size and attachment area due to expression of a mutant SWI/SNF chromatin remodeling enzyme J. Cell Sci., November 15, 2004; 117(24): 5847 - 5854. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
