HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 48, 37937-37944, December 1, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/48/37937    most recent M000989200v1 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 Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Shirakawa, H. Articles by Postnikov, Y. Search for Related Content PubMed PubMed Citation Articles by Shirakawa, H. Articles by Postnikov, Y. Social Bookmarking                      What's this?

Targeting of High Mobility Group-14/-17 Proteins in Chromatin Is Independent of DNA Sequence^*

Hitoshi Shirakawa, Julio E. Herrera, Michael Bustin, and Yuri Postnikov

From the Protein Section, Division of Basic Sciences, NCI, National Institutes of Health, Bethesda, Maryland 20892

Received for publication, February 7, 2000, and in revised form, August 31, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS REFERENCES

Chromosomal proteins high mobility group (HMG)-14 and HMG-17 are nucleosomal-binding proteins that unfold the chromatin fiber and enhance transcription from chromatin templates. Their intracellular organization is dynamic and related to both cell cycle and transcription. Here we examine possible mechanisms for targeting HMG-14/-17 to specific regions in chromatin. Chromatin immunoprecipitation assays indicate that HMG-17 protein is not preferentially associated with chromatin regions containing transcriptionally active genes, or any type of specific DNA. We used a modification of the random amplified polymorphic DNA method to analyze DNA in various HMG-14/-17·nucleosome complexes. We found that although HMG-14 or HMG-17 proteins preferentially associate with core particles in which the DNA has a low frequency of CG dinucleotides, the genome does not contain consensus sequences that serve as specific targeting sites for the binding of either HMG-14 or HMG-17 proteins to nucleosomes. We used size exclusion and ion exchange chromatography to demonstrate that nuclei contain a large portion of HMG-17 associated with other proteins in a multiprotein complex. We suggest that these complexes regulate the dynamic organization of HMG-14/-17 in the nucleus and serve to target the proteins to specific sites in chromatin.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS REFERENCES

The binding of specific proteins to their appropriate target sites in chromatin facilitates the orderly progression of various DNA-dependent activities such as transcription and replication. Many of the proteins regulating these activities recognize their chromatin-binding targets in a DNA sequence-dependent manner. In addition, the nucleus contains numerous proteins, such as HMG¹ proteins, that bind to their chromatin targets without apparent specificity for the underlying DNA sequence. The high mobility group (HMG) proteins affect the expression of many genes, most probably by altering the local structure of DNA or chromatin and by inducing a conformation that facilitates the binding of specific regulatory factors (1, 2). Each type of HMG subgroup has a characteristic binding site; however, the main mechanisms for targeting the proteins to these sites or to specific regions in chromatin are not fully understood.

The HMG-14/-17 proteins are the only nuclear proteins known to preferentially bind to the 147-base pair nucleosome core particles, i.e. to the building block of the chromatin fiber (3). These proteins can function as architectural elements that unfold the chromatin fiber (4, 5). All the nucleosomes in the chromatin can bind these proteins; however, the amount of HMG-14/-17 in the nucleus is limited (6). Therefore, in vivo, only a small fraction of the nucleosomes in cellular chromatin contains these HMGs. The HMG-17 nucleosomes are clustered into domains and can serve to decompact the structure of the 30-nm chromatin fiber, thereby reducing the repressive activity of chromatin.

The intranuclear distribution of both HMG-14 and HMG-17 is dynamic and not uniform (7, 8). The proteins are not associated with chromatin during metaphase and re-enter the nucleus, in a facilitated process, only after the formation of the nuclear membrane (8). Thus, their cellular location is cell cycle related. In addition, the intranuclear organization of the proteins depends on transcriptional activity. At high levels of transcriptional activity the proteins are dispersed into small foci, while at low levels of transcriptional activity they are organized into larger clusters (9).

The molecular mechanisms that regulate the intranuclear distribution and chromatin organization of these proteins are not known. One possibility is that certain nucleosomes contain unique sequences with high affinity for HMG-14/-17 proteins. Alternatively, the HMG-14/-17 proteins are targeted to specific regions with the assistance of other proteins. It is relevant that activities known to modify the chromatin structure such as histone acetylases, histone deacetylases, and chromatin remodeling complexes are present in the nucleus as high molecular weight multiprotein complexes.

Here we investigate several mechanisms that may determine the distribution of the HMG-14/-17 proteins in chromatin. First, we use chromatin immunoprecipitation assays (ChIP) to examine whether any DNA class, such as transcriptionally active chromatin regions, or repetitive DNA sequences, are significantly enriched in HMG-17 protein. We find that the HMG-17 content in chromatin regions containing the transcribed T-cell receptor  gene from mouse thymus is only 1.5 times higher than in chromatin regions containing a non-transcribed gene. Similar levels of enrichment were observed previously by other approaches. In all these studies the choice of gene examined is arbitrary, and does not exclude the possibility that chromatin regions containing a particular subset of genes are highly enriched in HMG-14/-17 proteins. We therefore used a modification of the random amplified polymorphic DNA (RAPD) method (10) to screen the nucleosomes obtained from the entire genome, and search for DNA sequence elements that would serve to target HMG-14/-17 proteins to particular nucleosomes. We find that HMG-14/-17 proteins preferentially bind to DNA regions depleted of CG and enriched in AT and TA; however, this preference is not sufficient to target the proteins to specific regions in chromatin. Our results clearly indicate that the genome does not contain sequences that serve as HMG-14/-17-binding sites. By fractionating a nuclear extract we obtained a high molecular weight multiprotein complex containing non-nucleosome-bound HMG-17. We therefore suggest that in the nucleus the proteins are present as multiprotein complexes, and that components of this complex affect the intranuclear organization and chromatin binding of HMG-14/-17 proteins.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS REFERENCES

Immunoprecipitation of HMG-Nucleosome Complexes and Slot-blot Hybridization Analysis of DNA Pools from Mouse Thymus Cells (ChIP, Chromatin ImmunoPrecipitation)

Mouse thymus mononucleosome preparation and immunoprecipitation with antibodies to HMG-17 was carried out as described before (5). Three DNA preparations (T-DNA for total mononucleosome preparation, IP-DNA for immunoprecipitate, and S-DNA for supernatant) were blotted to nylon membranes and probed with the following probes: 1) mouse high C_ot-1 DNA (Life Technologies); 2) mouse repetitive B1 element-containing plasmid derived from the -fetoprotein gene (11); and 3) a sequence specific to the T-cell receptor  gene (12). DNA was labeled by T7 QuickPrime random-priming kit (Amersham Pharmacia Biotech) and hybridized as described below.

Core Particle Preparation, Reconstitution with HMG Proteins in Vitro, Mobility Shift Assays, Analysis of the Protein Pattern, and Isolation of DNA

Recombinant human HMG-14 and HMG-17 proteins were expressed using the T7 expression system and vector pVEX11 and purified as described before (13). Nucleosome core particles were isolated from chicken red blood cells (14). The histone content of the nucleosomes was monitored by SDS-polyacrylamide gel electrophoresis and Triton-acid-urea gel electrophoresis (15). Gel mobility shift assays of HMG-nucleosome complexes were carried out in 2 × TBE buffer, (1 × TBE buffer = 90 mM Tris, 90 mM borate, 1 mM EDTA, pH 8.4). In brief, 50 pmol of core particles were incubated with 5-150 pmol of HMG protein, in a volume of 50 µl, at 4 °C for 10 min. 10 µl of 20% Ficoll 400 were dispensed to each tube and the reaction mixture was loaded on a 5% native polyacrylamide gel, run, and stained with ethidium bromide. The core particles that were shifted at a low HMG:core particle ratio (below 1) were denoted as high affinity (HA) core particles. Typically, these were less than 15% of the total DNA (as measured by ethidium bromide staining). Conversely, the nucleosome cores that have not been shifted at HMG:core particles ratio of about 3 (approximately 15% of total particle population) were denoted as low affinity (LA) cores. Bands of interest were excised from the gel, and the DNA extracted by diffusion from crushed gels. The preparations served as DNA templates for PCR with RAPD primers.

Immunoprecipitation of HMG-17-containing Mononucleosomes from HeLa Cells

Sucrose density gradient-purified HeLa mononucleosomes were prepared as described previously (16), except that buffer A contained a complete mixture of protease inhibitors (Roche Molecular Biochemicals). Immunoprecipitation was done using affinity pure antibodies and immobilized Protein A. DNAs were extracted from the immunoprecipitates and the supernatants, and denoted as IP-DNA and S-DNA, respectively.

RAPD Reaction, Cloning, and Sequencing

Primers-- For each RAPD reaction, a set of three 10-base oligonucleotide primers was used to amplify various preparations of nucleosomal DNA. About 100 primers for RAPD reactions were purchased from Operon Technologies. They were selected with the requirements that the G+C content should be 60 to 70%, and no self-complementary ends were allowed. By using three primers per one reaction an average of 60 PCR products were observed. This number of fragments can be reliably resolved and analyzed using short sequencing gels. The primers were grouped into 30 sets of 3 primers each, ensuring no internal priming sites for the two other primers in a set (the complete list is available upon request).

DNA Amplification and Cloning-- Pre-mixed sets of primers were end-labeled with [-³²P]ATP by T4 polynucleotide kinase and used to amplify 100 ng of DNA. The cycling reactions were performed at 94 °C for 1 min, 45 °C for 1 min, 72 °C for 1 min, 30 cycles. After PCR, gel loading dye was added to the reaction mixtures and a 3-µl aliquot loaded onto 40-cm sequencing gels and run for 1.5 h, at constant power (30 W). The gel was transferred onto Whatman 3MM paper and the PCR products visualized by PhosphorImager (Storm Instrument, Molecular Dynamics). The rest of the reaction mixture was loaded on preparative sequencing gels. The bands were excised from the preparative gels and subjected to a second round of PCR using the original set of RAPD primers, which were phosphorylated with cold dATP. The ends of the PCR products were blunted using Klenow and the reaction mixture loaded on 8% PAGE in 1 × TBE buffer. The bands were visualized by staining with ethidium bromide, and the DNAs eluted and ligated into the SmaI site of a pUC18 cloning vector.

Hybridization of the Cloned Fragments-- The DNA preparations (the same as were used for RAPD reaction) were extracted from the native gel, denatured, and slot blotted onto a nylon membrane (S&S, Nytran, 0.45 µm) in three dilutions (150, 50, and 15 ng). The DNA probes were labeled with the T7 QuickPrime Kit (Amersham Pharmacia Biotech). Hybridizations were done in 5 × Denhardt, 6 × SSC, 0.5% SDS, salmon sperm DNA (20 µg/ml) at 65 °C for 4 h. The blots were washed once with 2 × SSC, 0.1% SDS at 65 °C for 30 min and 3 times with 0.1 × SSC, 0.1% SDS at 65 °C for 5 min.

Reconstitution of HMG-Nucleosome Complexes on Cloned Sequences-- Cloned RAPD fragments were excised from vectors and then purified by polyacrylamide gel electrophoresis. The ends of the DNA fragment were filled-in with Klenow enzyme and [-³²P]dNTP. Nucleosome reconstitutions were carried out by the exchange technique (17), using chicken blood core particles as the histone octamer donors. HMG-14/-17 proteins were added to the reconstituted nucleosomes and HMG·CP complexes were analyzed as described above. These reaction mixtures contained 100 M excess of cold competitor DNA to minimize the effects of the slight fluctuations in concentration of the purified cloned fragment.

DNA Sequence Analysis-- The cloned RAPD bands were sequenced using USB Sequenase version 2.0 kit and 40-cm long 8% electrolyte-gradient polyacrylamide gel. GenBank homology searches were done using the Blast program. Multiple alignments were done using Clustal W (version 1.74). Di- and trinucleotide frequency analysis was done as described by others (18). The average roll angle, tilt angle, and twist angle of each cloned DNA sequence predicted from the di- and trinucleotide frequency were calculated using available software DNA tools (19, 20).

Reconstitution of Nucleosomes on Methylated Cloned Sequences

Selected cloned sequences were treated with Sss I methylase (CpG methylase, New England BioLabs), and an aliquot was digested by HpaII to verify the effectiveness of methylation as described previously (21). Then, the DNA was end-labeled, reconstituted into nucleosomes by the exchange method as described above, and titrated with increasing amounts of HMG-14 or HMG-17. The reaction mixture was loaded on native 5% polyacrylamide gel run in 2 × TBE. The gels were dried, autoradiographed, and the resulting bands were quantified with a PhosphorImager.

Isolation of a Multiprotein Complex Containing HMG-17 Protein

The presence of HMG-17 during the various purification steps was followed by Western analysis. After every round of chromatography, the HMG-17 containing fractions were dialyzed against the initial buffer of the next chromatography step and concentrated as required. Nuclear extracts from exponentially growing HeLa S3 cells were prepared as described (22). The nuclear extract was applied to a Superose 6 column eluted with in 100 mM sodium phosphate, pH 7.2. The fraction containing HMG-17 was dialyzed against 100 mM sodium phosphate, pH 6, loaded onto a Mono S HR 5/5 column and the proteins were eluted using linear 0 to 1 M NaCl gradient. The HMG-17 containing fraction was concentrated and size fractionated on BioSep SEC S3000 (Phenomenex) in 50 mM sodium phosphate, pH 7.2. The HMG-17 containing fraction was concentrated, and fractionated on a Mono Q HR 5/5 column eluted with a sodium chloride gradient from 0.1 to 0.6 M in 50 mM Tris-HCl, pH 8.8. The proteins in the various fractions were analyzed by electrophoresis in SDS-containing 15% polyacrylamide gels and visualized by silver staining of 15% SDS-polyacrylamide gel electrophoresis and silver staining.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS REFERENCES

Hybridization Analysis of the DNA from the Mouse Thymus HMG-17-containing Mononucleosomes (ChIP)-- Initially, we examined whether in a differentiated tissue, such as mouse thymus, HMG-17 protein is associated with a specific class of DNA sequence. Chromatin isolated from mouse thymus nuclei was digested with micrococcal nuclease, fractionated on sucrose gradients, and the mononucleosomes containing HMG-17 protein were immunoprecipitated using affinity pure antibodies. Approximately 1% of the DNA was recovered in the immunoprecipitates.

Membranes with equal amounts of total input DNA (T-DNA), immunoprecipitated DNA (IP-DNA), and DNA from the mononucleosomes that remained in the supernatant (S-DNA), were hybridized with the following ³²P-labeled probes: T-DNA, IP-DNA, B1-DNA (a highly Alu-like repetitive element found in >10⁵ copies per genome), mouse high C_ot-1 DNA, -satellite DNA, and a probe for the T-cell receptor  gene, known to be expressed in the mouse thymus. All the probes hybridize with equal efficiency to the T-DNA and S-DNA (Fig. 1). The IP-DNA was depleted of the repetitive B1 sequences and slightly enriched in high C_ot-1 DNA. The IP-DNA, i.e. the HMG-containing mononucleosomes, was also enriched (1.5 times) in  gene sequences. The differences in the amount of HMG-17 protein associated with various types of DNA were not significant. Therefore, we conclude that this protein is not permanently and exclusively associated with any specific class of DNA, including the DNA in transcriptionally active chromatin regions. This conclusion is in full agreement with previous information obtained by various immunofractionation approaches (23, 24), and also with the more recent finding that the intranuclear organization of the proteins is dynamic (7, 8), i.e. their interaction with any particular chromatin region is temporary rather than permanent.

View larger version (52K): [in this window] [in a new window]   Fig. 1.   Slot blot analysis of HMG-17-associated mononucleosomes from mouse thymus. DNA isolated from either the total population of mononucleosomes (T-DNA), from mononucleosomes specifically immunoprecipitated with anti-HMG-17 (IP-DNA), or from the non-precipitated mononucleosomes remaining in the supernatant (S-DNA), were immobilized and probed with the DNA indicated on the left of the blots. The radioactivity signals were quantitated with a PhosphorImager (bars on the right).

Usually, results obtained by ChIP assays are dependent on the choice of the probes and do not provide information on potential DNA sequence elements that may regulate the binding of HMG-14/-17 proteins to nucleosomes. Therefore we used an alternative approach, a modification of the RAPD method, to scan all of the possible HMG:core particles. This approach does not require the use of arbitrarily chosen probes.

The RAPD Technique Differentiates between Nucleosome Cores with Low Affinity and High Affinity for HMG-14/-17 Proteins-- The specific binding of HMG-14/-17 to nucleosome core particles (Fig. 2A) can be detected by mobility shift assays, in which the HMG·CP complex migrates slower than the free CP. A gradual increase in the HMG:CP molar ratio results in a concomitant increase in the relative amount of the HMG·CP complex (Fig. 2B). The nucleosomes with the highest affinity for HMG-14/-17 will form complexes with the protein first. We reconstituted the total population of chicken blood CP with various amounts of HMG-14 or HMG-17 and separated the HMG·CP complexes from free CP by electrophoresis on native nucleoprotein gels. The CP that formed complexes with either HMG-14 or HMG-17 at a low protein:CP ratio (less than 1), when approximately 15% of the total CP were shifted, were denoted as "high affinity HMG-14/-17 CP" (HA-CP in Fig. 2B), and the DNA excised from these bands was denoted as "HMG-14/-17 high affinity DNA" (HA-DNA). The CP that remained uncomplexed at relative high protein:CP ratios (molar ratio 3:1), when over 80% of the CP were shifted, were denoted as "low affinity HMG-14/-17 CP" (LA-CP in Fig. 2B), and the DNA extracted from these bands was denoted as "HMG-14/-17-low affinity DNA" (LA-DNA). The DNA, extracted from the total population of CP (CP in lane 1 in Fig. 2B), was denoted as "total CP DNA" (T-DNA). The average length of the nucleosomal DNA was the same in all the DNA populations, as assessed by electrophoresis of ³²P-end labeled DNA on sequencing gels and autoradiography (not shown). Likewise, electrophoresis in Triton-acid-urea gels did not reveal significant differences in the acetylation state of the core histones in all the populations of core particles (not shown).

View larger version (33K): [in this window] [in a new window]   Fig. 2.   Experimental design for using RAPD to examine the DNA sequence preference for the interaction of HMG-14/-17 proteins with nucleosome core particles (see "Experimental Procedures" for details). A, core particles (CP) were obtained from chicken erythrocytes and reacted with recombinant HMG-14/-17 proteins. B, mobility shifts of HMG-14/-17 with CP. In lane 2 approximately 15% of core particles were shifted. HA-CP denotes the CP with high affinity for HMG. In lane 4 over 80% of CP were shifted. LA-CP denotes CP with low affinity for HMG. C, the DNAs from the various bands were purified and PCR-amplified with 10-nucleotide long primers (RAPD procedure). D, PCR products were fractionated on sequencing gels. E, bands of interest were excised, re-amplified, re-purified, and cloned as "HA clones," LA clones," or "AA clones" (high, low, and average affinity, correspondingly). The cloned DNAs were examined by blot hybridization, gel mobility assays, and sequence analysis.

To examine whether the binding of HMG-14/-17 proteins to CP is affected by the sequence of the nucleosomal DNA, we used a modification of the RAPD procedure to search for differences between the HA-(14 or 17)-DNA, LA-(14 or 17)-DNA, and T-DNA preparations. The DNAs were amplified using different sets of ³²P-labeled RAPD primers. The radioactive PCR products were resolved by electrophoresis on sequencing gels and then visualized by autoradiography (Fig. 3A). Control reactions, in which DNA was not added, contained only short 15-25-nucleotide long PCR bands, which could be the products of primer dimerization at low annealing temperature. Therefore, in all the lanes, the bands migrating at these positions were ignored. For analysis, each lane was scanned, the background subtracted, and the scans superimposed.

View larger version (75K): [in this window] [in a new window]   Fig. 3.   Analysis of RAPD-generated products. A, various DNA preparations (labels on the left) have been amplified with more than 30 primer sets and resolved on sequencing gels. A representative autoradiogram is shown. High affinity band was 116 nucleotides long. B, slot-blot hybridization. HA-DNA, LA-DNA, and T-DNA were obtained by preparative mobility shifts of CP with either HMG-14 or HMG-17 protein, and slot blotted. The membrane was probed with DNAs cloned from specific RAPD-generated bands indicated above the blot. C, mobility shift assays. HMG-14 and HMG-17 bind better to core particles, reconstituted on HA clone sequences 40.1a.5 and H015-3 (lanes 1-6 in both panels) than to core particles reconstituted on LA clone sequences 40.1c.7 and 48.5c.1 (lanes 7-12). Each reaction contained 1 ng of reconstituted core particles, 100 ng of cold competitor DNA, and none (lanes 1, 4, 7, and 10), or 2.5 ng (lanes 2, 5, 8, and 11), and 5 ng (lanes 3, 6, 9, and 12) of HMG protein.

The results clearly indicate that most of the bands with identical mobility have the same intensity in all the tracks (Fig. 3A). There was no marked difference in the RAPD pattern generated from CP complexed with either HMG-14 or HMG-17 proteins. In more than 1000 bands generated from 30 RAPD reactions we detected only one specific band generated from the DNA isolated from the CP that preferentially bound to HMG-17, and one specific band generated by the CP that preferentially bound HMG-14. However, the autoradiography signals of about 4% of the bands in the HA-DNA tracks, and of 2% of bands in the LA-DNA tracks, were at least twice more intense than that of the corresponding bands in the T-DNA tracks. These results suggest that a small fraction of the total core particles binds HMG-14/-17 protein with either a higher (bands enriched in the HA-DNA tracks) or lower (bands enriched in the LA-tracks) affinity than the average particle.

To examine the DNA in the nucleosome particles that bind HMG-14/-17 with an affinity that is different from the average, we isolated, cloned, and sequenced all the PCR fragments that were highly enriched in either HA-DNA or LA-DNA (HA bands and LA bands). A total of 83 clones were analyzed (Table I). Since HMG-14 and HMG-17 proteins generated very similar RAPD patterns we did not distinguish between those originating from the HMG-14·CP complexes from those originating form HMG-17·CP complexes. As controls, we cloned and sequenced 41 fragments, that were equally intense in both HA- and LA- DNAs. These are named AA clones (AA stands for "average affinity").

We performed two kinds of tests to verify that the DNAs identified by RAPD indeed had some property that was different from the average DNA. First, we tested whether the DNA rehybridized to the selected bands with the expected intensity. Slot-blot analysis, in which the cloned DNAs were used as probes for HA-DNA, LA-DNA, or T-DNA indicated that the cloned sequences do indeed hybridize with the expected intensity (Fig. 3B). Thus, HA clone 42.1a.7 that was enriched in an RAPD reaction with HA-DNA, preferentially hybridized to HA-DNA on a membrane. As expected, AA clone 48.6b.2 that was cloned as one of the control sequences, did not show any preference for either LA-DNA or HA-DNA, while LA clone 40.1c.6 preferentially hybridizes to LA-DNA. The results corroborate the RAPD data.

Second, we tested whether the cloned DNAs can be complexed with histones to reconstitute nucleosomes with the expected affinity for HMG-14/-17 proteins. Clones with inserts longer than 140 bp were reconstituted into nucleosomes and tested for affinity for HMG-14 or HMG-17 proteins by gel mobility shift assays (Fig. 3C). The affinity constant of HMG-14 and HMG-17 for nucleosomes is 1.1 and 0.5 × 10⁷, respectively (15). We found that nucleosomes reconstituted using the DNA from HA clones (40.1a.5 and H015-3, lanes 1-6) have higher affinity for HMG proteins (~2.05 × 10⁷ for HMG-14 and 0.95 × 10⁷ for HMG-17), while the nucleosomes reconstituted using the DNA from LA clones (40.1c.7 and 48.5c.1, lanes 7-12) had a lower affinity for HMG-14/-17 (~0.25 × 10⁷ and 0.15 × 10⁷, respectively) (Fig. 3C). Thus, in lanes 2, 3, 5, and 6 most of the CPs, reconstituted on HA clone DNAs, were associated with HMG proteins while in lanes 8, 9, 11, and 12 most of the CPs, reconstituted on LA clone DNAs, were not associated with HMG proteins at similar protein concentrations. Thus, the RAPD reaction is suitable for isolating the DNA from nucleosomes that vary in their affinity for HMG-14/-17 proteins

Search for HMG-14/-17 Protein DNA-binding Sites-- To examine whether the sequences from the various DNA pools contain common sequence motifs, we performed multiple alignment analysis of all the sequences in both orientations. The analysis did not detect any common motifs in any of the DNA pools, even though we limited the number of aligned sequences to five at a time, checked all the possible combinations of five sequences, and also calculated the average pairwise alignment per each nucleotide. A Blast search of the GenBank data base with each of the sequences detected seven clones having 85-100% identity with known sequences. One of the HA clones was homologous to a region located 300 bp downstream from the protein phosphatase inhibitor 2 gene, and another HA clone to a region located 500 bp upstream from the acyl CoA-binding protein/diazepam-binding inhibitor gene. Three AA clones contained regions homologous to the transcribed portion of the chicken neuropeptide Y gene, to chicken endogenous proviral avian retroviral LTR and to Mus musculus uroplakin II gene. One LA clone was similar to a known sequence of a CpG island, and the other one contained an Alu-like repeat. We conclude that in vitro, the binding of HMG-14/-17 to nucleosome cores is not dependent on a specific DNA sequence element.

We next performed a similar analysis with HMG-17 containing nucleosomes isolated by immunoprecipitation of HeLa chromatin. These particles should be representative of the in vivo HMG·CP complexes. We immunoprecipitated HMG:CP particles as described before (see previous section) and analyzed the DNA isolated from both the immunoprecipitated nucleosomes and from the nucleosomes that remained in the supernatant, by the RAPD procedure using the primer sets that were used in the in vitro experiments.

Most of the bands generated by the entire sets of primers were common to two DNA pools; however, several primer sets generated bands that were specific to either the immunoprecipitated DNA (IP-DNA) or the DNA that remained in the supernatant (S-DNA, Fig. 4). We reasoned that the RAPD bands enriched in the IP-DNA lanes were preferentially associated with HMG-17 in vivo. Conversely, the bands were enriched in the S-DNA lanes, originated from the core particles that in vivo had low affinity for HMG-17. We isolated and sequenced 6 clones from each type of DNA (IP clones and S clones, respectively) and analyzed them by the same parameters as the CP isolated from the in vitro reactions (Fig. 3A). We did not obtain a consensus sequence and a Blast search did not yield any significant hits.

View larger version (24K): [in this window] [in a new window]   Fig. 4.   Analysis of PCR-generated bands from DNA pools isolated from HeLa cells by immunoprecipitation. DNAs for RAPD reaction were isolated either from the immunoprecipitated CP (IP-DNA) or from the DNA that remained in the supernatants (S-DNA). Arrowheads point to bands that were unique or enriched in one of the DNAs.

We conclude therefore that the genome does not contain unique DNA sequence motifs that target HMG-14/-17 proteins to specific regions in chromatin. It is important to note that in these studies we analyzed individual RAPD fragments with total length of more than 10⁵ base pairs (35 RAPD reactions, 50 bands in each, with an average length of 80 bp per band). This number of base pairs includes all possible genomic hexamers 25 times, all possible heptamers 7 times, and all possible octamers 1.5 times. Consensus sequences for abundant regulatory factors are of limited length. For example, the consensus sequence for the TATA-binding protein (TBP) is shorter than 7 bp. Therefore, any putative short consensus sequence regulating the binding of HMG-14/-17 proteins (either positively or negatively) to core particles would have been present in the samples that we analyzed.

Thus, our studies indicate that the genome does not contain specific short sequence elements that target HMG-14/-17 proteins to specific sites. Both in vivo and in vitro HMG-14/-17 proteins bind to nucleosome cores to form complexes containing either two molecules of HMG-14 or two molecules of HMG-17. Our finding that the RAPD patterns generated by the nucleosome cores interacting with HMG-14 are indistinguishable from those generated by HMG-17, provides additional support for the conclusion that the DNA sequence is not a significant factor in determining the binding of these proteins to core particles.

Characterization of the DNA in the Various Types of Nucleosome Core Particles-- Nevertheless, the RAPD analysis suggests that the affinity of about 6% of the core particles is either higher (4%), or lower (2%) than that of the bulk core particles. To further characterize the DNA in the various nucleosome core pools, we grouped all HA, AA, and LA clones into separate pools and calculated the observed versus expected frequencies for mono-, di-, and trinucleotides. As indicated in Fig. 5A, the 3 pools had a very similar content of mononucleotides. However, the 3 DNA pools differed in the frequency of some of the dinucleotides. Most strikingly, the frequency of the dinucleotide CG (arrowhead) was 3 times lower in HA clones (observed versus expected frequency 0.22) than that in the LA clones. Indeed, a plot of the distribution of CG occurrences in all the individual clones indicates that the sequences derived from HA-DNA have a significantly lower content of CG than those derived from LA-DNA (Fig. 5C). In addition, the frequencies for GG, CC, TA, and AT dimers in HA-DNA were 25-40% higher than in LA-DNA.

View larger version (30K): [in this window] [in a new window]   Fig. 5.   Frequency distribution of nucleotides in RAPD clones. A, observed versus expected frequency distribution of nucleotides in HA clones, LA clones, and AA clones obtained from in vitro reconstitution experiments. Mononucleotide (left side) and dinucleotide (right side) distribution. B, observed versus expected frequencies for the distribution of dinucleotides in the pooled clones prepared from IP- and S-DNA obtained by immunoprecipitation of HeLa chromatin, i.e. from HMG·CP complexes formed in vivo. C, histogram of the distribution of CG dinucleotides in individual clones. The abscissa denotes the ratio of observed frequency to expected frequency for CG in a sequence. The ordinate represents the number of sequences with similar CG frequency.

The trinucleotide frequency distribution of the clones reflected their dinucleotide content. Thus, the majority of the triplets were uniformly dispersed between all DNA pools (not shown). All the trinucleotides that were strongly under-represented in HA-DNA contained the dinucleotide CG. The 4 dinucleotides CC, GG, AT, or TA were not present in these CG-containing trinucleotides, but were over-represented in other trinucleotides that were prevalent in HA-DNA. We note that all these differences between the HA-DNA and LA-DNA are internally consistent, since the frequency distribution of the dinucleotides within each DNA pool was very close to the inverted complementary dinucleotide. For example, the dinucleotide frequency of GG is equal to that of CC. Likewise, the frequency of almost every dinucleotide, if increased for HA clones, was decreased in LA clones, and vice versa (in 13 out of 16 combinations). We also note that the dinucleotide frequencies of all the clones, i.e. the overall low content of CG (and not GC) and TA, and the high frequencies of AG, CT, TG, and CA, are characteristic of the chicken genome.

We also analyzed certain structural characteristics of the various sequences. We calculated the local maximum, minimum, and average bendability for each clone, and averaged the values for the HA-, AA-, and LA-DNA pools. Although the DNA in the HA clones was relatively untwisted, none of these calculations showed statistically significant differences between the various DNA pools.

The distribution frequency of the di- and trinucleotides and the physical characteristics of the DNA in the clones obtained from the mononucleosomes isolated from HeLa were the same as in the clones obtained from the in vitro reactions. Thus, the clones generated by the IP-DNA, i.e. from mononucleosomes that were associated with HMG-17, had both a lower frequency of CG (observed versus expected ratio 0.23), and a higher frequency of CC, GG, AT, and TA than the clones generated from the mononucleosomes that remained in the supernatant (Fig. 5B).

An obvious explanation of the dinucleotide frequency data would be that HMG-14/-17 proteins bind weaker to nucleosomes containing methylated DNA, since CG sequences are known methylation sites in DNA. We therefore reconstituted nucleosomes on Sss I-methylated cloned DNAs from either HA-DNA or LA-DNA. Mobility shift assays indicated that methylation did not affect the binding of HMG-14/-17 proteins to either DNA or cores particles (not shown). Thus, we conclude that the under-representation of CG in the HA-DNA does not reflect a lower affinity of HMG-14/-17 proteins to methylated DNA.

In summary, the characteristics of the nucleosomal DNA in the HMG-17-containing mononucleosomes isolated from HeLa cells by the ChIP procedure were similar to those in the chicken CP that associated with HMG-14/-17 in vitro. HMG-14/-17 proteins preferentially bind to nucleosomes in which the DNA has a relatively low content of CG dinucleotides and is somewhat enriched in TA and AT dinucleotides. However, based on the measurements of the affinity constants for the binding of HMG-14/-17 to sequences with different dinucleotide frequencies, we conclude that the dinucleotide composition is not a major factor in regulating the interaction of the proteins with chromatin. Thus, our major conclusion is that the genome does not contain specific DNA sequence elements that serve to target these proteins to nucleosomes, i.e. the organization of HMG-14/-17 in chromatin is not dependent on direct, specific interactions between the nucleosomal DNA and the HMG-14/-17 proteins.

Nuclear HMG-17 Protein Is in a Multiprotein Complex-- The absence of DNA sequence elements that specifically bind HMG-14/-17 protein raises the possibility that the proteins are actively targeted to specific regions, perhaps in association with other proteins, i.e. as multiprotein complexes. Indeed other nuclear activities known to target and modify chromatin have been isolated as large complexes containing several proteins. We searched for the presence of such tentative complexes, by fractionating a nuclear extract of HeLa S3 cells on size exclusion columns. Western analysis of fractions obtained from a BioSep SEC S3000 column revealed the presence of two major HMG-17-containing peaks (Fig. 6B). The first peak eluted with the void volume, i.e. contained complexes larger than 700 kDa, while the second peak contained free HMG-17 protein. High molecular weight complex containing HMG-17 protein has been partially purified by several chromatographic steps (Fig. 6A). We used Western blotting to confirm the presence of HMG-17 protein and silver staining of 15% polyacrylamide gels to analyze the protein composition of the peaks. We observed 8-10 protein bands that consistently co-purified with HMG-17 protein (Fig. 6C). These finding strongly suggest that HeLa nuclei contain a large portion of HMG-17 associated with other proteins in a multiprotein complex.

View larger version (24K): [in this window] [in a new window]   Fig. 6.   HMG-17 is a part of multiprotein complex. A, flow diagram of the fractionation steps. B, size exclusion chromatography of HeLa nuclear extract (100 µg of the protein). 1-ml fractions have been collected and 20-µl aliquots were loaded on 15% SDS-polyacrylamide gel. Western blotting shows binary distribution of HMG-17 protein into free and complexed form. C, SDS-polyacrylamide gel of the partially purified HMG-17-containing multiprotein complex.

Conclusions-- This article contains two major new findings. First, we conclusively demonstrate that the genome does not contain DNA sequences that target the HMG-14/-17 proteins to specific nucleosomes. Thus, the organization of HMG-14/-17 proteins is not dependent on direct interactions between the sequence of the nucleosomal DNA and HMG-14/-17 proteins. We suggest that the association of the proteins with a specific region in chromatin may be transient, dependent on the metabolic state of a cell, and independent of the sequence of the nucleosomal DNA. This observation fully agrees with recent photobleaching experiments indicating that in living cells, both HMG-14 and HMG-17 move rapidly and constantly throughout the entire nucleus in a diffusion driven, Brownian-type motion. Second, we demonstrate that in the nucleus HMG-17 is associated with other proteins to form a high molecular weight, multiprotein complex. Although this complex is not fully characterized, its occurrence provides a conceptual framework to explain the dynamics of the intracellular organization of HMG-14/-17 proteins. We suggest that the intracellular trafficking and chromatin targeting of HMG-14/-17 proteins are regulated by proteins that interact with these HMGs and form a multiprotein complex.

	FOOTNOTES

^* The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: Bldg. 37, Rm. 3D-20, NCI, National Institutes of Health, 9000 Rockville Pike, Bethesda, MD 20892. Tel.: 301-496-2885; Fax: 301-496-8419; E-mail: yupo@helix.nih.gov.

Published, JBC Papers in Press, September 5, 2000, DOI 10.1074/jbc.M000989200

	ABBREVIATIONS

The abbreviations used are: HMG, high mobility group; ChIP, chromatin immunoprecipitation assays; RAPD, random amplified polymorphic DNA; HA, high affinity; LA, low affinity; PCR, polymerase chain reaction; T-DNA, total mononucleosome preparation DNA; IP-DNA, immunoprecipitated DNA; S-DNA, supernatant DNA; AA, average affinity; bp, base pair(s); CP(s), nucleosome core particle(s).

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				M. Amen, H. M. Espinoza, C. Cox, X. Liang, J. Wang, T. M. E. Link, R. G. Brennan, J. F. Martin, and B. A. Amendt Chromatin-associated HMG-17 is a major regulator of homeodomain transcription factor activity modulated by Wnt/{beta}-catenin signaling Nucleic Acids Res., February 2, 2008; 36(2): 462 - 476. [Abstract] [Full Text] [PDF]

				N. Zhu and U. Hansen HMGN1 Modulates Estrogen-Mediated Transcriptional Activation through Interactions with Specific DNA-Binding Transcription Factors Mol. Cell. Biol., December 15, 2007; 27(24): 8859 - 8873. [Abstract] [Full Text] [PDF]

				T. Furusawa, J.-H. Lim, F. Catez, Y. Birger, S. Mackem, and M. Bustin Down-Regulation of Nucleosomal Binding Protein HMGN1 Expression during Embryogenesis Modulates Sox9 Expression in Chondrocytes Mol. Cell. Biol., January 15, 2006; 26(2): 592 - 604. [Abstract] [Full Text] [PDF]

				D. A. Hill, C. L. Peterson, and A. N. Imbalzano Effects of HMGN1 on Chromatin Structure and SWI/SNF-mediated Chromatin Remodeling J. Biol. Chem., December 16, 2005; 280(50): 41777 - 41783. [Abstract] [Full Text] [PDF]

				K. L. West, M. A. Castellini, M. K. Duncan, and M. Bustin Chromosomal Proteins HMGN3a and HMGN3b Regulate the Expression of Glycine Transporter 1 Mol. Cell. Biol., May 1, 2004; 24(9): 3747 - 3756. [Abstract] [Full Text] [PDF]

				J.-H. Lim, M. Bustin, V. V. Ogryzko, and Y. V. Postnikov Metastable Macromolecular Complexes Containing High Mobility Group Nucleosome-binding Chromosomal Proteins in HeLa Nuclei J. Biol. Chem., May 31, 2002; 277(23): 20774 - 20782. [Abstract] [Full Text] [PDF]

				K. L. West, Y. Ito, Y. Birger, Y. Postnikov, H. Shirakawa, and M. Bustin HMGN3a and HMGN3b, Two Protein Isoforms with a Tissue-specific Expression Pattern, Expand the Cellular Repertoire of Nucleosome-binding Proteins J. Biol. Chem., July 6, 2001; 276(28): 25959 - 25969. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 275/48/37937    most recent M000989200v1 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 Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Shirakawa, H. Articles by Postnikov, Y. Search for Related Content PubMed PubMed Citation Articles by Shirakawa, H. Articles by Postnikov, Y. Social Bookmarking                      What's this?