Introduction

The critical role of transcription factors in the regulation of cell function has been undoubtedly established; their tasks in activation, repression, and/or modification of gene expression are necessary and required for growth, development, and differentiation (Shi et al., 1997). Several lines of evidences have demonstrated deleterious outcomes when transcription factors become dysfunctionally activated or inactivated, leading to cellular malfunction, instability, and in some cases, tumorigenesis. Yin Yang 1 (YY1) is one such ubiquitous transcription factor. It is important to divulge how complex factors such as YY1 function in diverse biological processes and ultimately shape the growth and viability of eukaryotic cells. Thus, the biology of YY1 and its fundamental properties that initiate proper cellular development and the recently expanded potential role for YY1 in cancer biology, specifically the regulation of and resistance to cancer therapeutics, will be highlighted in this review.

YY1 discovery

YY1 is a ubiquitous and multifunctional zinc-finger transcription factor (also known as δ, NF-E1, UCRBP, and CF1) member of the Polycomb Group protein family, a group of homeobox gene receptors that play critical roles in hematopoiesis and cell cycle control. YY1 was initially cloned and characterized simultaneously by two independent groups, Shi et al. (1991) and Park and Atchison (1991) who were inspired by the original observation by Berns and Bohenzky (1987) and Chang et al. (1989). While investigating the adeno-associated virus (AAV) P5 promoter region and its activation by E1A gene products, using systematic deletion analysis of the P5 promoter, Chang et al. (1989) identified two elements associated with basal and E1A-induced P5 activity: (1) the R1–R2 region (P5-60 site), a tandem repeat sequence of 10 base pairs, and (2) a binding site for the major late transcription factor (MLTF). Both elements had a negative effect in the absence of E1A oncoprotein, but converted to transcriptional activators in its presence. They theorized that the two trans-activators acted in concert to stimulate the P5 promoter and induce transcriptional activation in the presence of E1A. Noteworthy, simultaneous deletion of both elements reduced P5 promoter activity 25-fold, raising the possibility of the presence of the dual-acting transcriptional factor YY1 (Chang et al., 1989). A subsequent study reported by Shi et al. (1991) once again identified two cellular protein complexes interacting with the P5-60 site of AAV P5 promoter. Band shift assays using a 22-base pair oligonucleotide containing the P5-60 element, detected the formation of a protein complex that was competed out by addition of excess unlabeled P5-60 oligonucleotide but not by the oligonucleotide (P5ML) containing the binding site for the MLTF. The major protein component of this complex was termed YY1. A second P5-60-specific binding protein (termed factor 2) was identified but masked by a co-migrating complex formed by a nonspecific DNA-binding activity in HeLa cells. Shi et al. (1991) formally named the factor YY1. During the preparation of the manuscript, Shi et al. (1991) learned that YY1 had been cloned by two additional laboratories. Park and Atchison (1991) have identified and cloned the protein, which they termed NF-E1, based on its ability to bind within the Igκ 3′ enhancer (Park and Atchison, 1991). Hariharan et al. (1991) identified and cloned the protein, which they termed δ, based on its ability to bind to sequence elements downstream of the transcriptional start sites in the ribosomal protein L30 and L32 genes. Subsequently, YY1 has been identified in other species and has been assigned alternate nomenclature by other authors, including UCRBP (Flanagan et al., 1992), nuclear matrix protein NMP1 (Guo et al., 1995), and common factor 1 (Thomas and Seto, 1999) (Table 1).

Table 1 YY1 aliases by species
Full size table

YY1 structure

Chromosomal localization and molecular structure

A purified YY1 genomic DNA probe was used in FISH analysis to map the location of the YY1 gene to the telomere region of human chromosome 14 at segment q32.2 (Yao et al., 1998). The YY1 gene consists of five highly conserved exons encoding a protein of 414 amino acids in length, and an estimated molecular weight of 44 kDa. However, due to the structure of the protein, SDS–polyacrylamide gel analysis reveals its weight to be 68 kDa (Shi et al., 1997). Figure 1 illustrates significant similarities between human and mouse YY1 nucleotide sequences. According to AceView database (http://www.ncbi.nih.gov/IEB/Research/Acembly/index.html), the sequence of the YY1 gene is supported by 850 sequences from 781 cDNA (accessed November 2004). The human YY1 gene produces eight different transcripts (a, b, c, d, e, f, g, and h) generated by alternative splicing, encoding eight different putative protein isoforms (three complete, three COOH-complete, and two partial). The functional significance of these isoforms remains elusive. There are two alternative promoters. Different transcripts differ by truncation of the 5′ end, truncation of the 3′ end, presence or absence of four cassette exons, and different boundaries on common exons due to variable splicing of an internal intron (Figure 2).

Figure 1
figure 1

YYl DNA Sequence homology. Sequence comparison of the main open reading frame of the human (M77698) YYl gene (Top sequence) and the mouse (M73963) YYl gene (Bottom sequence) showing 94.9% of identity and similarity determined by the Smith-Waterman alignment of nucleic acids. Both coding regions encode for a putative protein of 414 amino acids with a predicted molecular weight of approximately 44 kDa.

Full size image
Figure 2
figure 2

Gene organization and alternative splice variants of the YY1 gene. This figure schematically illustrates the localization of YY1 to chromosome 14. Eight different transcripts (a–h) generated by alternative slicing, encode eight different putative protein motifs.

Full size image

Biochemical and crystal structure

The YY1 protein contains four C2H2-type zinc-finger motifs with two specific domains that characterize its function as an activator or repressor. Analysis of GAL4 fusion protein revealed repression of transcription by the C-terminus domain (aa 298–397) (Shi et al., 1991, 1997) using a chloramphenicol acetyl transferase (CAT)-based reporter system driven by a promoter rich in GAL-4-binding sites. Two other domains contributing to its repression include sequences within the zinc-finger motifs and a glycine-rich residue between amino acids 157 and 201. The N-terminus region (aa 43–53), however, acts as a potent activation domain (Shi et al., 1997; Nguyen et al., 2004). This region is followed by a glycine-rich domain and 11 consecutive histidine residues (aa 70–80). The role of this sequence remains elusive (Shi et al., 1997). The cocrystal structure of YY1 is shown in Figure 3 (Houbaviy et al., 1996).

Figure 3
figure 3

YY1 Cocrystal structure. The cocrystal structure of YY1 is shown (Houbaviy et al., 1996, Research Collaboratory for Structural Bioinformatics Protein Data Bank http://www.pdb.org, accessed January 2005). The protein contains four C2H2-type zinc-finger motifs with two specific domains that characterize its function as an activator or repressor. Transcriptional repression is known to occur at the C-terminus (aa 298–397) directed by a promoter rich in GAL-4-binding sites. The N-terminus (aa 43–53), acts as a potent activation domain. Evidence that the zinc-fingers and glycine-rich regions of YY1 are instrumental in YY1 repression has been provided by deletional experiments of both regions, which render the protein incapable of transcriptional repression.

Full size image

Family/homology

Family

The YY1 sequence homology to the Drosophila Krüppel protein, a peptide initially described in 1984 and shown to be necessary for embryogenesis and normal morphology via transcriptional activation and repression, identifies it as a member of the GLI-Krüppel gene family (Wieschaus et al., 1984; Shi et al., 1991, 1997).

Homology

The fundamental role of YY1 in development and cellular propagation is supported by studies demonstrating mammalian cDNA encoding a YY1-binding protein possessing sequence homology with the yeast transcription factor reduced potassium dependence 3 protein (RPD3) (Yang et al., 1996). Thus, critical sequences reveal a high degree of interspecies homology for this transcriptionally active gene. More recent DNA and amino acid sequence database analyses show striking similarities in structure and function of YY1 to a newly discovered sister protein, Yin Yang 2 (YY2) (Nguyen et al., 2004). Deletion analysis reveals that YY2, like YY1, contains both activation and repression domains (N-terminus and C-terminus, respectively). In fact, it is heavily involved in gene regulation controlled by YY1. YY2 has been shown to interact with most, but not all promoter-binding sites associated with YY1 and has an almost identical YY1-like cDNA with slight nucleotide differences. The aa 256–365 sequence reveals an 86% homology to the zinc fingers of YY1 and 62% homology to the spacer regions. (Nguyen et al., 2004). The functions of these regions may have possible implications for the activities of YY1 and YY2, but are yet to be elucidated.

Role of YY1 in transcriptional regulation

Activation versus repression

By tethering to DNA promoters, YY1 regulates a variety of cellular and viral genes (Donohoe et al., 1999; Nguyen et al., 2004). What distinguishes this protein from other transcription factors is its ability to not only initiate transcription but also regulate it through activation or repression. Studies have repeatedly shown the association and modulation of YY1 by adenovirus-derived E1A, a protein that activates the AAV P5 promoter (Chang et al., 1989). The presence of E1A induces YY1-mediated activation of transcription. In its absence, the role of YY1 is reversed, converting to a transcriptional repressor (Shi et al., 1997); hence the name Yin Yang 1.

To clarify the process by which activation is favored in the presence of E1A, but switched to repression in its absence, studies were designed to test its functional status by masking and/or exposing the binding sites of YY1. In the absence of E1A, the AAV virus fails to undergo transcription, most likely due to YY1 binding to the P5 promoter (Shi et al., 1991). Mechanisms that have been proposed to explain this phenomenon include the possibility of a conformational change in YY1 through covalent modification, or a direct interaction between YY1 and an E1A-type accessory protein. The mechanisms by which this occurs remain unclear.

Embryogenesis, growth, and differentiation

YY1 also plays pivotal roles in mammalian biological processes. Donohoe et al. (1999) examined genotyped mouse embryos at different gestational stages. Mouse embryos made homozygous for the mutated YY1 allele did not survive; after uterine implantation they failed to develop beyond the blastula stage. YY1 heterozygotes survived, but displayed significant growth retardation and neurological defects, suggesting the significance of functional YY1 activity during later stages of mouse embryogenesis. Similar results were seen during the development of the African clawed frog, Xenopus laevis (Morgan et al., 2004).

Kurisaki et al. (2003) recently identified YY1 as a nuclear factor that interacts with mothers against dPP (MAD) and Mad-related (Smad) complexes, the principal signaling proteins of intracellular factors including transforming growth factor beta (TGF-β) and bone morphogenic protein (BMP), both of which are responsible for cell growth and differentiation. YY1 was found to interact with and repress Smad-specific transcriptional activity, suggesting its essential function in cell differentiation stimulated by TGF-β1 and other nuclear factors.

Proliferation and response to genotoxic stimuli

The ubiquitous presence of YY1 suggests important roles for cellular stability and normal functioning. YY1 has recently been found to activate DNA repair. Studies have shown an enhanced stimulation of poly (ADP-ribose) polymerase-1 (PARP-1) in HeLa cells transfected to overexpress YY1 after exposure to methyl-N-nitro-N′-nitrosoguanidine, an agent known to cause transient cell arrest in the G1 or G2 phase of the cell cycle. PARP-1 modulates DNA repair via the base excision repair pathway to rejoin nicked strands of DNA. Overexpression of YY1 in HeLa cells stimulates catalysed PARP-1, resulting in accelerated DNA repair (Oei and Shi, 2001a, 2001b). This, however, seems to act as a negative feedback; continuous overexpression of PARP-1 decreases YY1 affinity by poly-ADP-ribosylation at its DNA-binding sites and induces transcriptional silencing (Oei and Shi, 2001a, 2001b). The system, therefore, functions to control gene modification and decrease production and overexpression of damaged genes. This process may have implications in the relief of genetic defects, senescence, and cancer.

Induction of YY1

Cellular localization and trafficking

Cellular localization of transcription factors to the nuclear matrix is essential for transcriptional regulation and control. McNeil et al. (1998) has identified specific sequences that lead YY1 to nuclear targets. Analysis of deletion constructs composed of Gal-4-tagged YY1 fusion proteins expressed in Hela cells and human Saos-2-osteosarcoma cells reveal the C-terminal domain (aa 256–341) as the chief constituent involved in high-affinity efficient targeting of YY1 to the nuclear matrix. The N-terminal domain of the protein permits a low-affinity association into the nucleus, but is not necessary, thus suggesting the significance of the C-terminus in nuclear localization as well as transcriptional repression.

Progression through the cell cycle also induces a DNA replication-associated switch in YY1 subcellular localization. As a DNA-binding protein, YY1 functions in the replication and regulation of the histone alpha complex, vital for proliferating cells (Palko et al., 2004). YY1 nuclear localization and activity is significantly increased during the onset of the G1/S phase, followed by increased cytoplasmic localization in the late S phase with increased DNA-binding activity of YY1 and YY1-dependent histone genes (Palko et al., 2004).

Molecular regulation of YY1

Indirect evidence

Little is known regarding the regulation of YY1 activity. Several investigators have been able to demonstrate increased activity at promoters for genes such as histone deacetylase complex (HDAC) 1 and 2 in correlation with increased YY1 activity. Thus, indirect evidence exists to suggest that promoters, often in conjunction with cofactors such as msin3A, nuclear receptor corepressor (NCOR) and Sin3-associated polypeptide (SAP) 18/30, may regulate YY1 (Thomas and Seto, 1999). At the protein level, Hiromura et al. (2003) have shown in a murine retinoblastoma (Rb) model that alteration in YY1 chemical structure by O-linked N-acetylglucosaminylation frees YY1 to bind DNA, resulting in transcriptional activation. These findings provide the basis for a mechanistic hypothesis published in 2001 demonstrating the net suppressive effects after inhibition of YY1 binding to HDAC promoter binding sites, also in an Rb model (Osborne et al., 2001). Lastly, two models of post-translational cytoplasmic proteolytic activation are revealed in studies of the regulation of muscle development in primary skeletal muscle and cardiac cell lines. Loss of activation is achieved with proteolytic inhibition; these models are proposed to explain modulation of YY1 regulation in myoblast differentiation (Walowitz et al., 1998). Human gene promoters that regulate YY1 are summarized in Table 2.

Table 2 Human gene promoters that regulate YY1
Full size table

Direct evidence

Direct activation by transcriptional activators has been shown only in a few models. Lee et al. (2004) suggests that bone morphogenic protein (BMP) induces GATA genes in an autocrine fashion and modulates YY1 transcriptional activity via the direct interaction of YY1 with BMP-activated SMADs. The transcription factor nuclear factor kappa B (NF-κB) has also been shown to regulate YY1. Elegant studies by Sepulveda et al. (2004) demonstrate concurrent direct binding of the Rel-B component of NF-κB to YY1 and sequences at the hs4 enhancer region of B-cell lymphoma Igh gene, thereby implicating this complex in the anti-apoptotic response and the upregulation of the proliferative potential of these lymphocytes in vivo.

YY1-mediated transcriptional regulation

Mechanisms of YY1-mediated transcriptional repression

It has been suggested that YY1 represses transcription using multiple mechanisms. Most frequently, these mechanisms involve the competition of YY1 with activating factors in overlapping binding sites, thereby decreasing promoter activity and resulting in transcriptional repression. Other hypotheses include the negative regulation of YY1 on neighboring promoter-bound activators (Shi et al., 1997). As delineated by both Shi et al. (1997) and Thomas and Seto (1999), there are three models that explain YY1 as a transcriptional activator.

The displacement model

(Figure 4) Accumulating evidence suggests the presence of many promoters with sequences of YY1 sites that overlap and compete with activating factors, including serum response element (SRE) of the cellular FBJ/FBR osteosarcoma (s-fos) gene, α-actin muscle regulatory elements (MREs), and the muscle creatine kinase CarG motif (Shi et al., 1997) have demonstrated overlapping sites that compete with YY1 for occupancy. YY1 competition with MREs suggests its significance in modulating myoblast maturation and differentiation. Likewise, competition of YY1 with the β-casein activating promoter of mammary epithelial cells, known as mammary gland factor (MGF), results in transcriptional repression. Transcriptional repression is reversed by the alternative competition/displacement model. In the MGF model, when lactation results in increased MGF concentration, YY1 is displaced from its overlapping site on the β-casein promoter, resulting in baseline activation (Shi et al., 1997). Expression of other transcription factors such as NF-κB may also increase, thereby displacing YY1 and relieving repression, such as that demonstrated in serum amyloid gene transcription in hepatoma cell lines (Lu et al., 1994).

Figure 4
figure 4

Activator-displacement-induced repression model of YY1-mediated transcriptional repression. YY1 can effect transcriptional repression at promoter binding sites following competition with activators with subsequent activator displacement. This has been demonstrated in the myoblast YY1/α-actin MRE interaction and the YY1/muscle creatine kinase CarG motif interaction. Both genes are known to possess overlapping binding sites that compete with YY1 for occupancy.

Full size image

Interference with the function of transcriptional activators

(Figure 5) It has been established that the c-fos promoter not only contains overlapping sites for YY1 and SRE, but also possesses two additional YY1 sites between the calcium/cyclic AMP response element (CRE) and the TATA box. YY1 adheres distally, resulting in repression of the upstream CRE promoter (Figure 5a, Direct inhibition). YY1 can repress the c-fos promoter in either a binding site-dependent or binding site-independent manner (Shi et al., 1997), both of which involve the interaction of the zinc-finger motifs on YY1 and the basic leucine zipper region (bZIP) on the cAMP response element binding (CREB) protein. The presumed manner by which YY1 and CREB interact in the nucleus (Guo et al., 1997) and lead to the transcriptional repression of CREB represents an example of a binding site-independent reaction (Figure 5b). Galvin and Shi (1997) argue against the DNA-binding model, demonstrating the ability of YY1 to interfere with the communication of CREB and consequently retard CREB-mediated activation (Figure 5b) independently of physical interactions with DNA. As a potent coactivator of YY1, E1A can block YY1-induced repression by disturbing the YY1–CREB interaction (Chang et al., 1989; Shi et al., 1991, 1997; Yao et al., 1998).

Figure 5
figure 5

Models of YY1-mediated inhibition of transcriptional activation. (a) YY1 directly represses transcriptional activation. Despite the presence of a bound transcriptional activator to a gene promoter site, YY1 may adhere distally, resulting in repression of the upstream promoter. Such is the case with the c-fos promoter, possessing two additional YY1 sites between the CRE and the TATA box. (b) YY1 exerts transcriptional activator inhibition via direct physical binding. YY1 interferes with the action of a transcriptional activator, thus causing transcriptional repression. An example of binding-site-dependent interference involves the c-fos promoter, stimulated to transcriptional activation by binding at the CREB site, but transformed to repression due to the interaction of the YY1 zinc-finger motifs and the basic leucine zipper region (bZIP) on the CREB protein. (c) YY1 exerts transcriptional activation: inhibition by interaction. The DNA-bending model, suggesting the ability of YY1 to interfere with the communication of CREB and consequently retard CREB-mediated activation, represents an example of a binding site-independent reaction independent of physical interactions with DNA and gene promoter sites.

Full size image

Interactions with corepressors

(Figure 6) A third model exemplifies the ability of YY1 to recruit corepressors that directly act to facilitate transcriptional repression or induce chromatin remodeling/condensing to further assist YY1-mediated DNA interaction and repression (Thomas and Seto, 1999). The zinc-finger and glycine-rich regions of YY1 are known to be instrumental in YY1 repression activity. Simultaneous deletions in both regions and/or each individual region render GAL4-YY1 fusion proteins insufficient for transcriptional repression. Additionally, YY1 often requires the help of cofactors that interact with its repression domains to facilitate repression. Such cofactors include mRPD3, a mouse homologue of RPD3 protein which was shown to enhance transcriptional repression when overexpressed in GAL4-YY1 fusion proteins; GATA-1, involved in the corepression of the ɛ-globin gene (Yang et al., 1996); and Smad family members (Kurisaki et al., 2003), involved in the inhibition of TGF-β-induced epithelial to mesenchyme transition.

Figure 6
figure 6

YY1-mediated repression of activators via corepressor complexes. YY1 mediates transcriptional activator repression by complexing with corepressors. YY1 may recruit corepressors that directly act to facilitate transcriptional repression or induce chromatin remodeling/condensing to further assist YY1-mediated DNA interaction and repression.

Full size image

YY1 is also capable of repressing cofactors pivotal to cellular activity and viral regulation, including interferon beta (IFN-β) and gamma (IFN-γ). YY1 binding to the IFN-β promoter may activate or repress transcriptional activation of IFN-β depending on its association with HDACs (Weill et al., 2003). It also associates with nuclear factor AP2 to form protein complexes that relieve transcriptional activation of IFN-γ (Ye et al., 1994). Two mechanisms that may therefore account for YY1-mediated transcriptional regulation of gene products such as interferons include (1) a similar displacement competition with YY1 and an AP-1 overlapping site and (2) YY1-mediated repression only in the presence of a neighboring site binding an AP-2-like protein (Shi et al., 1997).

Direct activation

(Figure 7) The first model proposes direct interaction of YY1 with transcription factors that stimulate YY1-mediated transcriptional activation, such as TATA-binding protein (TBP), (TBP)-associated factors (TAFs), and transcription factor IIB (TFIIB) (Nguyen et al., 2004). It is likely that YY1 uses two acidic activation domains to accomplish this. However, studies have shown interactions with cofactors that act distal to the N-terminus, suggesting a possibility of a complex regulation of repression activity exceeding activation (Thomas and Seto, 1999). This model may therefore represent an oversimplified model for YY1-mediated activation.

Figure 7
figure 7

Direct activation by YY1. YY1 may directly activate gene transcription by binding to gene promoters and/or transcription factors that stimulate YY1-mediated transcriptional activation. Known examples include TATA-binding protein, TBP-associated factor II55, and transcription factor IIB.

Full size image

Cofactor-induced inhibition of YY1 repression

(Figure 8) The second model proposes a mechanism that induces the masking and unmasking of repression (C-terminus) and activation (N-terminus) domains, respectively. It is possible that YY1 interacts with other cellular factors to unmask the N-terminal activation domain, perhaps by undergoing structural alterations including changes in the C-terminus, thereby inhibiting the constitutive YY1 repression (Thomas and Seto, 1999). In addition, the C-terminus may play an important role in masking the N-terminal activation region; studies in which the YY1 C-terminal domain was deleted resulted in the exposure of N-terminal sequences and a significant enhancement in the transcriptional activation of YY1 (Thomas and Seto, 1999).

Figure 8
figure 8

Indirect activation by YY1. Activation via cofactor-induced inhibition of YY1 repression. The net result of separate or combined induction of C-terminus (repression) domain masking or N-terminus domain (activation) unmasking of repression is transcriptional activation. It is possible that YY1 may interact with cellular factors to undergo structural alterations, thereby inhibiting constitutive YY1 repression. The C-terminus domain itself may play an important role in masking the N-terminal activation region in an autocrine fashion. Studies in which the YY1 C-terminal domain was deleted resulted in the exposure of N-terminal sequences and a significant enhancement in YY1-mediated transcriptional activation.

Full size image

Recruitment of coactivators

(Figure 9) Thirdly, YY1 may act as an indirect activator of transcription by recruiting other transcription activating factors. It primarily induces cofactors to tether directly to the target promoter and initiate activation (Thomas and Seto, 1999). As was seen in the repression models, YY1 has also been found to interact with coactivators with histone acetyltransferase (HAT) activity, such as CBP and p300 (Lee et al., 1998). YY1 may recruit p300, thereby facilitating chromatin expansion to provide better DNA interactions. This mechanism allows for an easier manner by which YY1 can carry out transcriptional activation (Lee et al., 1995). These findings provide a strong consensus that supports the direct interaction of cofactors for YY1-induced activation.

Figure 9
figure 9

YY1-mediated activation via recruitment of coactivators (Co-A). YY1 may act as an indirect activator of transcription by inducing cofactors to tether directly to the target promoter and initiate activation. Coactivators with histone acetyltransferase (HAT) activity such as CBP and p300 are likely candidates. YY1 may also recruit p300, thereby facilitating chromatin expansion to provide better DNA interactions.

Full size image

Cofactors involved in YY1-mediated transcriptional regulation

Coactivators of YY1

YY1 may act independently. However, as noted previously, many of the effects of YY1 on gene transcription are executed via cofactors (Table 2). The AAV protein E1A is an example of a coactivator of YY1. Possible mechanisms involved in such collaborative activation might include protein–protein interactions with E1A/p300 complexes required for E1A to relieve YY1-mediated repression (Wang et al., 1993), and DNA binding (Kim and Shapiro, 1996; Shi et al., 1997). In addition, Wang et al. (1993) and Shi et al. (1991) independently demonstrated that E1A may serve as an initiator of YY1-mediated transcriptional activation via attachment to the P5 promoter. More recent studies screening for cellular proteins in the HeLA cDNA library have identified a novel protein, YY1-associated protein (YY1AP), a ubiquitous protein expressed in normal human tissue and metastatic cell lines. YY1AP was shown to colocalize into the nuclear matrix with YY1 and enhance its transcriptional activation in vivo and in vitro (Wang et al., 2004).

Corepressors of YY1

Yang et al. (1996) was able to isolate and identify mRPD3 by yeast two-hybrid assay revealing an identical glycine-rich domain (a necessary component for transcriptional repression) to that of YY1. Overexpression of mRPD3 significantly increased the ability of Gal4-YY1 fusion proteins to repress transcription, suggesting a possible role of mRPD3 as a corepressor of YY1.

SAP30 is also required for the normal functioning of RPD3, an alternate corepressor of YY1. SAP30 alone, however, is not sufficient for transcriptional repression, and is thus dependent on transcription factors such as YY1 to tether it to the promoter (Huang et al., 2003). The in vivo presence of this complex suggests one of many different YY1-dependent mechanisms of transcriptional repression.

The function of YY1 in transcription is context-specific and requires interactions with many cellular factors. As a result, YY1 develops intracellular networks that allow it to induce multiple functions in transcriptional initiation, activation and repression, ultimately leading to the regulation of normal cell growth and survival. As previously noted, it is apparent that YY1 expression and localization can be coordinated with phases of the cell cycle (Palko et al., 2004); it is via the study of the putative interactions between YY1 and cell cycle regulators, death genes, and transcription factors and cofactors that mechanistic evidence has surfaced to support the role of this ubiquitous transcription factor in the suppression or progression of various malignancies. The human promoters/gene products regulated by YY1 are summarized in Table 3.

Table 3 Human promoters/gene products regulated by YY1
Full size table

YY1 and cancer biology

In order to determine the potential impact of YY1 activity on tumorigenesis, a brief review of key cell cycle regulators, patterns of cell cycle dysregulation resulting in cancer, and known interactions between these regulators and YY1 is warranted. The two cell division events that control progression to replication are (1) entry into the S phase during which time DNA is replicated (G1 (first gap)/S (DNA synthesis) checkpoint), and (2) entry into the M-phase when mitosis occurs (G2 (second gap)/M checkpoint). CDK4/6-cyclin D and CDK2-cyclin E and the transcription complex that includes retinoblastoma (Rb) and E2F control the G1/S checkpoint. Phosphorylation of Rb by CDK 4/6-cyclin D and CDK2 dissociates the Rb-repressor complex, permitting rapid and transient transcription of S-phase-promoting genes (Bartek and Lukas, 2001). Expression of the proto-oncogene c-myc can influence cyclin D activity at this checkpoint (Amati et al., 1998). In addition, G1/S progression is associated with maximal phosphorylation of the ubiquitous DNA-associated protein DEK, which is thought to function as a transcription factor modulator (Kappes et al., 2004). Alternatively, the tumor suppressor transcription factor p53 may be activated at the G1/S checkpoint in response to DNA damage. Activation of p53 results in CDK2 inhibition, allowing for delay in the progression of the cell cycle for purposes of repair via poly (ADP-ribose) polymerase-1 (PARP-1), p21, and/or the negative regulator murine double minute 2 (Mdm2), an oncoprotein whose amplification and/or overexpression occurs in a wide variety of human cancers. Once repair is complete, progression to S phase proceeds. After S phase transcriptional activity is complete, dephosphorylation of a tyrosine residue within cell division control 2 (CDC2), the catalytic subunit of the cyclin/CDK heterodimer, signals activation late in G2 to activate G2/M progression (Zhao and Elder, 2005). HDAC's are critical at this juncture; they may enhance or inhibit progression at the G2/M checkpoint via DNA binding (Figure 10).

Figure 10
figure 10

Cell cycle, tumorigenesis and YY1. The cell cycle is a coordinated sets of events resulting in cell growth and cell division or proliferation. It can be described by four phases composed of M phase (mitosis), G1 phase (gap or growth 1), S phase (DNA synthesis) and G2 phase (gap or growth 2). Multiple mechanisms should act in concert in order to prevent uncontrolled cell division. Some of these are intrinsic molecules in the cell that regulate the transition from and between different phases of the cell cycle (i.e., CDKs, cyclins, pRB, p53, MDM2, c-Myc, etc.). Uncontrolled cell cycle progression is a major event in tumorigenesis. There is a dynamic interaction between the activity of the transcription factor YY1 and different components of the cell cycle and its check points. These interactions frequently result in a dysfunctional cell cycle progression and possibly tumorigenesis.

Full size image

Therefore, uncontrolled cell cycle progression is a major event in tumorigenesis. Multiple mechanisms should act in concert in order to prevent uncontrolled cell division. Some of these are intrinsic molecules in the cell that regulate the transition from and between different phases of the cell cycle (i.e., CDKs, cyclins, pRB, p53, MDM2, c-Myc, etc.), whereas others are signaling mechanisms sensing the environment that prompt a cell to remain in homeostatic balance with its surrounding tissue.

Regulation of the cell cycle by YY1

The putative role of YY1 in tumorigenesis is supported by its known interaction with the cell cycle regulation.

YY1 and cyclin D

The association of YY1 with cell cycle signaling pathways has been reported by Cicatiello et al. (2004), who noted that cyclin D1 gene promoter activation in estrogen-responsive human breast cancer is marked by release of the YY1 transcriptional repressor complex including HDAC 1 and is sufficient to induce the assembly of the basal transcription machinery on the promoter and to lead to initial cyclin D1 accumulation in the cell. Upon estrogen stimulation, the cyclin D1/CDK4 holoenzyme associates with the cyclin D1 promoter, where E2F and pRb can also be found, contribute to the long-lasting gene enhancement required to drive G1-phase completion.

YY1 and p53

Extensive evidence supports the role of YY1 in tumorigenesis via its association with the tumor suppressor gene product p53. Both expression and function of p53 are tightly regulated by post-translational modifications such as phosphorylation, ubiquitination, and acetylation with the goal of preservation of genomic integrity. In transfection experiments, YY1 inhibits p53-activated transcription from the p53-binding site that contains the ACAT sequence. Furthermore, YY1 and p53 are noted to be colocalized around the nucleoli and in discrete nuclear domains in an in vitro model of apoptosis in PC-12 rat adrenal tumor cells (Yakovleva et al., 2004). YY1 may attenuate p53-dependent transcription from a subset of p53 target genes, a hypothesis that may be relevant for defining the role of YY1 in directing cells either to growth arrest or apoptosis upon p53 binding. Further in-depth models, described below, are offered by several investigators:

YY1 disrupts p53/p300 interaction. Sui et al. (2004) demonstrated that YY1 interacts with p53 and inhibits its transcriptional activity by disrupting the interaction between p53 and its coactivator p300, thereby blocking p300-dependent p53 acetylation and stabilization, and disabling this checkpoint mechanism. Ablation of endogenous YY1 results in p53 accumulation due to a reduction in p53 ubiquitination and increased expression of p53 target genes in response to genotoxic stress in vivo. Conversely, YY1 overexpression stimulates p53 ubiquitination and degradation, thereby supporting the hypothesis that increased YY1 expression and activity inhibits the accumulation of p53.

p53 enhances murine double minute 2-mediated p53 inactivation. Alternatively, YY1 is known to interact with the negative regulator murine double minute 2 (mdm2), an oncoprotein whose amplification and/or overexpression occurs in a wide variety of human cancers. YY1 promotes the assembly of the p53–mdm2 complex, enhancing the mdm2-mediated ubiquitination and subsequent inactivation of p53 (Gronroos et al., 2004). Likewise, studies demonstrating the direct physical interaction between Hdm2 (the human homologue of Mdm2) and p53 show that the basis for YY1 regulation of p53 ubiquitination is its ability to facilitate Hdm2/p53 interaction (Sui et al., 2004). Significantly, recombinant YY1 is sufficient to induce Hdm2-mediated p53 polyubiquitination in vitro, suggesting that this function of YY1 is independent of its transcriptional activity. These findings identify YY1 as a potential cofactor for Mdm2 and Hdm2 in the regulation of p53 homeostasis and indicate a possible role for YY1 in tumorigenesis.

YY1 binds the neurofibromatosis promoter site. YY1 has been shown to bind p53 at the neurofibromatosis 1 (NF1)/YY1 promoter binding site (Nayak and Das, 1999). NF1 is the mutant gene mapped to chromosome 17 and known to cause the tumor von Recklinghausen neurofibromatosis. Interestingly, Nayak and Das (2002) have proven this binding to be absent in tumor cells expressing the proapoptotic gene Bax. Taken together or independently, the above models provide molecular clues as to the mechanism of YY1-mediated modulation of p53 with the potential for consequent regulation of resistance to apoptotic stimuli.

YY1 and c-myc

The proto-oncogene c-myc possesses a key role in cellular processes such as proliferation, differentiation, apoptosis, and transformation (Riggs et al., 1993; Shrivastava et al., 1996). C-myc activity has been implicated in the pathogenesis of malignancies such as breast, ovarian, prostate, hepatocellular, and colorectal carcinoma as well as lymphoma and plasma cell tumors. Riggs et al. (1993) and Shrivastava et al. (1996) independently demonstrated in murine and human tumor models that YY1 can activate both endogenous and exogenous c-myc promoters when overexpressed. In turn c-myc overexpression appears to alter the constitutive repressive role of YY1 by interfering with the association between YY1 and basal transcription proteins such as TATA-binding protein and transcription factor IIF, with altered transcription of target genes (Austen et al., 1998).

Translational evidence of the significance of these findings has been demonstrated in a murine model of hepatocellular carcinoma (HCC) whereby YY1 binding is blocked in a model of N-nitrosodiethylamine-induced hepatocarcinogenesis, with concomitant c-myc overexpression. Reversal of tumor formation with dodecanol-limonene, a monoterpene monocyclic compound with an unknown mechanism of cancer chemoprevention, is associated with a constitutive (high) level of YY1 binding along with inhibition of c-myc overexpression as seen in non-tumorous liver tissue. Hence, YY1 may constitutively serve to repress c-myc responsive antiapoptotic signals (Parija and Das, 2003).

YY1 and retinoblastoma Rb

It is known that YY1 binds with the retinoblastoma (Rb) protein to accelerate cellular progression to S phase, thereby potentiating cellular proliferation and tumorigenesis. Petkova et al. (2001) postulate that the responsible mechanism is the Rb-YY1 heterodimerization resulting in inhibition of the transcription factor due to binding destabilization. Conversely, both inhibition of the of YY1–Rb complex as well as inhibition of HDAC binding at the promoter reverses YY1 transcriptional activation, potentially altering acceleration of cell cycle mechanics with resultant loss of malignant potential (Osborne et al., 2001; Hiromura et al., 2003).

YY1 and HDACs

YY1 binds replication-dependent histone genes to effect proliferation and chromatin remodeling for accelerated replication via HDACs (Guo et al., 1997; Thomas and Seto, 1999; He and Margolis, 2002; Huang et al., 2003). The state of HDAC activation is dependent upon acetylation (active) versus deactylation (inactive). Studies by Palko et al. (2004) reveal changes in YY1 subcellular localization in CHO and HeLa cells, specifically, upregulation of the histone gene family at the G1/S boundary and subsequent downregulation at the mid-point of the S phase to correlate with YY1 activity. These data suggest an intimate association between YY1 activity and HDAC activation. Thus, YY1 localization appears to be coupled to DNA synthesis and responsive to cell cycle signaling pathways, lending credence to the hypothesis that proliferation and furthermore, resistance to apoptosis may be mediated via YY1 regulation of HDACs.

YY1 regulation of cell death

Cell proliferation and cell death are two functionally opposing cellular fates that paradoxically share many, nonoverlapping, molecular interdependent components and regulatory signals. The two processes are coupled at various levels through the individual molecular player responsible for orchestrating cell expansion. Moreover, the same molecular components are targets for oncogenic changes that frequently drive cell proliferation to cooperate with those that uncouple proliferation from apoptosis during transformation and tumorigenesis (Evan and Vousden, 2001; Fridman and Lowe, 2003). Consequently, alteration in either or both processes, uncontrolled cell proliferation and impaired cell death, might synergize toward tumorigenesis.

As we have discussed above, YY1 interacts with many elements involved in cell cycle with an overall outcome of regulation of positive signals promoting cell proliferation (i.e., p53, MDM2, cyclin D, etc.). In addition, YY1 has been implicated in the regulation of the activity and expression of apoptosis-related molecules (i.e., NF-κB, Fas, DR5, etc.). It would not be surprising that deregulated YY1 activity might serve as central molecule causing dysfunctional cell proliferation and increased resistance to cell death, therefore promoting tumorigenesis (Figure 10).

Apoptosis

Apoptosis is a cellular suicide program that eradicates excess or potentially dangerous cells. Its important physiological functions include terminating immune responses and eliminating infected or cancerous cells. The induction of apoptosis relies critically on the activation of caspases, a family of proteinases that kill the cell via proteolysis of key substrates. Two main pathways have been defined that initiate caspase activation. The first begins at the cell surface and involves ligand-induced activation of death receptors (e.g., Fas, TNF-R1, DR4, DR5), which then recruit and activate caspases. The second involves mitochondrial integration of cellular stress signals and mitochondrial dysregulation with release into the cytosol of cytochrome c, which activates caspase via the adaptor molecule, apoptotic protein factor 1 (APAF-1). In certain cells, the induction of apoptosis by death receptor signaling is complemented by mitochondrial activation. The apoptotic pathways are regulated at multiple levels such as by inhibitors of apoptosis family (IAPs) and the Bcl-2 family. Defects in the regulation of apoptosis contribute significantly into the pathogenesis and progression of most cancers. Apoptotic defects also contribute to tumor cell resistance to chemotherapy, radiotherapy, hormonal therapy, and immune-based treatments. Alterations in the expression and function of several apoptosis regulatory genes have been demonstrated in many cancers suggesting new targets for drug discovery (Wolf and Green, 2002; Reed, 2004). A few examples will be provided below in which YY1 is shown to play an important role in the regulation of apoptotic signaling pathways.

YY1 and NF-κB

The observation that the NF-κB family of transcription factors plays a key role in the regulation of immune and inflammatory responses as well as apoptosis has led several investigators to study the role of this factor and its regulated genes in the process of tumor progression and metastasis. Mounting evidence from several models, including metastatic murine colon cancer (Luo et al., 2004), B-cell lymphoma (Sepulveda et al., 2004; Jazirehi et al., 2005), murine breast cancer (Rahman and Sarkar, 2005), and murine cholangiocarcinoma (Chen et al., 2005) suggest that NF-κB may modulate the apoptotic response. Furthermore, Huber et al. (2004) noted that NF-κB is necessary for epithelial to mesenchymal transition, a process that facilitates tumor metastasis. Mechanisms for NF-κB regulation in tumorigenesis, however, remain unclear. It is now known that YY1 and NF-κB may interact in multiple ways: Sepulveda et al. (2004) describe binding of the Rel-B NF-κB subunit to YY1 with the subsequent complex binding an enhancer region of the genome of a B-cell lymphoma line to increase IgH chain expression. Chan et al. (1996) have identified a serum response factor/NF-κB-like element containing potential overlapping core recognition binding motifs for YY1 in the cytomegalovirus promoter, a virus associated with immunosuppression-related lymphoma. Lastly, a cytokine response unit within the serum amyloid A gene promoter that binds NF-κB contains an overlapping binding motif for YY1. YY1 binding at that site was shown to effectively inhibit NF-κB binding and transcriptional activity (Lu et al., 1994). Thus, indirect evidence exists that would suggest that NF-κB and YY1, when simultaneously transcriptionally active, may act in concert to exert synergistic or opposing effects, depending upon the cellular context and stimulus for gene regulation.

YY1 and proinflammatory cytokines

Reports of tumor and peritumoral proinflammatory cytokine production have been a recurring theme in the oncologic literature. Gene products such as inducible nitric oxide synthese (iNOS), tumor necrosis factor alpha (TNF-α), interleukin–1 (IL-1), and IFN-γ are elaborated in an autocrine and/or paracrine fashion in several cancer models and their expression has been shown to modulate response to apoptotic stimuli. Specifically, TNF-α is known to engage the TNF receptor and mediate apoptosis (Laster et al., 1988); NO has been shown in several systems to induce apoptosis both via mitochondrial and Fas death pathways as well as by p53 and NF-κB modulation (Li et al., 2004; Sagoo et al., 2004); IL-1 appears to play a critical role in models of lipopolysaccharide-induced apoptosis (Hilbi et al., 1997), and IFN-γ can also induce p53-independent apoptosis (Kano et al., 1997; Ossina et al., 1997).

Based on our knowledge of the ability of YY1 to interact, either directly or indirectly, with promoter regions for IFN-γ (Ye et al., 1994; Sweetser et al., 1998) and IL-1 (Patten et al., 2000), it may be possible for YY1 to modulate the transcriptional pattern of response to cytokine gene activation, thereby modifying expression of downstream effectors and the apoptotic response. A great deal of evidence exists for a direct relationship between NO and YY1 expression: the endothelial NO synthase (eNOS) promoter possesses a YY1-binding sequence first identified by Karantzoulis-Fegaras et al. (1999). Garban and Bonavida (2001) subsequently described inhibition of YY1-binding activity at the Fas promoter in ovarian carcinoma and prostate cancer cell lines in the presence of NO. Further, a possible mechanism was suggested by Hongo et al. (2004) in a PC-3 prostate carcinoma cell line in which introduction of an NO donor was found to inhibit NF-κB and YY1 expression through s-nitrosylation. Similarly, Vega et al. (2005b) have shown that YY1 repression in Ramos B-cell lymphoma via the chimeric anti-CD 20 mAb, rituximab, results in sensitization to Fas-mediated apoptosis. Thus, there is likely a significant association between YY1 activity and both cytokine and death receptor expression, particularly in tumor systems, with a potential for chemo- and immunoresistance and consequent survival via these pathways.

Role of YY1 in immune resistance

Resistance to Fas

Prostate cancer cell lines have been examined for sensitivity to Fas-induced apoptosis using the FasL agonist monoclonal antibody CH-11. The tumor cells were found to be resistant to CH-11-induced apoptosis. However, treatment of the tumor cells with IFN-γ, NO donors, or some chemotherapeutic drugs sensitized the tumor cells to CH-11-induced apoptosis. Sensitization was accompanied by increased expression of Fas, both at the cell surface and cytosolic. The mechanism of upregulation of Fas gene expression by these sensitizing agents was examined. We hypothesized that sensitization may be the result of de-repression of a transcription repressor. The examination of the Fas promoter revealed a silencer region with a putative binding site for YY1. Experiments were designed to test this hypothesis. We demonstrated that both IFN-γ and NO-releasing reagents treatment inhibited YY1 expression and DNA-binding activity. Further, deletion of the silencer region of the Fas promoter resulted in significant transcriptional activity assessed by a luciferase-based reporter system driven by different variations of the Fas promoter. These studies suggested strongly that YY1 negatively regulates Fas transcription and expression and, in addition, YY1 regulates tumor cell resistance to Fas-induced apoptosis (Garban and Bonavida, 2001). Furthermore, it was shown in prostate carcinoma cell lines that tumor-derived TNF-α regulates Fas resistance and that inhibition of TNF-α sensitized the cells to Fas-induced apoptosis (Huerta-Yepez et al., 2005). Recently, we have also shown that YY1 negatively regulates Fas expression in B-NHL cell lines and also regulates resistance to Fas-induced apoptosis (Vega et al., 2005a, 2005b). We further investigated the mechanism by which IFN-γ and NO inhibited YY1 expression and activity. We demonstrated that IFN-γ induces the expression iNOS and release of NO. NO inhibits both NF-κB and YY1 activity by s-nitrosylation (Hongo et al., 2004). Further, inhibition of NF-κB activity resulted in the concurrent inhibition of YY1, suggesting that YY1 transcription is under the regulation of NF-κB (Figure 11).

Figure 11
figure 11

Regulation by YY1 of tumor cell response to apoptotic stimuli. This figure schematically describes the role of YY1 in the regulation of both chemoresistance and immune resistance. Tumor cells constitutively express NF-κB and YY1 activities. It is known that NF-κB is a survival factor and is involved in the transcription regulation of several proapoptotic gene products. Likewise, YY1 has been shown to regulate negatively the expression of Fas and DR5 and hence, regulates resistance to both Fas-ligand and TRAIL-induced apoptosis. Inhibition of NF-κB and/or YY1, for example, by chemical inhibitors or by NO donors, results in the inhibition of NF-κB and YY1 and resulting in both chemosensitization and immunosensitization of drug/immune resistant tumors to apoptosis.

Full size image

Resistance to tumor necrosis factor-related apoptosis inducing ligand

Prostrate cancer cells were also found to be resistant to tumor necrosis factor-related apoptosis inducing ligand (TRAIL)-induced apoptosis. We examined the mechanism of resistance and demonstrated the treatment of tumor cells with NO or cytotoxic drugs (e.g., cis-diammine dichloroplatinum). CDDP sensitized the tumor cells to TRAIL-induced apoptosis (Huerta-Yepez et al., 2004). The mechanism of sensitization revealed that the tumor cells, following treatment with drug, overexpressed the TRAIL receptor DR5. The upregulation of DR5 was investigated and the findings revealed that the DR5 promoter contains one putative YY1-binding site. Inhibition of YY1 by NO or drugs resulted in inhibition of YY1 expression and activity. Further, using a DR5 reporter system, we demonstrated that deletion of the region containing the YY1-binding site and/or mutation of the YY1 site resulted in significant upregulation of luciferase activity over background level in a reporter system. This finding strongly suggested that YY1 negatively regulates DR5 transcription. The direct role of YY1 in DR5 expression was also corroborated by the use of siRNA YY1, which resulted in upregulation of DR5 expression and sensitization of the cells to TRAIL-induced apoptosis (Huerta-Yepez et al., 2005) (Figure 11).

Role of YY1 in chemoresistance

Prostate cancer cells are resistant to various chemotherapeutic drug-induced apoptosis. However, the cells can be sensitized following treatment with NO. The sensitization resulted in the downregulation of antiapoptotic gene product Bcl-2 and inhibition of both NF-κB and YY1 activity. The role of YY1 in chemosensitization was demonstrated by the use of siRNA YY1 whereby the transfectants were sensitive to CDDP-induced apoptosis in the absence of NO (Huerta-Yepez et al., 2005). These findings suggested that YY1 regulates drug resistance though the exact mechanism is not yet clear.

YY1 activity and metastatic potential

Beier and Gorogh (2005) provided evidence that downstream targets of YY1 activity are not limited to intracellular signaling. In a series of experiments, they demonstrate that YY1 and AP2 coactivation may potentiate their activity as galactocerebrosidase gene suppressors. Galactocerebrosidase is an enzyme overexpressed upon the cell surface of a variety of cancers. The accumulation of this protein promotes a reduction of cellular adhesion and inhibits apoptosis, leading to cellular proliferation, migration, and prolonged cell survival, all of which may contribute to carcinogenesis and metastasis.

In vivo and clinical evidence for the role of YY1 in the molecular regulation of tumorigenesis

Information gained from in vitro analyses has been applied to translational clinical models of carcinogenesis with tumor progression seen as the predominant effect of YY1 activity. Preclinical and clinical models that have been investigated include epithelial-based tumors (Sitwala et al., 2002; Seligson et al., 2005), hepatocellular carcinoma (Wang et al., 2001; Parija and Das, 2003), and breast cancer (Pilarsky et al., 2004).

Tumor suppression

YY1 and human papilloma virus. Studies in the elucidation of the regulation of human papilloma virus (HPV) 16 and 18 and the characterization of their roles in the progression of cervical carcinoma by both May et al. (1994) and Dong et al. (1994) implicate YY1-mediated repression when bound to viral oncogene promoters located in the long control region incorporated within the genome of cervical carcinoma cell lines. Likewise, Lichy et al. (1996) found a relatively marked increase in YY1 binding in nontumor lines of HeLa/fibroblast hybrids when compared to malignant cells. Taken together, these data suggest that YY1 acts to suppress tumor progression in HPV-infected cervical epithelial cells.

YY1 and basal cell carcinoma. Polymorphism studies of the gene encoding expression of glutathione S-transferase (GST) reveal that its variable expression may account for an individual's differential risk of developing malignancies such as colorectal, ovarian, and breast cancer as well as the potential aggressive nature of that tumor. Analysis of human basal cell carcinoma specimens reveals that absence of tumor progression of the GSTM3 genotype may be a result of YY1 repressive activity upon its recognition motif at the GST locus GSTM3*B (Yengi et al., 1996).

Tumor activation

Extensive evidence for the role of YY1 in the activation of malignant potential has surfaced. Clinical studies by Brankin et al. (1998) reveal that serum elevations of autoantibodies to nucleophosmin, an estrogen-regulated nucleolar phosphoprotein that suppresses YY1 transcriptional regulating activities, precede clinical evidence of recurrence in breast cancer patients. Interestingly, patients without recurrence demonstrated no change in serum levels. These data suggest that inhibition of nucleophosmin results in relief of repression of YY1 and is associated with clinical progression of breast carcinoma.

Prognostic/diagnostic significance

The above findings on YY1 expression in cancer cells suggested that YY1 overexpression may play an important role in the regulation of tumor cells' sensitivity and resistance to both chemotherapy and immunotherapy. Thus, tumor cells that overexpress YY1 may be selected during therapy and will exhibit drug/immune resistance and continued overexpression may correlate with disease progression and metastases. Initial studies were examined in prostate cancer using tissue microarrays, and YY1 expression and localization were examined by immunohistochemistry. Analysis of the data demonstrated that cancer cells show higher expression of YY1 than normal tissues. In addition, the data demonstrated that there was a subset of patients that can be identified whose YY1 expression predicted tumor recurrences. These studies suggested that YY1 expression may be considered a prognostic marker independent of circulating levels of prostate-specific antigen and other markers in prostate cancer (Seligson et al., 2005). Hence, it would be of interest to examine the prognostic and/or diagnostic significance of YY1 in other cancers.

Conclusion

YY1, a transcription factor that has been progressively characterized over the last 10 years, has become an intensive focus of study due to its ubiquitous nature, highly conserved molecular sequence, and increasingly apparent central role in embryologic development and differentiation as well as basic cellular functions such as replication, proliferation, senescence, and response to genotoxic stimuli.

Although its diverse functions allow for the context-specific paradoxical effects of transcriptional initiation, activation, and repression, the overwhelming evidence of the role of YY1 in tumor biology would support the theory that YY1 functions to promote carcinogenesis and perhaps even confer cells with a mechanism for evading cell death in the face of genotoxic stimuli including chemotherapy and/or immunotherapy (see Figures 10 and 11). Primary mechanisms appear to include perturbations in cellular surveillance systems as well as modulation of key genes involved in cell cycle regulation and programmed cell death. As enumerated herein, indirect and direct evidence exist to suggest a regulatory role for YY1 in the activation, progression, and/or maintenance of malignancy in multiple tumor models, both in vitro and in vivo. In addition, translational clincopathologic correlates have been documented in studies of human prostate, lymphoma, breast, and HCC tissues to corroborate these laboratory findings. As the clinical models noted herein are further elucidated and additional malignancies are studied, validation via transgenic mouse models with the use of xenografted tumors, inhibition via oligonucleotide testing in vitro, and in vivo oligonucleotide antisense testing, inhibitory RNA applications, and specific chemical inhibition will be necessary to clarify the role of YY1 in each instance and determine the possibility of utilization of the YY1 transcription factor as a target for antitumor therapy and reversal of drug-immune resistance.

Future directions

Diagnostic

As with other characterized protooncogenes, tumor suppressor genes, and tumor markers including p53 (Li-Fraumani Syndrome), RET oncogene (familial and sporadic medullary thyroid cancer), and Her-2/neu (breast cancer), YY1 may eventually serve as a diagnostic genetic marker for tumors proven to demonstrate a predictive pattern of expression in human tissue studies. Furthermore, predictive response to therapy may be extrapolated from these data and used to stratify patients.

Therapeutic

Lastly, as a potential modulator of cellular transformation and development of carcinogenesis, YY1 may serve as a target for cancer therapy with clinical application of therapeutics currently in use in other malignancies, including specific peptide or organic inhibitors, antisense therapies, and silencer RNA. These and other modalities may serve to enhance the spectrum of effective clinically available agents to choose from in the multimodal treatment of difficult tumors.