Oncogene (2009) 28, 1653–1668 & 2009 Macmillan Publishers Limited All rights reserved 0950-9232/09 $32.00 www.nature.com/onc REVIEW The SWI/SNF complex and cancer

D Reisman1, S Glaros1 and EA Thompson2

1Department of Internal Medicine, University of Michigan College of Medicine, Ann Arbor, MI, USA and 2Department of Cancer Biology, Mayo Clinic Comprehensive Cancer Center, Jacksonville, FL, USA

The mammalian SWI/SNF complexes mediate ATP- subset of the SWI genes are identical to those identified in dependent chromatin remodeling processes that are critical the SNF screen, and those genes that are involved both in for differentiation and proliferation. Not surprisingly, loss mating-type switching and sucrose fermentation have of SWI/SNF function has been associated with malignant come to be known as SWI/SNF genes (Peterson et al., transformation, and a substantial body of evidence 1994;Wolffe, 1994). Although SWI/SNF is a relatively indicates that several components of the SWI/SNF rare enzyme in yeast, present at only 100–500 copies per complexes function as tumor suppressors. This review nucleus (Cote et al., 1994), it has been estimated that summarizes the evidence that underlies this conclusion, 5–7% of all yeast genes require SWI/SNF activity for with particular emphasis upon the two catalytic subunits of expression (Sudarsanam et al., 2000;Zraly et al., 2006; the SWI/SNF complexes, BRM, the mammalian ortholog Monahan et al., 2008). Estimates of the role of SWI/SNF of SWI2/SNF2 in yeast and brahma in Drosophila, and in the control of Drosophila gene expression range from Brahma-related gene-1 (BRG1). 1–2% of the genome (based upon inactivation of SNR1/ Oncogene (2009) 28, 1653–1668;doi:10.1038/onc.2009.4; SNF5) to essentially the whole genome (based upon published online 23 February 2009 estimated chromosomal loss of RNA polymerase II in cells that express a dominant-negative brahma allele) Keywords: BRG1;BRM;BAF47;tumor suppressor; (Armstrong et al., 2002). In yeast, the SWI/SNF complex chromatin remodeling can both promote and suppress gene expression (Zhang et al., 2007);about a third of the yeast genes regulated by SWI/SNF are suppressed (Sudarsanam et al., 2000). The role of SWI/SNF genes in the differentiation of The SWI/SNF chromatin-remodeling complex plays multicellular organisms was initially demonstrated by essential roles in a variety of cellular processes including characterization of the Drosophila homeotic gene brahma differentiation, proliferation and DNA repair. Loss of or BRM (Tamkun, 1995). This gene was identified in SWI/SNF subunits has been reported in a number of screens for dominant suppressors of Polycomb mutations malignant cell lines and tumors, and a large number of (Tamkun et al., 1992). BRM is also a transcriptional experimental observations suggest that this complex activator of Hox genes and therefore required for the functions as a tumor suppressor. In this review, we specification of body segment identities (Armstrong et al., describe the characterization of this gene family and 2002). Sequence analysis established that BRM is a what is known about its many functions, with a focus on member of the Drosophila trithorax gene group and thus possible roles of SWI/SNF in cancer development, likely to be involved in chromatin remodeling, and BRM progression and therapy. is homologous to yeast SWI2/SNF2 (Tamkun, 1995; Papoulas et al., 1998). The Drosophila BRM amino sequence is 56% identical and 72% similar to the human Characterization of the SWI/SNF family of genes BRM protein (Elfring et al., 1994). Drosophila BRM-containing complexes bind the The components of the SWI/SNF chromatin-remodeling trithorax group Zeste to activate nucleosomal templates complex were initially identified in screens for genes that (Kal et al., 2000). Three other Drosophila trithorax genes regulate mating-type switching (SWI) and sucrose non- have been identified as SWI/SNF homologs, which, in fermenting (SNF) phenotypes in yeasts (Carlson et al., Drosophila, are known as BRM-associated proteins or 1981;Neigeborn and Carlson, 1984, 1987;Stern et al., BAPs: SNR1/BAP45 is homologous to the mammalian 1984;Abrams et al., 1986;Nasmyth and Shore, 1987; BAF47 and to yeast SNF5 (Dingwall et al., 1995); Osa is Carlson and Laurent, 1994). It was recognized that a homologous to the mammalian BAF250 and to yeast SWI1 (Collins et al., 1999;Nie et al., 2000);and Moira/ Correspondence: Dr D Reisman, Division of Hematology/Oncology, BAP155 is homologous to the mammalian BAF155/ Department of Internal Medicine, B-570B MSRBII, Campus Box BAF170 and to the yeast SWI3 (Neely and Workman, 0686, 1150 W Medical Center Drive, Ann Arbor, MI 48109-0686, 2002) (Table 1). Elucidation of the role of these genes in USA. E-mail: [email protected] pattern formation established a functional link between Received 26 September 2008;revised 1 December 2008;accepted 22 SWI/SNF activity and development in multicellular January 2009;published online 23 February 2009 organisms. SWI/SNF and cancer D Reisman et al 1654 Table 1 The different components in the yeast, Drosophila and mammalian SWI/SNF complex Mammalian SWI/SNF subunits Non-vertebrate orthologs

Subunits Gene Saccharomyces Drosophila BAF250A ARID1A SWI1 OSA BAF250B ARID1B BAF200 ARID2 BAP170 BRM SMARCA2 SWI2/SNF2 BRM BRG1 SMARCA4 SWI2/SNF2 BRM BAF180 PBRM1 RSC1 Moira/BAP155 BAF170 SMARCC2 SWI3 Moira/BAP155 BAF155 SMARCC1 SNF12 BAP60 BAF60A SMARCD1 SNF12 BAP60 BAF60B SMARCD2 SNF12 BAP60 BAF60C SMARCD3 BAF57 SMARCE1 Dalao/BAP111 BAF53A ACTL6A BAP55 BAF53B ACTL6B ARP4 BAF47 SMARCB1 SNF5 SNR1

Although some components are conserved across species, others can only uniquely be found in the higher order mammalian species.

Composition of the SWI/SNF complexes

Mammalian cells express BRM as well as a closely related protein called Brahma-related gene-1 (BRG1). Human BRG1 is approximately 74% identical to human BRM (Khavari et al., 1993), 52% identical to Drosophila BRM and 33% identical to yeast SWI2/ SNF2 (Fry and Peterson, 2001). The mammalian SWI/ SNF complexes contain, in addition to BRM or BRG1, Figure 1 The different domains that are contained in BRG1 and 8–10 subunits, which are referred to as BRM- or BRG1- BRM. Each of these proteins contain an Rb binding, ATPase and associated factors or BAFs (Wang et al., 1996a, b). Bromo domains. Despite these similarities, BRG1 and BRM are Table 1 lists all of the well-characterized mammalian only 75% similar at the proteins. BRM differs as it has a poly q BAFs and shows their non-vertebrate orthologs, the section that codes B33 repeating glutamine residues. Drosophila BAPs and yeast SWI/SNF gene family members. Two features of Table 1 warrant emphasis. First, no single mammalian SWI/SNF complex will functionally distinct is a topic of active investigation. contain all of these subunits;and second, not all The BRG1-containing complexes are further divided mammalian SWI/SNF subunits have invertebrate ortho- into those that contain the BAF250 (or OSA protein) or logs, suggesting that the mammalian SWI/SNF com- the BAF180 protein (Wang, 2003), which is thought to plexes are structurally and, perhaps functionally, more be a fusion protein of three separate RSC yeast proteins diverse than those of yeast or flies. The yeast SWI/SNF (Xue et al., 2000). BAF180 is the mammalian ortholog complex exhibits an apparent molecular mass of of Drosophila polybromo (Table 1), and the polybromo- 1.14 MDa (Smith et al., 2003), whereas the mammalian containing SWI/SNF complex has been designated as SWI/SNF complex has an apparent molecular mass of PBAF, to distinguish it from the BAF250-containing E2 MDa. The stoichiometry of the SWI/SNF com- BRM/BAF and BRG1/BAF complexes (Figure 1). plexes has not been unambiguously resolved, but it is Several other complexes have been identified that most likely that no single complex contains all of the contain BRG1 and a subset of SWI/SNF factors. These subunits listed in Table 1. Mammalian BAF proteins are complexes include the WINAC complex, which contains conventionally identified by their molecular size;hence the Williams syndrome transcription factor plus pro- BAF47 refers to a BRG1-associated protein with an teins involved in DNA replication and transcriptional apparent molecular mass of 47 kDa. BAF47 has been elongation (Aoyagi, 2005 #3;Kitagawa, 2003 #1), the shown to be homologous to both SNF5 in yeast and NUMAC complex, which contains coactivator-asso- SNR1 in Drosophila. In parallel, BAF47 was cloned as a ciated arginine methyltransferase 1 (Xu, 2004 #801), and gene that is required for lentiviral integration and was two repressor complexes that collaborate with Kap1 or thus originally referred to as INI1 (Kalpana et al., 1994). mSin3A (Underhill, 2000 #7) to recruit histone deace- It is believed that an individual SWI/SNF complex tylases (for review see Trotter and Archer (2008)). contains either BRM or BRG1, but not both (Wang Recently, a new complex has been discovered;it is in low et al., 1996a, b), such that BRM/BAF complexes abundance and contains an extra subunit, ENL, are structurally distinct from BRG1/BAF complexes that was conserved between human and yeast SWI/ (Figure 1). The extent to which these complexes are SNF (Nie et al., 2003). Interestingly, ENL is a fusion

Oncogene SWI/SNF and cancer D Reisman et al 1655 partner for the gene product of the MLL gene, which is a common target for chromosomal translocations in human acute leukemia. This MLL-ENL fusion protein associates and cooperates with SWI/SNF complexes to activate transcription of the promoter of HoxA7, a downstream target essential for the oncogenic activity of MLL-ENL.

SWI/SNF functions as an ATP-dependent chromatin-remodeling complex Figure 2 A complex set of expression pattern of the BRG1 and BRM genes that exist. Within BRM and BRG1, additional genes Genetic screens for SWI/SNF suppressor mutants exist that utilize the last 10–15 exons. The function of these identified several mutations in histones and other minigenes is not known. Moreover, both BRG1 and BRM have an chromatin-related proteins, suggesting that the SWI/ alternatively spliced exon(s). The expression of BRM is further SNF complex may act at the level of the chromatin complicated by having two different transcription start sites. (Kruger et al., 1995;Bortvin and Winston, 1996). A number of elegant biochemical studies have been Mammalian BRM and BRG1 share approximately carried out to define the functions of SWI/SNF proteins 75% identity at the amino-acid level. The first 60 amino in yeast and multicellular organisms (Hirschhorn et al., acids of BRG1 and BRM are divergent. BRM contains 1992;Kwon et al., 1994;Peterson and Tamkun, 1995; a polyQ expansion repeat that encodes approximately Tamkun, 1995;Peterson, 1998). These analyses revealed 33 glutamines (216–254 aa), whereas BRG1 contains no that the SWI/SNF proteins function as ATP-dependent such repeat (Figure 2). The length of the BRM polyQ chromatin-remodeling complexes. Several models have repeat varies among cell lines, but it is not known if been proposed to account for the ability of SWI/SNF to the expansion of this repeat affects BRM function modify chromatin structure (Peterson and Workman, (Muchardt and Yaniv, 1993). There are six conserved 2000;Hassan et al., 2001). Alternative models include domains in BRM and BRG1: QLQ domain, a proline- ATP-dependent movement of histone octamers in rich domain, a small helicase/SANT-associated domain, cis along the DNA, transfer of histone octamers from a DNA-dependent ATPase domain, an RB-binding one nucleosomal array to another or replacement of domain (LxCxE) and a Bromo domain. The carboxy- nucleosomal histones (Smith and Peterson, 2005; terminal Bromo domain [BRG1 (1455–1575);BRM Saha et al., 2006). Further studies revealed that (1380–1500)] binds acetylated histones (Winston and nucleosome remodeling by the yeast SWI/SNF is bi- Allis, 1999) and is necessary for the stable association of directional along the DNA, resulting in a continuous the SWI/SNF complex with chromatin in vitro (Martens positional re-distribution around a characteristic dis- and Winston, 2003). This binding, in conjunction with tance of motion of approximately 28 bp (Shundrovsky adherence to transcription factors, is thought to be how et al., 2006). The net result is an altered structure that SWI/SNF identifies the chromatin segment with which it is hypersensitive to nuclease digestion and exhibits will associate and subsequently elicit gene expression. increased affinity for transcription factors (Schnitzler BRM and BRG1 also contain an LxCxE motif et al., 1998) and basal transcriptional machinery, [BRG1: LTCEE (1356 aa);BRM: LTCEE (1292 aa)] such as the TATA box-binding protein, TBP (Imbal- that is essential for binding to members of the RB tumor zano et al., 1994). Such observations inform the suppressor family (Dahiya et al., 2000). Other regions prevailing view that SWI/SNF complexes regulate gene within BRG1 and BRM are also believed to facilitate expression, at least in part, by inducing a nucleosome RB family binding (Bourachot et al., 1999;Dahiya et al., conformation that is more accessible to the transcrip- 2000). The ATPase domains of BRM and BRG1 are tional machinery. composed of helicase and DEAD box domains, which The energy for SWI/SNF-mediated chromatin remo- function as the motor units that convert ATP energy to deling is transduced by the catalytic subunit, BRM or mechanical movement (Muchardt and Yaniv, 1999a). BRG1, both of which have DNA-dependent ATPase These regions are highly conserved and show a high activity (Muchardt and Yaniv, 1999a;Hassan et al., homology with both DEAD box proteins and helicase 2001). It was initially assumed that the intact SWI/SNF domain in other species (Chiba et al., 1994). Both BRG1 complex was required for chromatin remodeling. How- (210–345 aa) and BRM (255–375 aa) are highly enriched ever, this assumption has been challenged by the for proline residues, 35 and 21%, respectively, near their observation that both BRG1 and BRM have remodeling N termini. The proline-rich region in BRM is just distal activity in the absence of other subunits (Phelan et al., to the polyQ region, which is not contained in BRG1. 1999). Addition of other core subunits, such as BAF47, Both BRM and BRG1 contain B50 aa helicase/SANT- BAF155 and BAF170, increases remodeling activity to a associated domains [BRG1 (475–525);BRM (450–500)], level comparable with that of the whole SWI/SNF which are predicted to bind DNA and are often found complex, indicating that the entire complex is not associated with helicases. BRG1 (170–200 aa) and BRM absolutely necessary for chromatin remodeling, at least (168–208 aa) both contain a QLQ domain motif that has in vitro. beenshowntobeinvolvedinprotein–proteininteractions.

Oncogene SWI/SNF and cancer D Reisman et al 1656 regulate an exclusive signaling pathway;instead, it serves as a fundamental component of various essential, and often unrelated, pathways.

The SWI/SNF complex is involved in differentiation

Figure 3 Three major arranges of the SWI/SNF complex. There The SWI/SNF gene family was initially recognized are complexes that contain either BRG1 or BRM. The BRG1 because of its role in differentiation in yeast (mating- complexes are further subdivided based upon an additional type switching) and Drosophila (pattern formation). subunit: either BAF250 or BAF180. As in lung cancer, when both SWI/SNF has also been strongly linked to mammalian BRG1 and BRM are silenced, there can be no SWI/SNF activity as illustrated in this diagram. differentiation. Transforming growth factor-b is a master regulator of cellular function and differentiation, and most transforming growth factor-b responses in The BRM and BRG1 genes (SMARCA2 and human epithelial cells are dependent on BRG1 function SMARCA4, respectively) contain about 35 exons and (Xi et al., 2008). In the early developing mouse embryo, are subject to alternative splicing of a single exon near BRG1 is expressed throughout pre-implantation devel- their C termini (Figure 3). The BRG1 alternatively opment, whereas zygotic BRM is clearly detectable only spliced exon contains a high proportion of serine when differentiation first occurs, at the blastocyst stage. residues and may have some role in phosphorylation- BRM is restricted to the inner cell mass, and cell-type- dependent control of the splice variant that includes this specific expression of BRM is also observed after exon;however, the function of this exon remains differentiation of embryonic stem cells (LeGouy et al., unknown at this time. Alternative promoter utilization 1998). In adult tissue, we found that BRG1 expression is contributes additional regulatory variance to BRM, predominantly seen in cell types that constantly undergo which is transcribed from two different promoters proliferation or self-renewal, whereas BRM is prefere- (Figure 3). Both BRM and BRG1 genes have internal ntially expressed in the brain, liver, fibromuscular stroma promoters that give rise to carboxy-terminal truncated and endothelial cells, cell types not constantly engaged proteins (Figure 3). We have cloned and expressed in proliferation or self-renewal (Reisman et al., 2005). cDNAs that correspond to the BGR1 C-terminal Muscle development has been linked to the SWI/SNF fragments (unpublished data), but the function, if any, complex. Myo-D, a critical master regulator of of these proteins is unknown. mesenchymal cell differentiation, requires SWI/SNF activity (de La Serna et al., 2001a, b;Roy et al., 2002). BRG1 is required by geminin, Ngnr1 and Neuro-D for The SWI/SNF complex functions as a master regulator neuronal differentiation (Seo et al., 2005a, b). RNAi- of gene expression mediated knockdown of BAF60 in embryonic stem cells inhibits cardiovascular morphogenesis (Lickert et al., SWI/SNF is a master regulator of gene expression. 2004). A dominant-negative BRG1 allele blocks myeloid In mammalian cells, SWI/SNF has been linked to a differentiation of 32Dcl3 cells (Vradii et al., 2006), large number of transcription factors. The oncogenic whereas loss of BRG1 in keratinocytes results in defects transcription heterodimer activated protein-1 (AP-1) is in limb patterning (Indra et al., 2005). Studies in known to be SWI/SNF dependent (Ito et al., 2001); embryonic stem cells have revealed that lack of similarly, EKLF, which regulates b-hemogloblin synth- BAF250 can compromise cell pluripotency and severely esis, also requires this complex for its function (Arm- inhibit self-renewal, whereas functional BAF250 con- strong et al., 1998;Lee et al., 1999). All known steroid tributes to the proper expression of numerous genes receptors are functionally linked to SWI/SNF (Yoshi- involved in embryonic stem cell self-renewal, including naga et al., 1992;Sumi-Ichinose et al., 1997;Fryer and Sox2, Utf1 and Oct4 (Yan et al., 2008) (Gao et al., Archer, 1998;Belandia et al., 2002;Inoue et al., 2002; 2008). Furthermore, pluripotency defects in BAF250a Marshall et al., 2003;Flajollet et al., 2007). SWI/SNF mutant embryonic stem cells appear to be cell lineage has been linked to CD44, CEACAM1, E-cadherin and specific (Gao et al., 2008;Yan et al., 2008). Thus, SWI/ various integrins (Liu et al., 2001;Hendricks et al., 2004; SNF activity appears to be essential for differentiation Hill et al., 2004). It is tied to the expression of a large in yeasts, flies and mammals. It is therefore not number of interferon (IFN)-inducible genes (Wang surprising that SWI/SNF factors are also involved in et al., 2005;Yan et al., 2005). SWI/SNF has also been malignant transformation, an association most clearly shown to regulate the expression of an array of genes demonstrated with the SWI/SNF subunit BAF47. such as c-FOS, CSF-1, CRYAB, MIM-1, p21, HSP70, vimentin, cyclindromatosis and cyclin A (Murphy et al., 1999;Liu et al., 2001;Yamamichi-Nishina et al., 2003; The BAF47 subunit of the SWI/SNF complex Hendricks et al., 2004;Wang et al., 2005). SWI/SNF has is a bona fide tumor suppressor been shown to modulate alternative splicing by creating internal ‘roadblocks’ to transcriptional elongation INI1 was initially discovered as a protein that binds (Batsche et al., 2006). Hence, SWI/SNF does not HIV integrase (Kalpana et al., 1994) and was later

Oncogene SWI/SNF and cancer D Reisman et al 1657 shown to be identical to the BAF47 component of the the functions of steroid receptors (Inoue et al., 2002). SWI/SNF chromatin-remodeling complex (Wang et al., ARID1A is located in 1p36.11 (Kozmik et al., 2001), a 1996a). The gene that encodes BAF47 is located on region frequently deleted in human cancers (Huang chromosome 22q11, a region that is frequently rear- et al., 2007), and may be involved in cancer develop- ranged in pediatric rhabdoid tumors (Versteege et al., ment. BAF250A is lost in two cell lines, C33a and T47D 1998). BAF47 has been reported to undergo hetero- (DeCristofaro et al., 2001), and may be deficient in as zygous deletion in both the chronic and acute phase of many as 30% of renal carcinomas and 10% of breast CML, suggesting that it plays a role in this cancer carcinomas (Wang et al., 2004a). ARID1A- and (Grand et al., 1999). Alteration in BAF47 has also been ARID1B-containing SWI/SNF complexes appear to detected in a small subset (3–4%) of Hodgkin’s have antagonistic effects on cell cycle progression, with lymphoma patients (Yuge et al., 2000). ARID1A participating in repression and ARID1B in the The strongest evidence for the role of BAF47 in induction of key cell cycle regulators such as c-MYC cancer development comes from studies on rhabdoid (Nagl et al., 2006, 2007). tumors showing that one BAF47 allele is consistently BAF180 (polybromo, PB1, PBRM1)-containing SWI/ deleted, and the other allele is either mutated or silenced SNF complexes have different properties from those by methylation (Versteege et al., 1998;Rousseau-Merck that contain BAF250 subunits (Nie et al., 2000;Lemon et al., 1999;Sevenet et al., 1999a, b;Biegel et al., 2000, et al., 2001;Wang et al., 2004c) and are designated 2002;Biegel and Pollack, 2004). These data initially polybromo SWI/SNF complexes or PBAF (Figure 1). implicated BAF47 as a tumor suppressor gene. This BAF180 harbors a distinctive set of structural motifs, hypothesis was subsequently confirmed by the observa- characteristic of three components of RSC (Xue et al., tion that the heterozygous knockout of BAF47 (Snf5 þ /À) 2000), another chromatin-remodeling complex in yeast. in mice results in tumors that are histologically similar It also contains a Bromo domain that binds to to human malignant rhabdoid tumors. Roberts et al. acetylated histones. The BAF180 gene PBRM1 maps (2002) used a conditional inactivating allele to show that to chromosome 3p21, a frequent site of loss of loss of BAF47 results in lymphomas or rhabdoid tumors heterozygosity in human tumors;but BAF180 has not with 100% penetrance within a median time of 10 weeks yet been reported to be absent in cancer cells or primary from birth, the most tumorigenic mouse model pub- tumors (Sekine et al., 2005). lished to date. These observations strongly suggest that BAF200 is also believed to function as part of the BAF47 is a bona fide tumor suppressor. However, loss of PBAF complex, similar to BAF180 (Yan et al., 2005). BAF47 appears to be a very rare event in human tumors BAF200, like BAF250, is a member of the ARID gene (Roberts and Orkin, 2004). Its loss occurs in both family and is encoded by ARID2. In addition to the familial and sporadic renal and extra-renal rhabdoid ARID domain, BAF200 contains multiple LXXLL tumors, as well as in certain central nervous system motifs (which may mediate protein–protein interaction tumors (Biegel et al., 2002;Biegel and Pollack, 2004; between cofactors and nuclear hormone receptors), Bourdeaut et al., 2007);however, not more than 50 proline- and glutamine-rich regions, and two C2H2 Zn patients have been identified with this defect, and loss of fingers (which may interact with either DNA or BAF47 has very infrequently been observed in any adult proteins) and a DNA-binding domain conserved tumor (Grand et al., 1999;Manda et al., 2000;Yuge among the regulatory factor X family of proteins (Yan et al., 2000;DeCristofaro et al., 2001). Instead, in adult et al., 2005). Knockdown experiments indicate that tumors, the major cancer defect in SWI/SNF appears to BAF200 and BAF180 are required for PBAF-mediated be the loss of BRG1, BRM or both (Muchardt and transactivation of different sets of promoters (Yan et al., Yaniv, 2001). This appears to occur between 10 and 2005). 20% of a variety of tumor types (Reisman et al., 2003; BAF60 is a family of three separate proteins located Glaros et al., 2007). on different chromosomes: BAF60A, BAF60B and BAF60C (Wang et al., 1996a, b). Each has one or more alternative transcripts that add to the complexity of the SWI/SNF subunits and their role in gene regulation SWI/SNF complex. BAF60A has been linked to lung and cancer cancer risk (Gorlov, 2005), binds to p53 and is necessary for steroid receptor function (Hsiao et al., 2003). Baf60C The SWI/SNF complexes may be subdivided based is expressed specifically in the heart and somites in the upon their subunit composition (Figure 1). The BAF early mouse embryo (Lickert et al., 2004). Silencing of complexes contain BAF250 subunits, which belong to this gene using siRNA in mouse embryos causes defects the ARID (AT-rich DNA-interacting domain) gene in heart morphogenesis as well as results in abnormal family. Two different genes encode BAF250 proteins. cardiac and skeletal muscle differentiation (Lickert BAF250A (also known as OSA1 and p270/ARID1A) is et al., 2004) Moreover, BAF60C plays a critical role in encoded by ARID1A, whereas BAF250B (OSA2) is Notch-dependent transcriptional activation and in turn encoded by ARID1B (Hurlstone et al., 2002;Wang appears to be essential for the establishment of LR et al., 2004b). In addition to the ARID domains, asymmetry (Takeuchi et al., 2007). Like other SWI/SNF BAF250s also contain EHD1 and EHD2 domains, subunits, BAF60C has two isoforms, both of which which map to the C termini and mediate protein binding appear to be enriched in the central nervous system and (Hurlstone et al., 2002). BAF250A has been linked to also bind to and modulate the transcriptional activity of

Oncogene SWI/SNF and cancer D Reisman et al 1658 peroxisome proliferator-activated receptor-g and the shuttle between the cytoplasm and nucleus (Lee et al., retinoic acid-related orphan receptor-a1 (Debril et al., 2003). BAF53A has also been shown to bind to c-MYC 2004). Both receptors have anticancer mechanisms, and is critical for c-MYC oncogenic activity (Lee et al., suggesting that the loss of the SWI/SNF function during 2003). cancer development may impair the function of these receptors. BAF57 has a single high-mobility group (HMG) The SWI/SNF ATPase subunit genes are frequently domain that displays nonspecific DNA-binding char- silenced in cancer acteristics and contains a kinesin-like coiled-coil (KLCC) domain. BAF57 binds to and is required for The discovery that BAF47 is a bona fide tumor both estrogen and androgen receptor function (Belandia suppressor gene suggests that other SWI/SNF subunits et al., 2002;Link et al., 2005). Preliminary data suggest might also be tumor suppressors. We and others have that BAF57 may also regulate estrogen receptor examined a large number of human tumor-derived cell expression (unpublished data), but the expression of lines and primary tumors to determine the extent to BAF57 in primary tumors has not yet been investigated which loss of SWI/SNF proteins occurs during trans- and the role of BAF57 in the etiology of ER-negative formation (DeCristofaro et al., 2001) (Muchardt and breast cancer remains to be determined. Yaniv, 2001). Our laboratory examined a number of Homozygous BAF155 knockout mice develop to lung cancer and other cancer cell lines for the expression early implantation stage but undergo rapid degeneration of the various SWI/SNF subunits (DeCristofaro et al., thereafter (Han et al., 2008). About 20% of BAF155 2001;Reisman et al., 2002). We found that BRG1 and heterozygous mutant embryos display defects in brain BRM expressions are coordinately lost in about 30–40% development, angiogenesis and visceral endoderm devel- of lung cancer cell lines. Intriguingly, loss of either opment in the yolk sac (Han et al., 2008). BAF155 has BRG1 or BRM was observed at a much lower frequency been shown to stabilize BAF47, BRG1 and BAF60A by than loss of both ATPases. But both BRG1 and BRM attenuating their proteosomal degradation (Sohn et al., are concomitantly lost in about 15–20% of primary non- 2007). BAF155 may be phosphorylated by Akt, small-cell lung cancers (Reisman et al., 2003;Fukuoka suggesting that PI3K/Akt signaling may modulate et al., 2004). Analysis of more than 100 cell lines SWI/SNF function (Foster et al., 2006). Loss of revealed that both BRG1 and BRM are lost in about BAF155 has been reported in a small number of cancer 10% of established tumor cell lines (DeCristofaro et al., cell lines (DeCristofaro et al., 2001). Moreover, BAF155 2001). Immunohistochemical staining of tissue micro- maps to 3p21–23, which is not infrequency loss in array samples has shown that BRG1 or BRM is lost in human cancers (Ring et al., 1998). 10–20% of the bladder, colon, breast, melanoma, BAF170 has a significant homology with BAF155 esophageal, head/neck, pancreas and ovarian cancers (Wang et al., 1996b), 75% at the nucleotide level and (Glaros et al., 2007 and unpublished data). These data 66% at the protein level. In studies involving more than suggest that silencing of the SWI/SNF ATPases is 100 cell lines, this subunit has not been found to be involved in the etiology of a significant number and missing or altered, so any role in cancer development diversity of tumors. thus far has not been shown (DeCristofaro et al., 2001). However, BAF170 maps to 12q13–15, which is fre- quently altered in human cancers (Ring et al., 1998). Loss of heterozygosity occurs at the BRG1 and BRM loci BAF155 contains von Willebrand factor type A (vWA), SWIRM and Chromo domains, although the function BRG1 is located in or near several microsatellite of these domains in SWI/SNF activity is not well markers that demonstrate loss of heterozygosity in understood. BAF170 has been shown to regulate turn- human cancers (Medina et al., 2004;Gunduz et al., over of BAF57 (Chen and Archer, 2005). 2005), and the BRM locus is a site of loss of Several actin-related proteins were initially found to heterozygosity in human cancers as well (Eiriksdottir be constituents of the yeast SWI/SNF complex (Cairns et al., 1995;Neville et al., 1995;An et al., 1999;Jin et al., et al., 1998;Peterson et al., 1998). Mammalian SWI/ 1999;Sarkar et al., 2002;Tripathi et al., 2003;Sabah SNF complexes contain BAF53A (ACTL6A) and et al., 2005). About 26% of small-cell lung cancer cell BAF53B (ACTL6B). In addition to these actin-related lines and 76% of non-small-cell lung cancer cell lines proteins, b-actin has also been found to be associated have loss of heterozygosity at D9S288, which is in with the mammalian SWI/SNF complex (Zhao et al., proximity to the BRM gene (Girard et al., 2000). 1998). BAF53 binds the p53 tumor suppressor protein Another satellite marker, D19S221, is near the BRG1 (Lee et al., 2005;Wang et al., 2007), and the over- gene, and 23% of small-cell lung cancer cell lines and expression of an N-terminal truncated mutant of 77% of non-small-cell lung cancer cell lines exhibit loss BAF53A caused cell death (Choi et al., 2001a;Lee of heterozygosity at this marker (Girard et al., 2000). et al., 2005), implying an important role for BAF53 in Loss of heterozygosity is a hallmark of tumor suppres- cell survival. BAF53A is associated with mitotic sors, and the observation that both BRM and BRG1 chromosomes during mitosis and contributes to the loci exhibit loss of heterozygosity is consistent with the internal meshwork interactions of the chromatin fiber hypotheses that these two proteins function as tumor (Lee et al., 2007a, b). BAF53 has also been reported to suppressors.

Oncogene SWI/SNF and cancer D Reisman et al 1659 BRG1 is silenced by a number of mechanisms K-Ras-transformed fibroblasts and that introducing HDAC inhibitors blocks this inhibitory effect of Wong et al. (2000) sequenced BRG1 exons from B180 BRM. There are two carboxy-terminal acetylation sites cell lines, of which 18 cell lines were found to harbor in the BRM protein, and after the application of HDAC mutations in BRG1-coding sequences. However, most of inhibitors, they found that the resulting acetylation of the mutations were heterozygous or were missense theses site abrogated the function of BRM (Bourachot mutations and did not appreciably affect BRG1 expres- et al., 2003). Changing these sites so that they were non- sion. Similarly, Medina et al. (2008) detected mutations acetylatable blocked the inactivation of BRM by HDAC in 24% of lung cancer cell lines. Interestingly, this group inhibitors. Thus, nonspecific HDAC inhibitors, such as also examined primary lung tumors with loss of trichostatin A, butyrate and SAHA, effectively de- heterozygosity in the BRG1 locus and did not find any repress BRM expression but result in the accumulation appreciable mutations in these tumors (Medina et al., of inactive, acetylated BRM (Bourachot et al., 2003; 2004). Sentani et al. (2001) found no BRG1 mutations in Glaros et al., 2007), which does not restore BRM 8 gastric carcinoma cell lines and 33 primary gastric function. There is a great deal of interest in the carcinomas. Likewise, Valdman et al. (2003) reported no development of specific HDAC inhibitors, and it is somatic mutations in any of the 35 BRG1 exons in possible that different HDACs may be involved in samples from 21 prostate cancer patients. Gunduz et al. epigenetic silencing of the BRM gene and de-acetylation (2005), who examined loss of heterozygosity at the 19p13 of the BRM protein. If this proves to be the case, then it region in 39 oral cancers using six microsatellite markers, may be possible to use specific HDAC inhibitors to found allelic deletion in 25 of 39 (64%) samples; induce active BRM in tumor cells. however, they did not find any mutations in either the BRG1 genomic DNA or mRNA. Although mutations in BRG1 have been detected in a variety of established Transgenic knockout of BRM or BRG1 enhances cancer cell lines, such mutations have not been detected transformation in primary tumors. As only genomic DNA has been analysed from primary tumors, other mechanisms, such Transgenic knockout mouse models have been invalu- as epigenetic, protein stability or translational block, able in pre-clinical investigation of the role of SWI/SNF probably underlie the loss of BRG1 in primary tumors. in transformation. As cited above, the BAF47 knockout Further research will be needed to clarify this issue. mouse is the most tumorigenic model reported to date (Roberts et al., 2002). Knockout experiments with BRG1 and BRM have revealed that loss of either BRM is silenced by epigenetic mechanisms ATPase can potentiate cancer development in mice (Reyes et al., 1998;Bultman et al., 2000, 2008;Glaros Sequence analysis of BRM from 10 BRM-deficient cell et al., 2008). However, interpretation of the results of lines did not reveal any mutations or other alterations these experiments is complicated by the possibility that that could explain why BRM is silenced (Glaros et al., BRG1 and BRM can potentially compensate for each 2007). Unlike BRG1, experimental analysis by various other in certain circumstances. For example, in BRM groups has revealed that BRM is epigenetically silenced null mice, the expression level of BRG1 is threefold in 17 out of the 17 cell lines examined (Mizutani et al., higher than in wild-type BRM mice (Reyes et al., 1998). 2002;Bourachot et al., 2003;Yamamichi et al., 2005; Embryonic fibroblasts from BRM knockout mice Glaros et al., 2007). The observation that BRM is demonstrate striking abnormalities in cell cycle control, epigenetically silenced in cancer, rather than mutated, and BRM null mice are larger than wild-type littermates suggests that it might be possible to restore BRM (Reyes et al., 1998;Muchardt and Yaniv, 1999b;Coisy- function in tumor cells. Given that BRM is lost in Quivy et al., 2006). These data indicate that loss of B20% of a broad range of human cancers (Glaros BRM, although non-transforming, disrupts normal cell et al., 2007), and that the introduction of BRM into cycle control in a manner that cannot entirely be BRM-deficient cell lines causes growth arrest (Muchardt compensated for by BRG1 overexpression. In this et al., 1998;Muchardt and Yaniv, 1999b;Bourachot respect, BRM null mice are primed to undergo et al., 2003), it is plausible that restoration of BRM transformation, and we have reported that BRM expression might be clinically desirable. Several groups, knockout mice contain about 10 times more lung including our own, have observed that histone deacety- tumors than wild-type mice when tumors are induced lase (HDAC) inhibitors can restore BRM mRNA and by the lung carcinogen urethane (Glaros et al., 2007). protein expression in a variety of BRM null cell lines Knockout of a single allele of BRG1 results in (Mizutani et al., 2002;Bourachot et al., 2003;Yama- spontaneous tumor development in about 10% of michi et al., 2005;Glaros et al., 2007). Nuclear run-on BRG1 þ /À mice within a year (Bultman et al., 2000, transcription assays were used by Yamamichi et al. 2008). These tumors originated from the milk line and (2005) to show that BRM is strongly suppressed at the stained for histopathology markers indicative of mam- post-transcriptional level and that this suppression mary tumors. Interestingly, no alterations were noted in could be reversed by HDAC inhibitor treatment. the remaining BRG1 allele, suggesting that these tumors Furthermore, Bourachot et al. (2003) showed arose as a result of haplo-insufficiency. In addition, the that BRM overexpression inhibits the growth of absence of BRM did not change the overall penetrance

Oncogene SWI/SNF and cancer D Reisman et al 1660 of the Brg1 þ /À tumor phenotype but may have changed 2001;Glaros et al., 2007) and re-expression of either the types of tumor that occur. Biallelic knockout of BRG1 or BRM inhibits growth of such cells in culture BRG1 is embryonically lethal (Sumi-Ichinose et al., (Khavari et al., 1993;Muchardt et al., 1998). The growth 1997) (Bultman et al., 2000), and until recently it has inhibitory effects of BRG1 are attenuated by the been impossible to determine the effects of BRG1 adenovirus E1A protein, which blocks the cell cycle silencing in tumorigenesis. We have now developed a checkpoint functions of RB family members RB1, p107 novel mouse model that permits conditional biallelic and p130. The observation that E1A blocks BRG1- knockout of BRG1 in lung epithelial cells (Glaros et al., mediated growth inhibition strongly suggests that SWI/ 2008). Our data indicate that the loss of both BRG1 SNF activity is required for normal growth regulation by alleles induces apoptosis in non-transformed lung RB family members (Dunaief et al., 1994;Strober et al., epithelial cells. BRG1 is required for the establishment 1996), and BRG1 and BRM contain the RB-binding and maintenance of epithelial polarity (Indra et al., motif LxCxE and bind RB (Dahiya et al., 2000) as well as 2005), an observation that is consistent with both the RB family members p107 and p130 (Dunaief et al., 1994; pro-apoptotic effects of BRG1 knockout in the lung as Strober et al., 1996;Dahiya et al., 2000). Constitutively well as with the early embryonic lethality of BRG1 active RB does not induce G1 arrest in cells that lack knockout in mice. On the other hand, knockout of both BRG1 and BRM expression (Strobeck et al., 2000, 2001, BRG1 alleles in urethane-induced tumors promotes 2002;Zhang et al., 2000;Reisman et al., 2002). However, tumorigenesis (Glaros et al., 2008). BRG1-null, ur- restoration of BRG1 or BRM expression reconstitutes RB ethane-induced lung tumors are more abundant, larger growth inhibition in such cells. and have a higher index of proliferation than BRG1- SWI/SNF is required for E2F-dependent transcrip- positive tumors, demonstrating that the loss of BRG1 tion (Trouche et al., 1997;Kang et al., 2004;Liu et al., promotes tumorigenesis. Urethane is known to induce 2004), although all of the effects of SWI/SNF on RB K-Ras mutations, and it is possible that loss of BRG1 in checkpoint control may not require binding of BRM or cells with oncogenic K-Ras mutations promotes tumor BRG1 to RB. For example, re-expression of BRG1 development, whereas loss of BRG1 in cells with wild- induces p21, which inhibits RB phosphorylation by type K-Ras promotes apoptosis. However, this hypoth- cyclin dependent kinases (CDKs) (Kang et al., 2004). esis remains to be tested. Cyclin E has been shown to bind and, in conjunction The phenotypes of BRMÀ/À, BRG1À/À and BAF47À/À with CDK2, phosphorylate BRG1 and BAF155 (Sha- mice are strikingly different, indicating that there is nahan et al., 1999;Brumby et al., 2002). This observa- much to learn about SWI/SNF function in both normal tion raises the interesting possibility that not only does development and transformation. Our data on lung SWI/SNF regulate the cell cycle checkpoint apparatus carcinogenesis in transgenic mice suggest that BRM and (through RB), but also the cell cycle control machinery BRG1 may be involved in different stages of carcino- (through cyclin E/CDK2) may feed back to modulate genesis (initiation versus progression) (Glaros et al., SWI/SNF activity. 2008). The unresolved question is the effect of silencing The p53 tumor suppressor has also been functionally or mutating the genes that encode both of the catalytic linked to SWI/SNF (Lee et al., 2002;Wang et al., 2007). SWI/SNF subunits. Loss of both BRM and BRG1 p53 binds BAF53, and SWI/SNF activity is necessary occurs with significant frequency in many different types for p53-mediated transcription activation and p53- of solid tumors, and in lung cancer, loss of both subunits mediated cell cycle control (Bochar et al., 2000;Wang is more common than loss of either, alone (Reisman et al., 2007). It has been reported that p53-mediated cell et al., 2003;Fukuoka et al., 2004). As noted above, cycle arrest is dependent upon the RB family member BAF47 knockout is the most tumorigenic defect that has p130 (Claudio et al., 2000;Gao et al., 2002;Kapic et al., been engineered to date. However, SWI/SNF activity is 2006), which is known to bind BRG1 and BRM, and, by maintained at some level even in the absence of BAF47 analogy with RB1, is likely to require SWI/SNF activity (Doan et al., 2004), which seems to function more as an for function. Thus, the role of SWI/SNF in p53- enhancer of remodeling activity, rather than an ob- mediated checkpoint control is likely to involve both ligatory component of the chromatin-remodeling com- p53 itself as well as downstream effectors of p53 plex (Phelan et al., 1999). Conversely, concomitant loss signaling such as p130. Irrespective of the mechanism, of both BRG1 and BRM should result in the complete it is clear that loss of SWI/SNF activity, by silencing loss of SWI/SNF-mediated ATP-dependent chromatin- BRM and BRG1, induces a phenocopy of a p53 loss-of- remodeling activity, with attendant degradation of the function mutation. several tumor suppressor functions that require SWI/ SNF. Presumably, therefore, cells that have lost both BRM and BRG1 should be highly tumorigenic. BRM and BRG1 are linked to DNA repair

SWI/SNF complex plays an important role in DNA BRM and BRG1 play critical roles in the control repair (Wuebbles and Jones, 2004;Morrison and Shen, of cell proliferation 2006;Menoni et al., 2007;Osley et al., 2007). DNA repair proteins such as p53, BRCA1 and Fanconi A large number of tumor cells have lost both BRM and anemia proteins associate with the SWI/SNF complex BRG1 expression (Wong et al., 2000;DeCristofaro et al., and are functionally dependent on it (Bochar et al.,

Oncogene SWI/SNF and cancer D Reisman et al 1661 2000;Otsuki et al., 2001;Lee et al., 2002;Wang et al., SNF function will form an important focus of future 2007). BRCA1 binds BRG1, and p53-mediated stimula- investigations in our laboratory and others. tion of transcription by BRCA1 is dependent on BRG1 (Bochar et al., 2000). BAF47-deficient cells are hyper- sensitive to genotoxic stress (Klochendler-Yeivin et al., BRM and BRG1 are required for the activity 2006), and re-expression of BRG1 in BRG1-deficient of some nuclear receptors SW13 cells renders the cells resistant to ultraviolet- induced DNA damage (Gong et al., 2008). SWI/SNF Steroid receptors have been functionally linked to the facilitates the repair of cyclobutane pyrimidine dimers SWI/SNF complex (Yoshinaga et al., 1992). In Droso- (Gaillard et al., 2003) and acetylaminofluorene–guanine phila, SWI/SNF has been shown to regulate hormone- adducts (Hara and Sancar, 2002) in a nucleosomal responsive ecdysone-induced genes (Schwartz et al., context. Studies carried out in vivo indicate that SWI/ 2004;Zraly et al., 2006). In mammalian cells, the SNF is recruited to double-strand break sites (Chai glucocorticoid receptor has been functionally linked to et al., 2005), and the inactivation of the SWI/SNF the SWI/SNF complex (Muchardt and Yaniv, 1993; complex results in inefficient DNA double-strand break Ostlund Farrants et al., 1997;Trotter and Archer, 2004), repair in vivo (Park et al., 2006). These observations and the re-expression of either BRG1 or BRM is suggest that SWI/SNF activity (or expression of critical sufficient to restore glucocorticoid receptor function in SWI/SNF subunits such as BRM and BRG1) cells that lack both BRG1 and BRM (Glaros et al., may predict tumor cell susceptible to radiation and 2007). Although it has not been investigated, the loss of chemotherapy with DNA-damaging agents. BRG1 and BRM could very well underlie the resistance to glucocorticoids in hematopoietic malignancies such as myeloma, lymphoma and leukemia, as well as in lung BRM and BRG1 control the expression of genes cancer (Choi et al., 2001b;Ko et al., 2004;Pottier et al., that are involved in cellular adhesion 2007). The estrogen receptor is also functionally linked to the SWI/SNF complex (Ichinose et al., 1997; When either BRG1 or BRM is restored in BRG1/BRM- Belandia et al., 2002), and the androgen receptor is also deficient cell lines, cells show an increase in cell volume, SWI/SNF dependent (Inoue et al., 2002;Marshall et al., area of attachment, nuclear size and growth arrest (Hill 2003;Hong et al., 2005;Link et al., 2005;Dai et al., et al., 2004). Examination of focal adhesions reveals 2008). We have found that BRG1 and BRM are lost in changes in paxillin distribution, whereas increases in cell human prostate cancer samples, suggesting that a loss of size and shape correlate with the overexpression of two SWI/SNF function may play a role in the development integrins and the urokinase-type plasminogen activator of hormone-refractory prostate cancer. It has been receptor (Hill et al., 2004). These findings suggest that shown that the retinoic acid receptor is also tied to SWI/SNF regulates the expression of cell-adhesion SWI/SNF (Sumi-Ichinose et al., 1997;Flajollet et al., proteins and cytoskeleton structure changes. The exact 2007), suggesting that the inactivation of SWI/SNF genes underlying these phenomena are not completely would make tumor cells refractory to clinical interven- known. However, the cell-adhesion proteins CD44 and tion with this reagent. As nuclear receptors are E-cadherin are regulated by SWI/SNF (Banine et al., important regulators of both proliferation and differ- 2005). CD44 transcription appears to require SWI/SNF entiation, the abrogation of the SWI/SNf complex activity, but the regulation of E-cadherin expression would block hormone-sensitive pathways and thereby appears to be more complex and may involve splicing lead to cancer progression. (Batsche et al., 2006). Our preliminary data indicate that BRG1 regulates the expression of several genes that are involved in controlling Rho family GTPase activity BRM and BRG1 play important roles in the immune (including RhoGDIA and IQGAP), and that these system proteins may, in combination with CD44 and ERM family members, affect the stability of adherens junc- SWI/SNF plays an important role in immune responses, tions. As loss of CD44 and E-cadherin are commonly T-cell development and recombination events (Golding associated with transformation of epithelial cells, this et al., 1999;Agalioti et al., 2000;Kwon et al., 2000; aspect of SWI/SNF function warrants additional Spicuglia et al., 2002;Morshead et al., 2003). The SWI/ investigation. We and others have carried out prelimin- SNF complexes have been shown to remodel chromatin ary microarray experiments to identify BRG1-regulated in such a way as to drive consecutive developmental genes, and our data suggest that SWI/SNF regulates the transitions in T cells in response to external stimuli. expression of a cohort of important cell-adhesion Specific inactivation of SWI/SNF complexes in T cells proteins such as CEACAM1, as well as the extracellular demonstrates that such complexes are essential for matrix proteins such as Sparc, MMP1, MMP2, transge- thymocyte development. SWI/SNF further contributes lin, GALS3BP P8, PODXL, Integrin A5, LGALS1, to CD4/CD8 T-cell lineage divergence by repressing Integrin A3, TIMP3 and PLAU (Liu et al., 2001; CD4 receptor expression while activating the expression Hendricks et al., 2004;Wang et al., 2005). All of these of CD8 (Chi et al., 2002, 2003;Gebuhr et al., 2003). interactions are known to be involved in tumor Variable-diversity-joining (V(D)J) recombination has progression, and we anticipate that this aspect of SWI/ also been shown to be mediated by SWI/SNF. BRG1

Oncogene SWI/SNF and cancer D Reisman et al 1662 stimulates in vitro recombination of chromatin and is from transgenic knockout mice indicate that these two also bound to regions in the T-cell receptor (TCR) and complexes are not entirely redundant in their functions immunoglobulin loci that take part in recombination in vivo. BRG1 knockout mice are embryonically lethal, events (Golding et al., 1999;Kwon et al., 2000;Spicuglia whereas BRM mice develop more or less normally. et al., 2002;Morshead et al., 2003;Patenge et al., 2004). BRM knockout mice express very low levels of CD44, More recent studies have shown SWI/SNF to be even though BRG1 levels are increased threefold in fundamental to the promoter-directed assembly of BRM null mice (Reisman et al., 2002). Paradoxically, TCR-b (TCRB) genes (Osipovich et al., 2007);SWI/ the re-expression of either BRG1 or BRM into BRG1/ SNF is recruited to promoters and then facilitates BRM-deficient cell lines induces CD44 expression in recombination by exposing segments of genomic DNA culture (Strobeck et al., 2001;Reisman et al., 2002). This to V(D)J recombinase. The fact that loss of this paradox emphasizes the point that one may not be able chromatin-remodeling complex in thymocytes inacti- to use transient overexpression of individual ATPase vates recombinase targets at the endogenous TCRB subunits to discriminate between functions that are non- locus (Osipovich et al., 2007) provides further evidence redundant when BRM or BRG1 is expressed at for the role of SWI/SNF in recombination. physiological levels. Hence caution must be used in SWI/SNF also plays an important role in orchestrat- interpreting the results of experiments in which BRG1 ing IFN-induced responses. BRG1 regulates the expres- or BRM is transiently overexpressed. Additional studies sion of IFN-b, which is induced upon viral infection of will be required to dissect the redundant and non- many cell types, and is essential for innate viral redundant functions of the individual ATPase subunits, immunity (Agalioti et al., 2000). Microarray experi- which may vary in different cells depending upon the ments have further revealed that many IFN-inducible relative abundance of BRM and BRG1. proteins are SWI/SNF dependent (Yan et al., 2005). BRG1 induces the expression of a subset of IFN-a- activated genes in HeLa cells (Huang et al., 2002) Perspectives (Liu et al., 2002) and has been shown to be required for IFN-a to inhibit viral replication. Forced expression of Broadly speaking, it is clear that the SWI/SNF BRG1 in BRM/BRG1-deficient SW13 cells leads to the complexes are required for a number of processes that upregulation of various IFN-a target genes (Liu et al., are critical for cell cycle checkpoint control and 2002). Finally, yeast two-hybrid studies have shown that differentiation. Abrogation of the normal control BRG1 binds to signal transducer and activator of processes is essential for tumor growth and progression, transcription 2 (STAT2), an IFN-a-activated transcrip- and loss-of-function mutations in the RB and p53 tion factor (Huang et al., 2002). Importantly, the pathways are among the most common oncogenic induction of all of these genes has been shown to be events. Mutations that affect cellular adhesion are inhibited when BAF47 is blocked by RNAi interference essential for tumor progression and metastases, and (Coisy-Quivy et al., 2006). As a consequence of BAF47 other mutations are required to suppress the process of inhibition, cellular response to viral infections and anoikis, which is normally triggered by the disruption of cellular antiviral activity are significantly inhibited cellular adhesion and polarity. There is evidence that (Cui et al., 2004). Thus, the tumors that lack a links all of these events to SWI/SNF activity, and the functional SWI/SNF may be unable to suppress viral preponderance of data suggests that the SWI/SNF infection and replication. As such, these tumors might complexes function as tumor suppressors. From a become viral factories, leading to a prolonged cytokine mechanistic standpoint, it is clear that we have much surge by surrounding normal cells. This may be one to learn about how the SWI/SNF complexes control mechanism of tumor-induced cachexia. these processes. A critical question relates to the Major histocompatibility complex class I and II gene numerous observations that suggest that all SWI/SNF expression is also regulated by SWI/SNF. IFN-g components are not equivalent in this respect. For induction of CIITA, the master regulator of major example, loss of BAF47 is rare and is associated with a histocompatibility complex class II expression (Mudha- relatively small subset of tumor types, whereas loss of sani and Fontes, 2002), requires SWI/SNF. In addition, BRM and BRG1 is relatively common and is observed in BRG1 also associates with the activation of the a wide variety of solid tumors. Loss of BRM appears to enhancer A of the major histocompatibility complex predispose to tumor formation, whereas loss of BRG1 class I promoter (Brockmann et al., 2001). Major appears to be lethal in non-transformed cells, but to histocompatibility complex class I and II proteins are promote tumor progression in cells that have acquired essential for immune surveillance, and loss of SWI/SNF oncogenic mutations. These observations seem to function could conceivably make tumors invisible to the provide a rational basis for understanding why loss of immune system. both BRG1 and BRM is more common in tumors than loss of either alone. Furthermore, the connection between BRG1/BRM and tumor suppressors such as BRM and BRG1 exhibit partial functional redundancy RB and p53 suggests that the loss of the SWI/SNF catalytic subunits represents an alternative route to Biochemical evidence indicates that BRM and BRG1 inactivation of checkpoint control in the absence of nucleate different SWI/SNF/BAF complexes, and data loss-of-function mutations to the RB and p53 pathways.

Oncogene SWI/SNF and cancer D Reisman et al 1663 Data from our laboratory and several others suggest that HDAC inhibitors that will induce BRM but will not the etiology of 10–20% of all solid tumors is dependent result in the accumulation of inactive, acetylated BRM upon loss of both BRM and BRG1. Such observations protein. Such drugs could have wide applicability in the emphasize the need for additional studies to determine treatment of tumors that have lost BRM and BRG1, how these proteins individually and jointly impinge upon and the number of such tumors is quite significant. The proliferation, survival and progression of tumor cells. link between SWI/SNF activity and DNA repair is also From a translational perspective, one is drawn to the of potential clinical interest. As loss of BRG1 appears to observations (1) that the re-introduction of BRM into increase radiosensitivity of tumor cells, it is plausible cells that have lost both BRM and BRG1 appears to that inhibitors of this ATPase could be used as cause terminal differentiation, or at least irreversible adjuvants in radiotherapy. We are only now beginning withdrawal from the cell cycle and (2) that BRM is to recognize the clinical potential of BRM/BRG1 as epigenetically silenced and can, in theory, be induced in either therapeutic targets or biomarkers of chemo- or tumor cells. The initial work with HDAC inhibitors radiotherapy, and the promise of these two gene indicates that it may be possible to develop specific products in individualized medicine is significant.

References

Abrams E, Neigeborn L, Carlson M. (1986). Molecular analysis of SNF2 group protein I/Y-like DNA binding domain. Mol Cell Biol 19: and SNF5, genes required for expression of glucose-repressible genes in 3931–3939. Saccharomyces cerevisiae. MolCellBiol6: 3643–3651. Bourachot B, Yaniv M, Muchardt C. (2003). Growth inhibition by the Agalioti T, Lomvardas S, Parekh B, Yie J, Maniatis T, Thanos D. mammalian SWI-SNF subunit Brm is regulated by acetylation. (2000). Ordered recruitment of chromatin modifying and general EMBO J 22: 6505–6515. transcription factors to the IFN-beta promoter [In Process Bourdeaut F, Freneaux P, Thuille B, Lellouch-Tubiana A, Nicolas A, Citation]. Cell 103: 667–678. Couturier J et al. (2007). hSNF5/INI1-deficient tumours and An HX, Claas A, Savelyeva L, Seitz S, Schlag P, Scherneck S et al. rhabdoid tumours are convergent but not fully overlapping entities. (1999). Two regions of deletion in 9p23–24 in sporadic breast J Pathol 211: 323–330. cancer. Cancer Res 59: 3941–3943. Brockmann D, Lehmkuhler O, Schmucker U, Esche H. (2001). The Aoyagi S, Trotter KW, Archer TK. (2005). ATP-dependent chromatin histone acetyltransferase activity of PCAF cooperates with the remodeling complexes and their role in nuclear receptor-dependent brahma/SWI2-related protein BRG-1 in the activation of the transcription in vivo. Vitam Horm 70: 281–307. enhancer A of the MHC class I promoter. Gene 277: 111–120. Armstrong JA, Bieker JJ, Emerson BM. (1998). A SWI/SNF- Brumby AM, Zraly CB, Horsfield JA, Secombe J, Saint R, Dingwall related chromatin remodeling complex, E-RC1, is required for AK et al. (2002). Drosophila cyclin E interacts with components of tissue-specific transcriptional regulation by EKLF in vitro. Cell 95: the Brahma complex. EMBO J 21: 3377–3389. 93–104. Bultman S, Gebuhr T, Yee D, La Mantia C, Nicholson J, Gilliam A Armstrong JA, Papoulas O, Daubresse G, Sperling AS, Lis JT, Scott et al. (2000). A Brg1 null mutation in the mouse reveals functional MP et al. (2002). The Drosophila BRM complex facilitates global differences among mammalian SWI/SNF complexes. Mol Cell 6: transcription by RNA polymerase II. EMBO J 19: 5245–5254. 1287–1295. Banine F, Bartlett C, Gunawardena R, Muchardt C, Yaniv M, Bultman SJ, Herschkowitz JI, Godfrey V, Gebuhr TC, Yaniv M, Knudsen ES et al. (2005). SWI/SNF chromatin-remodeling factors Perou CM et al. (2008). Characterization of mammary tumors from induce changes in DNA methylation to promote transcriptional Brg1 heterozygous mice. Oncogene 27: 460–468. activation. Cancer Res 65: 3542–3547. Cairns BR, Erdjument-Bromage H, Tempst P, Winston F, Kornberg Batsche E, Yaniv M, Muchardt C. (2006). The human SWI/SNF RD. (1998). Two actin-related proteins are shared functional subunit Brm is a regulator of alternative splicing. Nat Struct Mol components of the chromatin-remodeling complexes RSC and Biol 13: 22–29. SWI/SNF. Mol Cell 2: 639–651. Belandia B, Orford RL, Hurst HC, Parker MG. (2002). Targeting of Carlson M, Laurent BC. (1994). The SNF/SWI family of global SWI/SNF chromatin remodelling complexes to estrogen-responsive transcriptional activators. Curr Opin Cell Biol 6: 396–402. genes. EMBO J 21: 4094–4103. Carlson M, Osmond BC, Botstein D. (1981). Mutants of yeast Biegel JA, Fogelgren B, Zhou JY, James CD, Janss AJ, Allen JC et al. defective in sucrose utilization. Genetics 98: 25–40. (2000). Mutations of the INI1 rhabdoid tumor suppressor gene in Chai B, Huang J, Cairns BR, Laurent BC. (2005). Distinct roles for the medulloblastomas and primitive neuroectodermal tumors of the RSC and Swi/Snf ATP-dependent chromatin remodelers in DNA central nervous system. Clin Cancer Res 6: 2759–2763. double-strand break repair. Genes Dev 19: 1656–1661. Biegel JA, Kalpana G, Knudsen ES, Packer RJ, Roberts CW, Thiele Chen J, Archer TK. (2005). Regulating SWI/SNF subunit levels CJ et al. (2002). The role of INI1 and the SWI/SNF complex in the via protein-protein interactions and proteasomal degradation: development of rhabdoid tumors: meeting summary from the BAF155 and BAF170 limit expression of BAF57. Mol Cell Biol workshop on childhood atypical teratoid/rhabdoid tumors. Cancer 25: 9016–9027. Res 62: 323–328. Chi TH, Wan M, Lee PP, Akashi K, Metzger D, Chambon P et al. Biegel JA, Pollack IF. (2004). Molecular analysis of pediatric brain (2003). Sequential roles of Brg, the ATPase subunit of BAF tumors. Curr Oncol Rep 6: 445–452. chromatin remodeling complexes, in thymocyte development. Bochar DA, Wang L, Beniya H, Kinev A, Xue Y, Lane WS et al. Immunity 19: 169–182. (2000). BRCA1 is associated with a human SWI/SNF-related Chi TH, Wan M, Zhao K, Taniuchi I, Chen L, Littman DR et al. complex: linking chromatin remodeling to breast cancer. Cell 102: (2002). Reciprocal regulation of CD4/CD8 expression by SWI/ 257–265. SNF-like BAF complexes. Nature 418: 195–199. Bortvin A, Winston F. (1996). Evidence that Spt6p controls chromatin Chiba H, Muramatsu M, Nomoto A, Kato H. (1994). Two human structure by a direct interaction with histones. Science 272: homologues of Saccharomyces cerevisiae SWI2/SNF2 and Droso- 1473–1476. phila brahma are transcriptional coactivators cooperating with the Bourachot B, Yaniv M, Muchardt C. (1999). The activity of estrogen receptor and the retinoic acid receptor. Nucleic Acids Res mammalian brm/SNF2alpha is dependent on a high-mobility- 22: 1815–1820.

Oncogene SWI/SNF and cancer D Reisman et al 1664 Choi EY, Park JA, Sung YH, Kwon H. (2001a). Generation of the Foster KS, McCrary WJ, Ross JS, Wright CF. (2006). Members of the dominant-negative mutant of hArpNbeta: a component of human hSWI/SNF chromatin remodeling complex associate with and are SWI/SNF chromatin remodeling complex. Exp Cell Res 271: phosphorylated by protein kinase B/Akt. Oncogene 25: 4605–4612. 180–188. Fry CJ, Peterson CL. (2001). Chromatin remodeling enzymes: who’s Choi YI, Jeon SH, Jang J, Han S, Kim JK, Chung H et al. (2001b). on first? Curr Biol 11: R185–R197. Notch1 confers a resistance to glucocorticoid-induced apoptosis on Fryer CJ, Archer TK. (1998). Chromatin remodelling by the developing thymocytes by down-regulating SRG3 expression. Proc glucocorticoid receptor requires the BRG1 complex. Nature 393: Natl Acad Sci USA 98: 10267–10272. 88–91. Claudio PP, Howard CM, Fu Y, Cinti C, Califano L, Micheli P et al. Fukuoka J, Fujii T, Shih JH, Dracheva T, Meerzaman D, Player A (2000). Mutations in the retinoblastoma-related gene RB2/p130 in et al. (2004). Chromatin remodeling factors and BRM/BRG1 primary nasopharyngeal carcinoma. Cancer Res 60: 8–12. expression as prognostic indicators in non-small cell lung cancer. Coisy-Quivy M, Disson O, Roure V, Muchardt C, Blanchard JM, Clin Cancer Res 10: 4314–4324. Dantonel JC. (2006). Role for brm in cell growth control. Cancer Gaillard H, Fitzgerald DJ, Smith CL, Peterson CL, Richmond TJ, Res 66: 5069–5076. Thoma F. (2003). Chromatin remodeling activities act on Collins RT, Furukawa T, Tanese N, Treisman JE. (1999). Osa UV-damaged nucleosomes and modulate DNA damage accessibility associates with the Brahma chromatin remodeling complex to photolyase. J Biol Chem 278: 17655–17663. and promotes the activation of some target genes. EMBO J 18: Gao CF, Ren S, Wang J, Zhang SL, Jin F, Nakajima T et al. (2002). 7029–7040. P130 and its truncated form mediate p53-induced cell cycle arrest in Cote J, Quinn J, Workman JL, Peterson CL. (1994). Stimulation of Rb(À/À) Saos2 cells. Oncogene 21: 7569–7579. GAL4 derivative binding to nucleosomal DNA by the yeast SWI/ Gao X, Tate P, Hu P, Tjian R, Skarnes WC, Wang Z. (2008). ES cell SNF complex. Science 265: 53–60. pluripotency and germ-layer formation require the SWI/SNF Cui K, Tailor P, Liu H, Chen X, Ozato K, Zhao K. (2004). The chromatin remodeling component BAF250a. Proc Natl Acad Sci chromatin-remodeling BAF complex mediates cellular antiviral USA 105: 6656–6661. activities by promoter priming. Mol Cell Biol 24: 4476–4486. Gebuhr TC, Kovalev GI, Bultman S, Godfrey V, Su L, Magnuson T. Dahiya A, Gavin MR, Luo RX, Dean DC. (2000). Role of (2003). The role of Brg1, a catalytic subunit of mammalian the LXCXE binding site in Rb function. Mol Cell Biol 20: chromatin-remodeling complexes, in T cell development. J Exp 6799–6805. Med 198: 1937–1949. Dai Y, Ngo D, Jacob J, Forman LW, Faller DV. (2008). Prohibitin Girard L, Zochbauer-Muller S, Virmani AK, Gazdar AF, Minna JD. and the SWI/SNF ATPase subunit BRG1 are required for effective (2000). Genome-wide allelotyping of lung cancer identifies androgen antagonist-mediated transcriptional repression of andro- new regions of allelic loss, differences between small cell lung gen receptor-regulated genes. Carcinogenesis 29: 1725–1733. cancer and non-small cell lung cancer, and loci clustering. Cancer de La Serna IL, Carlson KA, Imbalzano AN. (2001a). Mammalian Res 60: 4894–4906. SWI/SNF complexes promote MyoD-mediated muscle differentia- Glaros S, Cirrincione GM, Muchardt C, Kleer CG, Michael CW, tion. Nat Genet 27: 187–190. Reisman D. (2007). The reversible epigenetic silencing of BRM: de La Serna IL, Roy K, Carlson KA, Imbalzano AN. (2001b). MyoD implications for clinical targeted therapy. Oncogene 26: 7058–7066. can induce cell cycle arrest but not muscle differentiation in the Glaros S, Cirrincione GM, Palanca A, Metzger D, Reisman D. (2008). presence of dominant negative SWI/SNF chromatin remodeling Targeted knockout of BRG1 potentiates lung cancer development. enzymes. J Biol Chem 276: 41486–41491. Cancer Res 68: 3689–3696. Debril MB, Gelman L, Fayard E, Annicotte JS, Rocchi S, Auwerx J. Golding A, Chandler S, Ballestar E, Wolffe AP, Schlissel MS. (1999). (2004). Transcription factors and nuclear receptors interact with the Nucleosome structure completely inhibits in vitro cleavage by the SWI/SNF complex through the BAF60c subunit. J Biol Chem 279: V(D)J recombinase. EMBO J 18: 3712–3723. 16677–16686. Gong F, Fahy D, Liu H, Wang W, Smerdon MJ. (2008). Role of the DeCristofaro MF, Betz BL, Rorie CJ, Reisman DN, Wang W, mammalian SWI/SNF chromatin remodeling complex in the Weissman BE. (2001). Characterization of SWI/SNF protein cellular response to UV damage. Cell Cycle 7: 1067–1074. expression in human breast cancer cell lines and other malignancies. Gorlov I. (2005). Identification of a novel lung cancer candidate J Cell Physiol 186: 136–145. susceptibility gene using a familial aggregation approach. Proceed- Dingwall AK, Beek SJ, McCallum CM, Tamkun JW, Kalpana GV, ings of the 97th Annual AACR Meeting, Washington, DC 3: 41. Goff SP et al. (1995). The Drosophila snr1 and brm proteins are Grand F, Kulkarni S, Chase A, Goldman JM, Gordon M, Cross NC. related to yeast SWI/SNF proteins and are components of a large (1999). Frequent deletion of hSNF5/INI1, a component of the protein complex. Mol Biol Cell 6: 777–791. SWI/SNF complex, in chronic myeloid leukemia. Cancer Res 59: Doan DN, Veal TM, Yan Z, Wang W, Jones SN, Imbalzano AN. 3870–3874. (2004). Loss of the INI1 tumor suppressor does not impair the Gunduz E, Gunduz M, Ouchida M, Nagatsuka H, Beder L, Tsujigiwa expression of multiple BRG1-dependent genes or the assembly of H et al. (2005). Genetic and epigenetic alterations of BRG1 promote SWI/SNF enzymes. Oncogene 23: 3462–3473. oral cancer development. Int J Oncol 26: 201–210. Dunaief JL, Strober BE, Guha S, Khavari PA, Alin K, Luban J et al. Han D, Jeon S, Sohn DH, Lee C, Ahn S, Kim WK et al. (2008). (1994). The retinoblastoma protein and BRG1 form a complex and SRG3, a core component of mouse SWI/SNF complex, is cooperate to induce cell cycle arrest. Cell 79: 119–130. essential for extra-embryonic vascular development. Dev Biol 315: Eiriksdottir G, Sigurdsson A, Jonasson JG, Agnarsson BA, Sigurdsson 136–146. H, Gudmundsson J et al. (1995). Loss of heterozygosity on Hara R, Sancar A. (2002). The SWI/SNF chromatin-remodeling factor chromosome 9 in human breast cancer: association with clinical stimulates repair by human excision nuclease in the mononucleo- variables and genetic changes at other chromosome regions. Int J some core particle. Mol Cell Biol 22: 6779–6787. Cancer 64: 378–382. Hassan AH, Neely KE, Vignali M, Reese JC, Workman JL. (2001). Elfring LK, Deuring R, McCallum CM, Peterson CL, Tamkun JW. Promoter targeting of chromatin-modifying complexes. Front Biosci (1994). Identification and characterization of Drosophila relatives of 6: D1054–D1064. the yeast transcriptional activator SNF2/SWI2. Mol Cell Biol 14: Hendricks KB, Shanahan F, Lees E. (2004). Role for BRG1 in cell 2225–2234. cycle control and tumor suppression. Mol Cell Biol 24: 362–376. Flajollet S, Lefebvre B, Cudejko C, Staels B, Lefebvre P. (2007). The Hill DA, Chiosea S, Jamaluddin S, Roy K, Fischer AH, Boyd DD core component of the mammalian SWI/SNF complex SMARCD3/ et al. (2004). Inducible changes in cell size and attachment area due BAF60c is a coactivator for the nuclear retinoic acid receptor. Mol to expression of a mutant SWI/SNF chromatin remodeling enzyme. Cell Endocrinol 270: 23–32. J Cell Sci 117: 5847–5854.

Oncogene SWI/SNF and cancer D Reisman et al 1665 Hirschhorn JN, Brown SA, Clark CD, Winston F. (1992). Evidence repressing the SRG3 expression via Ras activation. J Biol Chem that SNF2/SWI2 and SNF5 activate transcription in yeast by 279: 21903–21915. altering chromatin structure. Genes Dev 6: 2288–2298. Kozmik Z, Machon O, Kralova J, Kreslova J, Paces J, Vlcek C. (2001). Hong CY, Suh JH, Kim K, Gong EY, Jeon SH, Ko M et al. (2005). Characterization of mammalian orthologues of the Drosophila osa Modulation of androgen receptor transactivation by the SWI3- gene: cDNA cloning, expression, chromosomal localization, and related gene product (SRG3) in multiple ways. Mol Cell Biol 25: direct physical interaction with Brahma chromatin-remodeling 4841–4852. complex. Genomics 73: 140–148. Hsiao PW, Fryer CJ, Trotter KW, Wang W, Archer TK. (2003). Kruger W, Peterson CL, Sil A, Coburn C, Arents G, Moudrianakis BAF60a mediates critical interactions between nuclear receptors EN et al. (1995). Amino acid substitutions in the structured and the BRG1 chromatin-remodeling complex for transactivation. domains of histones H3 and H4 partially relieve the requirement Mol Cell Biol 23: 6210–6220. of the yeast SWI/SNF complex for transcription. Genes Dev 9: Huang J, Zhao YL, Li Y, Fletcher JA, Xiao S. (2007). Genomic and 2770–2779. functional evidence for an ARID1A tumor suppressor role. Genes Kwon H, Imbalzano AN, Khavari PA, Kingston RE, Green MR. Chromosomes Cancer 46: 745–750. (1994). Nucleosome disruption and enhancement of activator Huang M, Qian F, Hu Y, Ang C, Li Z, Wen Z. (2002). Chromatin- binding by a human SW1/SNF complex [see comments]. Nature remodelling factor BRG1 selectively activates a subset of interferon- 370: 477–481. alpha-inducible genes. Nat Cell Biol 4: 774–781. Kwon J, Morshead KB, Guyon JR, Kingston RE, Oettinger MA. Hurlstone AF, Olave IA, Barker N, van Noort M, Clevers H. (2002). (2000). Histone acetylation and hSWI/SNF remodeling act in Cloning and characterization of hELD/OSA1, a novel BRG1 concert to stimulate V(D)J cleavage of nucleosomal DNA. Mol Cell interacting protein. Biochem J 364: 255–264. 6: 1037–1048. Ichinose H, Garnier JM, Chambon P, Losson R. (1997). Ligand- Lee CH, Murphy MR, Lee JS, Chung JH. (1999). Targeting a SWI/ dependent interaction between the estrogen receptor and the human SNF-related chromatin remodeling complex to the beta- homologues of SWI2/SNF2. Gene 188: 95–100. globin promoter in erythroid cells. Proc Natl Acad Sci USA 96: Imbalzano AN, Kwon H, Green MR, Kingston RE. (1994). 12311–12315. Facilitated binding of TATA-binding protein to nucleosomal Lee D, Kim JW, Seo T, Hwang SG, Choi EJ, Choe J. (2002). SWI/ DNA. Nature 370: 481–485. SNF complex interacts with tumor suppressor p53 and is necessary Indra AK, Dupe V, Bornert JM, Messaddeq N, Yaniv M, Mark M for the activation of p53-mediated transcription. J Biol Chem 277: et al. (2005). Temporally controlled targeted somatic mutagenesis in 22330–22337. embryonic surface ectoderm and fetal epidermal keratinocytes Lee JH, Chang SH, Shim JH, Lee JY, Yoshida M, Kwon H. (2003). unveils two distinct developmental functions of BRG1 in limb Cytoplasmic localization and nucleo-cytoplasmic shuttling morphogenesis and skin barrier formation. Development 132: of BAF53, a component of chromatin-modifying complexes. 4533–4544. Mol Cells 16: 78–83. Inoue H, Furukawa T, Giannakopoulos S, Zhou S, King DS, Tanese Lee JH, Lee JY, Chang SH, Kang MJ, Kwon H. (2005). Effects of Ser2 N. (2002). Largest subunits of the human SWI/SNF chromatin and Tyr6 mutants of BAF53 on cell growth and p53-dependent remodeling complex promote transcriptional activation by steroid transcription. Mol Cells 19: 289–293. hormone receptors. J Biol Chem 27: 27. Lee K, Kang MJ, Kwon SJ, Kwon YK, Kim KW, Lim JH et al. Ito T, Yamauchi M, Nishina M, Yamamichi N, Mizutani T, Ui M (2007a). Expansion of chromosome territories with chromatin et al. (2001). Identification of SWI/SNF complex subunit BAF60a decompaction in BAF53-depleted interphase cells. Mol Biol Cell as a determinant of transactivation potential of Fos/Jun dimers. 18: 4013–4023. J Biol Chem 276: 2852–2857. Lee K, Shim JH, Kang MJ, Kim JH, Ahn JS, Yoo SJ et al. (2007b). Jin YT, Myers J, Tsai ST, Goepfert H, Batsakis JG, el-Naggar AK. Association of BAF53 with mitotic chromosomes. Mol Cells 24: (1999). Genetic alterations in oral squamous cell carcinoma of 288–293. young adults. Oral Oncol 35: 251–256. LeGouy E, Thompson EM, Muchardt C, Renard JP. (1998). Kal AJ, Mahmoudi T, Zak NB, Verrijzer CP. (2000). The Drosophila Differential preimplantation regulation of two mouse homologues brahma complex is an essential coactivator for the trithorax group of the yeast SWI2 protein. Dev Dyn 212: 38–48. protein zeste. Genes Dev 14: 1058–1071. Lemon B, Inouye C, King DS, Tjian R. (2001). Selectivity of Kalpana GV, Marmon S, Wang W, Crabtree GR, Goff SP. (1994). chromatin-remodelling cofactors for ligand-activated transcription. Binding and stimulation of HIV-1 integrase by a human homolog Nature 414: 924–928. of yeast transcription factor SNF5 [see comments]. Science 266: Lickert H, Takeuchi JK, Von Both I, Walls JR, McAuliffe F, 2002–2006. Adamson SL et al. (2004). Baf60c is essential for function of BAF Kang H, Cui K, Zhao K. (2004). BRG1 controls the activity of the chromatin remodelling complexes in heart development. Nature 432: retinoblastoma protein via regulation of p21CIP1/WAF1/SDI. Mol 107–112. Cell Biol 24: 1188–1199. Link KA, Burd CJ, Williams E, Marshall T, Rosson G, Henry E Kapic A, Helmbold H, Reimer R, Klotzsche O, Deppert W, Bohn W. et al. (2005). BAF57 governs androgen receptor action and (2006). Cooperation between p53 and p130(Rb2) in induction of androgen-dependent proliferation through SWI/SNF. Mol Cell Biol cellular senescence. Cell Death Differ 13: 324–334. 25: 2200–2215. Khavari PA, Peterson CL, Tamkun JW, Mendel DB, Crabtree GR. Liu H, Kang H, Liu R, Chen X, Zhao K. (2002). Maximal induction (1993). BRG1 contains a conserved domain of the SWI2/SNF2 of a subset of interferon target genes requires the chromatin- family necessary for normal mitotic growth and transcription. remodeling activity of the BAF complex. Mol Cell Biol 22: Nature 366: 170–174. 6471–6479. Kitagawa H, Fujiki R, Yoshimura K, Mezaki Y, Uematsu Y, Matsui Liu K, Luo Y, Lin FT, Lin WC. (2004). TopBP1 recruits D et al. (2003). The chromatin-remodeling complex WINAC targets Brg1/Brm to repress E2F1-induced apoptosis, a novel a nuclear receptor to promoters and is impaired in Williams pRb-independent and E2F1-specific control for cell survival. Genes syndrome. Cell 113: 905–917. Dev 18: 673–686. Klochendler-Yeivin A, Picarsky E, Yaniv M. (2006). Increased DNA Liu R, Liu H, Chen X, Kirby M, Brown PO, Zhao K. (2001). damage sensitivity and apoptosis in cells lacking the Snf5/Ini1 Regulation of CSF1 promoter by the SWI/SNF-like BAF complex. subunit of the SWI/SNF chromatin remodeling complex. Mol Cell Cell 106: 309–318. Biol 26: 2661–2674. Manda R, Kohno T, Hamada K, Takenoshita S, Kuwano H, Ko M, Jang J, Ahn J, Lee K, Chung H, Jeon SH et al. (2004). T cell Yokota J. (2000). Absence of hSNF5/INI1 mutation in human receptor signaling inhibits glucocorticoid-induced apoptosis by lung cancer. Cancer Lett 153: 57–61.

Oncogene SWI/SNF and cancer D Reisman et al 1666 Marshall TW, Link KA, Petre-Draviam CE, Knudsen KE. (2003). Neville EM, Stewart M, Myskow M, Donnelly RJ, Field JK. (1995). Differential requirement of SWI/SNF for androgen receptor Loss of heterozygosity at 9p23 defines a novel locus in non-small cell activity. J Biol Chem 278: 30605–30613. lung cancer. Oncogene 11: 581–585. Martens JA, Winston F. (2003). Recent advances in understanding Nie Z, Xue Y, Yang D, Zhou S, Deroo BJ, Archer TK et al. (2000). chromatin remodeling by Swi/Snf complexes. Curr Opin Genet Dev A specificity and targeting subunit of a human SWI/SNF 13: 136–142. family-related chromatin-remodeling complex. Mol Cell Biol 20: Medina PP, Carretero J, Fraga MF, Esteller M, Sidransky D, Sanchez- 8879–8888. Cespedes M. (2004). Genetic and epigenetic screening for gene Nie Z, Yan Z, Chen EH, Sechi S, Ling C, Zhou S et al. (2003). Novel alterations of the chromatin-remodeling factor, SMARCA4/BRG1, SWI/SNF chromatin-remodeling complexes contain a mixed-line- in lung tumors. Genes Chromosomes Cancer 41: 170–177. age leukemia chromosomal translocation partner. Mol Cell Biol 23: Medina PP, Romero OA, Kohno T, Montuenga LM, Pio R, Yokota J 2942–2952. et al. (2008). Frequent BRG1/SMARCA4-inactivating mutations in Osipovich O, Cobb RM, Oestreich KJ, Pierce S, Ferrier P, Oltz EM. human lung cancer cell lines. Hum Mutat 29: 617–622. (2007). Essential function for SWI-SNF chromatin-remodeling Menoni H, Gasparutto D, Hamiche A, Cadet J, Dimitrov S, Bouvet P complexes in the promoter-directed assembly of Tcrb genes. Nat et al. (2007). ATP-dependent chromatin remodeling is required for Immunol 8: 809–816. base excision repair in conventional but not in variant H2A.Bbd Osley MA, Tsukuda T, Nickoloff JA. (2007). ATP-dependent nucleosomes. Mol Cell Biol 27: 5949–5956. chromatin remodeling factors and DNA damage repair. Mutat Mizutani T, Ito T, Nishina M, Yamamichi N, Watanabe A, Iba H. Res 618: 65–80. (2002). Maintenance of integrated proviral gene expression requires Ostlund Farrants AK, Blomquist P, Kwon H, Wrange O. (1997). Brm, a catalytic subunit of SWI/SNF complex. J Biol Chem 277: Glucocorticoid receptor-glucocorticoid response element binding 15859–15864. stimulates nucleosome disruption by the SWI/SNF complex. Mol Monahan BJ, Villen J, Marguerat S, Bahler J, Gygi SP, Winston F. Cell Biol 17: 895–905. (2008). Fission yeast SWI/SNF and RSC complexes show composi- Otsuki T, Furukawa Y, Ikeda K, Endo H, Yamashita T, Shinohara A tional and functional differences from budding yeast. Nat Struct et al. (2001). Fanconi anemia protein, FANCA, associates with Mol Biol 15: 873–880. BRG1, a component of the human SWI/SNF complex. Hum Mol Morrison AJ, Shen X. (2006). Chromatin modifications in DNA Genet 10: 2651–2660. repair. Results Probl Cell Differ 41: 109–125. Papoulas O, Beek SJ, Moseley SL, McCallum CM, Sarte M, Shearn A Morshead KB, Ciccone DN, Taverna SD, Allis CD, Oettinger MA. et al. (1998). The Drosophila trithorax group proteins BRM, ASH1 (2003). Antigen receptor loci poised for V(D)J rearrangement and ASH2 are subunits of distinct protein complexes. Development are broadly associated with BRG1 and flanked by peaks of 125: 3955–3966. histone H3 dimethylated at lysine 4. Proc Natl Acad Sci USA 100: Park JH, Park EJ, Lee HS, Kim SJ, Hur SK, Imbalzano AN et al. 11577–11582. (2006). Mammalian SWI/SNF complexes facilitate DNA double- Muchardt C, Bourachot B, Reyes JC, Yaniv M. (1998). ras strand break repair by promoting gamma-H2AX induction. transformation is associated with decreased expression of the brm/ EMBO J 25: 3986–3997. SNF2alpha ATPase from the mammalian SWI–SNF complex. Patenge N, Elkin SK, Oettinger MA. (2004). ATP-dependent EMBO J 17: 223–231. remodeling by SWI/SNF and ISWI proteins stimulates V(D)J Muchardt C, Yaniv M. (1993). A human homologue of Saccharomyces cleavage of 5 S arrays. J Biol Chem 279: 35360–35367. cerevisiae SNF2/SWI2 and Drosophila brm genes potentiates Peterson CL. (1998). SWI/SNF complex: dissection of a chromatin transcriptional activation by the glucocorticoid receptor. EMBO J remodeling cycle. Cold Spring Harb Symp Quant Biol 63: 545–552. 12: 4279–4290. Peterson CL, Dingwall A, Scott MP. (1994). Five SWI/SNF gene Muchardt C, Yaniv M. (1999a). ATP-dependent chromatin remodel- products are components of a large multisubunit complex required ling: SWI/SNF and Co. are on the job. J Mol Biol 293: 187–198. for transcriptional enhancement [see comments]. Proc Natl Acad Sci Muchardt C, Yaniv M. (1999b). The mammalian SWI/SNF complex USA 91: 2905–2908. and the control of cell growth. Semin Cell Dev Biol 10: 189–195. Peterson CL, Tamkun JW. (1995). The SWI–SNF complex: Muchardt C, Yaniv M. (2001). When the SWI/SNF complex a chromatin remodeling machine? Trends Biochem Sci 20: 143–146. remodels.the cell cycle. Oncogene 20: 3067–3075. Peterson CL, Workman JL. (2000). Promoter targeting and chromatin Mudhasani R, Fontes JD. (2002). The class II transactivator requires remodeling by the SWI/SNF complex. Curr Opin Genet Dev 10: brahma-related gene 1 to activate transcription of major histocom- 187–192. patibility complex class II genes. Mol Cell Biol 22: 5019–5026. Peterson CL, Zhao Y, Chait BT. (1998). Subunits of the yeast SWI/ Murphy DJ, Hardy S, Engel DA. (1999). Human SWI–SNF SNF complex are members of the actin-related protein (ARP) component BRG1 represses transcription of the c-fos gene. Mol family. J Biol Chem 273: 23641–23644. Cell Biol 19: 2724–2733. Phelan ML, Sif S, Narlikar GJ, Kingston RE. (1999). Reconstitution Nagl Jr NG, Wang X, Patsialou A, Van Scoy M, Moran E. (2007). of a core chromatin remodeling complex from SWI/SNF subunits. Distinct mammalian SWI/SNF chromatin remodeling complexes Mol Cell 3: 247–253. with opposing roles in cell-cycle control. EMBO J 26: 752–763. Pottier N, Cheok MH, Yang W, Assem M, Tracey L, Obenauer JC Nagl Jr NG, Zweitzig DR, Thimmapaya B, Beck Jr GR, Moran E. et al. (2007). Expression of SMARCB1 modulates steroid sensitivity (2006). The c-myc gene is a direct target of mammalian SWI/SNF- in human lymphoblastoid cells: identification of a promoter SNP related complexes during differentiation-associated cell cycle arrest. that alters PARP1 binding and SMARCB1 expression. Hum Mol Cancer Res 66: 1289–1293. Genet 16: 2261–2271. Nasmyth K, Shore D. (1987). Transcriptional regulation in the yeast Reisman DN, Sciarrotta J, Bouldin TW, Weissman BE, Funkhouser life cycle. Science 237: 1162–1170. WK. (2005). The expression of the SWI/SNF ATPase subunits Neely KE, Workman JL. (2002). The complexity of chromatin BRG1 and BRM in normal human tissues. Appl Immunohistochem remodeling and its links to cancer. Biochim Biophys Acta 1603: Mol Morphol 13: 66–74. 19–29. Reisman DN, Sciarrotta J, Wang W, Funkhouser WK, Weissman BE. Neigeborn L, Carlson M. (1984). Genes affecting the regulation of (2003). Loss of BRG1/BRM in human lung cancer cell lines and SUC2 gene expression by glucose repression in Saccharomyces primary lung cancers: correlation with poor prognosis. Cancer Res cerevisiae. Genetics 108: 845–858. 63: 560–566. Neigeborn L, Carlson M. (1987). Mutations causing constitutive Reisman DN, Strobeck MW, Betz BL, Sciariotta J, Funkhouser Jr W, invertase synthesis in yeast: genetic interactions with snf mutations. Murchardt C et al. (2002). Concomitant down-regulation of BRM Genetics 115: 247–253. and BRG1 in human tumor cell lines: differential effects on

Oncogene SWI/SNF and cancer D Reisman et al 1667 RB-mediated growth arrest vs CD44 expression. Oncogene 21: Smith CL, Horowitz-Scherer R, Flanagan JF, Woodcock CL, Peterson 1196–1207. CL. (2003). Structural analysis of the yeast SWI/SNF chromatin Reyes JC, Barra J, Muchardt C, Camus A, Babinet C, Yaniv M. remodeling complex. Nat Struct Biol 10: 141–145. (1998). Altered control of cellular proliferation in the absence of Smith CL, Peterson CL. (2005). A conserved Swi2/Snf2 ATPase motif mammalian brahma (SNF2alpha). EMBO J 17: 6979–6991. couples ATP hydrolysis to chromatin remodeling. Mol Cell Biol 25: Ring HZ, Vameghi-Meyers V, Wang W, Crabtree GR, Francke U. 5880–5892. (1998). Five SWI/SNF-related, matrix-associated, actin-dependent Sohn DH, Lee KY, Lee C, Oh J, Chung H, Jeon SH et al. (2007). regulator of chromatin (SMARC) genes are dispersed in the human SRG3 interacts directly with the major components of the SWI/ genome. Genomics 51: 140–143. SNF chromatin remodeling complex and protects them from Roberts CW, Leroux MM, Fleming MD, Orkin SH. (2002). proteasomal degradation. J Biol Chem 282: 10614–10624. Highly penetrant, rapid tumorigenesis through conditional Spicuglia S, Kumar S, Yeh JH, Vachez E, Chasson L, Gorbatch S et al. inversion of the tumor suppressor gene Snf5. Cancer Cell 2: (2002). Promoter activation by enhancer-dependent and -indepen- 415–425. dent loading of activator and coactivator complexes. Mol Cell 10: Roberts CW, Orkin SH. (2004). The SWI/SNF complex—chromatin 1479–1487. and cancer. Nat Rev Cancer 4: 133–142. Stern M, Jensen R, Herskowitz I. (1984). Five SWI genes are required Rousseau-Merck MF, Versteege I, Legrand I, Couturier J, Mairal A, for expression of the HO gene in yeast. J Mol Biol 178: 853–868. Delattre O et al. (1999). hSNF5/INI1 inactivation is mainly Strobeck MW, DeCristofaro MF, Banine F, Weissman BE, associated with homozygous deletions and mitotic recombinations Sherman LS, Knudsen ES. (2001). The BRG-1 subunit of the in rhabdoid tumors. Cancer Res 59: 3152–3156. SWI/SNF complex regulates CD44 expression. J Biol Chem 276: Roy K, De La Serna IL, Imbalzano AN. (2002). The myogenic basic 9273–9278. helix-loop-helix family of transcription factors show similar Strobeck MW, Knudsen KE, Fribourg AF, DeCristofaro MF, requirements for SWI/SNF chromatin remodeling enzymes during Weissman BE, Imbalzano AN et al. (2000). BRG-1 is required muscle differentiation in culture. J Biol Chem 8:8. for RB-mediated cell cycle arrest. Proc Natl Acad Sci USA 97: Sabah M, Cummins R, Leader M, Kay E. (2005). Leiomyosarcoma 7748–7753. and malignant fibrous histiocytoma share similar allelic imbalance Strobeck MW, Reisman DN, Gunawardena RW, Betz BL, Angus SP, pattern at 9p. Virchows Arch 446: 251–258. Knudsen KE et al. (2002). Compensation of BRG-1 function by Saha A, Wittmeyer J, Cairns BR. (2006). Mechanisms for nucleosome Brm: insight into the role of the core SWI/SNF subunits in RB- movement by ATP-dependent chromatin remodeling complexes. signaling. J Biol Chem 21: 21. Results Probl Cell Differ 41: 127–148. Strober BE, Dunaief JL, Guha, Goff SP. (1996). Functional Sarkar S, Roy BC, Hatano N, Aoyagi T, Gohji K, Kiyama R. (2002). interactions between the hBRM/hBRG1 transcriptional activators A novel ankyrin repeat-containing gene (Kank) located at 9p24 and the pRB family of proteins. Mol Cell Biol 16: 1576–1583. is a growth suppressor of renal cell carcinoma. J Biol Chem 277: Sudarsanam P, Iyer VR, Brown PO, Winston F. (2000). Whole- 36585–36591. genome expression analysis of snf/swi mutants of Saccharomyces Schnitzler G, Sif S, Kingston RE. (1998). Human SWI/SNF cerevisiae. Proc Natl Acad Sci USA 97: 3364–3369. interconverts a nucleosome between its base state and a stable Sumi-Ichinose C, Ichinose H, Metzger D, Chambon P. (1997). remodeled state. Cell 94: 17–27. SNF2beta-BRG1 is essential for the viability of F9 murine Schwartz YB, Boykova T, Belyaeva ES, Ashburner M, Zhimulev IF. embryonal carcinoma cells. Mol Cell Biol 17: 5976–5986. (2004). Molecular characterization of the singed wings locus of Takeuchi JK, Lickert H, Bisgrove BW, Sun X, Yamamoto M, Drosophila melanogaster. BMC Genet 5: 15. Chawengsaksophak K et al. (2007). Baf60c is a nuclear Notch Sekine I, Sato M, Sunaga N, Toyooka S, Peyton M, Parsons R et al. signaling component required for the establishment of left-right (2005). The 3p21 candidate tumor suppressor gene BAF180 asymmetry. Proc Natl Acad Sci USA 104: 846–851. is normally expressed in human lung cancer. Oncogene 24: Tamkun JW. (1995). The role of brahma and related proteins in 2735–2738. transcription and development. Curr Opin Genet Dev 5: 473–477. Sentani K, Oue N, Kondo H, Kuraoka K, Motoshita J, Ito R et al. Tamkun JW, Deuring R, Scott MP, Kissinger M, Pattatucci AM, (2001). Increased expression but not genetic alteration of BRG1, a Kaufman TC et al. (1992). brahma: a regulator of Drosophila component of the SWI/SNF complex, is associated with the homeotic genes structurally related to the yeast transcriptional advanced stage of human gastric carcinomas. Pathobiology 69: activator SNF2/SWI2. Cell 68: 561–572. 315–320. Tripathi A, Dasgupta S, Roy A, Sengupta A, Roy B, Roychowdhury S Seo S, Herr A, Lim JW, Richardson GA, Richardson H, Kroll KL. et al. (2003). Sequential deletions in both arms of chromosome 9 (2005a). Geminin regulates neuronal differentiation by antagonizing are associated with the development of head and neck squamous Brg1 activity. Genes Dev 19: 1723–1734. cell carcinoma in Indian patients. J Exp Clin Cancer Res 22: Seo S, Richardson GA, Kroll KL. (2005b). The SWI/SNF chromatin 289–297. remodeling protein Brg1 is required for vertebrate neurogenesis and Trotter KW, Archer TK. (2004). Reconstitution of glucocorticoid mediates transactivation of Ngn and NeuroD. Development 132: receptor-dependent transcription in vivo. Mol Cell Biol 24: 105–115. 3347–3358. Sevenet N, Lellouch-Tubiana A, Schofield D, Hoang-Xuan K, Gessler Trotter KW, Archer TK. (2008). The BRG1 transcriptional coregu- M, Birnbaum D et al. (1999a). Spectrum of hSNF5/INI1 somatic lator. Nucl Recept Signal 6: e004. mutations in human cancer and genotype- phenotype correlations. Trouche D, Le Chalony C, Muchardt C, Yaniv M, Kouzarides T. Hum Mol Genet 8: 2359–2368. (1997). RB and hbrm cooperate to repress the activation functions Sevenet N, Sheridan E, Amram D, Schneider P, Handgretinger R, of E2F1. Proc Natl Acad Sci USA 94: 11268–11273. Delattre O. (1999b). Constitutional mutations of the hSNF5/INI1 Underhill C, Qutob MS, Yee SP, Torchia J. (2000). A novel nuclear gene predispose to a variety of cancers. Am J Hum Genet 65: receptor corepressor complex, N-CoR, contains components of the 1342–1348. mammalian SWI/SNF complex and the corepressor KAP-1. J Biol Shanahan F, Seghezzi W, Parry D, Mahony D, Lees E. (1999). Cyclin Chem 275: 40463–40470. E associates with BAF155 and BRG1, components of the Valdman A, Nordenskjold A, Fang X, Naito A, Al-Shukri S, Larsson mammalian SWI–SNF complex, and alters the ability of BRG1 to C et al. (2003). Mutation analysis of the BRG1 gene in prostate induce growth arrest. Mol Cell Biol 19: 1460–1469. cancer clinical samples. Int J Oncol 22: 1003–1007. Shundrovsky A, Smith CL, Lis JT, Peterson CL, Wang MD. (2006). Versteege I, Sevenet N, Lange J, Rousseau-Merck MF, Ambros P, Probing SWI/SNF remodeling of the nucleosome by unzipping Handgretinger R et al. (1998). Truncating mutations of hSNF5/ single DNA molecules. Nat Struct Mol Biol 13: 549–554. INI1 in aggressive paediatric cancer. Nature 394: 203–206.

Oncogene SWI/SNF and cancer D Reisman et al 1668 Vradii D, Wagner S, Doan DN, Nickerson JA, Montecino M, Lian JB Xu W, Cho H, Kadam S, Banayo EM, Anderson S, Yates III JR et al. et al. (2006). Brg1, the ATPase subunit of the SWI/SNF chromatin (2004). A methylation-mediator complex in hormone signaling. remodeling complex, is required for myeloid differentiation to Genes Dev 18: 144–156. granulocytes. J Cell Physiol 206: 112–118. Xue Y, Canman JC, Lee CS, Nie Z, Yang D, Moreno GT et al. (2000). Wang L, Baiocchi RA, Pal S, Mosialos G, Caligiuri M, Sif S. (2005). The human SWI/SNF-B chromatin-remodeling complex is related The BRG1- and hBRM-associated factor BAF57 induces apoptosis to yeast rsc and localizes at kinetochores of mitotic chromosomes. by stimulating expression of the cylindromatosis tumor suppressor Proc Natl Acad Sci USA 97: 13015. gene. Mol Cell Biol 25: 7953–7965. Yamamichi-Nishina M, Ito T, Mizutani T, Yamamichi N, Wang M, Gu C, Qi T, Tang W, Wang L, Wang S et al. (2007). BAF53 Watanabe H, Iba H. (2003). SW13 cells can transition between interacts with p53 and functions in p53-mediated p21-gene two distinct subtypes by switching expression of BRG1 and Brm transcription. J Biochem 142: 613–620. genes at the post-transcriptional level. J Biol Chem 278: 7422–7430. Wang W. (2003). The SWI/SNF family of ATP-dependent chromatin Yamamichi N, Yamamichi-Nishina M, Mizutani T, Watanabe H, remodelers: similar mechanisms for diverse functions. Curr Top Minoguchi S, Kobayashi N et al. (2005). The Brm gene suppressed Microbiol Immunol 274: 143–169. at the post-transcriptional level in various human cell lines is Wang W, Cote J, Xue Y, Zhou S, Khavari PA, Biggar SR et al. inducible by transient HDAC inhibitor treatment, which exhibits (1996a). Purification and biochemical heterogeneity of the mamma- antioncogenic potential. Oncogene 24: 5471–5481. lian SWI–SNF complex. EMBO J 15: 5370–5382. Yan Z, Cui K, Murray DM, Ling C, Xue Y, Gerstein A et al. (2005). Wang W, Xue Y, Zhou S, Kuo A, Cairns BR, Crabtree GR. (1996b). PBAF chromatin-remodeling complex requires a novel specificity Diversity and specialization of mammalian SWI/SNF complexes. subunit, BAF200, to regulate expression of selective interferon- Genes Dev 10: 2117–2130. responsive genes. Genes Dev 19: 1662–1667. Wang X, Nagl Jr NG, Flowers S, Zweitzig D, Dallas PB, Moran E. Yan Z, Wang Z, Sharova L, Sharov AA, Ling C, Piao Y et al. (2008). (2004a). Expression of p270 (ARID1A), a component of human BAF250B-associated SWI/SNF chromatin-remodeling complex is SWI/SNF complexes, in human tumors. Int J Cancer 112: 636. required to maintain undifferentiated mouse embryonic stem cells. Wang X, Nagl NG, Wilsker D, Van Scoy M, Pacchione S, Stem Cells 26: 1155–1165. Yaciuk P et al. (2004b). Two related ARID family proteins are Yoshinaga SK, Peterson CL, Herskowitz I, Yamamoto KR. (1992). alternative subunits of human SWI/SNF complexes. Biochem J 383: Roles of SWI1, SWI2, and SWI3 proteins for transcriptional 319–325. enhancement by steroid receptors. Science 258: 1598–1604. Wang Z, Zhai W, Richardson JA, Olson EN, Meneses JJ, Firpo MT Yuge M, Nagai H, Uchida T, Murate T, Hayashi Y, Hotta T et al. et al. (2004c). Polybromo protein BAF180 functions in mammalian (2000). HSNF5/INI1 gene mutations in lymphoid malignancy. cardiac chamber maturation. Genes Dev 18: 3106–3116. Cancer Genet Cytogenet 122: 37–42. Winston F, Allis CD. (1999). The bromodomain: a chromatin- Zhang B, Chambers KJ, Faller DV, Wang S. (2007). Reprogramming of targeting module? [news]. Nat Struct Biol 6: 601–604. the SWI/SNF complex for co-activation or co-repression in prohibi- Wolffe AP. (1994). Transcriptional activation. Switched-on chromatin. tin-mediated estrogen receptor regulation. Oncogene 26: 7153–7157. Curr Biol 4: 525–528. Zhang HS, Gavin M, Dahiya A, Postigo AA, Ma D, Luo RX et al. Wong AK, Shanahan F, Chen Y, Lian L, Ha P, Hendricks K et al. (2000). Exit from G1 and S phase of the cell cycle is regulated (2000). BRG1, a component of the SWI–SNF complex, is by repressor complexes containing HDAC-Rb-hSWI/SNF and mutated in multiple human tumor cell lines. Cancer Res 60: Rb-hSWI/SNF. Cell 101: 79–89. 6171–6177. Zhao K, Wang W, Rando OJ, Xue Y, Swiderek K, Kuo A et al. Wuebbles RD, Jones PL. (2004). DNA repair in a chromatin (1998). Rapid and phosphoinositol-dependent binding of the SWI/ environment. Cell Mol Life Sci 61: 2148–2153. SNF-like BAF complex to chromatin after T lymphocyte receptor Xi Q, He W, Zhang XH, Le HV, Massague J. (2008). Genome-wide signaling. Cell 95: 625–636. impact of the BRG1 SWI/SNF chromatin remodeler on the Zraly CB, Middleton FA, Dingwall AK. (2006). Hormone-response transforming growth factor beta transcriptional program. J Biol genes are direct in vivo regulatory targets of Brahma (SWI/SNF) Chem 283: 1146–1155. complex function. J Biol Chem 281: 35305–35315.

Oncogene