ReviewUnlocking the mechanisms of transcription factor YY1: are chromatin modifying enzymes the key?
Introduction
Transcriptional activation and repression of a given gene are critical for its proper regulation. Early models of eukaryotic transcriptional regulation often portray a trans-activator or trans-repressor interacting with the general transcription factors or with each other to regulate their respective target genes. Over the years, however, we have learned that the mechanisms of activation and repression are far more complicated. For example, we have begun to appreciate that many accessory factors, known as coactivators and corepressors, are required for most activators and repressors to function properly. Another layer of complexity was recently added to the understanding of transcriptional regulation when it was discovered that some of these cofactors contain catalytic activities that can modify histones. Specifically, some coactivators have been found to contain histone acetyltransferase (HAT) activities while some corepressors contain histone deacetylase (HDAC) activities. This intriguing discovery is important because the structure of chromatin can be modulated by the post-translational modification of histones (Grunstein, 1997, Roth and Allis, 1996, Wade et al., 1997). The acetylation state of histone tails affects their affinity with DNA in vitro (Hong et al., 1993). In addition, acetylated histones destabilize higher-order folding of chromatin (Hansen et al., 1998). Hyperacetylation of histones is believed to cause the chromatin to be more accessible for interaction with DNA binding proteins (Bauer et al., 1994, Garcia-Ramirez et al., 1995, Norton et al., 1989). Consistent with this view is the observation that histones localized to transcriptionally active chromatin typically have a higher level of acetylation compared with histones localized to inactive chromatin (Struhl, 1998, Wade et al., 1997).
The focus of this review is the transcription factor Yin Yang 1 [YY1 (also known as δ, NF-E1, UCRBP, and CF1)] that may use HAT and HDAC cofactors to exert its transcriptional functions. YY1 is a 65 kDa member of the GLI-krüppel family of zinc finger transcription factors. The expression of YY1 is ubiquitous and the protein is highly conserved among human, mouse and Xenopus (Shi et al., 1997). In Drosophila, a sequence-specific DNA binding protein, PHO (a member of the polycomb group proteins), contains a remarkable 112 out of 118 amino acid identity with YY1 over the region encoding the zinc fingers (Brown et al., 1998). However, no other region of similarity exists between PHO and YY1 outside the zinc fingers. A recent survey on the number of promoters that contain potential YY1 binding sites is overwhelming, and reports on the number of promoters that can be regulated by YY1 are equally stunning. Some examples include, but are in no way limited to, c-Myc, c-Fos, p53, α-actin, surf-2, grp78, IgH enhancer, β-casein, Igκ 3′ enhancer, ρ-globin, ε-globin, IFN-γ, rpL30, rpL32, P5 of AAV, E6 and E7 of HPV, BZLF1 of EBV, P6 of B19-parvovirus, VP5 of HSV-1, and a number of other viral LTRs (Hyde-DeRuyscher et al., 1995, Shi et al., 1997, Shrivastava and Calame, 1994, Yant et al., 1995). More recently, studies suggest that YY1 is critical in the regulation of histone genes (Eliassen et al., 1998, Last et al., 1999). In addition, YY1 recognition sites have been identified as the initiator element of some promoters (Basu et al., 1993, Seto et al., 1991, Usheva and Shenk, 1994). Originally, YY1 was isolated as a repressor of the P5 promoter of AAV, as well as the MuLV LTR, and the immunoglobulin κ 3′ enhancer (Flanagan et al., 1992, Park and Atchison, 1991, Shi et al., 1991). It was also found that the adenovirus E1A protein could alter YY1's function from a repressor to a trans-activator (Shi et al., 1991). In contrast, work by other groups, around the same time, found that YY1 acted as an activator of such genes as c-Myc and rpL30 (Hariharan et al., 1991, Riggs et al., 1993). This YY1 mediated activation occurred in the absence of E1A, demonstrating that YY1 is able to activate certain promoters in the absence of viral factors. A possible explanation for the dual nature of YY1 is that the promoter context and factors already present at a promoter dictate which function of YY1 is displayed at that promoter. Alternatively, it is conceivable that the dual nature depends on the different pre-existing YY1–cofactor complexes being recruited to a promoter under different conditions.
The corepressors most relevant to understanding the mechanism of YY1 mediated repression are the mammalian members of the RPD3 family of HDACs. Three such proteins have been cloned to date and share extensive homology to the yeast global regulator RPD3 (Hassig and Schreiber, 1997, Pazin and Kadonaga, 1997, Roth and Allis, 1996). All three are known to interact with YY1 in vitro, and at least HDAC2 potentially interacts with YY1 in vivo (Yang et al., 1996, Yang et al., 1997). Two additional classes of deacetylases have recently been cloned in maize (HD2) and mouse (mHDA1 and mHDA2) that show little or no homology to the RPD3 related deacetylases (Lusser et al., 1997, Verdel and Khochbin, 1999). The in vivo function of these new deacetylases and whether they interact with YY1 is unknown at this time.
HDAC1 was first purified through the use of a deacetylase inhibitor and subsequently cloned by reverse genetics, while HDAC2 was cloned in a two-hybrid screen with YY1 as the bait (Taunton et al., 1996, Yang et al., 1996). A complete HDAC3 cDNA clone was isolated using a probe derived from an EST sequence that shares significant homology to HDAC1 and HDAC2 (Emiliani et al., 1998, Yang et al., 1997). The HDAC3 cDNA was also independently identified from PHA-activated immune cells (Dangond et al., 1998). The HDACs exist as components of multi-subunit complexes with HDAC1 and HDAC2 often isolated in the same complex, while HDAC3 exists in a distinct complex (Hassig et al., 1998, Zhang et al., 1997). The identification and cloning of members of these complexes are active areas of investigation. In the case of the HDAC1/2 complexes, a few members have been identified and include mSin3A, N-CoR, SAP18/30, RbAp46/48, CHD3/CHD4, and MeCP2 (Heinzel et al., 1997, Jones et al., 1998, Laherty et al., 1997, Laherty et al., 1998, Nan et al., 1998, Tong et al., 1998, Xue et al., 1998, Zhang et al., 1997, Zhang et al., 1998a, Zhang et al., 1998b). mSin3A and N-CoR interact with Mad and nuclear hormone receptors, respectively, which may help direct HDAC1/2 to a given promoter. In contrast, CHD3 and CHD4 are two proteins that are autoimmune antigens associated with the connective tissue disease Mi-2 dermatomyositis and are thought to contain chromatin remodeling activities. RbAp48 is a subunit of the chromatin assembly factor 1 (Tyler et al., 1996), and MeCP2 is a protein that selectively binds to methylated DNA sequences (Lewis et al., 1992).
Since YY1 can also act as an activator, it is conceivable that it may require a coactivator to function. In this regard, CBP and p300, two highly homologous proteins that serve as coactivators for many transcriptional activators (Boyes et al., 1998, Goldman et al., 1997, Kwok et al., 1994, Sartorelli et al., 1997, Yuan et al., 1996), have been shown to interact with YY1 (Austen et al., 1997b, Galvin and Shi, 1997, Lee et al., 1995a). CBP is required for activation of CREB dependent promoters, while p300 is involved in the proper function of the adenovirus E1A protein. Recently, it was shown that these two proteins are interchangeable for interactions with CREB and are thought to be functionally homologous (Lundblad et al., 1995). Interestingly, both of these proteins have been demonstrated to contain HAT activity (Ogryzko et al., 1996). This result suggests that modification of histones and chromatin may be an important mechanism of activation for factors utilizing CBP/p300 as coactivators.
With recent advances in understanding the characteristics of coactivators and corepressors, the opportunity to understand how YY1 might function through recruitment of cofactors is becoming increasingly apparent. In this review, we will present an overview of the data, much of it complex and often contradictory, that has been collected over the last eight years on YY1. We will then explore in a critical manner the recent findings concerning cofactors that interact with YY1. Finally, we will present models and potential explanations of how YY1 might activate and repress transcription. As with most reviews of this kind, due to space limitations it is impractical and quite impossible to cite all of the works pertaining to this field. For further details on YY1, therefore, the reader should consult an excellent comprehensive review by Shi et al. (1997).
Section snippets
Functional domains of YY1
As may be expected from the multi-functional nature of YY1, structural analyses of the activation and repression domains of YY1 proved to be quite complicated. A number of groups have analyzed YY1's structure/function relationships through the expression of YY1 deletion mutants in cotransfection assays with reporter constructs. Many of these studies used the DNA binding domain of Gal4 fused to YY1 to eliminate the need for an intact YY1 DNA binding domain. Results from these studies are
Protein/protein interaction domains of YY1
Unfortunately, the complexity of YY1 does not end with its functional domains. Attempts to identify YY1 domains that are responsible for protein/protein interactions have resulted in equally complicated findings. The constellation of proteins that interact with YY1 is staggering in number, and interestingly, most of the proteins that interact with YY1 are either coactivators/corepressors or transcription factors that are important in regulating a number of diverse promoters. Also, a recent
Models of YY1 mediated repression
Several proposed models of YY1 mediated repression are diagrammed in Fig. 2 [also discussed by Shi et al. (1997)]. Earlier, it was thought that YY1 acts by sterically hindering the binding of trans-activators to DNA through overlapping DNA recognition sites. A prime example of this is the α-actin promoter where the YY1 site occludes the serum response factor binding element (SRE) (Lee et al., 1992). However, numerous examples now exist where a YY1 binding site is nowhere near another trans
Models of YY1 mediated activation
Early observations that indicated YY1 could function as a transcriptional activator came from three separate studies. First, it was shown that the adenovirus E1A protein could relieve repression exerted by YY1 and further activate transcription through a YY1-binding site (Shi et al., 1991). It was thought that E1A could somehow convert YY1 from a repressor (or non-activator) to an activator. Second, soon after the cloning of YY1, it was found that YY1 could activate transcription when bound to
Conclusion
YY1 is a fascinating example of the complexity inherent in gene regulatory systems. Much of the recent work on YY1 has gone into understanding its domain structure, identifying the factors with which it interacts, and attempting to understand its mechanisms of action. As with studies on almost any biological molecule, analysis of YY1 raises many more questions than it answers. Of immediate interest is the question of whether recruitment of chromatin remodeling proteins in general, and histone
Acknowledgements
We thank Ya-Li Yao, Jennifer Westling, Wen-Ming Yang, Yang Shi, Michael Atchison, and Rosalind Jackson for helpful suggestions on the manuscript. Work related to YY1 and histone deacetylation in our laboratory is supported by grants from the National Institutes of Health and the National Science Foundation.
References (90)
- et al.
Characterization of the transcriptional regulator YY1. The bipartite transactivation domain is independent of interaction with the TATA box-binding protein, transcription factor IIB, TAFII55, or cAMP-responsive element-binding protein (CBP)-binding protein
J. Biol. Chem.
(1997) - et al.
Identification of a transcriptional initiator element in the cytochrome c oxidase subunit Vb promoter which binds to transcription factors NF-E1 (YY-1, delta) and Sp1
J. Biol. Chem.
(1993) - et al.
Nucleosome structural changes due to acetylation
J. Mol. Biol.
(1994) - et al.
The Drosophila Polycomb group gene pleiohomeotic encodes a DNA binding protein with homology to the transcription factor YY1
Mol. Cell
(1998) - et al.
Characterization of functional domains within the multifunctional transcription factor, YY1
J. Biol. Chem.
(1995) - et al.
Differential display cloning of a novel human histone deacetylase (HDAC3) cDNA from PHA-activated immune cells
Biochem. Biophys. Res. Commun.
(1998) - et al.
Modulation of chromatin folding by histone acetylation
J. Biol. Chem.
(1995) Yeast heterochromatin: regulation of its assembly and inheritance by histones
Cell
(1998)- et al.
Activation of p53 sequence-specific DNA binding by acetylation of the p53 C-terminal domain
Cell
(1997) - et al.
Nuclear histone acetylases and deacetylases and transcriptional regulation: HATs off to HDACs
Curr. Opin. Chem. Biol.
(1997)