- Email: [email protected]
pbX Marks the Spot
Developmental Cell, Vol. 6, 737–748, June, 2004, Copyright 2004 by Cell Press
Previews
pbX Marks the Spot Pbx and Meis proteins act as cofactors to various transcription factors, but their exact functions have been unclear. A report by Berkes et al. in Molecular Cell now demonstrates that Pbx and Meis may penetrate repressive chromatin to mark specific genes for activation. Members of the pbx and meis gene families were initially identified as genes involved in embryogenesis and as protooncogenes in various forms of leukemia. Subsequent biochemical and genetic studies revealed that Meis and Pbx act as cofactors to Hox proteins. Hox proteins are homeodomain transcription factors that regulate cell fate decisions in many tissues during early embryonic development and organogenesis. Pbx and Meis form complexes with Hox proteins and enhance Hox function by improving the affinity and specificity of Hox proteins for specific promoter elements (reviewed in Mann and Affolter, 1998). However, Meis and Pbx likely contribute additional, as yet unknown, activities. For instance, Meis proteins contain conserved domains in addition to those required for DNA binding and for forming Hox complexes (e.g., Choe et al., 2002). Pbx and Meis proteins also form complexes with other homeodomain transcription factors (e.g., Engrailed and Pdx), as well as with bHLH proteins (e.g., MyoD), suggesting that they serve a broader role as transcription cofactors. A report in the May 21 issue of Molecular Cell (Berkes et al., 2004) now demonstrates an essential role for Pbx and Meis as cofactors to MyoD. Like Hox proteins, MyoD requires Pbx and Meis for optimal binding to DNA. However, their results also demonstrate a novel activity of Pbx and Meis. In particular, Pbx and Meis appear to be constitutively bound to MyoD-dependent promoters (such as myogenin) prior to initiation of muscle differentiation and in the absence of MyoD. Since MyoD binding is essential for chromatin remodeling and transcription of myogenin (Gerber et al., 1997), this suggests that Pbx and Meis may act as “pioneer” transcription factors that penetrate repressive chromatin and mark specific genes for activation by MyoD. Might Pbx and Meis serve a similar role at Hox-dependent promoters? This appears likely since the ability of Pbx and Meis to access promoter elements in silent chromatin might be of particular importance for “master regulators,” such as MyoD and Hox proteins, that initiate novel differentiation programs and hence must activate transcription of genes that were previously silent. Notably, even if Pbx and Meis penetrate compacted chromatin and mark genes for activation by Hox proteins, this function may not be required for all hox-regulated genes. For instance, hox genes activate each other’s expression, but hox genes are located in genomic clusters and the chromatin state of these clusters seems to be
regulated on a greater scale, possibly obviating the need for Pbx and Meis in penetrating condensed chromatin. Pbx and Meis are still required for expression of genes in these clusters, however, likely due to their role in enhancing binding of Hox proteins to DNA. The mechanism whereby Pbx and Meis proteins penetrate repressive chromatin is not known, but based on other systems, this likely occurs either because their binding sites are in a nucleosome-free region (such as in the IFN- promoter [Agalioti et al., 2000]) or because Pbx and/or Meis have intrinsic activities enabling them to directly interact with and displace histones (similar to HNF3 [Cirillo et al., 2002]). It is also not clear if Pbx and Meis must act together to access silenced promoters, but since these proteins bind DNA more effectively as a heterodimer than as monomers, they likely act as a heterodimeric complex. One concern with a general role for Pbx and Meis in marking genes for activation by various transcription factors is that several pbx genes, as well as at least two meis family members (prep1 and prep2), are expressed ubiquitously in the embryo. Hence, Pbx and Prep proteins would be expected to mark Hox- and MyoDdependent promoters (and very likely additional ones) for activation in most cells of the embryo, seemingly increasing the risk of inappropriately initiating transcription. There are several possible explanations for how this lack of specificity might be counteracted. First, all Pbx and Meis sites may in fact not be occupied. For instance, nuclear localization of Meis and Pbx proteins is regulated (Mann and Affolter, 1998), suggesting that “marking” can be prevented, or perhaps even erased, by restricting access of Meis or Pbx to the nucleus. Second, although the prep genes are ubiquitously expressed, other meis genes show restricted expression patterns and Meis and Prep proteins may differ functionally (Thorsteinsdottir et al., 2001). Thus, it is possible that replacing one Meis protein with another dictates whether a Pbx/Meis complex can penetrate repressed chromatin or not. Third, there may be additional safeguards that prevent inadvertent gene activation even when a Pbx/Meis complex is bound. For instance, Pbx proteins bind several corepressors (Asahara et al., 1999; Saleh et al., 2000). Thus, Pbx and Meis may penetrate inactive chromatin, but once bound to DNA, these proteins may exist primarily in complexes with corepressors, which then ensures that the marked promoter remains inactive. In this scenario (Figure 1), transcription is not activated until a sequence-specific transcription factor is recruited to the marked gene where it must bind both Pbx/Meis and an adjacent promoter element. This ensures that genes are only transcribed if they are both marked for activation and contain the correct DNA motif. Notably, Hox proteins bind coactivators (Saleh et al., 2000), as does MyoD (Sartorelli et al., 1997), suggesting that sequence-specific transcription factors may recruit coactivators to relieve any repressor activity associated with the Pbx/Meis “mark.” This recruitment is likely via direct binding to coactivators, but it is not clear how other chromatin remodeling factors might be
Developmental Cell 738
and hence may mark genes for activation by Pdx during pancreas development. Further, if Pbx and Meis proteins act as markers that enable transcriptional activation by sequence-specific factors, this may have relevance to the role of Pbx and Meis in cancer. For instance, fusion of Pbx1 to the activation domain of E2A, as a result of chromosomal translocations, leads to preB-cell leukemia. Since E2A binds histone acetyl transferases (HATs), the E2APbx1 fusion protein might be particularly detrimental because it can penetrate silent chromatin, recruit HATs, and activate transcription without a requirement for other sequence-specific transcription factors. Charles G. Sagerstro¨m Department of Biochemistry and Molecular Pharmacology University of Massachusetts Medical School LRB822, 364 Plantation Street Worcester, Massachusetts 01605 Selected Reading
Figure 1. Model Outlining a Hypothetical Scenario for Regulation of a Hox-Dependent Gene Pbx/Meis complexes penetrate inactive chromatin (illustrated as DNA [diagonal black lines] wound around nucleosomes [light blue rectangles]) and make it accessible to Hox proteins. The promoter remains inactive, perhaps as a result of corepressors recruited by Pbx/Meis, until the Hox protein binds. The Hox protein recruits coactivators (e.g., histone acetyl transferases, HATs) as well as ATPdependent chromatin remodeling machines to activate transcription.
recruited. For instance, there are no known direct interactions between Meis, Pbx, or Hox and components of the ATP-dependent chromatin remodeling complexes such as SWI/SNF. Importantly, Pbx and Meis bind proteins in addition to Hox and MyoD. For instance, Pbx and Meis bind Pdx
Colinearity Loops Out
Modulation of chromatin structure has long been proposed to underlie the colinear regulation of Hox genes during animal development. In a recent paper, Chambeyron and Bickmore explore this possibility in retinoic acid-induced ES cells. They show that, while chromatin remodeling confers transcriptional competence to the gene cluster, subsequent sequential extrusion of genes from their chromosome territory may determine their coordinated expression in time.
Agalioti, T., Lomvardas, S., Parekh, B., Yie, J., Maniatis, T., and Thanos, D. (2000). Cell 103, 667–678. Asahara, H., Dutta, S., Kao, H.Y., Evans, R.M., and Montminy, M. (1999). Mol. Cell. Biol. 19, 8219–8225. Berkes, C.A., Bergstrom, D.A., Penn, B.H., Seaver, K.J., Knoepfler, P.S., and Tapscott, S.J. (2004). Mol. Cell 14, 465–477. Choe, S.-K., Vlachakis, N., and Sagerstro¨m, C.G. (2002). Development 129, 585–595. Cirillo, L.A., Lin, F.R., Cuesta, I., Friedman, D., Jarnik, M., and Zaret, K.S. (2002). Mol. Cell 9, 279–289. Gerber, A.N., Klesert, T.R., Bergstrom, D.A., and Tapscott, S.J. (1997). Genes Dev. 11, 436–450. Mann, R.S., and Affolter, M. (1998). Curr. Opin. Genet. Dev. 8, 423–429. Saleh, M., Rambaldi, I., Yang, X.J., and Featherstone, M.S. (2000). Mol. Cell. Biol. 20, 8623–8633. Sartorelli, V., Huang, J., Hamamori, Y., and Kedes, L. (1997). Mol. Cell. Biol. 17, 1010–1026. Thorsteinsdottir, U., Kroon, E., Jerome, L., Blasi, F., and Sauvageau, G. (2001). Mol. Cell. Biol. 21, 224–234.
Animals displaying a bilateral symmetry need a genetic system to specify structures along their various body axes. For example, rib-carrying vertebrae are necessary at the thoracic level, but would not be appropriate in the neck. Likewise, digits should be positioned at the distal end of the limbs, rather than at their beginning. This important task is regulated by the members of the Hox gene family. In mammals, these genes are grouped into four genomic clusters and encode proteins whose various combinations will instruct different body levels as to the type of structures to be generated (e.g., Krumlauf, 1994). Such a combinatorial system calls for precise protein distribution, hence the underlying transcriptional control must be well orchestrated. An important aspect of this control relies upon an enigmatic property, referred to as “colinearity” (Lewis, 1978), whereby
Previews
pbX Marks the Spot Pbx and Meis proteins act as cofactors to various transcription factors, but their exact functions have been unclear. A report by Berkes et al. in Molecular Cell now demonstrates that Pbx and Meis may penetrate repressive chromatin to mark specific genes for activation. Members of the pbx and meis gene families were initially identified as genes involved in embryogenesis and as protooncogenes in various forms of leukemia. Subsequent biochemical and genetic studies revealed that Meis and Pbx act as cofactors to Hox proteins. Hox proteins are homeodomain transcription factors that regulate cell fate decisions in many tissues during early embryonic development and organogenesis. Pbx and Meis form complexes with Hox proteins and enhance Hox function by improving the affinity and specificity of Hox proteins for specific promoter elements (reviewed in Mann and Affolter, 1998). However, Meis and Pbx likely contribute additional, as yet unknown, activities. For instance, Meis proteins contain conserved domains in addition to those required for DNA binding and for forming Hox complexes (e.g., Choe et al., 2002). Pbx and Meis proteins also form complexes with other homeodomain transcription factors (e.g., Engrailed and Pdx), as well as with bHLH proteins (e.g., MyoD), suggesting that they serve a broader role as transcription cofactors. A report in the May 21 issue of Molecular Cell (Berkes et al., 2004) now demonstrates an essential role for Pbx and Meis as cofactors to MyoD. Like Hox proteins, MyoD requires Pbx and Meis for optimal binding to DNA. However, their results also demonstrate a novel activity of Pbx and Meis. In particular, Pbx and Meis appear to be constitutively bound to MyoD-dependent promoters (such as myogenin) prior to initiation of muscle differentiation and in the absence of MyoD. Since MyoD binding is essential for chromatin remodeling and transcription of myogenin (Gerber et al., 1997), this suggests that Pbx and Meis may act as “pioneer” transcription factors that penetrate repressive chromatin and mark specific genes for activation by MyoD. Might Pbx and Meis serve a similar role at Hox-dependent promoters? This appears likely since the ability of Pbx and Meis to access promoter elements in silent chromatin might be of particular importance for “master regulators,” such as MyoD and Hox proteins, that initiate novel differentiation programs and hence must activate transcription of genes that were previously silent. Notably, even if Pbx and Meis penetrate compacted chromatin and mark genes for activation by Hox proteins, this function may not be required for all hox-regulated genes. For instance, hox genes activate each other’s expression, but hox genes are located in genomic clusters and the chromatin state of these clusters seems to be
regulated on a greater scale, possibly obviating the need for Pbx and Meis in penetrating condensed chromatin. Pbx and Meis are still required for expression of genes in these clusters, however, likely due to their role in enhancing binding of Hox proteins to DNA. The mechanism whereby Pbx and Meis proteins penetrate repressive chromatin is not known, but based on other systems, this likely occurs either because their binding sites are in a nucleosome-free region (such as in the IFN- promoter [Agalioti et al., 2000]) or because Pbx and/or Meis have intrinsic activities enabling them to directly interact with and displace histones (similar to HNF3 [Cirillo et al., 2002]). It is also not clear if Pbx and Meis must act together to access silenced promoters, but since these proteins bind DNA more effectively as a heterodimer than as monomers, they likely act as a heterodimeric complex. One concern with a general role for Pbx and Meis in marking genes for activation by various transcription factors is that several pbx genes, as well as at least two meis family members (prep1 and prep2), are expressed ubiquitously in the embryo. Hence, Pbx and Prep proteins would be expected to mark Hox- and MyoDdependent promoters (and very likely additional ones) for activation in most cells of the embryo, seemingly increasing the risk of inappropriately initiating transcription. There are several possible explanations for how this lack of specificity might be counteracted. First, all Pbx and Meis sites may in fact not be occupied. For instance, nuclear localization of Meis and Pbx proteins is regulated (Mann and Affolter, 1998), suggesting that “marking” can be prevented, or perhaps even erased, by restricting access of Meis or Pbx to the nucleus. Second, although the prep genes are ubiquitously expressed, other meis genes show restricted expression patterns and Meis and Prep proteins may differ functionally (Thorsteinsdottir et al., 2001). Thus, it is possible that replacing one Meis protein with another dictates whether a Pbx/Meis complex can penetrate repressed chromatin or not. Third, there may be additional safeguards that prevent inadvertent gene activation even when a Pbx/Meis complex is bound. For instance, Pbx proteins bind several corepressors (Asahara et al., 1999; Saleh et al., 2000). Thus, Pbx and Meis may penetrate inactive chromatin, but once bound to DNA, these proteins may exist primarily in complexes with corepressors, which then ensures that the marked promoter remains inactive. In this scenario (Figure 1), transcription is not activated until a sequence-specific transcription factor is recruited to the marked gene where it must bind both Pbx/Meis and an adjacent promoter element. This ensures that genes are only transcribed if they are both marked for activation and contain the correct DNA motif. Notably, Hox proteins bind coactivators (Saleh et al., 2000), as does MyoD (Sartorelli et al., 1997), suggesting that sequence-specific transcription factors may recruit coactivators to relieve any repressor activity associated with the Pbx/Meis “mark.” This recruitment is likely via direct binding to coactivators, but it is not clear how other chromatin remodeling factors might be
Developmental Cell 738
and hence may mark genes for activation by Pdx during pancreas development. Further, if Pbx and Meis proteins act as markers that enable transcriptional activation by sequence-specific factors, this may have relevance to the role of Pbx and Meis in cancer. For instance, fusion of Pbx1 to the activation domain of E2A, as a result of chromosomal translocations, leads to preB-cell leukemia. Since E2A binds histone acetyl transferases (HATs), the E2APbx1 fusion protein might be particularly detrimental because it can penetrate silent chromatin, recruit HATs, and activate transcription without a requirement for other sequence-specific transcription factors. Charles G. Sagerstro¨m Department of Biochemistry and Molecular Pharmacology University of Massachusetts Medical School LRB822, 364 Plantation Street Worcester, Massachusetts 01605 Selected Reading
Figure 1. Model Outlining a Hypothetical Scenario for Regulation of a Hox-Dependent Gene Pbx/Meis complexes penetrate inactive chromatin (illustrated as DNA [diagonal black lines] wound around nucleosomes [light blue rectangles]) and make it accessible to Hox proteins. The promoter remains inactive, perhaps as a result of corepressors recruited by Pbx/Meis, until the Hox protein binds. The Hox protein recruits coactivators (e.g., histone acetyl transferases, HATs) as well as ATPdependent chromatin remodeling machines to activate transcription.
recruited. For instance, there are no known direct interactions between Meis, Pbx, or Hox and components of the ATP-dependent chromatin remodeling complexes such as SWI/SNF. Importantly, Pbx and Meis bind proteins in addition to Hox and MyoD. For instance, Pbx and Meis bind Pdx
Colinearity Loops Out
Modulation of chromatin structure has long been proposed to underlie the colinear regulation of Hox genes during animal development. In a recent paper, Chambeyron and Bickmore explore this possibility in retinoic acid-induced ES cells. They show that, while chromatin remodeling confers transcriptional competence to the gene cluster, subsequent sequential extrusion of genes from their chromosome territory may determine their coordinated expression in time.
Agalioti, T., Lomvardas, S., Parekh, B., Yie, J., Maniatis, T., and Thanos, D. (2000). Cell 103, 667–678. Asahara, H., Dutta, S., Kao, H.Y., Evans, R.M., and Montminy, M. (1999). Mol. Cell. Biol. 19, 8219–8225. Berkes, C.A., Bergstrom, D.A., Penn, B.H., Seaver, K.J., Knoepfler, P.S., and Tapscott, S.J. (2004). Mol. Cell 14, 465–477. Choe, S.-K., Vlachakis, N., and Sagerstro¨m, C.G. (2002). Development 129, 585–595. Cirillo, L.A., Lin, F.R., Cuesta, I., Friedman, D., Jarnik, M., and Zaret, K.S. (2002). Mol. Cell 9, 279–289. Gerber, A.N., Klesert, T.R., Bergstrom, D.A., and Tapscott, S.J. (1997). Genes Dev. 11, 436–450. Mann, R.S., and Affolter, M. (1998). Curr. Opin. Genet. Dev. 8, 423–429. Saleh, M., Rambaldi, I., Yang, X.J., and Featherstone, M.S. (2000). Mol. Cell. Biol. 20, 8623–8633. Sartorelli, V., Huang, J., Hamamori, Y., and Kedes, L. (1997). Mol. Cell. Biol. 17, 1010–1026. Thorsteinsdottir, U., Kroon, E., Jerome, L., Blasi, F., and Sauvageau, G. (2001). Mol. Cell. Biol. 21, 224–234.
Animals displaying a bilateral symmetry need a genetic system to specify structures along their various body axes. For example, rib-carrying vertebrae are necessary at the thoracic level, but would not be appropriate in the neck. Likewise, digits should be positioned at the distal end of the limbs, rather than at their beginning. This important task is regulated by the members of the Hox gene family. In mammals, these genes are grouped into four genomic clusters and encode proteins whose various combinations will instruct different body levels as to the type of structures to be generated (e.g., Krumlauf, 1994). Such a combinatorial system calls for precise protein distribution, hence the underlying transcriptional control must be well orchestrated. An important aspect of this control relies upon an enigmatic property, referred to as “colinearity” (Lewis, 1978), whereby