Distribution of somatic H1 subtypes is non-random on active vs. inactive chromatin II: distribution in human adult fibroblasts.

For nearly twenty years researchers have observed changes in the histone H1 subtype content of tissues as an organism develops into an adult. To better understand the consequences of such changes, immunofractionation of chromatin using previously characterized antibodies specific for human H1 subtypes was employed in the analysis of a fibroblast cell strain derived from a 37-year-old individual. DNAs isolated from immunoprecipitates were probed for the existence of a variety of DNA sequences. The results presented lend further support to a previously-proposed model (Parseghian et al. [2000] Chromosome Res 8:405-424) in which transcription of a sequence is accompanied by the selective depletion of subtypes. The data also suggest that there is more total H1 on actively transcribed sequences in these cells as compared to fetal fibroblasts and that there is less difference in the subtype compositions of active genes vs. inactive sequences in this strain. Specifically, the consequences of these changes appear to correlate with the attenuation of the heat shock response in aging fibroblasts. In a broader context, these results could explain why there are reductions in transcription in cells from mature tissue that approach senescence.

A compendium of the histone H1 family of somatic subtypes: an elusive cast of characters and their characteristics.

The last 35 years has seen a substantial amount of information collected about the somatic H1 subtypes, yet much of this work has been overshadowed by research into highly divergent isoforms of H1, such as H5. Reports from several laboratories in the past few years have begun to call into question some of the traditional views regarding the general function of linker histones and their heterogeneity. Hence, the impression in some circles is that less is known about these ubiquitous nuclear proteins as compared with the core histones. The goal of the following review is to acquaint the reader with the ubiquitous somatic Hls by categorizing them and their characteristics into several classes. The reasons for our current state of misunderstanding is put into a historical context along with recent controversies centering on the role of H1 in the nucleus. Finally, we propose a model that may explain the functional role of H1 heterogeneity in chromatin compaction.

The distribution of somatic H1 subtypes is non-random on active vs. inactive chromatin: distribution in human fetal fibroblasts.

Chromatin immunoprecipitation was employed to determine whether or not the previously reported depletion of histone H1 on actively transcribed sequences was selective with respect to H1 subtypes. DNA of immunofractionated chromatin was analyzed by slot-blots for repetitive sequences and PCR for single and low-copy sequences. Based on the analysis of a diverse set of sequences, we report distinct differences in subtype distributions. Actively transcribed chromatin, as well as chromatin poised for transcription, is characterized by a relative depletion of somatic H1 subtypes 2 and 4 (H1s-2 and H1s-4),whereas facultative and constitutive heterochromatin contain all four somatic subtypes. These results support a model in which subtypes are selectively depleted upon gene expression. In turn, the data also support the possibility that the somatic subtypes have different functional roles based on their selective depletion from different classes of DNA sequences.

Identification of histone H1 as a cognate antigen of the ulcerative colitis-associated marker antibody pANCA.

Perinuclear anti-neutrophil cytoplasmic antibody (pANCA)(4)is a predominant serum marker of ulcerative colitis (UC), and a familial trait associated with disease susceptibility and disease associated MHC haplotypes. This study characterizes the pANCA antigen defined by representative UC-pANCA human monoclonal antibodies, Fab 5-3 and 5-2. Western blot analysis probed with Fab 5-3 revealed specific binding to a nuclear protein doublet (apparent MW=32-33 kDa) expressed in several cell types. Purification and tryptic peptide sequencing identified the protein as histone H1, and this specificity was confirmed by Fab 5-3 binding to purified H1. Rabbit anti-histone H1 immunostaining and Western blot analysis confirmed that the pANCA epitope is preferentially immunoaccessible in polymorphonuclear neutrophils (PMN). The epitope was localized to the COOH-terminal region by site-specific proteolysis, and recombinant deletants further localized binding activity for both Fab 5-2 and 5-3 to two non-overlapping segments (AA 69-171 and 172-226) associated with a recurring PKKAK motif. Serum IgG binding was detectable to these segments, but was not significantly correlated with pANCA titer or disease status. These findings indicate that histone H1 bears a recurring COOH-terminal epitope recognized by monoclonal ulcerative colitis-associated pANCA marker antibodies, but this epitope is not a predominant specificity of serum pANCA.

Purification and initial characterization of primate satellite chromatin.

Nucleoprotein hybridization, a method for the purification of specific DNA sequences as chromatin, was employed to fractionate primate centromeric alpha satellite chromatin as a first step in the identification and analysis of novel centromere-enriched proteins. In order to optimize the amount of material available for further study, cultured African green monkey cells were employed because satellite DNA represents approximately 25% of the genome. Two chromatin preparations were compared for the yield and total protein content of purified material. Regardless of the preparation, alpha satellite sequences were enriched to near purity. Since intact satellite chromatin is relatively refractile to the enzymatic digestion steps in the method, the total amount of solubilized material available for purification is rather low. In contrast, nuclei treated with acidic washes to extract histone H1 provided solubilized material enriched in satellite sequences. In addition, this material is more efficiently utilized in an affinity chromatography step. However, the extraction of many non-histones at low pH resulted in very low yields of protein in the purified fraction. Two-dimensional gel comparisons of proteins associated with H1-containing satellite chromatin after iodination of total chromatin proteins revealed a number of polypeptides enriched to varying degrees in the purified fraction. The electrophoretic mobilities of a few enriched polypeptides corresponded to previously identified heterochromatin-associated proteins while many others appear to be novel. The work presented validates nucleoprotein hybridization as a purification method for highly repeated sequences as chromatin in analytical amounts. The fact that a number of the enriched proteins are visible in stained gels of bulk chromatin proteins suggests that further biochemical analysis can be carried out on these polypeptides directly.

Chromosomal locations of major tRNA gene clusters of Xenopus laevis.

In Xenopus laevis eight tRNA genes are located in a 3.18 kb tandemly repeated unit. There are 150 copies of the unit at a single locus near the long arm telomere of one of the acrocentric chromosomes in the 14-17 group. Two additional classes of tRNA gene-containing repeats have been isolated (defined by clones p3.1 and p3.2) that have structures related to that of the 3.18 kb unit. Using in situ hybridization at the electron microscopic level, the p3.2 repeats are found clustered at a single locus in the subtelomeric region on one of the submetacentric chromosomes, whereas the p3.1 repeats are clustered at a locus indistinguishable from that containing the 3.18 kb repeats. This suggests that these tDNA tandem repeats can diverge in sequence from each other without being at distantly separated loci.

Characterization of a set of antibodies specific for three human histone H1 subtypes.

A series of human histone H1 subtype-specific antibodies are described that were generated for localization and functional studies. Since our previous attempts to produce such antibodies against intact subtypes met with limited success, resulting in one antibody against a subtype we have designated H1-3, the approach used in the work presented is based on the production of antibodies against synthetic peptides or peptide fragments encompassing the variant NH2-terminal region of each protein. Subtype-specific antibodies were obtained against synthetic peptides derived from subtypes designated H1-1 and H1-2 and the NH2-terminal fragment from an N-bromosuccinimide digest of H1-4. Antibody specificities were documented in all cases by enzyme-linked immunosorbent and protein immunoblot assays against the purified subtypes as well as immunoblots against whole cell and nuclear extracts. In addition, the in vivo distribution of each antibody was determined by indirect immunofluorescence. H1-1 appears to be distributed in parallel with DNA concentration, similar to the results with an antibody that recognizes all subtypes. However, H1-2 and H1-4 are non-uniformly distributed, exhibiting similar punctate staining patterns. The staining patterns described are different from the pattern described for the distribution of H1-3, suggesting that several subtypes are concentrated in distinct regions of the nucleus and, therefore, may be associated with distinct regions of the genome.

A proposal for a coherent mammalian histone H1 nomenclature correlated with amino acid sequences.

Bio-Rex 70 chromatography was combined with reverse-phase (RP) HPLC to fractionate histone H1 zero and 4 histone H1 subtypes from human placental nuclei as previously described (Parseghian MH et al., 1993, Chromosome Res 1:127-139). After proteolytic digestion of the subtypes with Staphylococcus aureus V8 protease, peptides were fractionated by RP-HPLC and partially sequenced by Edman degradation in order to correlate them with human spleen subtypes (Ohe Y, Hayashi H, Iwai K, 1986, J Biochem (Tokyo) 100:359-368; 1989, J Biochem (Tokyo) 106:844-857). Based on comparisons with the sequence data available from other mammalian species, subtypes were grouped. These groupings were used to construct a coherent nomenclature for mammalian somatic H1s. Homologous subtypes possess characteristic patterns of growth-related and cAMP-dependent phosphorylation sites. The groupings defined by amino acid sequence also were used to correlate the elution profiles and electrophoretic mobilities of subtypes derived from different species. Previous attempts at establishing an H1 nomenclature by chromatographic or electrophoretic fractionations has resulted in several misidentifications. We present here, for the first time, a nomenclature for somatic H1s based on amino acid sequences that are analogous to those for H1 zero and H1t. The groupings defined should be useful in correlating the many observations regarding H1 subtypes in the literature.

Fractionation of human H1 subtypes and characterization of a subtype-specific antibody exhibiting non-uniform nuclear staining.

Four histone H1 subtypes and H1(0) were fractionated from human placental nuclei and purified to homogeneity by a combination of Bio-Rex 70 chromatography and reverse-phase high-performance liquid chromatography (RP-HPLC). Polyclonal antibodies were generated in rabbits against one of these subtypes designated H1-3. Antibodies reacted only against this subtype in enzyme-linked immunosorbent assays and Western assays; subtype specificity was documented further by Western blotting of cell and nuclear extracts. They crossreacted with monkey H1, but not with H1 from other vertebrates tested. The epitope(s) recognized were mapped by immunoblotting against peptides prepared by cleavage with N-bromosuccinimide (NBS) and alpha-chymotrypsin; it includes the variant amino-terminal tail of the protein as well as a portion of the globular domain. The antibody stains mitotic chromosomes weakly but uniformly and, unlike antibodies that recognize total H1 which show uniform nuclear staining after indirect immunofluorescence localization, anti-H1-3 exhibits preferential labelling of the nuclear periphery. This non-uniform staining suggests compartmentalization of this subtype which may have functional significance with respect to differential chromatin condensation.

Hoechst 33258, distamycin A, and high mobility group protein I (HMG-I) compete for binding to mouse satellite DNA.

The experiments described were designed to test the hypothesis that the (A+T)-specific DNA binding ligands Hoechst 33258 and distamycin A affect the condensation of mouse centromeric heterochromatin by competing for binding to satellite DNA with one or more chromosomal proteins. The studies focused on the nonhistone chromosomal protein HMG-I since its binding properties predict it would be a target for competition. Gel mobility shift assays show that HMG-I forms specific complexes with satellite DNA and that the formation of these complexes is competed for by both Hoechst and distamycin. In addition, methidium propyl EDTA Fe(II) [MPE Fe(II)] footprints of ligand-satellite DNA complexes showed essentially the same protection pattern for both drugs and a similar, but not identical, HMG-I footprint. If these in vitro results reflect the in vivo situation then the incomplete condensation of centromeric heterochromatin observed when mouse cells are grown in the presence of either chemical ligand could be a consequence of competition for binding of HMG-I (and possibly other proteins) to satellite DNA.

Cytological and molecular characterization of centromeres in Mus domesticus and Mus spretus.

We have applied EM in situ hybridization (EMISH) and pulsed field gel electrophoresis (PFGE) to samples from diploid primary cell cultures and an established cell line to examine in detail the relative organization of the major and minor satellite DNAs and telomere sequences in the genomes of Mus domesticus and Mus spretus. EMISH localizes the Mus domesticus minor satellite to a single site at the centromere-proximal end of each chromosome. Double label hybridizations with both minor satellite and telomere probes show that they are in close proximity and possibly are linked. In fact, PFGE of M. domesticus DNA digested with Sal I and Sfi I reveals the presence of fragments which hybridize to both probes and is consistent with the physical linkage of these two sequences. The M. domesticus minor satellite is the more abundant satellite in Mus spretus. Its distribution in M. spretus is characterized by diffuse labeling with no obvious concentration near chromosome ends. In addition to this repeat the M. spretus genome contains a small amount of DNA that hybridizes to a M. domesticus major satellite probe. Unlike the M. domesticus minor satellite, it is not telomere proximal but is confined to a domain at the border of the centromere and the long arm. Thus, although both species possess all three sequences, except for the telomeres, their distribution relative to one another is not conserved. Based on the results presented, we propose preliminary molecular maps of the centromere regions of Mus domesticus and Mus spretus.

Ultrastructural localization of nucleic acid sequences in Saccharomyces cerevisiae nucleoli.

The putative nucleolus in Saccharomyces cerevisiae is visible in electron micrographs as a darkly stained, crescent-shaped structure associated with the nuclear envelope. The haploid yeast genome contains 100-200 tandem copies of a 9.1 kb ribosomal DNA (rDNA) repeat predicted to reside in this structure. We combined in situ hybridization of non-isotopically labeled probes to isolated S. cerevisiae nuclei with immunogold detection to localize rDNA and rDNA precursor sequences in nuclei at the electron microscope (EM) level. Gold particles are restricted to defined regions of nuclei which appear more electron dense than the bulk of the nucleus and which generally exhibit the crescent shape typical of the structure thought to be the nucleolus. In addition, snR17, the yeast homolog of mammalian U3, a nucleolar-restricted small nuclear RNA (snRNA), was localized to the same electron dense region of the nucleus. These data, in conjunction with published immunofluorescent localizations of nucleolar-associated antigens, provide definitive proof that the dense crescent is the nucleolus. Finally, the technique described is applicable to probing nuclear organization in a genetically manipulable system.

Preferential distribution of active RNA polymerase II molecules in the nuclear periphery.

We have combined immunogold labeling with the Miller spreading technique in order to localize proteins at the electron microscope (EM) level in whole mount nuclei from mouse and human fibroblasts. Anti-histone H1 antibody labels nuclei uniformly, indicating that the nuclear interior is accessible to both antibodies and gold conjugates. Anti-topoisomerase I antibody labels nucleoli intensely, in agreement with previous immunofluorescent and biochemical data. Two different antibodies against the large subunit of RNA polymerase II (pol II) show preferential labeling of the nuclear periphery, as do antibodies against lamin, a known peripheral nuclear protein. Treatment of cells with alpha-amanitin results in loss of virtually all RNA polymerase II staining, supporting the specificity of labeling. Finally, when nuclei are incubated in the presence of biotin-UTP (bio-UTP) under run-off transcription conditions, incorporation is preferentially located at the nuclear periphery. These results support the conclusions that transcriptionally active pol II molecules are non-uniformly distributed in fibroblast nuclei, and that their differential distribution mirrors that of total pol II.

DNA sequence mapping using electron microscopy.

DNA sequences can be mapped on chromosomes at high resolution in the electron microscope after hybridization with a nonisotopically labeled probe followed by detection with a two-step antibody reaction employing a colloidal gold tag. Hybridization probes can be modified with biotin-dUTP, digoxigenin-dUTP, dinitrophenyl-dUTP, or N-acetoxy-2-acetylaminofluorene (AAF). The availability of different sizes of colloidal gold particles permits the simultaneous detection of several sequences. In addition, low signals can be amplified either with an antibody sandwich scheme or by silver intensification. This technology is applicable both to TEM and SEM preparations of chromosomes, and we have used it to map a number of highly and moderately repeated sequences on whole mount metaphase chromosomes.

High resolution mapping of Xenopus laevis 5S and ribosomal RNA genes by EM in situ hybridization.

We have developed a modification of in situ hybridization at the electron microscope level that permits simultaneous detection of at least two sequences. Probes are labelled with either biotin or AAF and detected with two distinct sizes of colloidal gold. This protocol has been applied to map the positions of Xenopus laevis oocyte-type 5S genes relative to ribosomal precursor genes in several independently derived cell lines. The results for the line TRXO, which expresses some oocyte 5S RNA, indicate that this inappropriate expression is not due to translocation from telomeric sites into the nucleolus organizer, as previously hypothesized. In addition we found that four other Xenopus cell lines, none of which express these genes, also contain distinct 5S oocyte translocations. These results suggest that an alteration in chromosome position is insufficient to result in gene activation and that sequences which are telomeric-proximal are exceptionally prone to translocation.

Localization of nucleic acid sequences by EM in situ hybridization using colloidal gold labels.

It is possible to combine hybridization to specimens on electron-microscope grids of nucleic-acid probes labelled nonisotopically with immunogold detection of hybrid sites to map the position of target sequences rapidly and precisely. The basic technique is described, and examples are provided to illustrate the types of questions which can be approached in the general area of higher-order chromosome organization and function. A combination of two differentially labelled probes and two different-sized gold particles permits the simultaneous detection of closely linked or interspersed sequences.

Deoxyribonucleic acid sequence mapping on metaphase chromosomes by immunoelectron microscopy.

Nucleic acid sequences can be localized on chromosomes in the electron microscope after hybridization with a biotinylated DNA probe followed by detection with a primary antibiotin antibody and a secondary antibody coupled to colloidal gold. Hybridization probes can also be labelled with alternative ligands such as N-acetoxy-2-acetylaminofluorene (AAF), Dinitrophenyl-dUTP and Digoxigenin-dUTP. Multiple labelling is possible if these differently modified DNA probes are used in conjunction with colloidal gold preparations of varying particle sizes. A substantial signal amplification can be achieved by incubating preparations with successive cycles of primary antibiotin antibody followed by a biotinylated secondary antibody. Detection is with Streptavidin-gold, and in the case of highly and moderately repeated sequences, the signal is visible in the light microscope. Detailed protocols are given for EM in-situ hybridization to whole mount metaphase chromosomes and include instructions necessary to perform multiple sequence localization and signal amplification.

Curvature of mouse satellite DNA and condensation of heterochromatin.

Cloned, sequenced mouse satellite DNA exhibits properties characteristic of molecules that possess a stable curvature. Circularly permuted fragments containing the region predicted to bend were used to map the curvature relative to DNA sequence. The altered mobility of these fragments in polyacrylamide gels is reversed when gels are run in the presence of distamycin A, a drug that binds preferentially to AT-rich DNA. Treatment of living mouse cells with this drug dramatically reduces the condensation of centromeric heterochromatin, the exclusive location of satellite sequences. In situ hybridization of satellite probes to extended chromosomes at the electron microscope level shows that satellite does not comprise a single block but is distributed throughout the centromere region. Based on these experiments, we hypothesize that the structure of mouse satellite DNA is an important feature of centromeric heterochromatin condensation.