|Home | About | Journals | Submit | Contact Us | Français|
Nuclear actin is involoved in transcription of all three RNA polymerases, chromatin remodeling, and formation of hnRNP complexes as well as recruitment of histone modifier to the active gene. In addition, actin-binding proteins (ABPs) control actin nucleation, bundling, filament capping, fragmentation, and monomer availability in the cytoplasm. In recent years, more and more attention is on the role of actin and ABPs in the modulation of the subcellular localization of transcriptional regulators. This review focuses on the recent developments about transcription and transcriptional regulation by nuclear actin, regulation of muscle-specific gene expression, nuclear receptor and transcription complexes by ABPs. Among them, STARS and ABLIM regulate actin dynamics and SRF-dependent muscle-specific gene expression. Functionally and structurally unrelated cytoplasmic ABPs interact cooperatively with nuclear receptor and regulate its transactivation. Furthermore, ABPs also participate in the formation of transcription complexes.
Actin is one of the major components of the cytoskeleton and plays a critical role in all eukaryotic cells. The actin cytoskeleton functions in diverse cellular processes, including cell motility, contractility, mitosis and cytokinesis, intracellular transport, endocytosis, and secretion [1, 2]. Besides these mechanical functions, actin has also been implicated in the regulation of gene transcription, either through cytoplasmic changes in cytoskeletal actin dynamics  or through the assembly of transcriptional regulatory complexes . Cytoskeletal actin dynamics, i.e., actin polymerization by which monomeric actin (globular actin [G-actin]) is assembled into long actin polymers (filamentous actin [F-actin]) and actin deploymerization by which F-actin is severed into G-actin, is the key for these diverse functions. The dynamic nature of the actin cytoskeleton is directly determined spatiotemporally by the actions of numerous actin-binding proteins (ABPs). The activity of different classes of ABPs controls actin nucleation, bundling, filament capping, fragmentation, and monomer availability. Transcriptional regulation mediated by cytoskeletal actin dynamics can be ascribed to the modulation of the subcellular localization of transcriptional regulators by ABPs . In addition, some of mechanisms by which actin affects transcription and transcription regulation depend on molecular interactions of actin with RNA polymerases and components of the transcription machinery in the nucleus.
Actin is not only a major cytoskeletal component in all eukaryotic cells, but also a constitutent of nuclear protein complexes. Nuclear actin plays a role in many nuclear functions [6–8]. First, nuclear actin is required for transcription by all three nuclear RNA polymerases. Second, nuclear actin associates with small nuclear ribonucleoproteins (snRNPs), which have a major role in mRNA processing [8, 9], and plays a direct part in the nuclear export of RNA and cellular proteins [10, 11]. Third, nuclear actin also forms complexes with certain heterogeneous nuclear ribonucleoproteins (hnRNPs) that bind to and accompany mRNA from the nucleus to the cytoplasm [12–14]. Fourth, nuclear actin and actin-related proteins have been found in association with chromatin-remodeling and histone acetyl transferase complexes, suggesting a role for actin in chromatin remodeling . Together, recent investigations suggest that nuclear actin plays a role in gene transcription associated with three main entities: components of the three RNA polymerases, ATP-dependent chromatin remodeling complexes, and RNP particles in the eukaryotic cell nucleus.
Nuclear actin is required for transcription of all three RNA polymerases. Specifically, β-actin has been identified as a component of RNA pol II pre-initiation complexes (PICs). Injection of anti-actin antibodies into the nuclei of salamander oocytes results in a contraction of the lateral loops and inhibits transcription . Furthermore, Hofmann et al.  found that actin is associated with actively transcribed genes and plays an essential role in transcriptional activation. In addition, actin is required for the initiation of transcription through participation in the formation of pre-initiation complexes . These conclusions are based on the following data: (1) β-actin participates in RNA polymerase II transcription through a direct effect using only purified transcription factors[18, 19]; (2) nascent RNA molecules are associated with actin in the nuclear matrix and antibodies to β-actin inhibits the synthesis of nascent transcripts and RNA polymerase II transcription[17, 19]; (3) adding actin to a highly purified RNA polymerase II fraction stimulates transcription; (4) actin co-localizes with transcription sites in early mouse embryos[4, 17]; (5) actin is recruited to the promoter region of transcribing genes in vivo[19, 20]; (6) antibodies to β-actin inhibits the production of a 15-nucleotide transcript that is a pre-requisite for the commitment to elongation[19, 21]; (7) actin is a component of premRNP particles that is incorporated into pre-mRNAs by binding to a specific subset of RNA-binding proteins[4, 22]; and (8) actin is a component of PICs and depletion of actin prevents the formation of PICs[19, 23]. In conclusion, the above evidence suggests that there is a strong and specific interaction between actin and RNA polymerase II, and actin participates in RNA polymerase II transcription. Then, what is the function of actin in RNA polymerase II transcription? From the above data, we conclude that the following: (1) based on the ChIP assays results, i.e., actin is recruited to genes that are poised to start transcribing, it is known that actin is involved in recruiting RNA polymerase II to the PIC; (2) decreased actin levels resulting from the anti-actin antibodies inhibit PIC formation by preventing the binding of TBP to the TATA box, indicating that PIC formation is required for the association of actin with promoter DNA ; (3) antibodies to β-actin prevent PIC formation, suggesting that actin acts as a bridge between the polymerase and the other constituents of the PIC; and  (4) actin and nuclear myosin 1 (NM1), an isoform of myosin 1, are involved in transcription elongation[6, 25, 26]. Together, these data suggest that actin is involved in multiple stages of the transcription process.
β-actin also has an important role in RNA polymerase III transcription . First, β-actin is tightly associated with RNA polymerase III through direct protein–protein interactions with one of more of the RPC3, RPABC2, and RPABC3 subunits, and constitutes part of the active RNA polymerase III . Photochemical cross-linking experiments performed with a transcription initiation complex indicated that actin makes complex contact with DNA . Second, the ChIP assay identified that β-actin is located at promoter sequences of an actively transcribed U6 gene in vivo, which suggest that it participates in the transcription process of transcription of RNA polymerase III [27, 29, 30]. Upon treatment with methane methylsulfonate (MMS), a drug that represses RNA polymerase III transcription, the U6 initiation complex and β-actin are largely dissociated from promoter sequences [27, 29, 31]. Notably, there is a much larger decrease in β-actin association with the U6 promoter region compared to dissociation of RNA polymerase III, which suggests that β-actin dissociates from the RNA polymerase III complex. In addition, Hu et al.  also identified that RNA polymerase III is localized in vivo on a gene known to be transcribed by RNA polymerase III. Third, many experiments have shown that β-actin is required for RNA polymerase III transcription [27, 29, 32]. The form of actin required for RNA polymerase III transcription is the monomeric actin, suggesting that β-actin is essential for basal RNA polymerase trabscription.
Actin and NM1 interact with different components of the RNA polymerase I machinery, and together serve as a nucleolar motor involved in the transcription of ribosomal RNA genes [26, 33]. Recent studies have revealed that actin is associated with rDNA genes, and micro-injection of anti-actin antibodies into the nuclei of HeLa cells inhibits pre-rRNA synthesis in vivo [25, 34]. The interaction of NM1 with actin in the initiation complex may trigger a conformational change that favors the transition of RNA polymerase I from the initiation phase into the elongation phase [25, 33]. NM1 mutants that lack ATPase activity or actin binding are incapable of associating with RNA polymerase I , and their association with rDNA is greatly impaired. Moreover, the association of actin and NM1 with RNA polymerase I is abolished in the presence of ATP and stabilized by ADP, further suggesting that nuclear actomyosin complexes act as a molecular motor that facilitates transcription . NM1 binds the DNA backbone through its positively charged tail domain, while the head interacts with actin bound to RNA polymerase . It has been suggested that by anchoring NM1 to DNA and actin to RNA polymerase, an auxiliary motor is generated that works in concert with nuclear RNA polymerases to drive transcription . This suggests that the cooperative action of actin and myosin in the nucleus is required for RNA polymerase I transcription and reveals an actomyosin-based mechanism in transcription.
Actin is essential for the function of chromatin remodeling complexes in transcriptional activation. Nuclear actin is an ATPase that cycles between monomeric (G-actin or β-actin) and polymerized (F-actin) states . Eukaryotic cells have several ATP-dependent chromatin remodeling complexes, depending on the ATPase in the complex, as follows: SWI/SNF complexes, ISWI (imitation of SWI)-containing complexes, Mi-2 complexes, histone acetyltransferase complexes, such as the Nu4A and TIP60 complexes, and INO80 complexes
β-actin is an integral component of chromatin remodeling complexes, such as the BAF, BAP, and INO80 complexes, as well as Nu4A and TIP60 complexes [24, 27, 29, 35–38]. It is generally accepted that chromatin-remodeling complexes contain actin, actin-related proteins (ARPs), and/or ABPs. Nuclear ARPs (ARP5–9) are associated with actin in chromatin remodeling complexes of the SWI/SNF family, such as those containing the ATPase subunits INO80 or SWR1 [15, 24, 39]. In the SWI/SNF-like BAF complex, β-actin binds directly to the BRG1 ATPase subunit of BAF and stimulates BRG1 ATPase activity, and this interaction is necessary for the binding of the BAF complex to chromatin [27, 29, 40]. Actin binding to BRG1 is required for stable association of the complex and provides a link between the chromatin remodeling complex and the nuclear matrix [5, 41]. In the INO80 complex, actin is required for efficient DNA binding, ATPase activity, and nucleosome mobilization, as INO80 complexes lacking actin as well as the ARPs, ARP4 and ARP8, are deficient for these activities . BAF53 and β-actin have also been identified as subunits of the human TIP60 HAT complex, which is involved in DNA repair and apoptosis, and BAF53 is found in a distinct HAT complex involved in c-myc activation, whereas Act3/ARP4 and actin are components of the yeast Nu4A HAT complex [38, 42]. In the yeast Nu4A HAT complex, actin and Act3/ARP4 are essential for the structural integrity and activity of the complex .The presence of actin in chromatin remodeling complexes suggests that there is a functional link between actin and the regulation of chromatin structure, and a major function of actin is to act as an allosteric regulator in the remodeling of some macromolecular assemblies, such as chromatin remodeling factors or transcription complexes.
The hnRNP U, a component of pre-mRNP particles, has been shown to interact directly with actin through a specific and conserved actin-binding site located in the hnRNP U C terminus and associate with the phosphorylated C-terminal domain (CTD) of RNA polymerase II . Injection of a peptide acting as a competitive inhibitor of a protein–protein contact involving actin and the hnRNP protein, HRP36, into the salivary glands of Chironomus tentans disrupts global RNA polymerase II transcription as measured by bromo-UTP incorporation, an effect that is due at least in part to a decrease in elongation as measured by run-on assays . A recent study has shown that actin binds directly to the Chironomus tentans hnRNP, HRP65-2, which is a molecular platform for recruitment of the HAT histone H3-specific acetyltransferase p2D10 on active genes. Both actin and the pre-mRNP protein, HRP65, are complexed in situ with p2D10, and disruption of the actin–HRP65 interaction causes release of p2D10 from RNA polymerase II-transcribing genes coincident with reduced H3 acetylation and diminished transcription . The mechanism is that HRP65-2 binds directly to p2D10, and the interaction between actin and HRP65-2 is required for p2D10 to associate with the transcribed chromatin . Moreover, the association of p2D10, actin, and HRP65-2 with chromatin is sensitive to ribonuclease digestion, which indicates that these proteins are tethered to the transcribed genes by binding to the nascent transcript. These new findings considerably bolster the notion of a link among nuclear actin, chromatin remodeling, and RNA polymerase II transcription [43, 44]. Obrdlik [45, 46] identified that the histone acetyltransferase (HAT), PCAF, associates with actin and hnRNP U. Moreover, it has been shown that actin, hnRNP U, and PCAF associate with the Ser2/5- and Ser2-phosphorylated RNA polymerase II carboxy-terminal domain. hnRNP U and PCAF are present at the promoter and coding regions of constitutively-expressed RNA polymerase II genes and they are associated with ribonucleoprotein complexes . In summary, these finding suggest that actin, HRP65-2, and HAT (p2D10 or PCAF) become assembled into nascent pre-mRNPs during transcription. Based on these evidences, it may be proposed that the actin-HRP65-2-HAT complex is part of the nascent pre-mRNP, and it can travel along the transcribed gene, allowing HAT to acetylate histones. According to this proposal, the actin–HRP65-2–HAT complex maintains the chromatin in a transcription competent conformation. This model is supported by the observations that H3 acetylation is reduced and transcription is inhibited when the interaction between actin and HRP65-2 is disrupted . In addition, Actin-mediated RNA polymerase II transcriptional control may be sensitive to the different polymerization states of actin . Transcriptionally-competent actin may be present in a monomeric or oligomeric form which is different from the canonical actin filaments. The polymerization states of actin involved in the initiation or elongation phases are different (Fig. 1) .
The cytoplasmic dynamics of the actin cytoskeleton have been shown to regulate the subcellular localization of some transcription factors, such as the myocardin-related transcription factors (MRTFs), MRTF-A (also referred to as MAL, MKL1, and BSAC) and MRTF-B (also referred to as MKL2 or MAL16) [47, 48], the developmentally regulated PREP2 homeoprotein, and the transcriptional repressor, YY1[49, 50]. Because actin dynamics are regulated by a number of ABPs, ABPs can play a critical role in the regulation of transcription and gene expression . Studies to date have established that some ABPs induce the formation of actin filaments by their ability to nucleate actin filament polymerization; other ABPs promote breakdown of filaments by a mechanism referred to as severing. Still other ABPs crosslink or bundle actin filaments or prevent filament formation by their so-called sequestering activity. Among the notable transcription factors controlled by ABPs are MRTFs, which associate with SRF and stimulate SRF-dependent transcription [47, 52, 53]. In addition, actin dynamics are regulated by several signal transduction cascades that converge on ABPs .
MRTF-A associates with G-actin, and is predominantly localized to the cytoplasm of NIH 3T3 cells in the absence of serum and accumulates in the nucleus in response to serum stimulation. MRTF-B also undergoes nuclear translocation in response to serum stimulation, although it shows less responsiveness to serum than MRTF-A . Upon activation of RhoA, actin becomes polymerized and releases MRTF-A, which in turn translocates to the nucleus to associate with SRF . Striated muscle activator of Rho signaling (STARS) is a muscle-specific ABP capable of stimulating SRF-dependent transcription through a mechanism involving RhoA activation and actin polymerization . Recently, MRTF-A and -B were shown to serve as a link between STARS and SRF. In NIH 3T3 cells co-transfected with expression plasmids encoding MRTFs and STARS, the MRTFs are translocated to the nucleus in the absence of serum. The nuclear localization of myocardin is unchanged in the absence or presence of STARS . Thus, STARS can substitute for serum stimulation and promote the nuclear translocation of MRTFs with consequent activation of SRF-dependent transcription. Kuwahara et al.  found that the co-expression of STARS with a dominant-negative myocardin mutant, which can inhibit the transcriptional activities of myocardin and MRTF-A and -B, can completely block the ability of STARS to induce SRF-dependent transcription in NIH 3T3, COS1, and 293T cells. However, STARS does not alter the level of expression of MRTFs. These observations suggest that STARS stimulates SRF-dependent transcription solely by promoting the nuclear translocation of MRTF-A and -B.
The STARS protein contains 375 amino acids with the conserved actin-binding domain contained within the C-terminal 142 residues . The STARS C-terminal deletion mutant, N233, which cannot bind actin or activate SRF, fails to induce the nuclear accumulation of MRTF-A and -B. In contrast, the C-terminal 142 amino acids of STARS, which bind actin and stimulate SRF, induce the nuclear accumulation of MRTFs as efficiently as full-length STARS. STARS N233 fails to enhance MRTF-dependent activation of SRF-dependent reporters, whereas STARS C142 synergistically enhances MRTF-mediated transcription to the same level as full-length STARS . These results demonstrate that the actin-binding domain of STARS is both necessary and sufficient for the nuclear accumulation and transcriptional activation of MRTFs by STARS.
The activity of STARS involves actin dynamics. Treatment of NIH 3T3 cells with latrunculin B (LB), which sequesters actin monomers and prevents Rho-dependent nuclear accumulation of MRTF-A and SRF activation , blocks nuclear accumulation of MRTF-A and -B in the presence of STARS. Conversely, cytochalasin D (CD), which dimerizes actin, but prevents actin polymerization and activates SRF, strongly induces the nuclear translocation of MRTFs, even in the absence of STARS . Consistent with these effects on MRTF nuclear import, LB significantly blocks the stimulatory effect of STARS on MRTF-dependent transcription and CD enhances the activity of MRTFs alone. These results indicate that actin dynamics are involved in the STARS-induced nuclear accumulation of MRTFs and transcriptional activation of SRF via MRTFs.
MRTF-A is recently reported to interact directly with G-actin . The unpolymerized G-actin controls MRTF activity  and STARS induces actin polymerization . Kuwahara et al.  demonstrated that expression of wild-type actin, which increases the amount of G-actin, but does not alter the F-actin/G-actin ratio, reduced the ability of STARS to activate MRTF-dependent transcription, though it did not significantly alter the activity of MRTF in the absence of STARS. The actin mutant that favors F-actin formation and increases the F-actin/G-actin ratio  stimulates MRTF activity, even in the absence of STARS and abolishes further activation of MRTFs by STARS. In contrast, the actin mutant that is unable to polymerize and decreases the F-actin/G-actin ratio inhibits MRTF activity and also reduces the ability of STARS to enhance MRTF activity. These results suggest that STARS stimulates MRTF activity by inducing the dissociation of MRTFs from actin by depleting the G-actin pool.
The N-terminal regions of MRTFs contain three RPEL motifs, which have been shown to sequester the MRTFs in the cytoplasm by association with actin [47, 57]. Consistent with the possibility that STARS promotes the nuclear import of MRTFs by displacing them from monomeric G-actin, the RPEL motifs are required for the effects of STARS on MRTFs. MRTFs are cytoplasmic, accumulating in the nucleus upon activation of Rho GTPase signaling, which alters interactions between G-actin and the RPEL domain. Guettler et al.  showed that the RPEL domain of MRTF-A binds actin more avidly than that of myocardin and that the RPEL motif itself is an actin-binding element. RPEL1 and RPEL2 of myocardin bind actin weakly compared with MRTF-A, while RPEL3 is of comparable and low affinity in the two proteins. Actin binding by all three motifs is required for MRTF-A regulation. The differing behaviors of MRTF-A and myocardin are specified by the RPEL1-RPEL2 unit, while RPEL3 can be exchanged between them. They proposed that differential actin occupancy of multiple RPEL motifs regulates nucleocytoplasmic transport and activity of MRTF-A. Because myocardin is insensitive to the effects of STARS, its target genes would be expected to be highly active irrespective of the polymerization state of actin, although STARS would be expected to further augment the expression of these genes through its actions on MRTF-A and -B, which are also expressed in cardiac muscle, and which form heterodimers with myocardin.
Recently, Barrientos et al.  identified two novel members of the actin-binding LIM protein (ABLIM) family, ABLIM-2 and -3, as STARS-interacting proteins by a yeast two-hybrid screen of a skeletal muscle cDNA library using STARS as bait. These novel proteins contain four LIM domains and a carboxyl-terminal villin headpiece domain, which mediates actin-binding in several proteins, such as villin and dematin . Both ABLIM-2 and -3 have high homology with ABLIM-1. ABLIM-1 was originally found in the human retina, as well as in the sarcomeres of murine cardiac tissue and was postulated to regulate actin-dependent signaling . Similarly, ABLIM-2 and -3 are expressed in a tissue-specific pattern. In adult human multiple tissues, ABLIM-2 is highly expressed in skeletal muscle and at lower levels in brain, spleen, and kidney. No significant expression was detected in the heart. In contrast to ABLIM-2, ABLIM-3 is predominantly expressed in human heart and brain, whereas the murine ABLIM-3 homologue displays a somewhat broader tissue distribution that also includes lung and liver .
Both ABLIM-2 and -3 strongly bind F-actin and co-localize with actin stress fibers. The interaction of STARS with ABLIM-2 and -3 was confirmed by co-immunoprecipitation and was further supported by the co-localization of STARS and ABLIM-2, as detected by immunofluorescence . The complementary expression patterns of ABLIM-2 and -3 in striated muscle imply that STARS interacts in vivo with ABLIM-2 in skeletal muscle and ABLIM-3 in cardiac muscle. Consistent with the notion that STARS activates SRF-dependent transcription via stabilization of the actin cytoskeleton , both ABLIM-2 and -3 modulate STARS-dependent activation of a luciferase reporter construct controlled by the SM22 promoter, which contains two essential SRF-binding sites and is highly sensitive to STARS activity. These data suggest that ABLIM-2 and -3 stimulate STARS activity. ABLIM-2 and -3 enhance STARS-dependent SRF-transcription in COS cells in a dose-dependent manner , suggesting that STARS and ABLIMs not only physically interact, but also functionally synergize to deliver activating signals to SRF. These data imply that in striated muscle STARS plays a critical role in the MRTF-A-nuclear translocation process; STARS promotes nuclear translocation of MRTFs, and thereby SRF-dependent transcription (Fig. 2).
STARS activation of SRF-dependent transcription is mediated, in part, by a Rho-dependent mechanism, since the Rho inhibitor C3 transferase reduces SRF activation by STARS. The ability of the Rho kinase inhibitor, Y-27632, to diminish SRF activation by STARS also implicates Rho kinase as a downstream effector of STARS . The Rho family of GTPases, including the best characterized members, Rho, Rac, and Cdc42, serve as molecular switches in the regulation of a wide variety of signal transduction pathways [62, 63], in particular actin polymerization and stress fiber formation . RhoA signaling has been shown to induce the nuclear import of MRTF-A in smooth muscle cells, thereby triggering smooth muscle gene activation . It is well-known that actin dynamics and Rho signaling are involved in STARS-induced nuclear translocation and transcriptional activation of MRTFs, and Rho activity is crucial for actin dynamics. Kuwahara et al.  showed that the dominant-negative RhoA mutant inhibits the nuclear accumulation of MRTFs and the stimulatory effect of STARS on the transcriptional activity of MRTFs. Although STARS requires Rho activity to induce actin treadmilling and MRTF nuclear translocation and the inhibition of Rho activity blocks STARS activity, assays of RhoA activity in STARS-transfected cells showed no difference from that in untransfected cells. Thus, STARS does not appear to function as an upstream activator of Rho, but requires Rho-actin signaling and changes in actin dynamics to evoke its stimulatory effects on MRTFs and SRF activity. Taken together, the small GTPase acts downstream of STARS, and it seems possible that ABLIM integrates signals from the small GTPases, Rac and RhoA (via STARS) toward the actin cytoskeleton.
Nuclear receptors regulated by ABPs include the glucocorticoid receptor (GR), the estrogen receptor (ER), the androgen receptor (AR), the thyroid receptor (TR), and peroxisome proliferator-activated receptor-γ (PPAR-γ). Among these nuclear receptors, AR is the most widely studied and well-characterized. AR is a ligand-activated transcription factor which controls the expression of genes involved in functions, such as cell proliferation, cell growth, differentiation, and cell death [66, 67]. AR contains an N-terminal domain (NTD) harboring activation function 1 (AF-1), a central DNA-binding domain (DBD), and a C-terminal ligand-binding domain (LBD) containing activation function 2 (AF-2) [68–71]. Upon binding androgens, AR LBD undergoes conformational changes leading to dissociation from chaperones and translocation to the nucleus [72–75]. AR binding to DNA facilitates the recruitment of general transcriptional machinery and ancillary factors that results in the activation or repression of specific genes in targeted cells and tissues . In the last decade, an increasing number of proteins have been proposed to possess AR co-activating or co-repressing characteristics [77, 78]. Co-factors facilitate AR transcription function by histone modifications, chromatin remodeling, and regulation of AR N-terminal domain and the ligand-binding domain interaction (N/C interaction) [79–83]. All available data suggest that no single AR-binding protein completely defines AR multiple functions in controlling cellular growth and differentiation in normal and malignant cells . Alternatively, AR pleiotropic activities are probably mediated through its binding to specific functional protein complexes to carry out its broad biological functions in mammalian cells. Greater than 200 nuclear receptor co-regulators have been identified since the isolation of the first nuclear receptor co-activator, SRC-1, in 1995 . Among the nuclear receptor co-regulators, ABPs and actin monomers bind to the AR, indicating that they also play an important role in AR-mediated transcription (Fig. 3) [5, 85]. For example, supervillin, a nuclear/cytoplasmic F-actin-bundling protein, is able to interact with the AR NTD and DBD-LBD. This association is enhanced in the presence of androgens . In recent years, ABPs have been shown to elicit increased activity in regulating AR than previously thought (Table 1).
Filamin, which was originally identified as a protein that facilitates nuclear transport of the AR, interacts with the AR DBD-LBD in a ligand-independent manner [78, 87, 88]. The absence of filamin hampers androgen-induced AR transactivation. The mechanism is that in the absence of filamin, the receptor-Hsp90 (Hsp90, a chaperone protein that plays a key role in the conformational change and transcriptional activity of the AR) complex may remain anchored in an inactive stateto the actin filaments, even in the presence of steroid and anavailable nuclear localization sequence on the receptor . The function of filamin may be acting as a mediator between the receptor and the Hsp90 and controlling the release of activated receptor after ligand binding in the AR cytoplasmic trafficking [88, 89]. Filamin-A (FLNa) interferes with AR interdomain interactions and competes with the co-activator transcriptional intermediary factor 2 (TIF2) to specifically down-regulate AR function . Full-length FLNa, when cleaved at the protease-cleavage site between repeats 15 and 16, releases the FLNa (16–24) [87–91]. This naturally-occurring filamin fragment, the C-terminal 100-kDa fragment, interacting with the motor protein dynein, can exert its inhibitory effect by interfering with interactions between the N- and C-terminal domains, and co-activator functions of the AR [87, 92]. Full-length FLNa being bound to the actin cytoskeleton on the cell surface and perinuclear areas of the cell through its N-terminal actin-binding domain. In the absence of ligand, AR is localized predominantly in the cytoplasm whose hinge domain and the LBD are tethered to the C-terminal end of FLNa . FLNa (16–24) co-localizes with liganded AR to the nucleus. In the nucleus, FLNa (16–24) disrupts interactions between the N- and C-termini of the AR and interferes with the binding of the co-activator TIF2 [87, 92]. There is evidence that interaction between the FXXLF (X=any amino acid) motif of the TAD and the LBD reduces co-activator recruitment and the binding of the LXXLL motif of TIF2 . Alternatively, FLNa (16–24) can also directly recruit transcriptional repressors onto the target promoter or possess intrinsic histone deacetylase activity to inhibit transcription initiation . In addition, the recent report of the Rho-regulated PAK6 as an AR hinge-interacting kinase  suggests that the FLNa (16–24)-AR hinge complex may serve as an integrator for the many cytoskeletal signaling cascades converging on the AR.
Supervillin (SV) was initially identified from blood cells as an ABP and found to be expressed in skeletal muscles and in several cancer cell lines . SV is localized to the plasma membrane at sites of intracellular contact. The nuclear localization signal is located in the middle of this protein . At low density, SV shows a punctate distribution localized to the cytoplasm and nucleus, whereas at high density, SV is localized almost exclusively to the plasma membrane. SV has been identified as an AR-interacting protein, which can interact with both N-terminal activation function-1 and C-terminal activation function-2 of AR and plays a role in the AR dimerization . The functional co-regulator domain of SV is located in amino acids 831–1281 of bovine origin, which has putative actin-binding sites and NLS . Ting et al.  showed that SV (amino acids 831–1281) has a better enhancing effect on AR transactivation than full-length SV and SV (amino acids 1010–1792). It is possible that by remaining within the nucleus, SV may increase the interaction frequency with AR, hence resulting in changing the AR conformation to an activated form to facilitate the binding of the androgen response element located in the target genes. SV is relatively weak in promoting non-androgenic steroid-mediated AR transactivation, but is capable of coordinating with other co-regulators, including ARA55 and ARA70, to enhance AR transactivation [97, 98]. These results suggest that the final AR activity may be the balance and coordination of multiple co-regulators in any given cell. In addition, previous experiments reported that actin and SV potentiate each other in promoting AR activity . Since several putative actin-binding sites and functional NLS of SV are important for the AR transactivation function, the minimal functional fragment of SV, and which only contains one actin-binding site and is located in the nucleus, recruiting actin into the chromatin remodeling complex may be a potential mechanism of co-regulators . On the other hand, the actin chelator, latrunculin B, which attenuates the co-regulator function of both the full-length and minimal functional fragment of SV, further identifies this potential mechanism. Furthermore, Rac signaling stimulates membrane ruffling that further attenuates the co-regulator activity of SV. There are two possible explanations for this outcome, as follows: 1) the accumulation of SV in the membrane refraining it from associating with AR, and 2) the decrease in the amount of actin monomer affecting SV co-regulator activity which demands actin monomers . However, SV has no effect on cytoplasmic-nuclear translocation of the AR, nor does it affect the half-life of the AR .
Gelsolin is a multifunctional ABP, well-characterized as having implications in cell signaling, cell motility, apoptosis, and carcinogenesis [99, 100]. Gelsolin regulates actin polymerization and depolymerization by sequestering actin monomers, and it can sever and cap actin filaments . Nishimura et al.  has identified that gelsolin is an AR-interacting protein that can interact with AR and enhance its transactivation in prostate cancer cells. As gelsolin lacks the nuclear localization signal, it is possible that it can be co-translocated into the nucleus with binding to other proteins . Like filamin, gelsolin is able to interact with AR at the time of its nuclear localization to facilitate the nuclear translocation of AR . Increased expression of gelsolin can enhance AR activity under HF with low levels of androgen treatment to maintain AR-mediated growth and survival of tumor cells. Gelsolin itself interacts with AR LBD via an FXXFF and an FXXMF motif and enhances its activity in the presence of androgen. The interaction between the N and C termini of the AR does not affect gelsolin FXXFF binding to AR LBD, indicating that the gelsolin FXXFF motif has a higher affinity for AR LBD . Two peptides, D1 (amino acids 551–600) and H1–2 (amino acids 665–695) located within AR DBD and LBD, respectively, can block gelsolin-enhanced AR activity . Altogether, gelsolin interacts with the AR during nuclear translocation and enhances ligand-dependent AR activity.
Transgelin, also named SM22α, was first isolated from chicken gizzard as a transformation-and shape change-sensitive ABP . Recently, Yang et al.  characterized transgelin as a potential prostate cancer suppressor via inhibition of ARA54-enhanced AR transactivation. ARA54, a RING finger protein, interacts with AR and enhances its transcriptionalactivity in a ligand-inducible manner. Transgelin does not directly interact with the AR, but exerts its effects through recruitment to ARA54. ARA54 can interact with transgelin both in vitro and in vivo in an androgen-independent manner . These data suggest that transgelin might need the specific interaction with ARA54 to suppress AR transactivation. In contrast, transgelin shows little interaction with AR, ARA70, ARA55, SRC-1, supervillin, gelsolin, and CBP. Silencing of endogenous ARA54 via its siRNA can abolish transgelin’s suppressive effect on AR function . Therefore, this suggests that transgelin may be able to suppress ARA54-enhanced AR transactivation via interruption of the interaction between AR and ARA54, as well as ARA54 homodimerization, resulting in the enhanced cytoplasmic retention and impaired nuclear translocation of ARA54 and AR.
Flightless-1 (Fli-I) is an ABP that can be either associated with the cytoskeleton or in the nucleus, but its exact physiologic functions have not been elucidated . Fli-I can associate with AR directly and function in cooperation with specific combinations of other AR co-activators to enhance the ability of AR to activate transcription of AR-regulated genes . Since Fli-I by itself does not enhance the activity of AR, but requires the presence of a p160 co-activator, the binding of Fli-I to AR is apparently not sufficient for Fli-I co-activator function . The contacts between Fli-I and multiple components in the transcription complex (AR, glucocorticoid receptor-interacting protein 1 [GRIP1], p160 and co-activator-associated arginine methyltransferase 1 [CARM1]) may result in a more efficient recruitment of Fli-I to the promoter, a more stable co-activator complex, or a more highly functional conformation of the co-activator complex. So, Fli-I is a secondary co-activator in AR transcription activation .
α-actinin-2 is a major structural component of sarcomeric Z-lines in skeletal muscle, where they function to anchor actin-containing thin filaments in a constitutive manner . α-actinin-2 enhances the transactivation activity of SRC-2 and serves as a primary co-activator for the AR, acting in synergy with SRC-2 to increase AR transactivation function . Huang et al.  indicated that wild-type α-actinin-2 (containing a LXXLL motif) and mutant α-actinin-2 (a mutation of the LXXLL motif to LXXAA) both bind to AR, but the mutant form has a much weaker ability than wild-type α-actinin-2. That is to say, LXXLL motif in α-actinin-2 plays a major role in the interaction with AR. However, the LXXLL motif of α-actinin-2 is shown to be a dispensable motif for its primary co-activator role in NR functions, as two truncated α-actinin-2 (encoding 281–700 and 701–894) lacking the LXXLL motif or the mutant α-actinin-2 (LXXAA) retains the primary and secondary co-activator functions of wild-type α-actinin-2. In addition, α-actinin-2 not only serves as a primary co-activator in AR, but also synergistically interacts with GRIP1 and enhances GRIP1-induced AR co-activator functions in the presence of cognate ligands . Furthermore, α-actinin-4 also binds to the AR and exhibits co-regulating properties. α-actinin-4 may target AR for degradation and/or antagonize AR synthesis upon the addition of androgen. In addition, α-actinin-4 also negatively regulatesAR-mediated transcription .
More and more experiments have identified that proteins traditionally thought of as strictly cytoplasmic structural factors can influence gene regulation. ABPs transduced the changes in cell structure that occur during morphogenesis to the nucleus, which result in changes in gene expression either through the nuclear shuttling of transcription factors or through the assembly of transcriptional regulatory complexes .
ABPs can recruit the multiple components to transcription complexes through different types of interactions. Fli-I not only binds actin, but also binds the actin-like BAF53 (BAF complex 53kDa subunit, BRG1 associated factor), as well as p160 co-activator [105, 109]. Fli-I can help to secure the association of an SWI/SNF (switching/sucrose non-fermenting) complex to a p160 co-activator complex, and thus help to coordinate the complementary ATP-dependent nucleosome-remodeling activity of the SWI/SNF complex with the histone acetylating (e.g., from CBP and p300) and methylating [e.g., from CARM1 and protein arginine methyltransferase 1 (PRMT1)] activities of the p160 co-activator complex . In addition, Fli-I and Fli-I LRR-associated protein 1 (FLAP1) have an important role in regulating transcriptional activation by β-catenin and lymphoid enhancer factor/T-cell factor (LEF1/TCF). FLAP1 is a key activator, cooperating synergistically with p300 to enhance LEF1/TCF-mediated transcription by β-catenin. This Fli-I negatively regulates the synergy of FLAP1 and p300 . Lee et al.  found that Fli-I does not bind well to the p300 KIX domain and does not appear to inhibit FLAP1-p300 binding, suggesting that Fli-I does not interfere with the binding of FLAP1 to p300. The possible mechanism is that Fli-I may exert its negative influence by squelching the activity of FLAP1 and other essential factors that bind to Fli-I (Fig. 4). It is also possible that Fli-I may recruit negative regulators, such as HDACs, CtBP, Groucho, and Chibby, to the β-catenin/LEF1/TCF transcription complex. Both the leucine rich repeat (LRR) and gelsolin-like domains of Fli-I are required for the negative regulation of β-catenin function. Increased nuclear levels of Fli-I presumably favor NR-mediated transcription, while lowered nuclear levels of Fli-I or increased FLAP1 levels probably result in the release of FLAP1 and activate β-catenin/LEF1/TCF-mediated transcription through the synergy of FLAP1 and p300. Since Fli-I acts positively on NR-mediated transcription and negatively on β-catenin/LEF1/TCF-mediated function, Fli-I may help to determine the balance between NR and β-catenin/LEF1/TCF activity .
FLNa interacts with transcription factor forkhead box C1 (FOXC1) and serves as a transcriptional barrier for FOXC1 activity . The possible proposed mechanism of transcriptional regulatory activity by FLNa is as follows: (1) In the cytoplasm, FLNa crosslinks with actin filaments to regulate actin cytoskeletal integrity. Full-length FLNa can be localized to the nucleus. (2) Nuclear import of transcriptional regulatory molecules, such as pre-B-cell leukemia transcription factor 1 (PBX1), is regulated by FLNa. Such regulation may be achieved by FLNa association with protein kinases (PK). That is to say, efficient nuclear localization of PBX1 and the formation of a transcriptionally inactive FOXC1-PBX1 complex required the FLNa. (3) In response to cell stimuli and cytoskeletal reorganization, the expression of FLNa increases and raises the levels of the nuclear FLNa pool. In the nucleus, FLNa acts as a scaffold whereby FOXC1 and PBX1 transcriptional inhibitory complexes are assembled. Interaction of FOXC1 and FLNa partitions FOXC1 to HP1α-rich condensed heterochromatin in the nucleus and promotes an inhibitory interaction between FOXC1 and PBX1, reducing FOXC1 transactivity. Furthermore, FOXC1-PBX1 complexes are unable to recruit co-activator complexes and are targeted to transcriptionally inactive, HP1α-rich heterochromatin regions of the nucleus [108, 111]. That is to say, FLNa can promote the active repression of FOXC1 activity through the association with inhibitory proteins rather than simply the prevention of FOXC1 activation . FLNa also interacts with polyoma enhancer-binding protein (PEBP2β). FLNa retains PEBP2β in the cytoplasm, thereby hindering its engagement as a Runx1 (runt-related transcription factor 1) partner. On the other hand, PEBP2β is translocated into the nuclei in cells lacking FLNa, which enhances the transcriptional activity of PEBP2/CBF. The interaction with FLNa is mediated by a region within PEBP2β that includes amino acid residues 68–93. The deletion of this region enables PEBP2β to translocate to the nucleus [112, 113].
α-actinin-4 is capable of interacting with class II HDACs and other transcription factors and potentiates transcription activity by MEF2 . First, transient transfection data indicate that α-actinin-4 potentiates transcriptional activity by MEF2. Second, overexpression of α-actinin-4 decreases the interaction of MEF2A and HDAC7. Third, knockdown of α-actinin-4 decreases expression of TAF55. Fourth, MEF2C, α-actinin-4, and HDAC7 associate with the TAF55 promoter. Furthermore, HDAC7 binds to amino acids 1–86 of MEF2A, suggesting that MEF2 cannot bind HDAC7 and α-actinin-4 simultaneously. Thus, a possible competition model is that MEF2 may directly recruit α-actinin-4 to displace HDAC7 from MEF2. Alternatively, HDAC7 may recruit α-actinin in response to stimuli followed by association of α-actinin-4 with MEF2 and activation of transcription [78, 114].
FHL (four and a half LIM domains) family members also belong to the family of ABPs and are directly involved in the differentiation of muscle cells. The best-characterized member of this family is FHL2/DRAL. FHL2 has potential transcriptional activity and participates in a number of transcription regulations . Labalette et al.  identified that FHL2 cooperates with CBP/p300 and activates β-catenin/TCF target gene cyclin D1. FHL2 also interacts with myocardin and enhances myocardin and myocardin-related transcription factor (MRTF)-A-dependent transactivation of smooth muscle α-actin, SM22α, and cardiac atrial natriuretic factor (ANF) promotersin 10T1/2 cells . Hamidouche et al.  demonstrated that FHL2 interacts with β-catenin, a key player involved in bone formation induced by Wnt signaling, which potentiates β-catenin nuclear translocation and TCF/LEF transcription, resulting in increased Runx2 and alkaline phosphatase expression.
Human heart LIM protein (hhLIM) participates in remodeling of the actin cytoskeleton, possibly by promoting actin bundling . hhLIM has a feature of dual subcellular location depending on context. In the cytoplasm, hhLIM increases actin cytoskeleton stability by promoting bundling of actin filaments . In the nucleus, hhLIM interacts with Nkx2.5 (a cardiac-restricted transcription factor) via its N-terminal LIM domain and enhances the binding ability of Nkx2.5 to the NKE (Nkx2.5-binding element) boxes in the ANF promoter. These results suggest that hhLIM promotes the specific expression of the ANF gene by cooperating with Nkx2.5 . Muscle LIM protein (MLP) has been found in the nucleus during early development , where it is a potent activator of the myogenic regulatoryfactor myoD [122, 123]. Lu et al.  identified that MLP promotes the specific expression of the AChR gamma-subunit gene cooperatively with the myogenin-E12 complex during myogenesis.
In addition, two ABPs, RPABC-2 and -3, are present in all three RNA polymerases, and the solution of the crystal structure of pol-II shows that these two subunits are located close to each other at the surface of the polymerase [29, 125] and participate in the transcription initiation. RPABC-2 and -3 form an actin-binding patch that is common to all three RNA polymerases and identify the same function.
The findings reported above clearly show that ABPs participate in a wide range of muscle-specific gene expression, androgen receptor transport, and formation of transcription complexes. This aspect of ABPs is entirely novel and would never have been predicted 10 years ago. As an interesting note, modulation of nuclear ABPs on target gene expression offers a feasible target for developing new therapeutic agents. For example, because ABPs physically interact with AR to modulateits transcriptional activity, disruption of the AR-ABPs interaction could be an important strategy to regulate AR-mediated growth of prostate cancer cells. The expression of selective ABPs may offer a growth advantage to tumor cells in androgen ablation and/or anti-androgen therapy. We also predict that future work in this field will continue to uncover new properties of ABPs, not only revealing unexpected roles in the nucleus, but also the way they shuttle between cell compartments. This exciting area of research will require more detailed investigation.
This work was supported by the Program for Major State Basic Research Development Program of China (No. 2008CB517402), the National Natural Science Foundation of the People’s Republic of China (No. 30770787, 30670845, 30871272), the New Century Excellent Talents in University (No. NCET-05-0261), the Key Project of Chinese Ministry of Education (No.206016), and the Hebei Natural Science Foundation of the People’s Republic of China (No. C2008001049). This research was supported in part by the Intramural Research Program of the NIH, National Institute on Aging.
Conflict of interest: None declared.