This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Schaufele, F. Articles by Day, R. N. Search for Related Content PubMed PubMed Citation Articles by Schaufele, F. Articles by Day, R. N. Pubmed/NCBI databases Substance via MeSH

CCAAT/Enhancer Binding Protein Assembles Essential Cooperating Factors in Common Subnuclear Domains

Metabolic Research Unit and Department of Medicine (F.S., X.W., C.T., R.S.), University of California, San Francisco, California 94143-0540; Departments of Medicine and Cell Biology (J.F.E., R.N.D.), National Science Foundation Center for Biological Timing, University of Virginia Health Sciences Center, Charlottesville, Virginia 22908; and Department of Physiology (R.E., O.A.M.), University of Michigan Medical School, Ann Arbor, Michigan 48109

Address all correspondence and requests for reprints to: Fred Schaufele, University of California, San Francisco, California 94143-0540. E-mail: freds{at}metabolic.ucsf.edu

	ABSTRACT

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
The transcription factor CCAAT/enhancer binding protein (C/EBP) is the DNA binding subunit of a multiprotein complex that regulates the pituitary-specific GH promoter. C/EBP is absent from the GHFT1–5 pituitary progenitor cell line in which ectopic C/EBP expression leads to activation of the otherwise dormant GH promoter. Transcriptional regulatory complexes are commonly envisaged as assembling from components that evenly diffuse throughout the nucleoplasm. We show that C/EBP, expressed in GHFT1–5 cells as a fusion with color variants of the green fluorescent protein (GFP), concentrated specifically at peri-centromeric chromosomal domains. Although we found the CREB-binding protein (CBP) to activate C/EBP-dependent transcription, CBP was absent from the pericentromeric chromatin. C/EBP expression was accompanied by the translocation of endogenous and ectopically expressed CBP to pericentromeric chromatin. The intranuclear recruitment of CBP required the transcriptional activation domains of C/EBP. C/EBP also caused GFP-tagged TATA binding protein (TBP) to relocate to the Hoechst-stained domains. The altered intranuclear distribution of critical coregulatory factors defines complexes formed upon C/EBP expression. It also identifies an organizational activity, which we label "intranuclear marshaling," that may regulate gene expression by determining the cooperative and antagonistic interactions available at specific nuclear sites.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
THE FAMILY OF CCAAT/enhancer binding protein (C/EBP) transcription factors are key regulators of cellular differentiation, and function in the control of many genes involved in energy metabolism (1, 2, 3). For example, C/EBP is required for both adipose and liver cell terminal differentiation. In these tissues, C/EBP controls the synthesis of proteins that are necessary for lipid metabolism and gluconeogenesis (4, 5, 6, 7). Gene knockout and gene transfer experiments have linked C/EBP to the control of developmental programs for a number of other organ systems, including hematopoietic cells (8, 9), lung (5), and the ovary (10).

GH is one of the most powerful regulators of energy metabolism. Our previous studies have implicated C/EBP as an activator of pituitary-specific GH gene expression (11, 12). C/EBP is present in GH-secreting pituitary cell lines, but absent from immortalized pituitary GHFT1–5 cells, which do not express GH. The GHFT1–5 cell line was derived by targeted transformation of embryonic pituitary cells and has characteristics of the progenitor for the GH-secreting, pituitary somatotrope cell lineage (13). Expression of exogenous C/EBP in GHFT1–5 cells leads to activation of a cotransfected GH gene promoter (12) and blockage of proliferation (Liu, W., W. Hyun, R. N. Day, and F. Schaufele, submitted). This suggested that C/EBP might play a role in somatotrope cell differentiation, analogous to its role in the regulation of gene expression and proliferation during adipocyte cell differentiation (7, 14).

Recently, it was shown that, during adipocyte cell differentiation, C/EBP became localized to specific regions of the cell nucleus that stained preferentially with DNA binding dyes that associated with markers for centromeres (15). Here, we demonstrate that C/EBP, when expressed as a fusion protein with GFP, also localizes to intranuclear sites associated with pericentromeric chromatin in pituitary progenitor GHFT1–5 cells. We extend these observations to demonstrate that the CREB binding protein (CBP), which we show to enhance C/EBP gene regulatory activity in GHFT1–5 cells, does not localize to pericentromeric chromatin in these pituitary cells. The paradox of differing intranuclear locations for cooperating C/EBP and CBP was resolved by finding that C/EBP expression caused CBP to translocate to the pericentromeric chromatin and colocalize with C/EBP. Similarly, the basal factor TATA-binding protein (TBP) was recruited to these intranuclear domains upon GFP-C/EBP expression. C/EBP truncated of its transcriptional activation functions still targeted to the Hoechst-stained chromosomal domains, but was incapable of reorganizing either CBP or TBP to these nuclear domains. Thus, C/EBP regulates the spatial positions of critical coregulatory factors within the nucleus. This alteration in the concentration of specific regulatory complexes at particular subnuclear structures may constitute a new means by which a transcription factor directs changes in patterns of gene expression.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
C/EBP-Dependent Transcription in Pituitary GHFT1–5 Cells
The GH gene is inactive in GHFT1–5 pituitary progenitor cells (12). Western blot analysis demonstrated that, in contrast to the GH-secreting pituitary GC cell line, GHFT1–5 cells do not express C/EBP (12) (Fig. 1A). When C/EBP was expressed in GHFT1–5 cells, transcription from the otherwise silent cotransfected rat GH promoter was induced (12). GH gene expression also is regulated by the pituitary-specific transcription factor Pit-1 (16, 17, 18), and GHFT1–5 cells express Pit-1 at a level much reduced when compared with pituitary GC cells (Fig. 1A). However, the expression of exogenous Pit-1 alone in GHFT1–5 cells does not lead to GH promoter activation (12).

View larger version (31K): [in this window] [in a new window]   Figure 1. GHFT1–5 Cells Are Deficient in C/EBP, Which When Expressed, Activates a Cotransfected Reporter Gene A, Western blot analysis of proteins prepared from nuclear extracts of GHFT1–5 cells or GH-secreting pituitary GC cells probed with antibodies that specifically recognized the indicated proteins. B, C/EBP activation of a minimal promoter containing the GH TATA box (-33 to +8) and the GH C/EBP binding site (C/EBP-TATA promoter) driving expression of the chloramphenicol acetyltransferase gene was dependent upon the dose of C/EBP expression vector. Transcriptional activation was abrogated by deletion of the amino-terminal 153-amino acid residues of C/EBP (C/EBP154). Data from three independent experiments were normalized to the activity of the GH promoter in the presence of 5 µg of C/EBP expression vector (100%) and plotted as the mean ± SD. Inset, Western blots of nuclear extracts prepared from the same transfected cells were stained with an antibody against C/EBP; appropriately sized C/EBP and C/EBP154 were expressed at similar levels in the transfected GHFT1–5 cells.

CBP Enhancement of C/EBP-Dependent Transcription
The CREB-binding protein (CBP) is a coactivator of Pit-1 (23) and some C/EBP family members other than C/EBP (24, 25). The CBP-related protein, p300, physically interacts with C/EBPß (24) and functionally interacts with both C/EBP (22) and C/EBPß (24). We found CBP to be present in nuclear extracts of GHFT1–5 cells at a level comparable to that in the GH-secreting GC cells (Fig. 1A). Adenovirus 12S E1a is an effective inhibitor of CBP coactivator function (26, 27). To investigate the potential role of CBP as a coactivator for C/EBP-dependent transcription from the GH promoter, we initially determined the effect of E1a coexpression on activation of the full-length (-237 to +8) rGH promoter and of the C/EBP-TATA promoter. Coexpression of the E1a protein blocked C/EBP-dependent transcription of the full-length rGH promoter (Fig. 2A) and the minimal C/EBP-TATA promoter (data not shown). Western blots confirmed that E1a expression did not affect the expression from the cotransfected C/EBP vector (inset, Fig. 2A). This result suggested that an E1a-sensitive coactivator, such as endogenous CBP/p300, enhanced C/EBP transcriptional activity.

View larger version (24K): [in this window] [in a new window]   Figure 2. CBP Enhances Transcriptional Activation by C/EBP in GHFT1–5 Cells A, Transfection of 5 µg of an expression plasmid for Adenovirus 12S E1a inhibited activation of the -237/+8 rat GH promoter by coexpressed, FLAG epitope-tagged C/EBP. Inset, Western blots of nuclear extracts prepared from the transfected cells showed that E1a expression had no effect on the amount of expressed C/EBP. B, Cotransfection of 10 µg of CBP expression vector (+CBP) enhanced the dose-dependent activation of the GH promoter by FLAG-tagged C/EBP. Cells were transfected with 0, 2.5, 5, or 10 µg of a FLAG epitope-tagged C/EBP expression vector. Inset, Western blots of nuclear extracts prepared from the same transfected cells were stained with an anti-FLAG antibody: C/EBP expression was not affected by CBP expression. Data from four (panel A) or five (panel B) independent experiments were normalized to the activity of the GH promoter in the presence of 5 µg of C/EBP expression vector and in the absence of E1a (100%), and plotted as the mean ± SD.

Transcriptionally Active Fusion of C/EBP with GFP
Using GFP as a label for the ER expressed in living cells, we (29) and others (30) recently demonstrated that expression of the ER dramatically affected the intranuclear organization of the coactivator proteins GRIP1 and SRC-1. Here, we studied whether C/EBP expression similarly affected the intranuclear redistribution of the coactivator CBP. Initially, C/EBP fusions with GFP were used to identify the intranuclear location of C/EBP in living cells. Expression vectors were constructed in which the cDNA for GFP was fused to either the amino terminus or the carboxy terminus of the cDNA for C/EBP (GFP-C/EBP or C/EBP-GFP, respectively). Western analysis showed that the expressed C/EBP-GFP and GFP-C/EBP were of the size expected for full-length GFP-C/EBP (Fig. 3). The C/EBP-GFP fusion was transcriptionally active at the C/EBP-TATA promoter in GHFT1–5 cells, whereas the GFP-C/EBP fusion was comparatively inactive (Fig. 3). Transfection of the C/EBP-GFP expression vector into GHFT1–5 cells resulted in a 12.01 ± 3.54 fold activation of the cotransfected C/EBP-TATA promoter, compared with a 2.28 ± 1.37 fold promoter activation by expression of GFP-C/EBP. On average, C/EBP-GFP was 42.00 ± 14.59% as effective in activating the C/EBP-TATA promoter as similarly expressed, unfused C/EBP in parallel experiments. Despite this transcriptional difference, the C/EBP-GFP and the GFP-C/EBP fusions, as well as ectopically expressed and antibody-stained C/EBP, all behaved similarly in the subsequent experiments described in this report.

View larger version (48K): [in this window] [in a new window]   Figure 3. An In-Frame Fusion of GFP to the Carboxy Terminus of C/EBP Is Transcriptionally Active Activation of the C/EBP-TATA promoter by the transfection of 10 µg C/EBP-GFP or GFP-C/EBP expression vector was compared with activation by 0, 1, 2, and 5 µg of C/EBP expression vector. Activity was compared with C/EBP expression levels determined by Western blots of nuclear extracts prepared from the same transfected cells that were stained for the FLAG epitope common to all three expression vectors. Whereas C/EBP-GFP was transcriptionally active at levels approaching that of unfused C/EBP, the GFP-C/EBP was comparatively inert. Data from three independent experiments were normalized to the activity of the C/EBP-TATA promoter in presence of 5 µg C/EBP expression vector and plotted as the mean ± SD.

GFP-C/EBP Concentrates at Discrete Intranuclear Structures

View larger version (59K): [in this window] [in a new window]   Figure 4. Targeting of GFP-C/EBP to A-T-rich Chromatin in Pituitary Progenitor GHFT1–5 Cells A, Cells transiently expressing GFP-C/EBP were stained, 20 min before imaging, with 0.5 µg/ml of the A-T-rich DNA binding dye H33342. Green fluorescence emitted by GFP was distinguished from the blue fluorescence emitted by H33342-stained chromatin using the filter sets described in Materials and Methods. Overlay of the green and blue images from the same cell showed the relative positions of GFP-C/EBP and H33342-stained chromatin with overlap appearing cyan in color. The scale bar indicates 10 µm. B, The intranuclear localization of GFP-C/EBP remained constant in cells expressing a 112-fold difference in GFP-C/EBP amount. Images of dim and bright cells on the same coverslip were acquired at comparable intensity levels by controlling the intensity of excitation and by varying the on-camera integration time. C, GFP-C/EBP deleted of all but amino acids 245–358, including the DNA-binding domain conserved in, and characteristic of, all bZIP transcription factors, still targeted to H33342-stained chromatin.

Targeting of C/EBP to Pericentromeric Chromatin
The H33342-stained foci have been previously described in other mouse cell-types as tracts of satellite DNA repeats located at centromeric regions of interphase chromosomes (31, 32, 33). We found that the H33342-stained chromatin was associated with the centromeres of interphase chromosomes in the GHFT1–5 cell nucleus (Fig. 5A). Nontransfected GHFT1–5 cells were fixed, and immunohistochemical staining was performed using a serum containing a human autoantibody that reacts with centromeric kinetochore proteins (34). The kinetochore signal was visualized with tetramethylrhodamine isothiocyanate-conjugated secondary antibody. Dual-color imaging of the fixed cells counterstained with H33342 showed a pair of kinetochores were typically associated with each stained chomatin focus. Thus, the H33342-stained chromatin surrounds the centromeres of the interphase GHFT1–5 cell nucleus.

View larger version (47K): [in this window] [in a new window]   Figure 5. The H33342-Stained DNA to Which GFP-C/EBP Localizes in GHFT1–5 Cells Consists of Transcriptionally Inactive Pericentromeric Chromatin A, GHFT1–5 cells grown on coverslips were methanol fixed and incubated with sera containing human autoantibodies against kinetochore proteins of the centromere. Antibody-stained structures were detected by incubation with an anti-human TRITC-conjugated secondary antibody. The slides were counterstained with H33342 and images collected using the indicated filter sets (see Materials and Methods). B, GHFT1–5 cells were permeabilized and incubated with Br-UTP, CTP, ATP, and GTP for 20 min. Cells were fixed and washed, and the bromouridine incorporated into RNA was detected with an antibromouracil antibody (see Materials and Methods). Br-UTP labeling (left panel) and H33342 (middle panel) were taken at the same focal plane. The overlay (right panel) shows that Br-UTP labeling did not coincide with pericentromeric chromatin.

It was previously speculated that the pericentromeric targeting of C/EBP in 3T3-L1 cells induced to differentiate into adipocytes was due to the presence of C/EBP binding sites within the repeated DNA sequences that comprise the bulk of pericentromeric chromatin (15). We examined whether the DNA binding domain of C/EBP was sufficient for targeting to pericentromeric chromatin in GHFT1–5 cells. A GFP-C/EBP fusion was constructed in which only amino acids 245–358 of C/EBP were retained. This encompassed the entire "bZIP" DNA binding domain located between amino acids 278–344 of C/EBP. By itself, this isolated bZIP region targeted specifically to H33342-stained DNA (Fig. 4C), as did a second fusion protein in which GFP was appended to the carboxy terminus of C/EBP amino acids 259–358 (data not shown). Moreover, C/EBP deleted of the leucine zipper component of the DNA binding domain no longer concentrated at the H33342-stained chromatin (Liu, W., W. Hyun, R. N. Day, and F. Schaufele, submitted). Thus, an intact DNA binding domain, which is critical for gene-specific transcriptional activation, is both sufficient and necessary for C/EBP targeting to pericentromeric chromatin.

The Intranuclear Distribution of CBP and C/EBP Are Distinct
Although DNA binding was sufficient for pericentromeric targeting, it was not sufficient for transcriptional activation. This implied that activities beyond DNA binding and/or pericentromeric targeting were required for C/EBP activity. We therefore examined the intranuclear position of CBP, which we had determined to enhance C/EBP activation (Fig. 2). In striking contrast to GFP-C/EBP, CBP expressed in GHFT1–5 cells as a fusion to GFP was distributed throughout the nucleus (Fig. 6A), similar to that previously shown in immunohistochemical staining of endogenous CBP in HEp-2 nuclei (37). Moreover, GFP-CBP was excluded from the pericentromeric chromatin preferentially labeled with H33342 (Fig. 6A, overlay). We then determined that endogenous CBP also was excluded from H33342-labeled pericentromeric chromatin; fixed GHFT1–5 cells were stained with a primary antibody directed against mouse CBP and a secondary antibody labeled with TRITC (Fig. 6B). Thus, the intranuclear localization of the expressed GFP-CBP accurately reflected the distribution of its endogenous counterpart, and both were absent from the pericentromeric sites to which GFP-C/EBP was preferentially localized.

View larger version (49K): [in this window] [in a new window]   Figure 6. CBP Does Not Associate with Pericentromeric Chromatin in GHFT1–5 Cells A, GHFT1–5 cells were transfected with an expression plasmid for GFP-CBP, and images of its subnuclear localization relative to chromatin stained with H33342 were acquired as described for Fig. 4. B, Endogenous GHFT1–5 cell CBP was detected by immunohistochemical staining of fixed cells using an anti-CBP antibody, followed by detection using the tyramide signal amplification technique (see Materials and Methods). H33342-stained DNA was obtained from each cell at the same focal plane using the blue filter set. The overlay shows the merged images from the same focal plane of the same cell and indicated that the H33342-stained chromatin occupies domains from which both GFP-CBP and endogenous CBP were excluded. The scale bar represents 10 µm.

C/EBP Expression Results in Coincident Intranuclear Positioning of C/EBP and CBP

To determine whether C/EBP expression had an effect on the intranuclear position of critical coregulatory factors including CBP, we first tagged C/EBP with the spectrally distinct blue color variant of GFP (BFP) (39, 40, 41). We then specifically detected the intranuclear positions of BFP-tagged C/EBP and GFP-tagged CBP expressed in the same cell by using BFP- and GFP-specific excitation and emission filter sets (see Materials and Methods). When expressed in GHFT1–5 cells, BFP-C/EBP (Fig. 7A, top left panel) and C/EBP-BFP (not shown) assumed the same distinctive intranuclear distribution of GFP-C/EBP described above. This was confirmed by coexpressing BFP-C/EBP and GFP-C/EBP in the same cells and observing that their intranuclear distributions overlapped (data not shown).

View larger version (36K): [in this window] [in a new window]   Figure 7. Recruitment of CBP and TBP to the Intranuclear Location of C/EBP Depends upon C/EBP Transactivation Domains A, Expression of BFP-C/EBP caused the redistribution of GFP-CBP to the intranuclear sites occupied by C/EBP (A, upper panels). The intranuclear localization of a BFP-fusion to the mutant C/EBP (BFP-C/EBP154) was identical to wild-type C/EBP (A, left panels), but the mutant protein no longer recruited GFP-CBP (A, lower panels). B, GHFT1–5 cells transfected with GFP-C/EBP (B, upper panels) or GFP-C/EBP154 (B, lower panels) were fixed and stained with anti-CBP antibody. The overlay shows the merged green and red images at the same focal plane and indicates overlapping regions of distribution as yellow. C, C/EBP influences the intranuclear distribution of TBP. Coexpression of GFP-TBP with BFP-C/EBP resulted in accumulation of TBP at the sites occupied by BFP-C/EBP (C, upper panels). In contrast, no accumulation of GFP-TBP was observed at the nuclear sites occupied by the deletion mutant, BFP-C/EBP154 (C, lower panels). The distribution of GFP-TBP, when expressed with BFP-C/EBP154, is the same as when GFP-TBP is expressed alone (not shown).

Expression of GFP-C/EBP also altered the subnuclear localization of the endogenous CBP protein. GHFT1–5 cells expressing either GFP-C/EBP (Fig. 7B, upper panels) or GFP-C/EBP154 (Fig. 7B, lower panels) were fixed and stained using an antibody directed against CBP. Dual-color imaging showed the antibody-labeled endogenous CBP colocalized with the full-length GFP-C/EBP. Endogenous CBP did not localize to these subnuclear sites in cells expressing the mutant GFP-C/EBP154 protein. Together, these results showed that the expression of exogenous C/EBP in GHFT1–5 cells caused a trans-activation domain-dependent recruitment of CBP to specific subnuclear sites.

The failure of the transcriptionally inactive 154 mutant of C/EBP to recruit CBP may suggest a role for CBP recruitment in transcriptional activation. Indeed, the expression of BFP-C/EBP is also associated with an enhanced concentration of a GFP fusion with TBP at the location of BFP-C/EBP (Fig. 7C, upper panels). The concentration of GFP-TBP was not seen with the transcriptionally inactive BFP-C/EBP154 (Fig. 7C, lower panels). However, the sites of active transcription in GHFT1–5 cell nuclei, detected by Br-UTP labeling (see Fig. 5B), were as absent from pericentromeric chromatin after C/EBP-GFP expression as they were in the absence of C/EBP (data not shown). Because Br-UTP labeling of nascent transcripts measures global transcription rather than C/EBP-regulated transcription, the transcriptional consequences of the change in the intranuclear distribution of CBP and TBP upon C/EBP expression may require mapping the intranuclear locations of transcripts, specifically activated or repressed upon C/EBP expression, relative to the locations of C/EBP and pericentromeric chromatin.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
C/EBP Alters the Intranuclear Compartmentalization of Transcription Coregulatory Factors
We demonstrated that C/EBP and CBP cooperate to activate transcription in pituitary GHFT1–5 cells (Fig. 2) and that C/EBP expression is accompanied by a reorganization of CBP in distinct chromosomal domains within the nucleus ( Figs. 4–7). When tagged with GFP, C/EBP localized in preadipocyte 3T3-L1 cells (our unpublished data) and in pituitary progenitor GHFT1–5 pituitary cells (Fig. 4) to a subnuclear domain preferentially stained by the A/T-rich DNA binding dye H33342. This chromosomal domain was associated with a marker for interphase nucleus centromeres (Fig. 5). The pericentromeric localization of GFP-C/EBP in both cell types was identical to the location of endogenous C/EBP induced during differentiation of mouse 3T3-L1 preadipocytes into mature adipocyte cells (15 and our unpublished data). The localization of GFP-C/EBP to 3T3-L1 cell pericentromeric chromatin in the absence of the chemical induction of differentiation suggests that intranuclear targeting is an intrinsic property of C/EBP rather than an event induced by differentiation.

The GFP-C/EBP and C/EBP-GFP fusions, which differed in their ability to activate transcription of a C/EBP-sensitive reporter gene, and the transcriptionally defective C/EBP154 mutant all localized specifically at pericentromeric chromatin. This indicated that pericentromeric targeting of C/EBP was not sufficient for transcriptional activation. Indeed, the DNA binding domain of C/EBP by itself was sufficient for pericentromeric targeting (Fig. 4C) but not transcriptional activation (22 and our unpublished data). In contrast, the disruption of CBP and TBP recruitment to the intranuclear location of C/EBP (Fig. 7) by the transcriptionally defective C/EBP154 mutant suggested that the C/EBP-mediated, intranuclear relocation of CBP may be associated with transcriptional activation. However, C/EBP expression is not associated with a global enhancement of nascent, Br-UTP-labeled transcripts at pericentromeric chromatin (Fig. 5B and data not shown) although the low abundance of genes within centromeric DNA (42) may have precluded our ability to detect transcription activation by a global labeling of transcripts. At a minimum, the data strongly suggest that C/EBP organizes CBP and TBP into macromolecular complexes that are readily visible because of the distinct intranuclear localization pattern of C/EBP. It is not yet known whether CBP and TBP recruitment are linked or are separate, unrelated consequences of C/EBP expression.

Highly Specific, Intranuclear Marshalling of CBP by C/EBP
We refer to the alteration in the intranuclear location of transcription coregulatory factors induced by a transcription factor as intranuclear marshalling (29). The specificity of the intranuclear marshalling of CBP and TBP by C/EBP was illustrated by investigating the consequences of C/EBP expression on the intranuclear locations of a number of other transcription factors and cofactors (data not shown). Most of these factors did not localize at pericentromeric chromatin, and their intranuclear positions were not affected by C/EBP expression. For instance, the coactivator GRIP1, which distributes throughout the nucleus (29) and, like CBP, contains histone acetyltransferase activity, was not affected by the expression of BFP-C/EBP (data not shown). However, GRIP1 was recruited to the subnuclear location occupied by the ER, but only if the cells were treated with estrogens (29). The ER did not recruit CBP to its intranuclear location, and the intranuclear distribution of the ER was not affected by C/EBP coexpression. Similarly, GFP fusions to the basal factor TFIIB or the Sin3A component of some histone deacetylase complexes distributed independently of coexpressed C/EBP in GHFT1–5 cells (not shown). The C/EBP- and ER-induced sequestrations of different histone acetyltransferase-containing factors to different regions of the cell nucleus may dramatically affect the balance of acetylation activities at discrete locations within the nucleus. Indeed, we have determined that the expression of C/EBP is associated with an increase in the amount of acetylated histone H3 present in pericentromeric chromatin relative to the amount of acetylated histone H3 outside of pericentromeric chromatin (our unpublished data).

Despite highly specific marshaling of CBP to pericentromeric chromatin by C/EBP, in vitro studies of CBP interactions with column-attached C/EBP, coimmunoprecipitation studies in cellular extracts, and fluorescence resonance energy transfer studies in living cells have, to date, failed to reliably detect any evidence of a strong physical interaction between CBP and C/EBP. The intracellular complexes detected by intranuclear marshalling may therefore reflect an association of CBP and C/EBP involving other factors within the complex. We found that GRIP1, which is known to interact with CBP (43, 44), bound in vitro to column-attached C/EBP (our unpublished data). However, we saw no evidence of transcriptional coactivation or intranuclear marshaling by C/EBP and GRIP1. Thus, the intranuclear marshaling of CBP by C/EBP correlated better with the observed functional interactions of C/EBP than did in vitro interaction assays. This may be because intranuclear marshalling and functional studies are conducted under the same cellular environments.

Pericentromeric Organization and Regulation of Gene Expression
Given the potential contribution of nuclear architecture to gene expression, there have been very few studies of the spatial organization of transcription-regulatory factors within the nucleus (29, 30, 36, 45, 46, 47, 48, 49). Specific intranuclear locations for transcription factors and coregulatory factors may allow productive interactions only between colocalized transcription-regulatory factors and gene sets. Perhaps as important, the formation of complexes between factors, cofactors, and genes sequestered in different compartments may be restricted. The role that C/EBP association with pericentromeric chromatin may play in any of the differentiative effects of C/EBP remains to be defined. However, pericentromeric targeting of C/EBP in GHFT1–5 cells required the bZIP domain of C/EBP (our unpublished data) essential for C/EBP dimerization and DNA binding, suggesting that pericentromeric targeting is associated with at least one activity important to transcriptional regulation.

We have found that expression of the transcription factor Pit-1, an important coregulator of pituitary differentiation (17), leads to a highly selective marshaling of C/EBP, and associated CBP, away from pericentromeric chromatin in GHFT1–5 cells (Enwright III, J. F., M. Kawecki, F. Schaufele, and R. N. Day, submitted). Thus, C/EBP targeting to pericentromeric chromatin may be an intermediate step in differentiation of GH-secreting cell types. In contrast, C/EBP remains targeted to the pericentromeric chromatin in differentiated adipocytes. The different final locations of C/EBP and associated CBP relative to pericentromeric chromatin may contribute to the cell-specific differences in the complement of genes expressed in these two different cell types. It will be important to identify the genes differentially expressed or repressed in both cell types and to compare the activity and locations of those genes relative to pericentromeric chromatin.

Historically, pericentromeric chromatin has been viewed as being devoid of expressed genes. More recent evidence suggests that these regions are actively involved in gene regulation (35). A few genes are even embedded within the centromeric DNA of Arabidopsis thaliana, the multicellular organism for which genome sequencing is most complete in the centromeric regions (42). This shows that the centromeres are not completely transcriptionally inert. Some centromere-associated factors, such as the zinc-finger protein Ikaros/Lyf-1, may play a role in silencing particular genes during lymphocyte activation (50, 51). The centromere also facilitates the initiation of chromatin condensation and decondensation and positions chromosome territories within the interphase nucleus (52) and may therefore play a structural role in both gene activation and repression. The transcriptional regulator ATRX, in association with the chromatin-binding protein HP1, interacts with the SWI/SNF complex at pericentromeric chromatin (53). The Polycomb group complex that, like the SWI/SNF complex, regulates higher order chromatin structure in Drosophila (54) also associates with centromeric chromatin in human cell lines (55). Thus, the centromere may be an important nexus at the interface of intranuclear architecture and gene regulation. The C/EBP-induced concentration of specific coregulatory factors at the centromere, or away from the centromere in the presence of Pit-1, may provide a molecular and cellular basis for regulating the transcriptional regulatory complexes available to these sites.

C/EBP Assembles Nucleoprotein Complexes
The intranuclear marshalling of CBP and TBP by the C/EBP activation domain (Fig. 7) is consistent with the results reported by others that the amino-terminal region of some C/EBP family members is involved in interactions with TBP and CBP (20, 24). Our finding that critical factors including CBP and TBP did not localize to pericentromeric chromatin unless C/EBP was coexpressed suggests that, although stable, these assemblies are not permanent structures (36, 45). This supports the view that certain architectural proteins can nucleate the assembly of transcription-coregulatory complexes within the nucleus (46, 56, 57, 58, 59).

Beyond simple recruitment, it is intriguing to speculate that the marshaling of CBP and TBP specifically to pericentromeric sites could globally influence gene expression by permitting CBP and TBP access to factors and genes that target to pericentromeric DNA. Alternatively, CBP sequestration at pericentromeric chromatin may restrict CBP access to factors and genes present in nonpericentromeric locations. At a minimum, the marshaling activity of C/EBP demonstrates that C/EBP promotes the assembly of specific multiprotein complexes, and it is conceivable that the relocation of these complexes to specific chromatin compartments may dramatically affect the cohort of genes expressed in a cell. Thus, the recruitment of CBP by C/EBP to pericentromeric regions in pituitary presomatotrope and in preadipocyte cells might reflect a general mechanism by which the cell controls the progression of specific programs of gene expression. Together, these results demonstrate that specific protein domains play critical roles in the assembly of cooperating factors at certain subnuclear sites. The remodeling of nuclear structure and organization are likely to be key components in the flow of regulatory information controlling cell type-specific gene expression in response to environmental cues or developmental programs.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
Construction of Expression Vectors and Transfection of GHFT1–5 Cells
The sequences encoding C/EBP, CBP, and TBP were fused in frame to either the GFP S65T or BFP Y66H, Y145F (60) mutant variants in vectors described previously (41). For transfection, GHFT1–5 cells were maintained as a monolayer of DMEM containing 10% FCS. 3T3-L1 cells were maintained and differentiated as described previously (22). Transfected GHFT1–5 cells were transfected with the indicated plasmid DNA(s) by electroporation as described previously (12, 41). For the reporter gene experiments, the rGH promoter (-237 to +8) or a single copy of the GH gene promoter C/EBP element (-239 to-219) linked to the rGH gene TATA box (-33 to +8) were coupled to the bacterial chloramphenicol acetyl transferase reporter gene. Transfected cell extracts were prepared for determination of reporter gene activity or for Western blot analyses using antibodies directed against C/EBP, GFP, Pit-1, E1a, or FLAG as described previously (29, 41).

Unlike wild-type E1a, E1a containing the RG2 mutant, which specifically disrupts E1a interaction with CBP (26, 27), did not block C/EBP transcriptional activation (data not shown), suggesting that CBP was the target of E1a-mediated inhibition. However, Western blots using three different commercially available E1a antibodies showed that this mutant protein was not present in nuclear extracts prepared from transfected GHFT1–5 cells. An informal survey of the literature shows that most publications have not included controls for expression of the RG2 mutant. We caution against future interpretations based upon this commonly used reagent in the absence of this expression control.

Immunohistochemistry
Nontransfected mouse pituitary GHFT1–5 cells, mouse 3T3-L1 cells or GHFT1–5 cells transfected with the indicated expression vectors, were cultured on glass cover slips. Cells were maintained in culture 24–48 h, and then fixed by a 5-min incubation in cold methanol and processed for immunohistochemical detection. Cells expressing only GFP fusion proteins were not fixed and were viewed live. CBP was detected using the tyramide signal amplification technique (61). The fixed cells were incubated with an anti-CBP primary antibody (1:200 dilution of sc-369, Santa Cruz Biotechnology, Inc., Santa Cruz, CA), followed by a biotinylated antirabbit secondary antibody and tertiary step using the Vectastain ABC kit (Vector Laboratories, Inc., Burlingame, CA). The target was detected using a horseradish peroxidase-catalyzed reaction of tyramide. After the tyramide reaction, the fixed cells were washed and then stained for 5 min with H33342 at a concentration of 0.2 µg/ml, and the coverslips were subsequently mounted using Vectashield (Vector Laboratories, Inc., Burlingame, CA). Kinetochore proteins were detected by incubating the fixed cells with sera containing the human nuclear centromere autoantibody (1:250 dilution of ANA, Cortex Biochem, San Leandro, CA), followed by incubation with an antihuman TRITC-conjugated secondary antibody. Endogenous C/EBP was detected in fixed adipocytes by incubation with a rabbit polyclonal C/EBP primary antibody (1:100 dilution of sc-61, Santa Cruz Biotechnology, Inc.) followed by incubation with an antirabbit rhodamine-conjugated secondary antibody.

Labeling of nascent mRNA transcripts was performed as previously described (36) except cells were exposed to Br-UTP for 20 min. Briefly, cells that had been plated on cover glasses the previous day were permeabilized with saponin, incubated with an in vitro transcription buffer containing Br-UTP, CTP, GTP, and ATP for 20 min at 33 C, and then fixed in paraformaldehyde. After fixation, cells were washed, and incubated overnight at 4 C with antibromouracil antibody to detect the nascently transcribed mRNA. The next day cells were washed followed by detection with a Texas Red-conjugated secondary antibody. Cells were washed again and stained with H 33342 at a concentration of 0.2 µg/ml, and the coverslips were subsequently mounted using Vectashield (Vector Laboratories, Inc.).

Microscopy and Image Analysis
Pituitary GHFT1–5 cells were typically transfected with 3–10 µg of expression plasmid DNA encoding the GFP-fusion proteins. The transfected cells were inoculated into culture dishes containing no. 1 borosilicate cover glasses. The cells were maintained in culture as described above, and then subjected to dual color fluorescence microscopy (39, 40, 41, 62). For experiments involving staining with H33342, the stain was added to a final concentration of 0.5 µg/ml approximately 20 min before imaging living cells or at 0.2 µg/ml for 5 min to image fixed cells. The fluorescence images were acquired with either an inverted IX-70 (Olympus Corp., Lake Success, NY) or Axioplan microscope (Carl Zeiss, Thornwood, NY) equipped with a 60x aqueous-immersion or a 63x oil-immersion objective lens, respectively. The filter combinations were 485/22 nm excitation and 535/50 nm emission for GFP images; 365/15 nm excitation and 460/50 nm emission for H33342 or BFP images; and Texas Red or rhodamine filter sets for immunohisochemical staining (Chroma Technology Corp., Brattelboro, VT). Grayscale images with no saturated pixels were obtained using a cooled digital interline camera (Orca-200, Hamamatsu, Bridgewater, NJ). All images were collected at a similar gray-level intensity by controlling the excitation intensity using neutral density filtration, and by varying the on-camera integration time. For the result shown in Fig. 4B, the relative illumination energy was calculated as the product of integration time and excitation intensity, with 1 sec at 0.1 excitation equal to 1. ISEE software (Inovision Corp., Raleigh, NC) or Metamorph software (Universal Imaging Corp., Downingtown, PA) was used to background subtract and then convert the digital images to red-green-blue images. The GFP signal was assigned to the green channel, H33342 or BFP signals to the blue channel, and the TRITC or rhodamine signals to the red channel of the red-green-blue digital image. Image files were processed for presentation using Adobe Photoshop 5.5 or 6.0 (Adobe Systems, Inc., San Jose, CA).

	ACKNOWLEDGMENTS

 
We thank John D. Baxter for critical reading of the manuscript, Meg Kawecki and Phat Tran for expert assistance, and Ammasi Periasamy of the W.M. Keck Center for Cellular Imaging and Bill Hyun of the University of California San Francisco Cancer Center for microscopy advice. We also thank Dr. David Allis for helpful discussions.

	FOOTNOTES

 
This work was supported by NIH Grant DK-47301 and the National Science Foundation Center for Biological Timing to R.N.D., and by NIH Grant DK-54345, the American Cancer Society Grant RPG-94–028-TBE, and the University of California San Francisco Academic Senate Committee on Research to F.S.

Abbreviations: BFP, Blue fluorescent protein; Br-UTP, bromo-uridine triphosphate; CBP, CREB binding protein; C/EBP, CCAAT/enhancer binding protein; CREB, cAMP response element binding protein; GFP, green fluorescent protein; GRIP, GR-interacting protein; H33342, Hoechst 33342; TBP, TATA binding protein; TRITC, tetramethylrhodamine isothiocyanate.

Received for publication April 3, 2001. Accepted for publication June 25, 2001.

	REFERENCES

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 

1. Darlington GJ, Wang N, Hanson RW 1995 C/EBP : a critical regulator of genes governing integrative metabolic processes. Curr Opin Genet Dev 5:565–570[CrossRef][Medline]
2. McKnight SL, Lane MD, Gluecksohn-Waelsch S 1989 Is CCAAT/enhancer-binding protein a central regulator of energy metabolism? Genes Dev 3:2021–2024[Free Full Text]
3. Wedel A, Ziegler-Heitbrock HW 1995 The C/EBP family of transcription factors. Immunobiology 193:171–185[Medline]
4. Lin FT, Lane MD 1992 Antisense CCAAT/enhancer-binding protein RNA suppresses coordinate gene expression and triglyceride accumulation during differentiation of 3T3–L1 preadipocytes. Genes Dev 9:533–544
5. Flodby P, Barlow C, Kylefjord H, Ahrlund-Richter L, Xanthopoulos KG 1996 Increased hepatic cell proliferation and lung abnormalities in mice deficient in CCAAT/enhancer binding protein . J Biol Chem 271:24753–24760[Abstract/Free Full Text]
6. Croniger C, Trus M, Lysek-Stupp K, et al. 1997 Role of the isoforms of CCAAT/enhancer-binding protein in the initiation of phosphoenolpyruvate carboxykinase (GTP) gene transcription at birth. J Biol Chem 272:26306–26312[Abstract/Free Full Text]
7. Darlington GJ, Ross SE, MacDougald OA 1998 The role of C/EBP genes in adipocyte differentiation. J Biol Chem 273:30057–30060[Free Full Text]
8. Zhang DE, Zhang P, Wang ND, Hetherington CJ, Darlington GJ, Tenen DG 1997 Absence of granulocyte colony-stimulating factor signaling and neutrophil development in CCAAT enhancer binding protein -deficient mice. Proc Natl Acad Sci USA 94:569–574[Abstract/Free Full Text]
9. Radomska HS, Huettner CS, Zhang P, Cheng T, Scadden DT, Tenen DG 1998 CCAAT/enhancer binding protein is a regulatory switch sufficient for induction of granulocytic development from bipotential myeloid progenitors. Mol Cell Biol 18:4301–4314[Abstract/Free Full Text]
10. Piontkewitz Y, Enerback S, Hedin L 1996 Expression of CCAAT enhancer binding protein-alpha (C/EBP ) in the rat ovary: implications for follicular development and ovulation. Dev Biol 179:288–296[CrossRef][Medline]
11. Schaufele F, Cassill JA, West BL, Reudelhuber T 1990 Resolution by diagonal gel mobility shift assays of multisubunit complexes binding to a functionally important element of the rat growth hormone gene promoter. J Biol Chem 265:14592–14598[Abstract/Free Full Text]
12. Schaufele F 1996 CCAAT/enhancer-binding protein activation of the rat growth hormone promoter in pituitary progenitor GHFT1–5 cells. J Biol Chem 271:21484–21489[Abstract/Free Full Text]
13. Lew D, Brady H, Klausing K, et al. 1993 GHF-1-promoter-targeted immortalization of a somatotropic progenitor cell results in dwarfism in transgenic mice. Genes Dev 7:683–693[Abstract/Free Full Text]
14. Lin FT, Lane MD 1994 CCAAT/enhancer binding protein is sufficient to initiate the 3T3–L1 adipocyte differentiation program. Proc Natl Acad Sci USA 91:8757–8761[Abstract/Free Full Text]
15. Tang QQ, Lane MD 1999 Activation and centromeric localization of CCAAT/enhancer-binding proteins during the mitotic clonal expansion of adipocyte differentiation. Genes Dev 13:2231–2241[Abstract/Free Full Text]
16. Schaufele F 1994 Regulation of the growth hormone and prolactin genes. In: Imura H, ed. The pituitary gland. New York: Raven Press; 91–116
17. Rhodes SJ, DiMattia GE, Rosenfeld MG 1994 Transcriptional mechanisms in anterior pituitary cell differentiation. Curr Opin Genet Dev 4:709–717[CrossRef][Medline]
18. Li S, Crenshaw III EB, Rawson EJ, Simmons DM, Swanson LW, Rosenfeld MG 1990 Dwarf locus mutants lacking three pituitary cell types result from mutations in the POU-domain gene pit-1. Nature 347:528–33[CrossRef][Medline]
19. Pei DQ, Shih CH 1991 An "attenuator domain" is sandwiched by two distinct transactivation domains in the transcription factor C/EBP. Mol Cell Biol 11:1480–1487[Abstract/Free Full Text]
20. Nerlov C, Ziff EB 1995 CCAAT/enhancer binding protein- amino acid motifs with dual TBP and TFIIB binding ability co-operate to activate transcription in both yeast and mammalian cells. EMBO J 14:4318–4328[Medline]
21. Lee YH, Williams SC, Baer M, Sterneck E, Gonzalez FJ, Johnson PF 1997 The ability of C/EBPß but not C/EBP to synergize with an Sp1 protein is specified by the leucine zipper and activation domain. Mol Cell Biol 17:2038–2047[Abstract]
22. Erickson RL, Hemati N, Ross SE, MacDougald OA 2001 p300 coactivates the adipogenic transcription factor CCAAT/enhancer-binding protein. J Biol Chem 276:16348–16355[Abstract/Free Full Text]
23. Xu L, Lavinsky RM, Dasen JS, et al. 1998 Signal-specific co-activator domain requirements for Pit-1 activation. Nature 395:301–306[CrossRef][Medline]
24. Mink S, Haenig B, Klempnauer KH 1997 Interaction and functional collaboration of p300 and C/EBPß. Mol Cell Biol 17:6609–6617[Abstract]
25. Leahy P, Crawford DR, Grossman G, Gronostajski RM, Hanson RW 1999 CREB binding protein coordinates the function of multiple transcription factors including nuclear factor I to regulate phosphoenolpyruvate carboxykinase (GTP) gene transcription. J Biol Chem 274:8813–8822[Abstract/Free Full Text]
26. Lee JS, Galvin KM, See RH, et al. 1995 Relief of YY1 transcriptional repression by adenovirus E1A is mediated by E1A-associated protein p300. Genes Dev 9:1188–1198[Abstract/Free Full Text]
27. Lundblad JR, Kwok RP, Laurance ME, Harter ML, Goodman RH 1995 Adenoviral E1A-associated protein p300 as a functional homologue of the transcriptional co-activator CBP. Nature 374:85–88[CrossRef][Medline]
28. Chuang FM, West BL, Baxter JD, Schaufele F 1997 Activities in Pit-1 determine whether receptor interacting protein 140 activates or inhibits Pit-1/nuclear receptor transcriptional synergy. Mol Endocrinol 11:1332–1341[Abstract/Free Full Text]
29. Schaufele F, Chang C-Y, Liu W, et al. 2000 Temporally distinct and ligand-specific recruitment of nuclear receptor interacting peptides and co-factors to subnuclear domains containing the estrogen receptor. Mol Endocrinol 14:2024–2039[Abstract/Free Full Text]
30. Stenoien DL, Mancini MG, Patel K, Allegretto EA, Smith CL, Mancini MA 2000 Subnuclear trafficking of estrogen receptor- and steroid receptor coactivator-1. Mol Endocrinol 14:518–534[Abstract/Free Full Text]
31. Amirand C, Viari A, Ballini JP, et al. 1998 Three distinct sub-nuclear populations of HMG-I protein of different properties revealed by co-localization image analysis. J Cell Sci 3:551–561
32. Miller OJ, Schnedl W, Allen J, Erlanger BF 1974 5-Methylcytosine localized in mammalian constitutive heterochromatin. Nature 251:636–637[CrossRef][Medline]
33. Stephanova E, Russanova V, Chentsov Y, Pashev I 1988 Mouse centromeric heterochromatin: isolation and some characteristics. Exp Cell Res 179:545–553[CrossRef][Medline]
34. Abend M, Blakely WF, van Beuningen D 1995 Simplified and optimized kinetochore detection: cytogenetic marker for late-G2 cells. Mutat Res 334:39–47[Medline]
35. Gasser SM 2001 Positions of potential: nuclear organization and gene expression. Cell 104:639–642[CrossRef][Medline]
36. Pombo A, Cuello P, Schul W, et al. 1998 Regional and temporal specialization in the nucleus: a transcriptionally-active nuclear domain rich in PTF, Oct1 and PIKA antigens associates with specific chromosomes early in the cell cycle. EMBO J 17:1768–1778[CrossRef][Medline]
37. LaMorte VJ, Dyck JA, Ochs RL, Evans RM 1998 Localization of nascent RNA and CREB binding protein with the PML-containing nuclear body. Proc Natl Acad Sci USA 95:4991–4996[Abstract/Free Full Text]
38. Chrivia JC, Kwok RP, Lamb N, Hagiwara M, Montminy MR, Goodman RH 1993 Phosphorylated CREB binds specifically to the nuclear protein CBP. Nature 365:855–859[CrossRef][Medline]
39. Ellenberg J, Lippincott-Schwartz J, Presley JF 1998 Two-color green fluorescent protein time-lapse imaging. Biotechniques 25:838–846[Medline]
40. Ellenberg J, Lippincott-Schwartz J, Presley JF 1999 Dual-colour imaging with GFP variants. Trends Cell Biol 9:52–56[CrossRef][Medline]
41. Day RN 1998 Visualization of Pit-1 transcription factor interactions in the living cell nucleus by fluorescence resonance energy transfer microscopy. Mol Endocrinol 12:1410–1419[Abstract/Free Full Text]
42. Initiative TAG 2000 Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature 408:796–815[CrossRef][Medline]
43. Voegel JJ, Heine MJ, Tini M, Vivat V, Chambon P, Gronemeyer H 1998 The coactivator TIF2 contains three nuclear receptor-binding motifs and mediates transactivation through CBP binding-dependent and -independent pathways. EMBO J 17:507–519[CrossRef][Medline]
44. Ma H, Hong H, Huang SM, et al. 1999 Multiple signal input and output domains of the 160-kilodalton nuclear receptor coactivator proteins. Mol Cell Biol 19:6164–6173[Abstract/Free Full Text]
45. Cook PR 1999 The organization of replication and transcription. Science 284:1790–1795[Abstract/Free Full Text]
46. Zeng C, McNeil S, Pockwinse S, et al. 1998 Intranuclear targeting of AML/CBF regulatory factors to nuclear matrix-associated transcriptional domains. Proc Natl Acad Sci USA 95:1585–1589[Abstract/Free Full Text]
47. van Wijnen AJ, Bidwell JP, Fey EG, et al. 1993 Nuclear matrix association of multiple sequence-specific DNA binding activities related to SP-1, ATF, CCAAT, C/EBP, OCT-1, and AP-1. Biochemistry 32:8397–8402[CrossRef][Medline]
48. Nardozza TA, Quigley MM, Getzenberg RH 1996 Association of transcription factors with the nuclear matrix. J Cell Biochem 61:467–