The Temporal and Spatial Subnuclear Organization of Skeletal Gene Regulatory Machinery: Integrating Multiple Levels of Transcriptional Control
Bone formation during development and skeletal remodeling throughout adult life necessitates a complex temporal and interdependent expression of cell growth and phenotypic genes. There is a requirement for responsiveness to a broad spectrum of regulatory cues that transduce physiological signals from the extracellular matrix to sites within the nucleus for control of gene expression. As our understanding of gene regulatory mechanisms expands, it becomes increasingly evident that unique parameters of transcriptional control support the transient activation and suppression of genes for skeletal development and bone homeostasis. Other mechanisms are invoked for long-term commitments to gene expression that sustain the specialized structural and functional properties of bone cells. As the repertoire of factors that influence gene expression for commitment to the osteoblast lineage expands, there is a need to understand how multiple levels of transcriptional control are integrated. We present a perspective on the molecular mechanisms that are necessary for precise skeletal cell type and gene-specific physiological regulation.
The Requirements for Physiological Control of Skeletal Gene Expression In Vivo
Selective utilization and convergence of multiple regulatory signals at promoter sequences
Mechanisms that render the promoters of cell growth and bone-specific genes competent for protein-DNA and protein-protein interactions in a physiologically responsive manner
The composition, organization, and assembly of regulatory complexes at sites within the nucleus that support transcription.
These parameters become functional within the three-dimensional context of nuclear architecture.
Key components of the basal transcription machinery and several tissue-specific transcription factor complexes are functionally compartmentalized as specialized subnuclear domains [6, 7, 8]. Such compartmentalization may, at least in part, accommodate biological constraints on the control of transcription in nuclei of intact bone cells. The low representation of promoter regulatory elements and cognate transcription factors also necessitates a subnuclear organization of nucleic acids and regulatory proteins that supports threshold concentrations for the activation and repression of gene expression. Lastly, the need for establishing boundaries within gene promoters for transcriptional control and instructions for directing signaling proteins to regulatory complexes requires gene expression within the structural organization of the nucleus.
The bone-specific osteocalcin gene and skeletal-restricted Runx2 (AML3/Cbfa1) transcription factor serve as paradigms for obligatory relationships linking nuclear structure with physiological control of skeletal gene expression. The modularly organized promoter of the bone-specific osteocalcin gene contains proximal and distal regulatory elements that support basal, tissue-specific as well as growth factor, homeodomain, proteins and steroid hormone mediated transcriptional control . Modulation of osteocalcin gene expression during bone formation and remodeling requires physiologically responsive accessibility of these proximal and upstream promoter sequences to regulatory proteins and for protein-protein interactions that integrate independent promoter domains. The Runx family of transcription factors contribute to control of gene expression by sequence-specific binding to promoter elements of target genes and by serving as scaffolds in the nuclear matrix for the architectural control of promoter activity and assembly and organization of coregulatory protein complexes [10, 11, 12, 13, 14, 15].
Nuclear Microenvironments: Accommodating the Rules that Govern In Vivo Transcriptional Control
There is increasing acceptance that components of nuclear architecture facilitate the organization and sorting of regulatory information in a manner that permits selective utilization of factors on gene promoters. The components of higher order nuclear architecture, which includes nuclear pores, the nuclear matrix (also called the nuclear scaffold), and subnuclear focal domains contribute to the organization and activities of nucleic acids and regulatory proteins.
Compartmentalization of these complexes is strikingly illustrated by focal organization of functional activities, including the ribosomal genes in the nucleolus, chromosomes and their reorganization during mitosis, PML bodies (in hematopoietic and pomyelocytic leukemic cells), AML bodies (in myeloid leukemias), as well as by the punctate intranuclear distribution of sites for replication, DNA repair, transcription (RNA polymerase II domains), and the processing of gene transcripts (SC35 domains) [7 and references therein]. Cellular, molecular, biochemical and genetic evidence indicates an obligatory relationship between sites within the nucleus where nucleic acids and regulatory complexes reside and fidelity of transcriptional control [6, 10, 16, 17]. In a broader context, the involvement of nuclear architecture facilitates a dynamic and bidirectional exchange of gene transcripts and regulatory factors between the nucleus and cytoplasm, as well as between regions and structures within the nucleus [8, 18]. The formation of transcriptionally active Runx foci associated with the nuclear matrix exemplifies a requirement for structural organization of regulatory complexes involved in tissue-specific gene activation.
The nuclear matrix is functionally involved in gene localization and in the concentration and subnuclear localization of regulatory factors [14, 19, 20, 21, 22, 23]. Genes and cognate factors associated with the nuclear matrix supports the formation and/or activities of nuclear domains that facilitate transcriptional control [24, 25, 26, 27, 28, 29]. However, there is a need to gain insight into mechanisms that vectorally direct skeletal regulatory factors to subnuclear sites where regulatory events occur. Both biochemical and immunofluorescence analyses have shown that Runx-transcription factors exhibit a punctate nuclear distribution with the nuclear matrix in situ [13, 14, 30]. The association of osteogenic and hematopoietic Runx proteins with the nuclear-matrix requires a nuclear matrix targeting signal, a 31 amino acid segment near the C-terminus that is distinct from nuclear localization signals and is independent of DNA binding and functions autonomously . Thus, at least two trafficking signals appear to be required for subnuclear targeting of Runx transcription factors; the first supports nuclear import (nuclear localization signal) and a second mediates association with the nuclear matrix (nuclear matrix targeting signal). The unique sequences and crystal structure of the nuclear matrix targeting signal of Runx transcription factors [13, 30, 31] supports specificity for localization at intranuclear sites where the regulatory machinery for gene expression is assembled, rendered operative, and/or suppressed. Insight is thereby provided into mechanisms linked to the assembly and activities of subnuclear domains where transcription occurs.
The intranuclear trafficking of skeletal regulatory factors to the nuclear matrix is important for their function in gene transcription [12, 26, 30]. However, components of the nuclear matrix that serve as the acceptor sites remains to be established. Characterization of such nuclear matrix components will provide an additional dimension to characterizing molecular mechanisms associated with gene expression, namely, the targeting of regulatory proteins to specific spatial domains within the nucleus.
Chromatin Remodeling Facilitates Promoter Accessibility and Integration of Regulatory Activities
It is well recognized that genomic DNA is packaged as chromatin [32, 33]. These “beads on a string” structures, designated nucleosomes, are structurally remodeled to accommodate requirements for transcription, emphasizing the extent to which architectural organization of genes is obligatory for functional activity. Proteins that catalyze histone acetylation, deacetylation, and phosphorylation, as well as the SWI/SNF-related proteins, mediate chromatin remodeling and potentially the accessibility of promoter sequences to regulatory and coregulatory factors . Multimeric transcription factor complexes often include co-regulatory factors with histone acetylase or deacetylase enzyme activity to enhance or dampen requirements for physiological expression. Chromatin structure and nucleosome organization of the DNA contribute to the overall three-dimensional architecture of promoters, with open loops of chromatin that render elements competent for interactions with positive and negative regulatory factors as well as reducing distances between regulatory sequences to facilitate crosstalk between promoter elements. Thus, gene activity is either structurally constrained or physiologically responsive through chromatin organized in a conformation that is conducive to requisite protein-DNA and protein-protein interactions.
The chromatin organization of the osteocalcin gene illustrates dynamic remodeling of a promoter to accommodate requirements for phenotype-related developmental and steroid hormone responsive activity [35, 36, 37, 38]. Nuclease hypersensitivity studies show that striking modifications in chromatin structure activate OC transcription by rendering the proximal promoter accessible to regulatory and coregulatory proteins that support basal level activity and responsiveness to skeletal regulatory signals. Vitamin D enhancement of osteocalcin gene transcription is associated with removal of the nucleosome at the upstream vitamin D-responsive element that permits binding of the vitamin D receptor-RXR heterodimer. The retention of a nucleosome between the proximal and upstream enhancer domain reduces distance between the basal and vitamin D-responsive element and supports a promoter conformation that is conducive to protein-protein interactions between the vitamin D receptor and the basal TFIIB transcription factor [20, 39]. Chromatin immunoprecipitation analyses have shown that developmental and vitamin D-linked reconfiguration of osteocalcin gene promoter organization is accompanied by acetylation of histones in the proximal basal and upstream vitamin D-responsive element domains . This posttranslational modification of histone proteins reduces the tenacity of histone DNA interactions in a manner that is conducive to an open chromatin organization with increased access to regulatory factors.
The most compelling evidence for a functional involvement of chromatin organization in osteoblast-specific gene expression is the obligatory relationship of dynamic changes in the biochemical and structural properties of the OC gene with competency for bone tissue-restricted transcription, mediated by Runx2 (Cbfa1) . Site-directed mutagenesis of osteocalcin genes that are genetically integrated in stable cell lines have established that Runx2 elements located in the proximal promoter and flanking the VDRE in the upstream promoter are responsible for chromatin reconfiguration of the osteocalcin gene promoter. The interaction of Runx factors with nucleosomal DNA, histone acetyl transferases (e.g., HATs, p300), and histone deacetylase enzymes (HDAC) supports the concept that Runx factors mediate chromatin remodeling [11, 35, 42]. Thus, a component of Runx2 function in activating bone-specific gene expression involves an initial chromatin remodeling for accessibility of transcription factors supporting developmental and steroid hormone-responsive transcription.
Despite the essential role of chromatin remodeling in transcriptional control of the osteocalcin gene, there are open-ended questions. It is not justifiable to extrapolate from these findings to conclude that all genes that are activated and suppressed during skeletogenesis employ identical mechanisms. More globally, there are multiple levels of control that must be mechanistically characterized to explain physiologically responsive regulation of chromatin remodeling within restricted and broader genomic contexts.
Runx2/Cbfa1 Coregulatory Factors Integrate Skeletal Signaling Pathways
Gene expression during skeletal development and bone remodeling is controlled by extracellular matrix (ECM) regulatory signals and physiological cellular responses that must converge at promoter elements to activate or repress transcription, in a biologically responsive manner. The subnuclear compartmentalization of transcription machinery necessitates a mechanistic explanation for directing these signaling factors to sites within the nucleus where gene expression occurs under conditions that support integration of regulatory cues. Runx2 interacts with several proteins that transduce signals from ECM. TLE/Groucho is a repressor protein that regulates Runx2 activity during early embryogenesis; the Yes-associated protein (YAP) mediates c-Src signaling and SMAD coregulatory proteins transduce BMP/TGFβ-mediated signals to skeletal target genes. These coregulatory factors interact with C-terminal segments of the Runx2 transcription factor that overlap the NMTS signal and thereby permits assessment of requirements for recruitment of TLE, YAP, and SMADs to Runx subnuclear foci [12, 43]. Our findings indicate that while nuclear import of YAP and SMAD coregulatory factors is agonist dependent, there is a stringent requirement for fidelity of Runx subnuclear targeting for recruitment of these signaling proteins to transcriptionally active subnuclear foci. Thus, interactions of Runx with TLE, SMAD, and YAP coregulatory proteins associated with the nuclear matrix sites are essential for assembly of transcription machinery that supports expression and attenuation of skeletal genes. Runx coregulatory proteins, including its DNA binding partner, CBFβ, can influence both enhancement and repression of gene transcription [43, 44, 45, 46]. These findings are consistent with Runx serving as a scaffold for protein-protein interactions that contribute to biological control. Furthermore, functional RUNX activity involves contributions by other transcription factors: ETS-1 [47, 48], AP-1 [49, 50], Rb , and C/EBP [52, 53]. For example, osteocalcin transcription becomes synergistically activated in the late stages of osteoblast differentiation through cooperativity between Runx and C/EBPβ or C/EBPδ . Runx may therefore facilitate recruitment of these factors to the nuclear matrix for control of gene expression.
Taken together these multifunctional properties of Runx factors provide insight into the ability of this class of transcription factors to function as essential components of organogenesis. Runx1 (Cbfa/AML1) is necessary for definitive hematopoiesis [54, 55]; Runx2 (Cbfa1/AML3) is required for osteogenesis [26, 56, 57] and Runx3 (Cbfa3/AML2) functions in gut development . Recent findings indicate both overlap in expression of the Runx factors in embryonic mesenchyme that will form skeletal elements, as well as the retention of specific Runx factors in distinct cell populations related to cartilage and bone formation [59, 60]. These findings reflect another level of complexity of a Runx-mediated transcriptional control of osteogenesis involving potential coordination of Runx activities. Furthermore, cross-regulation of Runx factors is likely to occur involving multiple Cbfa (Runx) regulatory sequences in the promoters of Runx genes . The roles of Runx factors in chromatin remodeling, in formation of multimeric DNA binding complexes in subnuclear domains with mediators of development signaling pathways, as well as with co-activator and co-repressor regulatory proteins, are mechanisms that support the induction and temporal modulation of tissue-specific genes necessary to promote cellular differentiation.
In Vivo Consequences of Aberrant Intranuclear Trafficking of Runx Transcription Factors
The biological relevance for the intranuclear distribution of regulatory complexes is directly reflected by aberrant nuclear structure-gene expression interrelationships that are associated with perturbations in skeletal development  and leukemia [16, 62, 63, 64]. Translocations of the Runx1 gene associated with different leukemias results in misdirecting Runx from its wildtype location in subnuclear domains [16, 17, 64]. Using Runx2 and its essential role in osteogenesis as a model, we investigated the fundamental importance of fidelity of subnuclear localization for tissue-differentiating activity by deleting the C-terminus of Runx2 (containing the intranuclear targeting signal and sites for coregulatory signaling factor interactions) via homologous recombination. Mice homozygous for the deletion (Runx2ΔC) do not form bone due to arrest of osteoblast differentiation . Heterozygotes do not develop clavicles. These phenotypes are indistinguishable from those of the homozygous and heterozygous null mutants , indicating that the intranuclear targeting signal and associated coregulatory interactions are a critical determinant for Runx2 function. The expressed truncated RUNX2ΔC protein enters the nucleus and retains normal DNA binding activity, but shows complete loss of intranuclear targeting with coregulatory factors that interact with C-terminus. These results demonstrate that the multifunctional N-terminal region of the Runx2 protein (containing the unique QA stretch and runt homology domain) is not sufficient for biological activity. On the other hand, our findings raise new questions. Is the subnuclear targeting of Runx transcription factors interdependent with the integration of signaling pathways and/or formation of regulatory complexes in foci? To address the unique functional contribution of the NMTS and coregulatory protein interactions, point mutations in the C-terminal domain must be evaluated. Nonetheless, our findings do show that subnuclear localization of Runx factors in specific foci together with associated regulatory functions is an essential component of control of Runx-dependent genes involved in tissue differentiation during embryonic development .
The New Frontiers in Osteobiology
With mounting evidence for organization of nucleic acids and regulatory proteins into subnuclear domains that are associated with components of nuclear architecture, the perception of a dichotomy between nuclear architecture and control of gene expression is difficult to justify. Rather, the challenges are to design experiments to define mechanisms that direct genes and regulatory factors to sites within the nucleus where localization integrates regulatory parameters of gene expression and establishes microenvironments with boundaries between regulatory complexes that are required for fidelity of activity.
It would be presumptuous to propose a single model to account for the specific pathways that direct transcription factors to sites within the nucleus that support transcription. However, present findings indicate that parameters of nuclear architecture functionally interface with components of transcriptional control. The full implications for the involvement of nuclear matrix-associated transcription factors with recruitment of regulatory components to modulate transcription remains to be defined. The diversity of targeting signals must be established to evaluate the extent to which regulatory discrimination is mediated by encoded intranuclear trafficking signals. It will additionally be important to biochemically and mechanistically define the checkpoints which are operative during subnuclear distribution of regulatory factors and the editing steps which are invoked to ensure the structural and functional fidelity of nuclear domains, where replication and expression of genes occur. It is realistic to anticipate that further understanding of mechanisms that position genes and regulatory factors for establishment and maintenance of the bone cell phenotype will clarify nuclear structure-function interrelationships that are operative during osteoblast differentiation.
Studies reported were supported in part by grants from the National Institutes of Health (AR39588, DE12528, PO1AR48818, PO1CA82834). The contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Institutes of Health.
- 1.Aubin, JE, Liu, F 1996
The osteoblast lineage.Bilezikian, JPRaisz, LGRodan, GA eds. Principles of bone biology.Academic PressSan Diego5168Google Scholar
- 3.Lian, JB, Stein, GS 2001
Osteoblast biology.Marcus, RFeldman, DKelsey, J eds. Osteoporosis.Academic PressSan Diego2171Google Scholar
- 4.Whitmarsh, AJ, Davis, RJ 1999Signal transduction by MAP kinases: regulation by phosphorylation-dependent switches.Sci STKE 1999: PE1Google Scholar
- 5.Xiao, G, Gopalkrishnan, RV, Jiang, D, Reith, E, Benson, MD, Franceschi, R 2002Bone morphogenetic proteins, extracellular matrix, and mitogen-activated protein kinase signaling pathways are required for osteoblast-specific gene expression and differentiation in MC3T3-E1 cells.J Bone Miner Res17101110Google Scholar
- 7.Stein, GS, Lian, JB, Montecino, M, van Wijnen, AJ, Stein, JL, Javed, A, Zaidi, K 2002
Involvement of nuclear architecture in regulating gene expression in bone cells.Bilezikian, JPRaisz, LGRodan, GA eds. Principles of bone biology.Academic PressSan Diego169188Google Scholar
- 9.Lian, JB, Stein, GS, Stein, JL, van Wijnen, AJ 1998Osteocalcin gene promoter: unlocking the secrets for regulation of osteoblast growth and differentiation.J Cell Biochem (suppl)30–316272Google Scholar
- 13.Zeng, C, McNeil, S, Pockwinse, S, Nickerson, JA, Shopland, L, Lawrence, JB, Penman, S, Hiebert, SW, Lian, JB, van Wijnen, AJ, Stein, JL, Stein, GS 1998Intranuclear targeting of AML/CBFa regulatory factors to nuclear matrix-associated transcriptional domains.Proc Natl Acad Sci USA9515851589CrossRefPubMedGoogle Scholar
- 14.Zeng, C, van Wijnen, AJ, Stein, JL, Meyers, S, Sun, W, Shopland, L, Lawrence, JB, Penman, S, Lian, JB, Stein, GS, Hiebert, SW 1997Identification of a nuclear matrix targeting signal in the leukemia and bone-related AML/CBFa transcription factors.Proc Natl Acad Sci USA9467466751CrossRefPubMedGoogle Scholar
- 16.McNeil, S, Zeng, C, Harrington, KS, Hiebert, S, Lian, JB, Stein, JL, van Wijnen, AJ, Stein, GS 1999The t(8;21) chromosomal translocation in acute myelogenous leukemia modifies intranuclear targeting of the AML1/CBFalpha2 transcription factor.Proc Natl Acad Sci USA961488214887CrossRefPubMedGoogle Scholar
- 26.Choi, J-Y, Pratap, J, Javed, A, Zaidi, SK, Xing, L, Balint, E, Dalamangas, S, Boyce, B, van Wijnen, AJ, Lian, JB, Stein, JL, Jones, SN, Stein, GS 2001Subnuclear targeting of Runx/Cbfa/AML factors is essential for tissue-specific differentiation during embryonic development.Proc Natl Acad Sci USA9886508655CrossRefPubMedGoogle Scholar
- 27.Merriman, HL, van Wijnen, AJ, Hiebert, S, Bidwell, JP, Fey, E, Lian, J, Stein, J, Stein, GS 1995The tissue-specific nuclear matrix protein, NMP-2, is a member of the AML/CBF/PEBP2/runt domain transcription factor family: interactions with the osteocalcin gene promoter.Biochemistry341312513132PubMedGoogle Scholar
- 30.Zaidi, SK, Javed, A, Choi, J-Y, van Wijnen, AJ, Stein, JL, Lian, JB, Stein, GS 2001A specific targeting signal directs Runx2/Cbfa1 to subnuclear domains and contributes to transactivation of the osteocalcin gene.J Cell Sci1143102Google Scholar
- 35.Gutierrez, J, Sierra, J, Medina, R, Puchi, M, Imschenetzky, M, Hiebert, S, van Wijnen, A, Lian, JB, Stein, G, Stein, J, Montecino, M 2000Interaction of CBFa/AML/PEBP2a transcription factors with nucleosomal sequences requires flexibility in the translational positioning of the histone octamer and exposure of the Cbfa site.Biochemistry391356513574CrossRefPubMedGoogle Scholar
- 38.Paredes, R, Gutierrez, J, Gutierrez, S, Allison, L, Puchi, M, Imschenetzky, M, van Wijnen, A, Lian, J, Stein, G, Stein, J, Montecino, M 2002Interaction of the 1alpha,25-dihydroxyvitamin D3 receptor at the distal promoter region of the bone-specific osteocalcin gene requires nucleosomal remodelling.Biochem J363667676CrossRefPubMedGoogle Scholar
- 43.Javed, A, Guo, B, Hiebert, S, Choi, J-Y, Green, J, Zhao, S-C, Osborne, MA, Stifani, S, Stein, JL, Lian, JB, van Wijnen, AJ, Stein, GS 2000Groucho/TLE/R-Esp proteins associate with the nuclear matrix and repress RUNX (CBFa/AML/PEBP2a-dependent activation of tissue-specific gene transcription.J Cell Sci11322212231PubMedGoogle Scholar
- 44.Javed, A, Barnes, GL, Jassanya, BO, Stein, JL, Gerstenfeld, L, Lian, JB, Stein, GS 2001runt homology domain transcription factors (Runx, Cbfa, and AML) mediate repression of the bone sialoprotein promoter: evidence for promoter context-dependent activity of Cbfa proteins.Mol Cell Biol2128912905CrossRefPubMedGoogle Scholar
- 52.Gutierrez, S, Javed, A, Tennant, D, van Rees, M, Montecino, M, Stein, GS, Stein, JL, Lian, JB 2002CCAAT/enhancer-binding proteins (C/EBP) b and d activate osteocalcin gene transcription and synergize with Runx2 at the C/EBP element to regulate bone-specific expression.J Biol Chem27713161323CrossRefPubMedGoogle Scholar
- 56.Komori, T, Yagi, H, Nomura, S, Yamaguchi, A, Sasaki, K, Deguchi, K, Shimizu, Y, Bronson, RT, Gao, Y-H, Inada, M, Sato, M, Okamoto, R, Kitamura, Y, Yoshiki, S, Kishimoto, T 1997Targeted disruption of Cbfa1 results in a complete lack of bone formation owing to maturational arrest of osteoblasts.Cell89755764PubMedGoogle Scholar
- 57.Otto, F, Thornell, AP, Crompton, T, Denzel, A, Gilmour, KC, Rosewell, IR, Stamp, GWH, Beddington, RSP, Mundlos, S, Olsen, BR, Selby, PB, Owen, MJ 1997Cbfa1, a candidate gene for cleidocranial dysplasia syndrome, is essential for osteoblast differentiation and bone development.Cell89765771PubMedGoogle Scholar
- 58.Li, QL, Ito, K, Sakakura, C, Fukamachi, H, Inoue, K, Chi, XZ, Lee, KY, Nomura, S, Lee, CW, Han, SB, Kim, HM, Kim, WJ, Yamamoto, H, Yamashita, N, Yano, T, Ikeda, T, Itohara, S, Inazawa, J, Abe, T, Hagiwara, A, Yamagishi, H, Ooe, A, Kaneda, A, Sugimura, T, Ushijima, T, Bae, SC, Ito, Y 2002Causal relationship between the loss of RUNX3 expression and gastric cancer.Cell109113124PubMedGoogle Scholar