Relocalizing genetic loci into specific subnuclear neighborhoods

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/50376689

Relocalizing Genetic Loci into Specific Subnuclear Neighborhoods ARTICLE in JOURNAL OF BIOLOGICAL CHEMISTRY · MARCH 2011 Impact Factor: 4.57 · DOI: 10.1074/jbc.M111.221481 · Source: PubMed

CITATIONS

READS

19

33

4 AUTHORS, INCLUDING: Sherry Lee

Emery Bresnick

Whitehead Institute for Biomedical Research

University of Wisconsin–Madison

7 PUBLICATIONS 338 CITATIONS

151 PUBLICATIONS 7,468 CITATIONS

SEE PROFILE

SEE PROFILE

Available from: Sherry Lee Retrieved on: 04 February 2016

Gene Regulation: Relocalizing Genetic Loci into Specific Subnuclear Neighborhoods Hsiang-Ying Lee, Kirby D. Johnson, Meghan E. Boyer and Emery H. Bresnick

Access the most updated version of this article at doi: 10.1074/jbc.M111.221481 Find articles, minireviews, Reflections and Classics on similar topics on the JBC Affinity Sites. Alerts: • When this article is cited • When a correction for this article is posted Click here to choose from all of JBC's e-mail alerts This article cites 50 references, 22 of which can be accessed free at http://www.jbc.org/content/286/21/18834.full.html#ref-list-1

Downloaded from http://www.jbc.org/ at Massachusetts Institute of Technology on November 5, 2014

J. Biol. Chem. 2011, 286:18834-18844. doi: 10.1074/jbc.M111.221481 originally published online March 11, 2011

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 286, NO. 21, pp. 18834 –18844, May 27, 2011 © 2011 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A.

Relocalizing Genetic Loci into Specific Subnuclear Neighborhoods* Received for publication, January 13, 2011, and in revised form, February 16, 2011 Published, JBC Papers in Press, March 11, 2011, DOI 10.1074/jbc.M111.221481

Hsiang-Ying Lee1, Kirby D. Johnson, Meghan E. Boyer, and Emery H. Bresnick2 From the Wisconsin Institutes for Medical Research, Paul Carbone Cancer Center, Department of Cell and Regenerative Biology, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin 53705

trans-Acting factors establish cell type-specific and developmental stage-specific transcriptional networks as a critical step in the development of multicellular organisms. Many questions remain unanswered regarding how genetic networks are established and maintained. An important parameter is the threedimensional organization of the cell nucleus, in which chromosomes occupy distinct territories, and regulatory factors reside in specific nuclear bodies that exert specialized functions (1–3).

The nuclear periphery is commonly associated with transcriptional repression (4) but also contains microdomains harboring active genes (5). Genes are recruited into transcription factories enriched in polymerase II and transcriptional machinery (6). Establishing and maintaining subnuclear positioning of genes is believed to be an important step in the regulation of transcription and fundamental biological processes, including development. Genes controlled by the master regulator of hematopoiesis GATA-1 (7, 8) segregate based on their residency sites in subnuclear domains (5). The cell type-specific coregulator Friend of GATA-1 (FOG-1) mediates GATA-1-dependent activation and repression of a large target gene ensemble (9, 10), although certain GATA-1 targets are FOG-1-independent (10, 11). FOG-1 facilitates GATA-1 chromatin occupancy at select sites (12, 13) and is required for chromatin looping (14) and GATA switches (12). FOG-1 interacts with other coregulators, including the NuRD3 chromatin remodeling complex (15) that is frequently implicated in repression (16) but also in activation (17), and the corepressor C-terminal binding protein. GATA-1 expels FOG-1-dependent, but not -independent, genes from the nuclear periphery during erythroid maturation (5). The finding that only FOG-1-dependent targets relocalize raises intriguing questions. How does a trans-acting factor differentially control subnuclear positioning of distinct members of its target gene ensemble? Is locus repositioning required for, or a consequence of, altered transcription? Studies in diverse systems indicate that genes can reside within or external to their chromosome territory (1). The human major histocompatibility complex (MHC) gene cluster spanning 3.6 Mb localizes away from its territory in a cell typespecific manner (18). During murine B lymphocyte development, immunoglobulin genes IgH and Ig␬ reside preferentially at the nuclear periphery in hematopoietic progenitors and pro-T cells, whereas they localize internally in pro-B cells (19). Interferon-␥ induces MHC II relocalization away from its territory in human fibroblasts (18). Retinoic acid induces activated Hoxb, but not Hoxd, to migrate away from its chromosome territory in mouse embryonic stem cells (20). The signaling environment can therefore be an important determinant of gene positioning. Although various studies imply important functional consequences of gene positioning, it is challenging to establish cau-

* This work was supported, in whole or in part, by National Institutes of Health Grants DK50190 and DK68634 (to E. H. B.). Recipient of a predoctoral fellowship from the American Heart Association. 2 To whom correspondence should be addressed. E-mail: [email protected]. 1

18834 JOURNAL OF BIOLOGICAL CHEMISTRY

3

The abbreviations used are: NuRD, nucleosome remodeling and deacetylase complex; EKLF, erythroid Kru¨ppel-like factor; BAC, bacterial artificial chromosome.

VOLUME 286 • NUMBER 21 • MAY 27, 2011

Downloaded from http://www.jbc.org/ at Massachusetts Institute of Technology on November 5, 2014

A poorly understood problem in genetics is how the threedimensional organization of the nucleus contributes to establishment and maintenance of transcriptional networks. Genetic loci can reside in chromosome “territories” and undergo dynamic changes in subnuclear positioning. Such changes appear to be important for regulating transcription, although many questions remain regarding how loci reversibly transit in and out of their territories and the functional significance of subnuclear transitions. We addressed this issue using GATA-1, a master regulator of hematopoiesis implicated in human leukemogenesis, which often functions with the coregulator Friend of GATA-1 (FOG-1). In a genetic complementation assay in GATA-1-null cells, GATA-1 expels FOG-1-dependent target genes from the nuclear periphery during erythroid maturation, but the underlying mechanisms are unknown. We demonstrate that GATA-1 induces extrusion of the ␤-globin locus away from its chromosome territory at the nuclear periphery, and extrusion precedes the maturation-associated transcriptional surge and morphological transition. FOG-1 and its interactor Mi-2␤, a chromatin remodeling factor commonly linked to repression, were required for locus extrusion. Erythroid Kru¨ppel-like factor, a pivotal regulator of erythropoiesis that often co-occupies chromatin with GATA-1, also promoted locus extrusion. Disruption of transcriptional maintenance did not restore the locus subnuclear position that preceded activation. These results lead to a model for how a master developmental regulator relocalizes a locus into a new subnuclear neighborhood that is permissive for high level transcription as an early step in establishing a cell type-specific genetic network. Alterations in the regulatory milieu can abrogate maintenance without reversion of locus residency back to its original neighborhood.

Subnuclear Neighborhoods

EXPERIMENTAL PROCEDURES Cell Culture—G1E cells expressing ER-GATA-1 (G1E-ERGATA-1) were maintained in Iscove’s modified Dulbecco’s medium (Invitrogen) containing 1% penicillin/streptomycin (Invitrogen), 2 units/ml erythropoietin, 120 nM monothioglycerol (Sigma), 0.6% conditioned medium from a Kit ligand-producing Chinese hamster ovary cell line, 15% fetal bovine serum (FBS) (Gemini BioProducts), and 1 ␮g/ml puromycin (Sigma). ER-GATA-1 was induced by 1 ␮M ␤-estradiol treatment. MAY 27, 2011 • VOLUME 286 • NUMBER 21

Small Interference RNA-mediated Knockdown—G1E-ERGATA-1 cells (3 ⫻ 106) were resuspended in 100 ␮l of Nucleofector solution R (Amaxa Biosystems) and transfected with 240 pmol of SMARTpool siRNA (Dharmacon/Thermo Fisher Scientific). siGenome nontargeting siRNA pool (Dharmacon/ Thermo Fisher Scientific) was used as a control. Cells were transfected twice with the Nucleofector II (Amaxa Biosystems) using the G-016 program, allowing 24 h between transfections, and treated with 1 ␮M ␤-estradiol after the second transfection. Cells were harvested 24 h after the second transfection. Protein and RNA were prepared from 1 ⫻ 106 and 2 ⫻ 106 cells, respectively. For analyses of transcriptional maintenance, cells were treated with ␤-estradiol, transfected twice after 24 and 36 h of ␤-estradiol treatment, and harvested 48 h post-treatment. Protein Analysis—Whole cell lysates were boiled for 10 min in SDS sample buffer (25 mM Tris, pH 6.8, 4% ␤-mercaptoethanol, 3% SDS, 0.05% bromphenol blue, and 5% glycerol). Lysates were resolved by SDS-PAGE, and proteins were measured by semi-quantitative Western blotting with ECL Plus (GE Healthcare). Quantitative Real Time RT-PCR—Total RNA was purified with TRIzol (Invitrogen) from identical cultures used for Western blotting and FISH analyses. cDNA was prepared by annealing RNA (1 ␮g) with 250 ng of a 5:1 mixture of random and oligonucleotide (deoxyribosylthymine) primers preheated at 68 °C for 10 min, followed by incubation with Moloney murine leukemia virus reverse transcriptase (RT; 50 units) (Invitrogen), 10 mM dithiothreitol (DTT), RNasin (Promega), and 0.5 mM deoxynucleoside triphosphates at 42 °C for 1 h. Reactions were diluted to a final volume of 130 ␮l and heat-inactivated at 98 °C for 5 min. Reactions (15 ␮l) contained 2.0 ␮l of cDNA, 7.5 ␮l of SYBR Green master mix (Applied Biosystems), and appropriate primers. Product was monitored by SYBR Green fluorescence. Control reactions lacking RT yielded little to no signal. Relative expression levels were determined from a standard curve of serial dilutions of cDNA samples and were normalized to Gapdh expression. RNA Stability Assay—G1E-ER-GATA-1 cells were pretreated with 1 ␮M ␤-estradiol for 24 h. Cultures were treated with 10 ␮g/ml actinomycin D (A1410; Sigma) or vehicle for up to 4 h. Total RNA was purified and quantitated by real time RT-PCR analysis. Microscopy and FISH Analysis—The three-dimensional immuno-FISH assay was conducted as described previously (5). In brief, G1E-ER-GATA-1 cells were stained with goat antilamin B (C-20) (Santa Cruz Biotechnology) and donkey antigoat Alexa Fluor 350 (Invitrogen) to detect the nuclear periphery prior to FISH. Biotinylated probes were prepared by nick translation (Roche Applied Science) of BACs obtained from BACPAC resources. The following BACs were used: RP23370E12 (␤-globin), RP24-330L18 (Epb4.9), and RP24-167E18 (Slc4a1); for sites on murine chromosome 7, RP23-238C3 (⫺28.5 Mb), RP23-368M23 (⫺25.6 Mb), RP24-232O8 (⫺5 Mb), RP24-272L15 (⫺0.8 Mb), RP24-344J7 (⫺0.3 Mb), RP24151B23 (⫺0.1 Mb), RP23-430G11 (⫹0.6 Mb), RP24-315I4 (⫹5.5 Mb), RP24-312B10 (⫹10 Mb), and RP23-447D22 (⫹18 Mb). Detection of probe was by sequential applications of Rhodamine600 Avidin D (Vector Laboratories), biotinylated goat JOURNAL OF BIOLOGICAL CHEMISTRY

18835

Downloaded from http://www.jbc.org/ at Massachusetts Institute of Technology on November 5, 2014

sality, and certain results are inconsistent with simple correlations (1). Repositioning of the lymphoid genes Rag and Tdt to centromeric subdomains in primary CD4⫹8⫹ T lymphocytes occurs after transcription silencing (21). Approaches to test whether nuclear periphery association instigates repression have involved fusing the Lac repressor to nuclear periphery constituents (lamin B, emerin, and Lap2␤) to tether enhanced GFP-containing reporter genes to the nuclear periphery (1). However, different modes of manipulating positioning yielded inconsistent conclusions. Although gene movement may be required for transcriptional activity in certain contexts, causality remains debatable, and the molecular underpinnings of how a locus is driven from one subnuclear compartment to another remain to be elucidated. In the context of erythropoiesis, in which billions of red blood cells are generated daily from specific progenitors, the FOG-1-dependent ␤-globin locus relocalizes from the nuclear periphery to internal sites as primary fetal liver erythroblasts mature (22). Relocalization correlates with high level transcription but not the transition from an inactive to an active state. Because considerable mechanistic insights have emerged regarding how the ␤-globin locus is regulated (23, 24), this system is attractive for dissecting mechanisms of how a transacting factor dynamically controls target gene positioning. Using a genetic complementation approach in GATA-1-null cells (25–27), we demonstrate that GATA-1-induced ␤-globin locus extrusion from its chromosome territory precedes induction of key markers of erythroid maturation. GATA-1-driven locus movement is therefore not merely a consequence of the surge of transcriptional activity nor of the gross morphological remodeling intrinsic to maturation. Because GATA-1 engages multiple trans-acting factors and coregulators to control hematopoiesis (28, 29), we assessed whether these factors interface with GATA-1 to drive chromosomal transitions. Finally, because many questions remain unanswered regarding similarities and differences in mechanisms of transcriptional activation versus maintenance, we used a novel paradigm to compare the molecular underpinnings of these processes. In aggregate, the results support a model in which multiple components involved in GATA-1 function exert critical activities to rapidly induce locus extrusion from the respective chromosome territory, yielding a new subnuclear residency site. Abrogating maintenance of transcriptional activity did not restore the locus to its original neighborhood, indicating that establishing the requisite nuclear topography is intrinsic to the process of building a cell type-specific genetic network, although network remodeling can occur independent of gross locus rearrangements.

Subnuclear Neighborhoods

RESULTS AND DISCUSSION trans-Acting Factor-driven Target Gene Relocalization in Subnuclear Domains—To elucidate how trans-acting factors instigate target gene relocalization during development, it is important to derive structural insights into the chromosomal transition. In principle, factors may expel target genes away from the nuclear periphery by repositioning the locus concomitantly with a major component of the chromosome (Fig. 1A, top) or by facilitating a higher order chromosomal transition, allowing for locus extrusion from its chromosome territory without relocalization of major flanking sequences (Fig. 1A, bottom). Although loci can relocalize away from their respective territories, given that GATA-1-induced transcriptional changes are accompanied by massive maturation-associated changes in cell morphology (5), it would not be surprising if entire, or major components of, chromosome territories relocalized during maturation. We tested whether GATA-1-dependent expulsion of the FOG-1-dependent ␤-globin locus from the nuclear periphery is highly restricted to the locus or involves concomitant repositioning of neighboring sequences. To distinguish between models, three-dimensional immuno-FISH was conducted with BAC probes to detect multiple regions of murine chromosome 7 containing the ␤-globin locus under conditions in which nuclear morphology was preserved. The subnuclear localization of target loci relative to nuclear periphery-associated lamin was quantitated as described (5) to yield the percentage of loci localizing at the periphery (outer 20% of the nuclear radius), as well as the inner four shells, each constituting 20% of the radius (Fig. 1B). BACs were used to detect regions ⫺27, ⫺5, ⫺0.8, ⫺0.3, ⫺0.1, ⫹0.6, ⫹5.5, ⫹10, and ⫹18 Mb relative to the ␤-globin locus (Fig. 1C). The subnuclear localization of these sites was compared with the ␤-globin locus in uninduced and ␤estradiol-treated G1E-ER-GATA-1 cells. The sites analyzed lacked GATA-1-regulated genes, based on transcription profil-

18836 JOURNAL OF BIOLOGICAL CHEMISTRY

ing and ChIP-seq analyses (30). The ␤-globin locus subnuclear localization was identical to our prior analysis (5), with 40 –50% of the loci at the nuclear periphery (shell 1) in immature cells in which ER-GATA-1 is inactive. ER-GATA-1 activation expelled ⬃20% of the loci from the periphery (p ⫽ 0.007), thereby relocating the loci into internal subnuclear compartments (Fig. 1D). Shell 1 constitutes 36% of the nuclear volume, a greater volume than any of the other four shells. This suggests that ER-GATA-1-mediated expulsion of loci from the periphery reflects the need for loci to escape from regulatory influences imposed by the periphery microenvironment or loci must rendezvous with regulatory factors residing in internal subnuclear compartments. Because genes are often expressed stochastically, with only certain alleles in the cell population active at any given time (31, 32), one would not expect the full complement of loci to relocalize in the genetic complementation assay. We also conducted a double-labeling three-dimensional immuno-FISH assay to detect the ␤-globin locus (Fig. 1E, red) and ⫹18-Mb site (green) simultaneously. This analysis confirmed that ␤-estradiol treatment significantly induced expulsion of the ␤-globin locus from the nuclear periphery (19.5% reduction in shell 1, p ⫽ 0.003) without significantly affecting the subnuclear localization of the ⫹18-Mb site (5.5% reduction, p ⫽ 0.478) (Fig. 1E). Upon maturation of primary fetal liver erythroblasts, the percentage of ␤-globin loci associated with the periphery progressively decreased from 66 to 27% (22). All of the other BACs, except the ⫺5-Mb BAC, detected 35– 45% of the loci at the periphery pre- and post-ERGATA-1 activation (Fig. 1D). ER-GATA-1 did not significantly affect the subnuclear localization of any of the other regions of chromosome 7 (Fig. 1D). The topography of these regions was therefore insensitive to ER-GATA-1-induced erythroid maturation. In immature cells, a greater percentage of the ⫺5-Mb loci localized internally versus ␤-globin, with ⬃25% loci residing at the periphery (Fig. 1D). The ⫺5-Mb region was immobile upon maturation (Fig. 1D). As GATA-1-driven ␤-globin locus repositioning is highly restricted to the locus, independent of major neighboring sequences, the results support a model (Fig. 1A, bottom) in which GATA-1 induces extrusion of FOG-1-dependent target loci from their chromosome territories without measurable repositioning of flanking chromosomal regions. Subnuclear Relocalization of an Erythroid Locus Precedes Hallmarks of Erythroid Maturation—As GATA-1 is essential for erythroid differentiation (33–35), repositioning of GATA-1 target genes might be a consequence of cell maturation and/or the associated transcriptional changes. To address this issue, we conducted kinetic analyses in G1E-ER-GATA-1 cells, systematically monitoring hallmark features of cell maturation, including reduced cell size, reduced nuclear size, and induction of ␤major and Slc4a1 primary transcripts. The kinetics of changes in these parameters were compared with that of ␤-globin subnuclear relocalization. Cell size and nuclear size progressively decreased concomitant with induction of ␤major transcription during erythroid maturation (Fig. 2A). 40 –50% of ␤-globin loci resided at the nuclear periphery in immature cells, whereas ␤-estradiol treatment rapidly expelled the loci from VOLUME 286 • NUMBER 21 • MAY 27, 2011

Downloaded from http://www.jbc.org/ at Massachusetts Institute of Technology on November 5, 2014

anti-avidin D (Vector Laboratories) and Rhodamine600 Avidin D. For microscopic analysis, images were captured by Olympus IX81 motorized inverted microscope, in which image stacks were captured from 0.25-␮m Z-sections and deconvoluted with Slidebook 5.0 software (Intelligent Imaging Innovations, Inc.). The image stacks were projected onto a single plane, and the distance of loci from nuclear periphery was measured with Slidebook 5.0 software. Measurements were normalized to nuclear size by dividing the distance by the average radius of the cell nucleus. The nucleus was divided into five concentric shells, each encompassing 20% of the radius. The subnuclear positioning of loci in shells 1–5 was scored. Shell 1 represents the nuclear periphery, and shell 5 represents the center of the nucleus. A minimum of 100 –200 loci was scored for each locus. Three-dimensional analyses, in which distances were measured with multiple single optical sections from a single cell, yielded identical conclusions. For statistical analysis, a two-sample binomial test was performed using the statistical software R. The proportion of loci localized at the nuclear periphery (shell 1) and in other nuclear compartments (shells 2–5) in control versus knockdown cells or untreated versus ␤-estradiol-treated cells was analyzed for statistical significance.

Subnuclear Neighborhoods

Downloaded from http://www.jbc.org/ at Massachusetts Institute of Technology on November 5, 2014

FIGURE 1. GATA-1-driven extrusion of a nuclear periphery-associated genetic locus out of its chromosome territory. A, models depicting GATA-1mediated locus relocalization. B, quantitative analysis of three-dimensional immuno-FISH data. Cells were divided into five concentric shells, each encompassing 20% of the radius. Shell 1 represents the nuclear periphery, whereas shell 5 represents the center of the nucleus. C, murine chromosome 7. D, subnuclear positioning of loci in untreated and ␤-estradiol-treated (24 h) G1E-ER-GATA-1 cells. The percent of loci in the nuclear periphery (shell 1) was compared in untreated versus ␤-estradiol-treated cells using a two-sample binomial test, and the respective p values are shown. Note that ␤-estradiol treatment only significantly affected the subnuclear positioning of the ␤-globin locus. Subnuclear positioning of the ⫺27 Mb site represents averaged results of probes detecting ⫺28.5 and ⫺25.6 Mb. E, three-dimensional immuno-FISH using two probes to detect the ␤-globin locus and ⫹18 Mb site simultaneously. Examples of untreated and ␤-estradiol-treated cells are shown. Red, ␤-globin locus; green, ⫹18 Mb site; yellow, overlapping red and green signals; blue, nuclear lamin immunofluorescence.

the periphery. Treatment of cells with ␤-estradiol for up to 6 h did not influence the subnuclear positioning of the ␤-globin locus. Maximal expulsion of the ␤-globin locus from the nuclear periphery was apparent by 12 h (Fig. 2B). The ␤-globin locus was expelled from the periphery considerably faster than MAY 27, 2011 • VOLUME 286 • NUMBER 21

the morphological remodeling and the surge of ␤major and Slc4a1 transcription coupled with maturation (Fig. 2C). As locus extrusion occurred early during maturation, the extrusion is not likely a consequence of maturation. These results support a model in which locus extrusion is a prerequisite for JOURNAL OF BIOLOGICAL CHEMISTRY

18837

Subnuclear Neighborhoods

high level transcription of the ␤-globin locus and is likely for a large cohort of FOG-1-dependent GATA-1 target genes. Molecular Determinants of GATA-1-instigated Locus Extrusion—Although GATA-1 interacts with a cohort of transacting factors and coregulators (29), how these factors contribute to GATA-1 function is incompletely understood. GATA-1 regulates target genes through multistep mechanisms (36, 37), and therefore individual GATA-1 interactors might have dedicated functions at specific steps. We established subnuclear relocalization of target loci as a rapid step in the GATA-1-dependent activation mechanism. To determine whether all or a subset of GATA-1-interacting factors mediate this relocalization, we used an RNAi-based loss-of-function approach to knock down trans-acting factors or coregulators in G1E-ER-

18838 JOURNAL OF BIOLOGICAL CHEMISTRY

GATA-1 cells, and the functional consequences were analyzed vis a` vis subnuclear localization and transcription. Because GATA-1 relocalizes FOG-1-dependent genes (5), we asked whether the GATA-1-dependent chromosomal transition requires FOG-1. After introduction of nontargeting control or anti-Fog1 siRNA into G1E-ER-GATA-1 cells, ERGATA-1 was activated by ␤-estradiol (Fig. 3A). Fog1 mRNA and protein were greatly reduced by the knockdown (Fig. 3, B and C). The FOG-1 knockdown significantly inhibited ␤-globin and ␣-globin loci expulsion from the periphery (p ⫽ 0.004 and 0.003, respectively) and locus relocalization internally (␤-globin: shell 4, p ⫽ 0.023; ␣-globin: shell 3, p ⫽ 0.007, shell 4, p ⫽ 0.016) (Fig. 3D), and it had a devastating consequence for ␤major transcription (Fig. 3B). By contrast, activated ERVOLUME 286 • NUMBER 21 • MAY 27, 2011

Downloaded from http://www.jbc.org/ at Massachusetts Institute of Technology on November 5, 2014

FIGURE 2. GATA-1-driven locus extrusion precedes major hallmarks of erythroid maturation. A–C, kinetic analyses of G1E-ER-GATA-1 cell maturation. A, top, representative photomicrographs of three-dimensional immuno-FISH analysis of ␤-globin locus subnuclear positioning. Blue, nuclear lamin immunofluorescence; red, probe detecting the ␤-globin locus. Bottom, Giemsa stain analysis of cell morphology. B, quantitation of ␤-globin locus subnuclear positioning (mean; three independent experiments). C, kinetics of nuclear periphery expulsion (maximal expulsion is designated as 100%) versus parameters representing hallmarks of erythroid maturation (mean ⫾ S.E.; three independent experiments).

Subnuclear Neighborhoods

GATA-1 expelled the ␤-globin locus from the nuclear periphery normally in cells containing control siRNA (Fig. 3D). ERGATA-1 did not significantly affect Epb4.9 subnuclear positioning (⬃50% of loci residing at the periphery) and was insensitive to the FOG-1 knockdown (Fig. 3D). These results reveal an essential role of FOG-1 in the GATA-1-driven expulsion of the ␤-globin locus from the nuclear periphery, and importantly, the expulsion was rapid relative to FOG-1-dependent changes in erythroid maturation hallmarks (Fig. 2C). Although FOG-1 mediates GATA-1 chromatin occupancy at select sites (12, 13) and local chromatin looping within restricted chromatin domains (14), as measured by chromosome conformation capture assay (38), its activity to mediate locus repositioning into a distinct subnuclear compartment was unexpected. A 12-amino acid sequence in the FOG-1 N terminus binds NuRD, and NuRD co-occupies chromatin sites with GATA-1 MAY 27, 2011 • VOLUME 286 • NUMBER 21

at GATA-1-activated and -repressed genes (15, 17, 39). Knock-in mice bearing mutations that disrupt the FOG-1NuRD interaction are anemic and exhibit macrothrombocytopenia (17, 39). These mice were bred to transgenic mice containing a human ␤-globin locus transgene to assess whether the FOG-1-NuRD interaction mediates developmental repression of fetal globin genes (40). Adult human and murine globin gene expression in bone marrow was considerably lower in mutant mice, suggesting a role for the FOG-1-NuRD interaction in ␤-globin gene activation. Given the FOG-1 requirement for GATA-1-dependent locus extrusion, we asked whether the NuRD complex also mediates this chromosomal transition. We used the RNAi strategy described above to knock down Mi-2␤, the critical ATPase component of the NuRD complex (16), which yielded a 57.8% reduction in Mi-2␤ mRNA (p ⫽ 0.0019) and nearly a complete loss of Mi-2␤ protein, without a reduction in ER-GATA-1 (Fig. JOURNAL OF BIOLOGICAL CHEMISTRY

18839

Downloaded from http://www.jbc.org/ at Massachusetts Institute of Technology on November 5, 2014

FIGURE 3. FOG-1 requirement for GATA-1-dependent locus extrusion. A, experimental strategy of RNAi assay in G1E-ER-GATA-1 cells. B, quantitative RT-PCR analysis of Fog1 and ␤major mRNA (mean ⫾ S.E.; three independent experiments). C, Western blot detection of endogenous FOG-1 in whole cell extracts. D, semiquantitative three-dimensional immuno-FISH analysis of ␤-globin, ␣-globin, and Epb4.9 subnuclear positioning in G1E-ER-GATA-1 cells transfected with control or Fog1 siRNA. The percent of loci in shell 1 was compared in Fog1 versus control siRNA conditions from ␤-estradiol-treated cells using a two-sample binomial test, and the respective p values are shown.

Subnuclear Neighborhoods

4, A and B). The knockdown significantly reduced ␤major expression (59.3%, p ⫽ 0.0001), and GATA-1-dependent ␤-globin locus expulsion from the nuclear periphery (shell 1) (p ⫽ 0.0002) was largely inhibited (Fig. 4C). By contrast, expressions of Fog-1 (p ⫽ 0.130) and the pivotal regulator of erythropoiesis Eklf (41, 42) (p ⫽ 0.584) were unaffected (Fig. 4A). Thus, both Mi-2␤ and FOG-1 are required for GATA-1-dependent locus repositioning. Because the FOG-1-NuRD interaction is required for maximal ␤-globin expression in vivo (40), it is attractive to propose that the FOG-1-NuRD-mediated subnuclear relocalization of the ␤-globin locus underlies this requirement for maximal activity in vivo. Similar to FOG-1-NuRD, the trans-acting factor EKLF colocalizes with GATA-1 on chromatin (43, 44). EKLF functionally interacts with GATA-1 in transfection assays (45), and EKLFregulated genes colocalize with GATA-1-regulated genes in specialized transcription factories (46). To test whether GATA1-mediated transcriptional regulation of endogenous loci requires EKLF, and if EKLF influences GATA-1-dependent locus repositioning, we knocked down EKLF in G1E-ERGATA-1 cells. The knockdown reduced Eklf mRNA by 71.2% (p ⫽ 0.0001), and EKLF protein was undetectable (Fig. 5, A and B), although Fog1 mRNA (Fig. 5A) and ER-GATA-1 (Fig. 5B) were unaffected. ␤major mRNA decreased by 74.2% (p ⫽ 0.0001) (Fig. 5A), and locus expulsion from periphery was sig-

18840 JOURNAL OF BIOLOGICAL CHEMISTRY

nificantly (p ⫽ 0.010) impaired versus cells treated with control siRNA (Fig. 5C). EKLF is therefore required for GATA-1 to maximally activate ␤major and to promote GATA-1-dependent locus extrusion. Activating Versus Maintaining Gene Activity and Subnuclear Positioning, Locus-specific Mechanisms, and Distinct Coregulator Requirements—Whereas major efforts have been made to unravel transcriptional activation mechanisms, considerably less is known about mechanisms that maintain specific functional states of genes. In the context of factor-driven locus movement into subnuclear domains, it is instructive to consider how mechanisms instigating relocalization relate to those ensuring that the relocalized locus maintains its new geography. An intriguing aspect of this problem is whether the new geography is required to maintain the active transcriptional state or if additional mechanisms emerge to increase or decrease transcription at the new site. We developed a system to selectively interrogate molecular requirements for activation versus maintenance. Because FOG-1 and the NuRD complex mediate specific GATA-1 activities, including nuclear periphery expulsion, we asked whether they are required to maintain GATA-1-dependent transcription and subnuclear positioning. The RNAi-based loss-of-function assay was modified such that factors are knocked down post-ER-GATA-1 activation in G1E-ER-GATA-1 cells (Fig. VOLUME 286 • NUMBER 21 • MAY 27, 2011

Downloaded from http://www.jbc.org/ at Massachusetts Institute of Technology on November 5, 2014

FIGURE 4. FOG-1-interactor and NuRD complex component Mi-2␤ promotes GATA-1-dependent locus extrusion. A, quantitative RT-PCR analysis of Mi-2␤, ␤major, Fog1, and Eklf mRNA (mean ⫾ S.E.; eight independent experiments). B, Western blot detection of endogenous Mi-2␤ in whole cell extracts. Top, anti-Mi-2␤; bottom, anti-GATA-1. C, semiquantitative three-dimensional immuno-FISH analysis of ␤-globin locus subnuclear positioning in G1E-ER-GATA-1 cells transfected with control or Mi-2␤ siRNA. The percent of loci in shell 1 was compared in Mi-2␤ versus control siRNA from ␤-estradiol-treated cells using a two-sample binomial test, and the respective p values are shown.

Subnuclear Neighborhoods

6A), and primary transcripts were quantitated as a metric of transcription. Surprisingly, although FOG-1 was crucial for GATA-1-mediated activation of ␤major (Fig. 3B) and Slc4a1 (47) transcription, it was only required to maintain Slc4a1 transcription (Fig. 6B). Newly synthesized ␤major and Slc4a1 primary transcripts persisted for an extended time when FOG-1 levels were greatly reduced post-ER-GATA-1 activation. Measurement of the t1⁄2 of the primary transcripts confirmed an intrinsic low stability (Fig. 6C), validating the maintenance assay. The FOG-1 interactor Mi-2␤ resembled FOG-1 in conferring ␤major and Slc4a1 activation (Fig. 4A). By contrast to FOG-1 function at Slc4a1, Mi-2␤ was not required to maintain Slc4a1 transcription (Fig. 6B). Thus, molecular determinants for activation versus maintenance can differ in a locus-specific manner. Despite strong evidence implicating FOG-1 collaboration with the NuRD complex (15, 17, 39), and the dual FOG-1/Mi-2␤ requirement to activate Slc4a1, FOG-1 and Mi-2␤ are differentially important for maintaining Slc4a1 transcription. These results illustrate a novel mechanism by which a cell type-specific trans-acting factor differentially utilizes coregulators to mediate activation versus maintenance of members of its target gene ensemble. Because FOG-1 mediates GATA-1-driven Slc4a1 expulsion from the nuclear periphery and subsequent activation (5), the MAY 27, 2011 • VOLUME 286 • NUMBER 21

FOG-1 activity to maintain transcription might also be linked to regulation of subnuclear positioning. We knocked down FOG-1 post-ER-GATA-1 activation, which impaired Slc4a1 transcriptional maintenance (Fig. 6B), and we asked whether Slc4a1 reverted to a nuclear periphery localization or if its internal localization was maintained. Quantitative analysis revealed that impairment of Slc4a1 transcriptional maintenance did not affect its subnuclear localization (Fig. 6D). Thus, although locus expulsion from the nuclear periphery occurs early in transcriptional activation, changes in the regulatory milieu can counteract maintenance mechanisms without a concomitant reversion to the preexisting subnuclear residency pattern. Loss of FOG-1 post-ER-GATA-1 activation did not abrogate maintenance of ␤major transcription, and analogous to Slc4a1, ␤major subnuclear localization was unaffected (Fig. 6D). Dynamic Subnuclear Transitions as Early Events in Establishing a Cell Type-specific Genetic Network—We describe herein the actions of a master regulator of hematopoiesis, in concert with interacting coregulators and another cell typespecific trans-acting factor, to drive the relocalization of a critical target gene from one subnuclear domain into another. Reducing the levels of these factors that mediate activation (and repression in other contexts) ablated the locus relocalization. Taken together with the rapidity of the relocalization, relative to induction of hallmark features of erythroid maturation, it JOURNAL OF BIOLOGICAL CHEMISTRY

18841

Downloaded from http://www.jbc.org/ at Massachusetts Institute of Technology on November 5, 2014

FIGURE 5. EKLF promotes GATA-1-dependent locus extrusion. A, quantitative RT-PCR analysis of Eklf, ␤major, and Fog1 mRNA (mean ⫾ S.E.; eight independent experiments). B, Western blot detection of endogenous EKLF in whole cell extracts. Top, anti-EKLF; bottom, anti-GATA-1. C, semiquantitative threedimensional immuno-FISH analysis of ␤-globin locus subnuclear positioning in G1E-ER-GATA-1 cells transfected with control or Eklf siRNA. The percent of loci in shell 1 were compared in Eklf versus control siRNA from ␤-estradiol-treated cells using a two-sample binomial test, and the respective p values are shown.

Subnuclear Neighborhoods

Downloaded from http://www.jbc.org/ at Massachusetts Institute of Technology on November 5, 2014

FIGURE 6. Differential mechanisms of transcriptional activation and maintenance. A, diagram depicting knockdown strategy. B, quantitative RT-PCR analysis of Fog1, Mi-2␤, ␤major, and Slc4a1 expression in cells transfected with control, Fog1, or Mi-2␤ siRNA. C, analysis of primary transcript stabilities. Cells were treated with vehicle or actinomycin D; RNA was isolated at various times thereafter, and primary transcripts were quantitated by real time RT-PCR (mean ⫾ S.E.; three independent experiments). D, semiquantitative three-dimensional immuno-FISH analysis in ␤-estradiol-induced G1E-ER-GATA-1 cells transfected with control, Fog1, or Mi-2␤ siRNA (mean ⫾ S.E.; three independent experiments). E, model of FOG-1-dependent activation and FOG-1-independent maintenance of ␤-globin locus transcription.

18842 JOURNAL OF BIOLOGICAL CHEMISTRY

VOLUME 286 • NUMBER 21 • MAY 27, 2011

Subnuclear Neighborhoods

Acknowledgment—We thank Dr. Ken Young for assisting with Wright-Giemsa staining.

REFERENCES 1. Takizawa, T., Meaburn, K. J., and Misteli, T. (2008) Cell 135, 9 –13 2. Zhao, R., Bodnar, M. S., and Spector, D. L. (2009) Curr. Opin. Genet. Dev. 19, 172–179 3. Schoenfelder, S., Clay, I., and Fraser, P. (2010) Curr. Opin. Genet. Dev. 20, 127–133 4. Deniaud, E., and Bickmore, W. A. (2009) Curr. Opin. Genet. Dev. 19, 187–191 5. Lee, H. Y., Johnson, K. D., Fujiwara, T., Boyer, M. E., Kim, S. I., and Bresnick, E. H. (2009) Mol. Cell 36, 984 –995 6. Osborne, C. S., Chakalova, L., Brown, K. E., Carter, D., Horton, A., Debrand, E., Goyenechea, B., Mitchell, J. A., Lopes, S., Reik, W., and Fraser, P. (2004) Nat. Genet. 36, 1065–1071 7. Evans, T., and Felsenfeld, G. (1989) Cell 58, 877– 885 8. Tsai, S. F., Martin, D. I., Zon, L. I., D’Andrea, A. D., Wong, G. G., and Orkin, S. H. (1989) Nature 339, 446 – 451

MAY 27, 2011 • VOLUME 286 • NUMBER 21

9. Tsang, A. P., Visvader, J. E., Turner, C. A., Fujiwara, Y., Yu, C., Weiss, M. J., Crossley, M., and Orkin, S. H. (1997) Cell 90, 109 –119 10. Crispino, J. D., Lodish, M. B., MacKay, J. P., and Orkin, S. H. (1999) Mol. Cell 3, 219 –228 11. Johnson, K. D., Boyer, M. E., Kang, J. A., Wickrema, A., Cantor, A. B., and Bresnick, E. H. (2007) Blood 109, 5230 –5233 12. Pal, S., Cantor, A. B., Johnson, K. D., Moran, T. B., Boyer, M. E., Orkin, S. H., and Bresnick, E. H. (2004) Proc. Natl. Acad. Sci. U.S.A. 101, 980 –985 13. Letting, D. L., Chen, Y. Y., Rakowski, C., Reedy, S., and Blobel, G. A. (2004) Proc. Natl. Acad. Sci. U.S.A. 101, 476 – 481 14. Vakoc, C. R., Letting, D. L., Gheldof, N., Sawado, T., Bender, M. A., Groudine, M., Weiss, M. J., Dekker, J., and Blobel, G. A. (2005) Mol. Cell 17, 453– 462 15. Hong, W., Nakazawa, M., Chen, Y. Y., Kori, R., Vakoc, C. R., Rakowski, C., and Blobel, G. A. (2005) EMBO J. 24, 2367–2378 16. Denslow, S. A., and Wade, P. A. (2007) Oncogene 26, 5433–5438 17. Miccio, A., Wang, Y., Hong, W., Gregory, G. D., Wang, H., Yu, X., Choi, J. K., Shelat, S., Tong, W., Poncz, M., and Blobel, G. A. (2010) EMBO J. 29, 442– 456 18. Volpi, E. V., Chevret, E., Jones, T., Vatcheva, R., Williamson, J., Beck, S., Campbell, R. D., Goldsworthy, M., Powis, S. H., Ragoussis, J., Trowsdale, J., and Sheer, D. (2000) J. Cell Sci. 113, 1565–1576 19. Kosak, S. T., Skok, J. A., Medina, K. L., Riblet, R., Le Beau, M. M., Fisher, A. G., and Singh, H. (2002) Science 296, 158 –162 20. Chambeyron, S., and Bickmore, W. A. (2004) Genes Dev. 18, 1119 –1130 21. Brown, K. E., Baxter, J., Graf, D., Merkenschlager, M., and Fisher, A. G. (1999) Mol. Cell 3, 207–217 22. Ragoczy, T., Bender, M. A., Telling, A., Byron, R., and Groudine, M. (2006) Genes Dev. 20, 1447–1457 23. Bresnick, E. H., Johnson, K. D., Kim, S. I., and Im, H. (2006) Prog. Nucleic Acids Res. Mol. Biol. 81, 435– 471 24. Dean, A. (2006) Trends Genet. 22, 38 – 45 25. Weiss, M. J., Yu, C., and Orkin, S. H. (1997) Mol. Cell. Biol. 17, 1642–1651 26. Welch, J. J., Watts, J. A., Vakoc, C. R., Yao, Y., Wang, H., Hardison, R. C., Blobel, G. A., Chodosh, L. A., and Weiss, M. J. (2004) Blood 104, 3136 –3147 27. Grass, J. A., Boyer, M. E., Pal, S., Wu, J., Weiss, M. J., and Bresnick, E. H. (2003) Proc. Natl. Acad. Sci. U.S.A. 100, 8811– 8816 28. Kim, S. I., and Bresnick, E. H. (2007) Oncogene 26, 6777– 6794 29. Bresnick, E. H., Lee, H. Y., Fujiwara, T., Johnson, K. D., and Keles, S. (2010) J. Biol. Chem. 285, 31087–31093 30. Fujiwara, T., O’Geen, H., Keles, S., Blahnik, K., Linnemann, A. K., Kang, Y. A., Choi, K., Farnham, P. J., and Bresnick, E. H. (2009) Mol. Cell 36, 667– 681 31. Larson, D. R., Singer, R. H., and Zenklusen, D. (2009) Trends Cell Biol. 19, 630 – 637 32. Raj, A., and van Oudenaarden, A. (2008) Cell 135, 216 –226 33. Pevny, L., Simon, M. C., Robertson, E., Klein, W. H., Tsai, S. F., D’Agati, V., Orkin, S. H., and Costantini, F. (1991) Nature 349, 257–260 34. Simon, M. C., Pevny, L., Wiles, M. V., Keller, G., Costantini, F., and Orkin, S. H. (1992) Nat. Genet. 1, 92–98 35. Weiss, M. J., Keller, G., and Orkin, S. H. (1994) Genes Dev. 8, 1184 –1197 36. Kim, S. I., Bultman, S. J., Jing, H., Blobel, G. A., and Bresnick, E. H. (2007) Mol. Cell. Biol. 27, 4551– 4565 37. Kim, S. I., Bultman, S. J., Kiefer, C. M., Dean, A., and Bresnick, E. H. (2009) Proc. Natl. Acad. Sci. U.S.A. 106, 2259 –2264 38. Dekker, J., Rippe, K., Dekker, M., and Kleckner, N. (2002) Science 295, 1306 –1311 39. Gao, Z., Huang, Z., Olivey, H. E., Gurbuxani, S., Crispino, J. D., and Svensson, E. C. (2010) EMBO J. 29, 457– 468 40. Miccio, A., and Blobel, G. A. (2010) Mol. Cell. Biol. 30, 3460 –3470 41. Miller, I. J., and Bieker, J. J. (1993) Mol. Cell. Biol. 13, 2776 –2786 42. Perkins, A. C., Sharpe, A. H., and Orkin, S. H. (1995) Nature 375, 318 –322 43. Im, H., Grass, J. A., Johnson, K. D., Kim, S. I., Boyer, M. E., Imbalzano, A. N., Bieker, J. J., and Bresnick, E. H. (2005) Proc. Natl. Acad. Sci. U.S.A.

JOURNAL OF BIOLOGICAL CHEMISTRY

18843

Downloaded from http://www.jbc.org/ at Massachusetts Institute of Technology on November 5, 2014

is unlikely that relocalization reflects gross morphological remodeling as a consequence of maturation. It is attractive to propose that the GATA-1-dependent chromosomal transition reflects a direct action of GATA-1 and its interactors to establish a specific subnuclear position as a pivotal step in assembling a cell type-specific genetic network. Because our prior work established that GATA-1 expels multiple FOG-1-dependent, but not -independent, genes from the nuclear periphery (5), the mechanistic insights derived from this study are likely to be broadly applicable to a large cohort of GATA-1 target genes. By contrast to the proposed functional importance of locus subnuclear positioning in transcriptional activation, abrogation of maintenance did not revert locus positioning to the nuclear periphery. Thus, once a locus becomes integrated into a new subnuclear compartment, transcriptional activity can be modulated without shuttling it back and forth between distinct subnuclear compartments (Fig. 6E), at least given the limits of resolution of the specific microscopy analysis conducted. In principle, there might be instances in which dynamic locus contact with specific subnuclear structures is crucial for maintenance, but we are unaware of examples in which the selective control of maintenance involves ensuring a specific subnuclear positioning, and our results are inconsistent with this potential mechanism, as least in the context of GATA factor mechanisms. Accordingly, overt subnuclear transitions might be instrumental for establishing cell type-specific genetic networks during the development of specific cell and tissue types, whereas such transitions might not be commonly required to remodel a preexisting network. Given the mechanistic insights described herein, it will be particularly important to devise novel strategies to further interrogate interconnections between GATA-1-driven locus repositioning and specific molecular steps intrinsic to the mechanism by which GATA-1 activates and represses specific target genes. Furthermore, because GATA-1 mutations can be leukemogenic (48, 49) and differentially dysregulate target genes (50), whether such mutants are impaired in their capacities to drive subnuclear transitions at select loci is of considerable interest.

Subnuclear Neighborhoods 102, 17065–17070 44. Tallack, M. R., Whitington, T., Yuen, W. S., Wainwright, E. N., Keys, J. R., Gardiner, B. B., Nourbakhsh, E., Cloonan, N., Grimmond, S. M., Bailey, T. L., and Perkins, A. C. (2010) Genome Res. 20, 1052–1063 45. Merika, M., and Orkin, S. H. (1995) Mol. Cell. Biol. 15, 2437–2447 46. Schoenfelder, S., Sexton, T., Chakalova, L., Cope, N. F., Horton, A., Andrews, S., Kurukuti, S., Mitchell, J. A., Umlauf, D., Dimitrova, D. S., Eskiw, C. H., Luo, Y., Wei, C. L., Ruan, Y., Bieker, J. J., and Fraser, P. (2010) Nat.

Genet. 42, 53– 61 47. Pope, N. J., and Bresnick, E. H. (2010) Nucleic Acids Res. 38, 2190 –2200 48. Wechsler, J., Greene, M., McDevitt, M. A., Anastasi, J., Karp, J. E., Le Beau, M. M., and Crispino, J. D. (2002) Nat. Genet. 32, 148 –152 49. Shimizu, R., Engel, J. D., and Yamamoto, M. (2008) Nat. Rev. Cancer 8, 279 –287 50. Johnson, K. D., Kim, S. I., and Bresnick, E. H. (2006) Proc. Natl. Acad. Sci. U.S.A. 103, 15939 –15944

Downloaded from http://www.jbc.org/ at Massachusetts Institute of Technology on November 5, 2014

18844 JOURNAL OF BIOLOGICAL CHEMISTRY

VOLUME 286 • NUMBER 21 • MAY 27, 2011