A Mouse Histone H1 Variant, H1b, Binds Preferentially to a Regulatory Sequence within a Mouse H3.2 Replication-dependent Histone Gene*

(Received for publication, March 14, 1997, and in revised form, April 8, 1997)

Nikola K. Kaludov Dagger , Lil Pabón-Peña §, Margaret Seavy , Gail Robinson and Myra M. Hurt

From the Department of Biological Science, Florida State University, Tallahassee, Florida 32306-3050

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES


ABSTRACT

H1 histones, found in all multicellular eukaryotes, associate with linker DNA between adjacent nucleosomes, presumably to keep the chromatin in a compact, helical state. The identification of multiple histone H1 subtypes in vertebrates suggests these proteins have specialized roles in chromatin organization and thus influence the regulation of gene expression in the multicellular organism. The mechanism by which the association of H1 with nucleosomal DNA is regulated is not completely understood, but affinity for different DNA sequences may play a role. Here we report that a specific H1 subtype in the mouse, namely H1b, selectively binds to a regulatory element within the protein-encoding sequence of a replication-dependent mouse H3.2 gene. We have previously shown that this coding region element, Omega , is the target of very specific interactions in vitro with another nuclear factor called the Omega factor. This element is required for normal gene expression in stably transfected rodent cells. The mouse H1b protein interacts poorly (100-fold lower affinity) with the comparable "Omega " sequence of a replication-independent mouse H3.3 gene. This H3.3 sequence differs at only 4 out of 22 nucleotide positions from the H3.2 sequence. Our findings raise the possibility that this H1b protein plays a specific role in regulation of expression of the replication-dependent histone gene family.


INTRODUCTION

Five major classes of histones occur in the mouse and most eukaryotes, H1 and the core histones H2a, H2b, H3, and H4. These basic proteins play key roles in eukaryotic chromatin structure and organization. The core histones, H2a, H2b, H3, and H4, are responsible for nucleosome formation. H1 histones associate with "linker" DNA between adjacent nucleosomes and are thought to play a role in folding eukaryotic DNA into condensed higher order chromatin structures (1-3). It has long been known that H1 can act as a repressor of transcription; considerable evidence indicates that chromatin structure is important in the regulation of transcription because it restricts the accessibility to DNA of both general and gene-specific transcription factors (3-8). Linker histones are also known to modulate nucleosome position (9-11). Further, depletion of H1 from chromatin has been shown to activate transcription in vitro (4, 12, 13). There is evidence, however, that H1 is not completely absent from transcribed genes (14, 15), which implies that linker histones may interact differently with transcriptionally active than with inactive regions of chromatin (16).

The H1 class of histones displays the most complex pattern of subtypes among the histone gene family, including differentiation-specific and tissue-specific variant proteins. The diversity of H1 histones, their differential state of phosphorylation (17, 18), and their varied distribution with respect to the stage of growth or differentiation (19) suggest that H1 subtypes also vary in their ability to confer repression throughout the genome. Both in vitro and in vivo studies of gene expression implicate the different H1 subtypes as part of a global regulatory process that is responsible for selectivity in repression of transcription (13, 20-22).

Histone genes of all classes are highly conserved at the nucleotide level and are among the most highly expressed mammalian protein-encoding genes (23). H1 is the least conserved class of histone proteins. There are seven H1 protein sequence variants in the mouse. H4 is the most conserved of the histone classes; there are no sequence variants. Histone genes can be classified as replication-dependent or -independent (24, 25) on the basis of their differential regulation of expression in the cell cycle. The expression of replication-dependent histone genes is tightly coupled with DNA synthesis in the eukaryotic cell. These genes are coordinately up-regulated at the G1-S boundary of the cell cycle, but the exact molecular mechanisms responsible remain to be elucidated (26, 27). We previously identified a coding region activating sequence (CRAS)1 in mouse H2a.2 and H3.2 replication-dependent histone genes that is involved in the up-regulation of these and possibly all replication-dependent histone genes (28, 29). Subsequently, we identified two elements within the H3.2 CRAS, the alpha  and Omega  elements, that interact with nuclear proteins in vitro and are required for normal gene expression in vivo in stably transfected Chinese hamster ovary cells (30, 31). Mutation of the 7 base pairs of either the alpha  or the Omega  element caused a 4-fold drop in expression in vivo.

Mouse nuclear factors, the alpha  and Omega , have been shown to interact very specifically with these H3.2 elements. Mutations that change the alpha  or Omega  element to the sequence found in a replication-independent H3.3 gene totally abolish DNA-protein interactions and reproduce the effects on gene expression in vivo caused by mutation of all seven nucleotides of the alpha  or Omega  element. Although the DNA-binding protein interacting with the Omega  element has not been purified, we have shown by UV cross-linking that it has an apparent molecular mass of 45 kDa (31) and must be phosphorylated on a tyrosine residue to be active in a DNA-binding assay (32).

The alpha  and Omega  elements are present in the coding region of all four nucleosomal (H2a, H2b, H3, and H4) classes of replication-dependent mouse histone genes, and the interactions of these elements with their respective nuclear factors are very similar if not identical (30-33). To date, no common elements, other than the CRAS alpha  and Omega  elements (30, 31), have been reported to be involved in regulating the coordinate expression of replication-dependent histone genes of all core histone classes.

Here we report that a second nuclear factor, a specific mouse H1 subtype, H1b, also interacts with the H3.2 CRAS Omega  element in a highly specific manner. There have been reports that total H1 shows preferred DNA binding toward AT-rich tracts because of the intrinsic properties associated with A-rich sequences (34, 35). Yaneva et al. (36) showed that DNA curvature in itself is not enough to promote strong H1 binding and that there are additional sequence requirements for high affinity H1 binding sites. Others have reported that H1 histones prefer some non-AT-rich eukaryotic sequences over others (37-41). All of these previous studies utilized total cellular H1. It should be remembered that, in asynchronously growing cells, a number of sequence variants may be present, disguising the preferential participation or nonparticipation in such interactions by specific H1 subtypes. Herein we show in in vitro experiments that the pure H1b protein interacts with 100-fold less affinity with the comparable sequence from a replication-independent H3.3 gene, differing at only 4 out of 22 nucleotides from the H3.2 sequence. The specificity of this interaction for sequences within the replication-dependent gene indicates that mouse H1b may play a distinct role in regulation of coordinate expression of histone genes in the eukaryotic cell cycle.


EXPERIMENTAL PROCEDURES

Nuclear Extracts

Mouse myeloma cells were grown in spinner cultures to a density of 5 × 105 cells/ml in Dulbecco's modified Eagle's medium, 10% horse serum, and 5% CO2, at 37 °C. Nuclear extracts were prepared by our modification of the method of Shapiro et al. (42). Briefly, cells were harvested, washed once with phosphate-buffered saline, and resuspended in hypotonic buffer (10 mM Hepes, pH 7.9, 0.75 mM spermidine, 0.75 mM spermine, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM DTT, 10 mM KCl). The cells were incubated for 15 min on ice (or until swollen), and their membranes were disrupted with 13 strokes in a homogenizer. The nuclei were pelleted at 3,000 rpm and resuspended in nuclear resuspension buffer (20 mM Hepes, pH 7.9, 0.75 mM spermidine, 0.15 mM spermine, 0.2 mM EDTA, 2 mM EGTA, 2 mM DTT, 25% glycerol). KCl was added to a final concentration of 0.6 M, and the mixture was rocked for 30 min at 4 °C. Chromatin was pelleted at 40,000 rpm for 45 min in a fixed angle rotor. The supernatant was transferred to autoclaved dialysis tubing, and the nuclear extract was dialyzed against 100-fold dialysis buffer (20 mM Hepes, pH 7.9, 0.2 mM EDTA, 0.2 mM EGTA, 100 mM KCl, 2 mM DTT, 20% glycerol) for 2 h at 4 °C. Protein concentration was measured by the Bio-Rad protein assay.

DNA-Cellulose Column Chromatography

Nuclear extracts were applied to DNA-cellulose resin (Pharmacia Biotech Inc., 40-ml bed volume) equilibrated with buffer (20 mM HEPES, pH 7.9, 0.2 mM EGTA, 0.2 mM EDTA, 100 mM KCl, 20% glycerol, and 2 mM DTT). Proteins were eluted with the same buffer containing 200 mM, 300 mM, 400 mM, and 1 M KCl. The collected fractions were analyzed for Omega -binding activity by electrophoretic mobility shift assay (EMSA) and Southwestern analysis.

EMSA

Synthetic oligonucleotides used in EMSA competition experiments and Southwestern assays were synthesized with the mouse H3.2 CRAS Omega  sequence (31) and that of the H3.3 "CRAS Omega ." The oligonucleotide sequences are
<UP>GTGCGCGAGATCGCGCAGGACT</UP>
<UP>              CGCTCTAGCGCGTCCTGAcacg</UP>
<UP><SC>Sequence</SC> 1</UP>
for the H3.2 CRAS Omega  box duplex and
<AR><R><C><UP>                 *       *       *       *           </UP></C></R><R><C>GTGCGAGAAATTGCTCAGGACT    </C></R><R><C></C></R><R><C>               CTCTTTAACGAGTCCTGAcacg</C></R></AR>
<UP><SC>Sequence</SC> 2</UP>
for the H3.3 CRAS Omega  box duplex. Lowercase letters at the 3' end of the duplex indicate nucleotides synthesized as complementary overhangs. The nucleotides in the 5' overhang are part of the nucleotide sequence of both genes. Asterisks indicate changed nucleotide positions in the H3.3 Omega  sequence. The sequences used for synthesis of the H3.3 oligonucleotides were obtained from published mouse H3.3 sequences (43-45). The duplex oligonucleotides were 3'-end-labeled with [alpha -32P]dCTP or [alpha -32P]dATP by means of the Klenow fragment of DNA polymerase I (29). The binding reactions proceeded at 4 °C for 20 min in the presence of 1 µg of the nonspecific competitor poly(dI-dC)-poly(dI-dC) and then were analyzed by electrophoresis on a 4% native acrylamide gel at 200 V (46). Radioactivity was quantified directly by analysis of the intact gel with a Betascope (Betagen).

DNase I Footprint Analysis

Footprinting of the H1b-CRAS Omega  complex was performed by the DNase I footprinting procedure previously described (31) with 40-mer duplex Omega  oligonucleotides as probe.
<UP>5′ - AGCTTGCCTGGTGCGCGAGATCGCGCAGGACTTCAAGACG - 3′</UP>
<UP>     ACGGACCACGCGCTCTAGCGCGTCCTGAAGTTCTGCTTAA - 5′</UP>
<UP><SC>Sequence</SC> 3</UP>
The 5' end of the appropriate strand was labeled with [gamma -32P]ATP by means of T4 polynucleotide kinase, then annealed to the other oligonucleotide. The annealed duplex was incubated with mouse nuclear extract, with subsequent addition of H1b, duplicating conditions used in EMSA (see Fig. 1), and then treated with DNase I (31). Free and bound complexes were then separated by EMSA. The supershifted H1b-Omega -DNA complexes were sliced from the gel and eluted at 37 °C overnight and then analyzed on an 8% polyacrylamide, M urea sequencing gel.


Fig. 1. The CRAS Omega  factor elutes in the 0.3 M fraction from DNA-cellulose. Mouse myeloma cell nuclear extract was applied to a DNA-cellulose column, and fractions were eluted as described under "Experimental Procedures." EMSA of fractions is as follows. Lane 1, crude nuclear extract (column load); lane 2, flow-through (F.T.) of the DNA-cellulose column; lanes 3-7, pooled 0.1, 0.2, 0.3, 0.4, and 1 M KCl DNA-cellulose fractions. All reactions contained the same amount of protein (6 µg) and radioactively labeled duplex H3.2 CRAS Omega  oligonucleotides.
[View Larger Version of this Image (45K GIF file)]

Southwestern Analysis

Southwestern blotting of SDS gels containing proteins from the DNA-cellulose and reversed-phase HPLC fractions was performed as described by Lenormand et al. (47) except that the nitrocellulose membrane was incubated for 24 h at 37 °C in 10 mM Tris-HCl, pH 7.5, 50 mM NaCl, 1 mM EDTA, 1 mM DTT (TNE-50) containing 2 × 106 cpm/ml end-labeled duplex Omega  oligonucleotides and the nonspecific competitor poly(dI-dC)-poly(dI-dC) (50 µg/ml). Membranes were then washed with excess TNE-50, and radioactive bands were detected by autoradiography.

Reversed-phase Chromatography

The 1 M KCl DNA-cellulose column fraction was subsequently further fractionated by reversed-phase (RP)-HPLC on a Brownlee Aquapore RP-300 analytical column (Rainin Instruments, MA) with 0.1% aqueous trifluoroacetic acid and 0.08% trifluoroacetic acid in acetonitrile as mobile phases A and B, respectively. The column was developed with a linear increase in acetonitrile concentration from 25 to 45% at 0.3%/min, with a constant flow rate of 1 ml/min. Column eluate was monitored at 214 nm, and fractions were collected by hand, lyophilized, and stored at -20 °C. The HPLC system included Beckman System Gold software, run on an IBM PS/2, the Beckman 126 solvent delivery system, an Altex 210A sample injection valve, and a Waters 441 absorbance detector.

Acid-Urea Gel Electrophoresis

Proteins in the DNA-cellulose 1 M fraction and RP-HPLC fractions were further examined by acid-urea gel electrophoresis. The extraction of total H1 from mouse myeloma cells utilized the method of R. D. Cole (48). Acid-urea gel electrophoresis followed the method of Lennox and Cohen (49). Glass plate dimensions used were 26 × 25 × 0.5 cm. The gels were conditioned as described previously (49). After samples were loaded, the proteins were separated at 250 V for 12 h. Gels were stained with Amido Black, destained, and then silver-stained (Bio-Rad).

Proteolytic Digests

Reversed-phase purified fractions were resuspended in 0.1 M Tris-HCl, pH 8.0, and the proteins were digested overnight at 37 °C with 1-2% (w/w) V8 protease. Peptides were separated by RP-HPLC on a Deltabond® Octyl analytical column (Keystone Scientific) with a linear acetonitrile gradient from 0 to 50% at 0.5%/min. Absorbance was monitored at 214 nm, and fractions were collected manually and dried for automated protein sequenation.

Amino Acid Composition Analysis

Amino acid composition data were obtained by the Pico-Tag® method (Waters Chromatography Division, MA). Briefly, dried protein samples were subjected to vapor phase hydrolysis with 6 N HCl in a MDS-200 microwave sample preparation system (CEM Corp.). Hydrolyzed samples were derivatized with phenylisothiocyanate according to standard methods and run on a Waters HPLC system controlled by a Maxima 820 Workstation for data acquisition and analysis. Amino acids were quantitated by comparison with Amino Acid Standard H (Pierce).


RESULTS

Identification of Nuclear Proteins Interacting Specifically with the CRAS Omega  Element

We utilized DNA-cellulose chromatography as the first step in the purification of the CRAS Omega -binding factor(s) from crude mouse myeloma cell nuclear extract. Fig. 1 shows a gel mobility shift assay of the DNA-cellulose fractions, using as probe radioactively labeled Omega  oligonucleotides that contain the CRAS Omega -binding site CGAGATC (31). The Omega -binding activity eluted in the 0.3 M fraction (lane 5). We have previously shown that the binding activity eluting in the 0.3 M fraction protects the Omega  sequence in a DNase I footprinting assay and that the H3.2 Omega  duplex oligonucleotides, but not the H3.3 Omega  duplex oligonucleotides, act as specific competitors for the Omega -binding activity.

We further examined the DNA-cellulose fractions by Southwestern analysis. The results are shown in Fig. 2A. Radioactively labeled duplex H3.2 Omega  oligonucleotides or, alternatively, the corresponding sequence from a replication-independent H3.3 gene that we have shown cannot act as a binding site for the CRAS Omega factor (31) was used as probes. The two oligonucleotide sequences are compared in panel C. Before hybridization with radioactive probe, the proteins in the column fractions were separated on SDS gels and blotted onto nitrocellulose membranes as described under "Experimental Procedures." In lane 1 of Fig. 2A, an intense signal is observed as a result of an interaction between the H3.2 probe and a protein with an apparent molecular mass of 30 kDa in the crude nuclear extract. The 0.3 M fraction showed no evidence of protein interaction with the H3.2 oligonucleotides (Fig. 2A, lane 3), although as shown above, this fraction contained the specific Omega -binding activity observed in EMSA (Fig. 1, lane 5). In Fig. 2A, the DNA-protein interaction observed in crude extract (lane 1) was also observed in lane 4, showing that the interacting protein (apparent molecular mass, 30 kDa) was contained in the M DNA-cellulose fraction. This interaction was not detected when the H3.3 Omega  oligonucleotides were used as probe (lanes 5 and 8). Direct quantitation of the radioactivity bound to the membranes showed that the H3.3 oligonucleotides did interact to some extent with the protein producing the strong signal seen in lanes 1 and 4, but the amount of bound probe was over 100-fold less than that bound in the comparable lanes when the H3.2 oligonucleotides were used. The H3.2 and H3.3 oligonucleotide sequences differ at only four nucleotide positions, three of which are in the Omega  element (see Fig. 2C).


Fig. 2. A protein(s) in the 1 M DNA-cellulose column fraction interacts with the Omega  element in a Southwestern assay. A, crude nuclear extract and DNA-cellulose fractions (30 µg of protein in all lanes) were loaded on a 10% SDS gel, electrophoretically separated, and electroblotted onto a nitrocellulose membrane. The blot was treated as described by Lenormand et al. (47), cut in half, and hybridized with concatamerized H3.2 Omega  oligonucleotides (lanes 1-4) or concatamerized H3.3 "Omega " oligonucleotides (lanes 5-8). B, the 1 M KCl DNA-cellulose fraction was loaded in all lanes, separated electrophoretically, and blotted onto a nitrocellulose membrane for treatment as described above. The membrane was cut into strips, and the strips were incubated with labeled concatamerized H3.2 Omega  oligonucleotide in the presence of unlabeled H3.2 or H3.3 competitor duplex oligonucleotides. All reactions contained poly(dI-dC)-poly(dI-dC) (50 µg/ml). Lane 1 shows the results of a control hybridization reaction containing no competitors; lanes 2 and 3, 10- or 100-fold molar excess of unlabeled concatamerized H3.2 Omega  oligonucleotides; lanes 4 and 5, 10- or 100-fold molar excess of unlabeled concatamerized H3.3 Omega  oligonucleotides. C, coding sequence comparison between the H3.2 Omega  and H3.3 Omega  oligonucleotides.
[View Larger Version of this Image (31K GIF file)]

The highly specific nature of the interaction between the protein(s) in the 1 M fraction and the H3.2 Omega  sequence was confirmed in a separate southwestern competition experiment, shown in Fig. 2B. Unlabeled duplex H3.2 Omega  oligonucleotides competed very efficiently with the radioactive H3.2 probe, eliminating the interaction when in 100-fold excess (lane 3). Conversely, the H3.3 Omega  competitor duplex did not compete with the H3.2 probe (lanes 4 and 5) even when present in 100-fold molar excess. Because the two oligonucleotide duplexes differ at only 4 nucleotide positions (Fig. 2C), these results show that the DNA-protein interactions detected in the Southwestern analysis are highly specific for the H3.2 CRAS Omega sequence.

A Unique H1 Subtype Interacts Preferentially with the H3.2 Omega  Element

Further purification of the proteins in the 1 M fraction was achieved by RP-HPLC. The RP-HPLC profile of the 1 M fraction is presented in Fig. 3A as peaks of absorbance at 214 nm. Subsequent SDS-polyacrylamide gel electrophoresis analysis of the collected fractions A, B, and C is shown in Fig. 3B, lanes 1-3. Silver staining of the SDS gel revealed that fractions A-C contain proteins with apparent molecular masses of about 30-35 kDa, whereas lane 4 (1 M fraction) contained multiple proteins of much higher molecular mass (data not shown).


Fig. 3. The Omega -interacting protein is purified to homogeneity by RP-HPLC. The proteins in the 1 M DNA-cellulose fraction were further separated by RP-HPLC. A, RP-HPLC profile of the 1 M KCl DNA-cellulose fraction. Retention times are recorded above the peaks. Peaks represent absorbance at 214 nm. B, silver stain of SDS gel, containing the RP-HPLC fractions as indicated above the lanes. Lane 4, the RP-HPLC load, the 1 M DNA-cellulose fraction. Molecular weight markers (M) are shown in lanes on each side of the gel. C, Southwestern analysis of the RP-HPLC fraction A. Proteins in RP-HPLC fraction A were separated on a 10% SDS gel, electroblotted, and treated as described under "Experimental Procedures." The same amount of protein was loaded in all lanes. Lane 1 was hybridized with concatamerized labeled H3.2 Omega  oligonucleotides, and lane 2 was incubated with concatamerized labeled H3.3 Omega  oligonucleotides.
[View Larger Version of this Image (15K GIF file)]

A repeat of the Southwestern assay in Fig. 2 using the H3.2 and H3.3 Omega  oligonucleotides as probes with the RP-HPLC fraction A is shown in Fig. 3C. The protein(s) contained in fraction A showed the same interaction with the H3.2 Omega  sequence as seen in Fig. 2, indicated by the intense radioactive band observed in lane 1. As was true for the proteins in the 1 M DNA-cellulose fraction (Fig. 2A, lane 8), the protein(s) in RP-HPLC fraction A showed 100-fold less affinity for the H3.3 Omega  sequence (lane 2).

The strong affinity of mouse nuclear proteins contained in the 1 M DNA-cellulose fraction for DNA (elution with 1 M salt) and their rapid mobility in SDS-polyacrylamide gel electrophoresis suggested that these proteins might be histones. Although SDS-polyacrylamide gel electrophoresis theoretically separates proteins according to their molecular weight, histones behave anomalously in such gels because of post-translational modifications such as phosphorylation (25). Acid-urea gels have been used successfully to separate histones on the basis of differences in net charge and mass (50). This method accomplishes the separation of modified and unmodified forms of the basic histone proteins and allows for the partial separation of the various H1 subtypes as well as their differentially phosphorylated forms (51). In fact, only highly basic proteins (like histones) enter these gels. In Fig. 4, acid-urea gel analysis of column fractions is shown. Total H1 prepared from mouse chromatin is shown in lane 1. The 1 M DNA-cellulose fraction was loaded in lane 2. The proteins observed in this lane migrate identically to those observed in lane 1 (total H1). Acid gel separation of proteins in RP-HPLC fraction A is shown in lane 3. The subset of bands seen in this lane is also observed in lane 1 (total H1). The results shown in Fig. 4 are consistent with the hypothesis that protein(s) in the 1 M fraction that interact preferentially with the H3.2 Omega  sequence are H1 histones.


Fig. 4. The proteins in RP-HPLC fraction A are H1 histones. Shown is acid-urea gel electrophoresis of mouse total H1 (lane 1), the 1 M DNA-cellulose fraction (lane 2), and RP-HPLC fraction A (lane 3). The gel was stained with Amido Black as described under "Experimental Procedures."
[View Larger Version of this Image (84K GIF file)]

To identify conclusively the proteins interacting in such a highly specific manner with the H3.2 Omega  sequence, we digested the protein contained in RP-HPLC fraction A with Staphylococcus aureus V8 protease. The resulting peptides were fractionated by RP-HPLC and analyzed by automated protein sequenation. Fig. 5A shows the RP-HPLC elution profile of these peptides. Retention times are recorded above the peaks, and the sequenced fragments are numbered at the bases of their respective peaks. Sequence was obtained for amino acids at positions 9-115. The amino acid sequence of the purified protein is shown in panel B, the underlined amino acids designating the primary sequence information obtained in this study. Comparison of the protein sequence we obtained by peptide sequenation to published H1 sequences identified the H1 protein in RP-HPLC fraction A as a single mouse H1 subtype, namely H1b. No peptides were recovered from the extremely basic carboxyl-terminal 100 amino acids, despite repeated attempts, but amino acid composition analyses of RP-HPLC fraction A, performed by the Pico-Tag® method, were consistent with the interpretation that the purified protein contained these amino acids (Table I). In addition, mass spectrometric analysis produced results consistent with that of full-length H1 protein (data not shown). The gene encoding this unique mouse H1 has since been cloned independently by two laboratories (see "Discussion"). Although several protein bands are observed in the SDS gel shown in Fig. 3B, lane 1 (RP-HPLC fraction A), these are very likely to be different post-translationally modified forms of H1b. The protein contained in the RP-HPLC fraction A, via automated peptide sequencing, produced sequence from over 100 amino acid positions, in some cases from two or more peptides (Fig. 5B). Every amino acid position sequenced matched the sequence of the mouse H1b sequence (52).2 This is unambiguous evidence that a single protein is contained in fraction A. 


Fig. 5. Amino acid sequence identifies the protein in RP-HPLC fraction A as mouse H1b. RP-HPLC fraction A was digested with V8 protease, and the resulting peptides were separated by RP-HPLC and sequenced. A, elution profile of V8 peptides. Retention times are recorded above the peaks, and the sequenced peptides are numbered at the bases of their respective peaks. B, complete protein sequence of mouse H1b. Numbers in boldface type under the sequence in single letter code represent the different peptides shown in panel A, and the lines extending from these numbers show the amino acid sequence obtained from these peptides.
[View Larger Version of this Image (19K GIF file)]

Table I. Comparison of the amino acid composition analysis of RP-HPLC fraction A with the predicted amino acid sequence of H1b


Amino acid Molar percentage
Fraction A H1b

%
Asx 2 2
Glx 4 3.5
Ser 7.2 7.6
Gly 7.8 7.6
Arg 2.1 1.5
Thr 6.2 5.1
Ala 22.5 25
Pro 9.4 8.6
Tyr 0.5 0.5
Val 5.6 5.1
Met 0.2 0.5
Ile 0.7 1.0
Leu 4.2 4.6
Phe 0.44 0.5
Lys 27.5 26.5

H1b Interaction with the H3.2 Omega  Oligonucleotides in EMSA

Next, we examined the interactions of this H1 subtype with the H3.2 Omega  sequence in EMSA. The results, using end-labeled duplex H3.2 Omega  oligonucleotides as probe, are shown in Fig. 6. Purified H1 protein (RP-HPLC fraction A) was added in increasing amounts (lanes 1-6). In lane 3 (35 ng of added pure H1b), there was no indication of any interaction, even on much longer exposures, but most of the probe was shifted in a diffuse smear from the position of free probe to the top of the gel when 50 ng of pure H1b was added (lane 4). This was also true at 55 ng (lane 5), and when 60 ng of H1b was added (lane 6), all of the mobility-shifted probe was at the top of the gel. A duplicate experiment performed with another pure H1 subtype (H1a) revealed no evidence of interaction with the H3.2 Omega  element under these conditions, even in the presence of Omega -binding activity (data not shown).


Fig. 6. Mouse H1b interactions with the Omega  sequence in EMSA. Radioactively end-labeled H3.2 Omega  oligonucleotides (1.5 ng in all lanes) were incubated on ice with increasing amounts of pure H1b histone in the presence of nonspecific competitor poly(dI-dC) (1 µg) and separated on a native gel as described under "Experimental Procedures." Pure H1b was added to the binding reaction with Omega oligonucleotides in the following amounts: 7 ng (lane 1), 20 ng (lane 2); 35 ng (lane 3); 50 ng (lane 4); 55 ng (lane 5); 60 ng (lane 6).
[View Larger Version of this Image (26K GIF file)]

Because another nuclear protein (the Omega  factor, Fig. 1, lanes 1 and 5) was identified on the basis of its interaction with the H3.2 Omega  sequence in EMSA, we examined the possibility that interactions of the H1b protein would be detectable in the presence of the Omega  factor. In Fig. 7, we show the results of such an experiment. The source of Omega -binding activity in this experiment was crude nuclear extract, added in identical amounts to all lanes. Crude nuclear extract and probe were incubated on ice for 15 min and then increasing amounts of H1b were added over the same range of concentrations used in Fig. 6, and the reactions were incubated an additional 15 min (lanes 2-8). The Omega  complex, the result of interaction between the Omega  factor and the H3.2 Omega  sequence, is observed in lane 1 (compare with Fig. 1, lane 1). There is some evidence of nuclear protein-DNA-H1b interactions at 10-20 ng of added H1b (lanes 3 and 4), but at 35 ng a "supershifting" of the Omega  complex is observed (lane 5). This supershifting increased with the addition of higher concentrations of H1b (lanes 6 and 7), and the Omega  complex was no longer observed at 100 ng of added H1b (lane 8).


Fig. 7. H1b binds noncompetitively with the Omega  factor to the CRAS Omega  element. End-labeled H3.2 Omega  duplex oligonucleotides were first incubated at 4 °C with crude mouse myeloma nuclear extract (6 µg, all lanes) for 15 min in the presence of 1 µg poly(dI-dC)-poly(dI-dC). Then increasing amounts of H1b were added, and the reactions were incubated for an additional 15 min. Pure H1b was added to the binding reaction with labeled Omega  oligonucleotides and crude extract in the following amounts: no added H1 (lane 1), 5 ng of H1b (lane 2), 10 ng (lane 3), 20 ng (lane 4), 35 ng (lane 5), 50 ng (lane 6), 70 ng (lane 7), 100 ng (lane 8).
[View Larger Version of this Image (43K GIF file)]

Two things are remarkable about these interactions. First, evidence of H1 binding to the Omega  DNA duplex is detected at a much lower H1 concentration than that required for detection of the interaction of the pure H1b protein with the Omega  sequence alone (compare Fig. 7, lanes 3-5 (10-35 ng) with Fig. 6, lane 4 (50 ng of H1b)). Second, the amount of H1 required to shift most of the probe to the top of the gel is also higher when the Omega  factor is present (compare lane 6 of Fig. 6 (60 ng of H1b) to lane 8 of Fig. 7 (100 ng)). When pure H1b was added to the binding reaction with the Omega  probe at a molar ratio of 1:1 (Fig. 6, lane 2), no evidence of interaction was detected. Only when the molar ratio of H1 exceeded that of DNA probe by greater than 2:1 (Fig. 6, lane 4) were interactions observed. Conversely, when the CRAS Omega -binding activity was present (Fig. 7, lane 1), H1-CRAS Omega  complex interactions were detected at a much lower molar ratio (Fig. 7, lane 3, 0.5:1; lane 4, 1:1), indicating that binding by H1b to the Omega  DNA sequence may be enhanced by the presence of the preincubated Omega  factor-DNA complex. We know that our crude nuclear extracts contain H1; these extracts are the source of our RP-HPLC-purified H1b. The apparent enhancement of H1b interaction with the Omega  complex might simply be due to the presence of H1b molecules in crude nuclear extract, but the molar amount of H1b present in the crude extract added to the EMSA reactions is quite low. Therefore, it is unlikely that this is the explanation for the differences observed between Figs. 6 and 7. As the next experiment will demonstrate (Fig. 8), it is possible that DNA structure plays a role in the H1b-Omega -DNA interaction. A change in the Omega  DNA structure upon interaction with the Omega -binding activity may be responsible for the apparent enhancement of H1b interactions observed in Fig. 7. Until the Omega  factor is purified, however, we cannot distinguish between these possibilities.


Fig. 8. The addition of H1b to the Omega -DNA complex produces specific DNase I-hypersensitive sites in DNase I footprint analysis. 40-mer duplex Omega  oligonucleotides end-labeled on one strand as probe were incubated with crude mouse nuclear extract for 15 min followed by the addition of pure H1b exactly duplicating reactions in Fig. 7 and then treated with DNase I as described previously (31). Free and bound probe were separated by EMSA (in the +H1b panel; see Fig. 7, lane 5), and the supershifted complex migrating behind the Omega  complex was removed from the gel and analyzed on an 8 M urea sequencing gel (see "Experimental Procedures"). G indicates that the lane contains the Maxam-Gilbert G reaction of the probe; F, free probe; B, bound probe complexes.
[View Larger Version of this Image (64K GIF file)]

Evidence of Specific Effects upon DNA Structure by H1b Participation in the Omega  DNA-Omega Factor Interaction

Elsewhere (30, 31), we have shown that two elements, alpha  and Omega , within the coding sequence of core histone genes interact with nuclear proteins in vitro and are required for normal gene expression in vivo in stably transfected Chinese hamster ovary cells. Mutation of the seven conserved nucleotides of the histone Omega  sequence caused a 4-fold drop in expression of a mouse H3.2 gene in stable transfectants and abolished the Omega  complex observed in gel mobility shift experiments. In Fig. 8, we compare the DNA-protein interactions at the Omega  element (shown in gel shift assay in Fig. 1, lane 1) by DNase I footprint analyses, performed in the absence and in the presence of purified H1b. The labeled strand in these experiments was the noncoding strand, and the Omega  sequence is indicated by a bracket. G indicates that the lane was loaded with the Maxam-Gilbert G reaction cleavage products of the probe; B indicates that the lane was loaded with the DNase I-treated complex of probe-Omega factor after elution from a native gel slice (see "Experimental Procedures"); and F indicates free probe after elution as described for the B lanes.

As previously shown (31), nucleotides within the histone Omega  sequence (GCTCTAG on the noncoding strand) show protection from cleavage by DNase I (Fig. 8, minus H1b (-H1b) panel, B lane). In the panel showing experiments performed in the presence of H1b (+H1b), labeled Omega  oligonucleotides were preincubated with crude nuclear extract before the addition of pure H1b exactly as described in Fig. 7. DNase I digestion followed, and then the reaction was loaded onto a native gel for separation of free and bound probe molecules by EMSA. For reactions shown in the +H1b panel, the supershifted Omega  DNA-protein complex, which migrates slightly behind the histone Omega  complex (Fig. 7, lanes 4 and 5), was sliced from the native gel and treated as described above (for details, see "Experimental Procedures").

Strong protection of the first two nucleotides (GC) in the Omega  element by protein interaction is observed in both +H1b and -H1b panels, but several DNase I-hypersensitive sites not seen in the -H1b panel are observed in the +H1b panel, lane B. The C nucleotide immediately outside the Omega  element shows hypersensitivity to DNase I cleavage, as do three more nucleotides in the nine nucleotides 5' of the Omega  element. These hypersensitive sites must result from the specific interactions of H1b molecules with the preincubated histone Omega  complex. The increased sensitivity of these nucleotides indicates that the addition of H1b to the Omega  DNA-protein complex has a very specific effect on the DNA structure immediately 5' of the Omega  element. The structure of the duplex Omega  oligonucleotides in solution, whether free or bound, is such that DNA protein interactions 3' of the seven nucleotides composing the Omega  element cannot be detected. More detailed analyses of the H1b-Omega -DNA interaction will be possible when the Omega -binding activity is purified to homogeneity.


DISCUSSION

We have previously shown that the H3.2 CRAS Omega  element is required for normal expression of the mouse H3.2 gene in vivo (31). We have also reported that this regulatory element is present in the coding regions of replication-dependent histone genes of all four nucleosomal classes and that the DNA-protein interactions are identical for these genes. In addition, we have reported that the Omega  factor (DNA-binding activity) is present in G1, but not in S or M phase, nuclear extracts (32). The fact that a replication-independent histone gene, an H3.3, has a mutated Omega  sequence that fails to bind or compete for binding of the Omega  factor implicates this element in the coordinate regulation of the replication-dependent histone genes (31, 32).

Here, we have reported the purification by DNA-cellulose chromatography and RP-HPLC of a second nuclear factor that interacts specifically with the Omega  element, a specific mouse H1 subtype, H1b. H1b does not interact with high affinity with the comparable sequence from a replication-independent H3.3 gene. In fact, the H3.2 DNA-H1 interactions are so specific that the H1b protein interacted with over 100-fold lower affinity with the H3.3 Omega  oligonucleotides (Figs. 2 and 3). Yet, the H3.2 and H3.3 oligonucleotides differ at only 4 of 22 nucleotide positions (Fig. 2C). These results, also demonstrated in a very different set of experiments (EMSA), clearly show the binding specificity of the H1b histone in these interactions.

In 1993, Brown and Sittman developed an HPLC protocol to separate mouse H1 protein sequence variants (54). On the basis of relative mobilities of purified HPLC peak fractions in SDS and acid gel electrophoresis, they determined which variant was represented in each HPLC peak. Their HPLC peaks had the overall mobility characteristics on gels of the major H1 variants previously demonstrated by others (25, 55). Our RP-HPLC purification (Fig. 3A) and acid-urea gel electrophoresis (Fig. 4) reproduced these results. Peptide sequenation yielded information on over 100 amino acid positions of RP-HPLC fraction A, identifying the protein as H1b. This H1 is highly related to rat H1b, rabbit H1.3, and human H1-3 variants (56). Comparison of the protein sequence to published H1 sequences identified the H1 subtype in RP-HPLC fraction A as a unique mouse H1 subtype (H1b), the cloning of which was recently reported by the laboratory of D. Doenecke (52) and which has also been independently cloned by the laboratory of W. F. Marzluff.2 The Doenecke group designated the H1b subtype H1var.4, but we have chosen to adhere to the original nomenclature introduced by Lennox and Cohen (25, 57). The mouse H1b subtype identification was confirmed by mass spectrometry of fraction A (data not shown), in the laboratory of A. I. Skoultchi.

The identification of numerous histone H1 variants in vertebrates suggests that these proteins accomplish specialized functions during development and that their relative abundances are perhaps dependent upon the differentiated cell type and whether cells are proliferating or growth-arrested. Also, H1 subtypes differ in their relative rates of synthesis and degradation in dividing and in nondividing cells (56). For example, H1b is unstable after cells cease to proliferate and is not synthesized in nonlymphoid, nondividing cells (25). In the newborn mouse, H1b declines after 4 weeks of age and cannot be detected by Coomassie Blue after 16 weeks (58). H1b is very low in adult mouse tissues, representing for example only 3% of total H1 in liver chromatin (59).

In addition, there are differences in the amount and pattern of phosphorylation among H1 subtypes in humans and mice (18, 25, 53, 56). H1b was observed as having the largest number of phosphorylated forms among the H1 subtypes (25). It has been known for many years that there are variations in the amount of phosphorylation of specific H1 subtypes during the cell division cycle (for a review, see Ref. 25), and a large amount of data shows that H1 dephosphorylation correlates with chromatin condensation (13). Consistent with these observations, phosphorylated H1 was not detected in nonproliferating adult tissues (58). It is completely consistent with these data to hypothesize that H1b histones play a key role in the regulation of genes expressed only in proliferating cells, i.e. the replication-dependent histone gene family.

Our results are in agreement with other reports that H1 histones prefer some non-A/T eukaryotic sequences over others (37-41). However, this may be the first set of experiments to examine site-specific DNA interactions of a single H1 subtype. Our data support the hypothesis that specific H1 subtypes have specific roles to play in the regulation of specific genes or gene families. We have shown that a specific H1 subtype, H1b, exhibits a highly specific affinity for a cis-acting DNA regulatory sequence, the CRAS Omega element, in the Southwestern assay. We have also shown by EMSA that H1b is capable of binding to the Omega  element in a fashion noncompetitive with the CRAS Omega  factor (Fig. 7). The fact that a specific H1 subtype is responsible for this high affinity interaction, observed in vitro, provides evidence that, although different H1 subtypes may be present in the same transformed cell line in asynchronous logarithmic growth (mouse myeloma cells), these H1 subtypes may vary in their abilities to interact with specific DNA sequences.

Our observations raise a number of interesting possibilities. First, in the interaction between the Omega  element and H1b, is the phosphorylated or nonphosphorylated form of H1b required? Second, is H1b involved in nucleosome positioning on replication-dependent histone genes by way of the Omega  sequence? Finally, does the Omega  factor play a role in this interaction? A very attractive hypothesis is that the mouse H1b protein is involved in repression of the nucleosomal replication-dependent histone genes. The involvement of a specific H1 subtype in regulation of all replication-dependent histone gene expression could provide at least part of a common mechanism for up-regulation of this large gene family at the G1-S boundary of the eukaryotic cell cycle. We will examine this possibility in future studies.


FOOTNOTES

*   This work was supported by National Institutes of Health Grant RO1-GM46768 (to M. M. H.).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.
Dagger    Present address: NICHD, National Institutes of Health, Bethesda, MD 20892.
§   Vanderbilt University Medical Center, Nashville, TN 37232-6300.
   To whom correspondence should be addressed: Dept. of Biological Science, Florida State University, Tallahassee, FL 32306-3050. Tel.: 904-644-1554; Fax: 904-644-5766; E-mail: mhurt{at}mailer.fsu.edu.
1   The abbreviations used are: CRAS, coding region activating sequence; DTT, dithiothreitol; EMSA, electrophoretic mobility shift assay; HPLC, high pressure liquid chromatography; RP-HPLC, reverse phase HPLC.
2   W. F. Marzluff, personal communication.

ACKNOWLEDGEMENTS

We thank Dr. Richard Cooke at the Core Protein Sequenation Facility at Baylor College of Medicine for expertise and Drs. Arthur I. Skoultchi and William F. Marzluff for making available unpublished results. Finally, we thank Dr. Allen M. Sirotkin in the laboratory of Dr. Arthur Skoultchi for assisting in the mass spectrometry of our RP-HPLC fractions.


REFERENCES

  1. Simpson, R. T. (1978) Biochemistry 17, 5524-5531 [Medline] [Order article via Infotrieve]
  2. Thoma, F., Koller, T., and Klug, A. (1979) J. Cell Biol. 83, 403-427 [Abstract]
  3. van Holde, K. E. (1989) Chromatin, Springer-Verlag KG, Berlin
  4. Bresnik, E. H., Bustin, M., Marsaud, V., Richard-Foy, H., and Hager, G. L. (1992) Nucleic Acids Res. 20, 273-278 [Abstract]
  5. Weintraub, H. (1985) Cell 42, 705-711 [Medline] [Order article via Infotrieve]
  6. Lee, H. H., and Archer, T. K. (1994) Mol. Cell. Biol. 14, 32-41 [Abstract]
  7. Archer, T. K., Cordingley, M. G., Wolford, R. G., and Hager, G. L. (1991) Mol. Cell. Biol. 11, 688-698 [Medline] [Order article via Infotrieve]
  8. Kadonaga, R. D., and Lorch, Y. (1992) Annu. Rev. Cell Biol. 8, 563-587 [CrossRef]
  9. Wolffe, A. P. (1994) Cell 77, 13-16 [Medline] [Order article via Infotrieve]
  10. Clark, K. L., Halay, E. D., Lai, E., and Burley, S. K. (1993) Nature 364, 412-417 [CrossRef][Medline] [Order article via Infotrieve]
  11. McPherson, C. E., Shim, E.-Y., Friedman, D. S., and Zaret, K. S. (1993) Cell 75, 387-398 [Medline] [Order article via Infotrieve]
  12. Wolffe, A. P., and Brown, D. D. (1987) Cell 51, 733-740 [Medline] [Order article via Infotrieve]
  13. Wolffe, A. P. (1995) Chromatin: Structure and Function, Academic Press, London
  14. Mendelson, E., Landsman, D., Druckman, S., and Bustin, M. (1986) Eur. J. Biochem. 160, 253-260 [Abstract]
  15. Ericsson, C., Grossbach, U., Bjorkroth, B., and Daneholt, C. (1990) Cell 60, 73-83 [Medline] [Order article via Infotrieve]
  16. Nacheva, G. A., Guschin, D. Y., Preobrazhenskaya, O. V., Karpov, V. L., Ebralidse, K. K., and Mirzabekov, A. D. (1989) Cell 58, 27-36 [Medline] [Order article via Infotrieve]
  17. Wolffe, A. P. (1991) Trends Cell Biol. 1, 61-66 [Medline] [Order article via Infotrieve]
  18. Ajiro, K., Borun, T. W., and Cohen, L. H. (1981) Biochemistry 20, 1445-1454 [Medline] [Order article via Infotrieve]
  19. Breneman, J. W., Yau, P., Teplitz, R. L., and Bradbury, E. M. (1993) Exp. Cell Res. 206, 16-26 [CrossRef][Medline] [Order article via Infotrieve]
  20. Bouvet, P., Dimitrov, S., and Wolffe, A. P. (1994) Genes & Dev. 8, 1147-1159 [Abstract]
  21. Laybourn, P. J., and Kadonaga, J. T. (1991) Science 254, 238-245 [Medline] [Order article via Infotrieve]
  22. Grunstein, M. (1990) Annu. Rev. Cell Biol. 6, 643-678 [CrossRef]
  23. Zweidler, A. (1984) in Histone Genes: Structure, Organization and Regulation (Stein, G. S., Stein, J., and Marzluff, W. F., eds), pp. 339-371, John Wiley and Sons, Inc., New York
  24. Wu, R. S., and Bonner, W. M. (1981) Cell 27, 321-330 [Medline] [Order article via Infotrieve]
  25. Lennox, R. W., and Cohen, L. H. (1984) in Histone Genes: Structure, Organization, and Regulation (Stein, G. S., Stein, J., and Marzluff, W. F., eds), pp. 373-395, John Wiley and Sons, Inc., New York
  26. Osley, M. A. (1991) Annu. Rev. Biochem. 60, 827-861 [CrossRef][Medline] [Order article via Infotrieve]
  27. Heintz, N. (1991) Biochim. Biophys. Acta 1088, 327-339 [Medline] [Order article via Infotrieve]
  28. Hurt, M. M., Pandey, N. B., and Marzluff, W. F. (1989) Proc. Natl. Acad. Sci. U. S. A. 86, 4450-4454 [Abstract]
  29. Hurt, M. M., Bowman, T. L., and Marzluff, W. F. (1991) Mol. Cell. Biol. 11, 2929-2936 [Medline] [Order article via Infotrieve]
  30. Bowman, T. L., and Hurt, M. M. (1995) Nucleic Acids Res. 23, 3083-3092 [Abstract]
  31. Kaludov, N. K., Pabón-Peña, L., and Hurt, M. M. (1996) Nucleic Acids Res. 24, 523-531 [Abstract/Free Full Text]
  32. Kaludov, N. K., Bowman, T. L., and Hurt, M. M. (1996) Proc. Natl. Acad. Sci. U. S. A. 93, 4465-4470 [Abstract/Free Full Text]
  33. Bowman, T. L., Kaludov, N. K., Klein, M., and Hurt, M. M. (1996) Gene (Amst.) 176, 1-8 [CrossRef][Medline] [Order article via Infotrieve]
  34. Zlatanova, J. (1990) Trends Biochem. Sci. 15, 273-276 [CrossRef][Medline] [Order article via Infotrieve]
  35. Zlatanova, J., and Yaneva, J. (1991) Mol. Biol. Rep. 15, 53-56 [Medline] [Order article via Infotrieve]
  36. Yaneva, J., Schroth, G., van Holde, K., and Zlatanova, J. (1995) Proc. Natl. Acad. Sci. U. S. A. 92, 7060-7064 [Abstract]
  37. Kandolf, H. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 7257-7261 [Abstract]
  38. Berent, S. L., and Sevall, J. S. (1984) Biochemistry 23, 2977-2983 [Medline] [Order article via Infotrieve]
  39. Ristiniemi, J., and Oikarinen, J. (1989) J. Biol. Chem. 264, 2164-2174 [Abstract/Free Full Text]
  40. Pauli, U., Chiu, J. F., Ditullio, P., Kroeger, P., Shalhoub, V., Rowe, T., Stein, G., and Stein, J. (1989) J. Cell Physiol. 139, 320-329 [Medline] [Order article via Infotrieve]
  41. Diez-Caballero, T., Aviles, F. X., and Albert, A. (1981) Nucleic Acids Res. 9, 1383-1393 [Abstract]
  42. Shapiro, D., Sharp, P. A., Wahli, W., and Keller, M. (1988) DNA 7, 47-55 [Medline] [Order article via Infotrieve]
  43. Wellman, S. E., Casano, P., Pilch, D., Marzluff, W. F., and Sittman, D. (1987) Gene (Amst.) 59, 29-39 [Medline] [Order article via Infotrieve]
  44. Krimer, D., Cheng, G., and Skoultchi, A. I. (1993) Nucleic Acids Res. 21, 2873-2879 [Abstract]
  45. Hraba-Renevy, S., and Kress, M. (1989) Nucleic Acids Res. 17, 2449-2461 [Abstract]
  46. Gilman, M., Wilson, R., and Weinberg, R. (1986) Mol. Cell. Biol. 6, 4305-4316 [Medline] [Order article via Infotrieve]
  47. Lenormand, J.-L., Guillier, M., and Leibovitch, S. A. (1993) Nucleic Acids Res. 21, 695-702 [Abstract]
  48. Cole, R. D. (1989) Methods Enzymol. 170, 524-532 [Medline] [Order article via Infotrieve]
  49. Lennox, R. W., and Cohen, L. H. (1989) Methods Enzymol. 170, 532-549 [Medline] [Order article via Infotrieve]
  50. Panyim, S., and Chalkey, R. (1969) Arch. Biochem. Biophys. 130, 337-343 [Medline] [Order article via Infotrieve]
  51. Lennox, R. W., Oshima, R. G., and Cohen, L. H. (1982) J. Biol. Chem. 257, 5183-5189 [Abstract/Free Full Text]
  52. Drabent, B., Franke, K., Kosciessa, U., Bouterfa, H., Hameister, H., and Doenecke, D. (1995) Mammal. Genome 6, 505-511 [Medline] [Order article via Infotrieve]
  53. Ajiro, K., Shibata, K., and Nishikawa, Y. (1990) J. Biol. Chem. 265, 6494-6500 [Abstract/Free Full Text]
  54. Brown, D. T., and Sittman, D. B. (1993) J. Biol. Chem. 268, 713-718 [Abstract/Free Full Text]
  55. Kistler, W. S., and Geroch, M. E. (1975) Biochem. Biophys. Res. Commun. 63, 378-384 [Medline] [Order article via Infotrieve]
  56. Parseghian, M. H., Henschen, A. H., Krieglstein, K. G., and Hamkalo, B. A. (1994) Protein Sci. 3, 575-587 [Abstract/Free Full Text]
  57. Lennox, R. W. (1984) J. Biol. Chem. 259, 669-672 [Abstract/Free Full Text]
  58. Lennox, R. W., and Cohen, L. H. (1983) J. Biol. Chem. 258, 262-268 [Abstract/Free Full Text]
  59. Sirotkin, A., Edelmann, W., Cheng, G., Klein-Szanto, A., Kucherlapati, R., and Skoultchi, A. (1995) Proc. Natl. Acad. Sci. U. S. A. 92, 6434-6438 [Abstract]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.