Parathyroid-responsive modifications in the nuclear matrix of ROS 17/2.8 rat osteosarcoma cells

PTH is a mediator of skeletal development and remodeling that influences gene expression in osteoblastic cells. It is well established that PTH modulates the activity of membrane-associated second messenger signal transduction pathways. In these studies we have addressed the potential contribution of components of cell structure to the integration of PTH-related regulatory signals that influence the expression of bone cell genes. Chronic treatment of ROS 17/2.8 rat osteosarcoma cells with PTH is accompanied by changes in gene expression that are at least in part transcriptionally controlled. To explore the involvement of nuclear architecture in PTH-responsive modifications in gene expression, we investigated changes in the nuclear matrix after PTH treatment. Consistent with a role for the nuclear matrix in determining spatial organization and topology of chromatin as well as in the localization and targeting of transcription factors, we observed PTH-associated changes in a 200-kilodalton nuclear matrix protein in response to PTH. A significant down-regulation of synthesis was observed when nuclear matrix proteins were resolved electrophoretically in two-dimensional gels. This protein was restricted to the nuclear matrix and was not detected in the chromatin or cytoskeletal cellular fractions. These alterations in nuclear matrix proteins that occur after PTH treatment in osteosarcoma cells were phenotype related. They did not occur in UMR-106 POL or H4 hepatoma cells. Our findings support a role for the nuclear matrix in transducing PTH-mediated regulatory signals to facilitate the extent to which genes in osteoblasts are transcribed.

Subnuclear distribution of the vitamin D receptor

The subnuclear distribution of the vitamin D receptor was investigated to begin addressing the contribution of nuclear architecture to vitamin D-responsive control of gene expression in ROS 17/2.8 rat osteosarcoma cells. The nuclear matrix is an anastomosing network of filaments that is functionally associated with DNA replication, transcription, and RNA processing. The representation of vitamin D receptor in the nuclear matrix and nonmatrix nuclear fractions was determined by the combined application of 1) sequence-specific interactions with the vitamin D receptor binding element of the rat bone-specific osteocalcin gene promoter and 2) Western blot analysis. Both methods confirmed the presence of vitamin D receptor in the nonmatrix nuclear fraction and the absence of detectable vitamin D receptors associated with the nuclear matrix. In contrast, these same nuclear matrix proteins preparations exhibited association with the general transcription factor AP-1 and a bone tissue-specific promoter binding factor NMP2. NMP-2 exhibits recognition for a promoter domain contiguous to the vitamin D-responsive element of the osteocalcin gene, although the vitamin D receptor does not appear to be a component of the nuclear matrix proteins. Interrelationships between nuclear matrix proteins and nonmatrix nuclear proteins, in mediating steroid hormone responsiveness of a vitamin D-regulated promoter, are therefore suggested.

The B1C8 protein is in the dense assemblies of the nuclear matrix and relocates to the spindle and pericentriolar filaments at mitosis

The B1C8 monoclonal antibody detects a 180-kDa nuclear matrix-specific protein. The protein is a component of the dense, metabolically active bodies or assemblies revealed by resinless section electron microscopy of the nuclear matrix. These assemblies are scattered through the nuclear interior, enmeshed in a complex network of 11-nm filaments. Resinless section electron microscopy of immunogold-stained nuclear matrix preparations shows B1C8 located in many but apparently not all the assemblies. In this regard, the B1C8 antigen resembles previously studied nuclear matrix proteins such as the H1B2 protein. The speckled pattern of nuclear immunofluorescence by B1C8 reflects this labeling of the dense assemblies in the nuclear matrix. Somewhat unusual is the faint staining of cytoplasmic microtubules by B1C8, which appears to be due to a weakly cross-reacting protein. During cell division, the B1C8 antigen redistributed drastically, showing the dispersion of nuclear matrix assemblies at mitosis. Speckles of B1C8 fluorescence first coalesced at prophase within the nuclear interior and then scattered into numerous cytoplasmic speckles by prometaphase. At metaphase, the B1C8 speckled cytoplasmic staining had become even more widely distributed and finely grained. Also, intense labeling appeared at the mitotic pole and on the spindle fibers themselves. The reassembly of B1C8 antigens into larger cytoplasmic speckles began at anaphase and finally, at telophase, most B1C8 labeling redistributed into speckles in the re-forming nuclei.

The retinoblastoma gene product is a cell cycle-dependent, nuclear matrix-associated protein

The retinoblastoma gene product (Rb) has been established as a tumor suppressor and cell cycle regulator, although its mechanism of action remains obscure. The observations that several Rb-binding viral oncoproteins all associate with the nuclear matrix suggest that these interactions may occur on this structure. To determine whether Rb itself is a component of the matrix, we extracted synchronized cultured cells to isolate matrix proteins while preserving nuclear architecture. Immunoblot and immunolabeling data show that a significant portion of hypophosphorylated Rb associates with the matrix only during early G1. Mutant Rb in tumor cells did not associate with the matrix, whereas Rb-reconstituted cells contained abundant matrix-bound Rb. Rb is distributed widely throughout the matrix, particularly concentrated at the nuclear periphery and in nucleolar remnants. Core filaments of the matrix contained no detectable Rb. Our screening of expression libraries for potential Rb-associated proteins has identified several that are part of the matrix. Specifically, the peripheral matrix proteins lamin A and C bound Rb in vitro. We therefore suggest that Rb interactions with the nuclear matrix may be important for its ability to regulate cell cycle progression.

Nuclear matrix proteins distinguish normal diploid osteoblasts from osteosarcoma cells

Interrelationships between nuclear architecture and gene expression were examined by comparing the representation of nuclear matrix proteins in ROS 17/2.8 rat and MG-63 human osteosarcoma cells with those in normal diploid osteoblasts. The tumor-derived cells coexpress genes which are expressed in a sequential and mutually exclusive manner during the progressive stages of osteoblast differentiation. In osteosarcoma cells two-dimensional electrophoretic analysis indicates a composite representation of nuclear matrix proteins characteristic of both the proliferative and postproliferative periods of osteoblast phenotype development. In addition, nuclear matrix proteins unique to the tumor cells and the absence of nuclear matrix proteins found only in normal diploid osteoblasts are observed. Tumor-specific nuclear matrix proteins include those expressed in a proliferation-dependent and independent manner. There is a parallel relationship between nuclear matrix proteins and the expression of cell growth and tissue-specific genes during osteoblast differentiation and in osteosarcoma cells where the developmental sequence of gene expression has been abrogated. Nuclear matrix proteins therefore provide markers reflecting defined periods of bone cell differentiation and phenotypic characteristics of an osteosarcoma.

Nuclear matrix association of multiple sequence-specific DNA binding activities related to SP-1, ATF, CCAAT, C/EBP, OCT-1, and AP-1

The association of DNA binding proteins with the nuclear matrix may be related to a functional role of this subcellular structure in chromatin organization and gene regulation. In this study, nuclear matrix preparations from human HeLa S3 cervical carcinoma and rat ROS 17/2.8 osteosarcoma cells were assayed for the presence of DNA binding activities using consensus binding sequences of well-characterized transcription factors as probes. Competition analysis shows that each probe interacts with different nuclear matrix proteins in a sequence-specific manner and that DNA binding activities related to or identical with SP-1, ATF, CCAAT, C/EBP, OCT-1, and AP-1 are present in the nuclear matrix fraction of different cell types. Comparison of the relative abundance of these transcription factor binding activities in nuclear matrix and nonmatrix nuclear fractions suggests that the distribution between these two fractions is cell type specific, cell growth dependent, or independent of these biological parameters. These results are consistent with the postulated role of the nuclear matrix in transcriptional regulation of gene expression.

Osteocalcin gene promoter-binding factors are tissue-specific nuclear matrix components

The nuclear matrix appears to play an important role in developmental gene expression during osteoblast differentiation. To better understand this role, we examined nuclear matrix DNA-binding proteins that are sequence-specific and interact with the osteocalcin gene promoter. Multiple protein-DNA interactions involving two distinct nuclear matrix proteins occur within the 5' regulatory sequences (nt -640 to -430). One of these proteins, NMP-1, is a ubiquitous, cell growth-regulated protein that is related to the transcription factor ATF and resides in both the nuclear matrix and the nonmatrix nuclear compartment. The other protein, NMP-2, is a cell type-specific, 38-kDa promoter factor that recognizes binding sites resembling the consensus site for the CCAAT/enhancer-binding protein C/EBP and is localized exclusively on the nuclear matrix. NMP-1 and NMP-2 each interact with two nuclear matrix protein-binding elements. These elements are present near key regulatory sites of the osteocalcin gene promoter, such as the principal steroid hormone (vitamin D)-responsive sequences. Binding in this region of the osteocalcin gene promoter suggests transient associations with the nuclear matrix that are distinct from the stable interactions of matrix attachment regions. Our results are consistent with involvement of the nuclear matrix in concentrating and/or localizing transcription factors that mediate the basal and steroid hormone responsiveness of osteocalcin gene transcription.

Localization of nuclear matrix core filament proteins at interphase and mitosis

The gentle removal of chromatin uncovers a nuclear matrix consisting of two parts: a nuclear lamina connected to the intermediate filaments of the cytoskeleton and an internal matrix of thick, polymorphic fibers connecting the lamina to masses in the nuclear interior. This internal nuclear matrix can be further fractionated to uncover a highly branched network of 9 nm and 13 nm core filaments retaining some enmeshed bodies. The core filament network retains most of the nuclear RNA, as well as the fA12RNP antigen, and may be the most basic or core element of internal nuclear structure. One high molecular weight protein component of the core filament network, the H1B2 antigen, is normally masked in the interphase nucleus and is uncovered as the chromatin condenses at mitosis. This protein is associated with a fibrogranular network surrounding and connected to the chromosomes. The core filament-associated fA12 antigen also becomes associated with this perichromosomal network. We propose that the core filament nuclear matrix structure may not completely disassemble at mitosis but, rather, that parts remain as a structural network connected to chromosomes and other mitotic structures. These mitotic networks may, in turn, serve as the core structures on which the nuclear matrices of daughter cells are built.

Sequence-specific DNA-binding proteins are components of a nuclear matrix-attachment site

We have identified a nuclear matrix-attachment region within an upstream element of a human H4 histone gene promoter. Nuclear matrix proteins, isolated and solubilized from HeLa S3 cells, were found to interact with sequence specificity at this matrix-attachment region. Several types of assays for protein-DNA interaction showed that the minimal sequence for the nuclear matrix protein-DNA interaction was 5'-TGACGTCCATG-3'; the underlined region corresponds to the core consensus sequence for ATF transcription factor binding. Two proteins with molecular masses of 43 and 54 kDa were identified by UV-crosslinking analysis as integral components of this protein-DNA complex. The molecular masses of these proteins and the ATF-binding site consensus sequence suggest that these proteins are members of the ATF family. Our results provide direct evidence for nuclear matrix localization of sequence-specific DNA-binding factors for an actively transcribed gene. The proximity of a strong positive transcriptional regulatory element to the matrix-attachment region of this gene suggests that the nuclear matrix may serve to localize and concentrate trans-acting factors that facilitate regulation of gene expression.

A normally masked nuclear matrix antigen that appears at mitosis on cytoskeleton filaments adjoining chromosomes, centrioles, and midbodies

mAbs were generated against HeLa nuclear matrix proteins and one, HIB2, which selectively stained mitotic cells, was selected for further study. Western blot analysis showed H1B2 antibody detected a protein of 240 kD in the nuclear matrix fractions. The H1B2 antigen was completely masked in immunofluorescently stained interphase cells. However, removing chromatin with DNase I digestion and 0.25 M ammonium sulfate extraction exposed the protein epitope. The resulting fluorescence pattern was bright, highly punctate, and entirely nuclear. Further extraction of the nuclear matrix with 2 M NaCl uncovers an underlying, anastomosing network of 9-13 nm core filaments. Most of the H1B2 antigen was retained in the fibrogranular masses enmeshed in the core filament network and not in the filaments themselves. The H1B2 antigen showed remarkable behavior at mitosis. As cells approached prophase the antigen became unmasked to immunofluorescent staining without the removal of chromatin. First appearing as a bright spot, the antibody staining spread through the nucleus finally concentrating in the region around the condensed chromosomes. The antibody also brightly stained the spindle poles and, more weakly, in a punctate pattern in the cytoskeleton around the spindle. As the chromosomes separated at anaphase, H1B2 remained with the separating daughter sets of chromosomes. The H1B2 antigen returned to the reforming nucleus at telophase, but left a bright staining region in the midbody. Immunoelectron microscopy of resinless sections showed that, in the mitotic cell, the H1B2 antibody did not stain chromosomes and centrioles themselves, but decorated a fibrogranular network surrounding and connected to the chromosomes and a fibrogranular structure surrounding the centriole.

Human papillomavirus 16 E7 protein is associated with the nuclear matrix

The cellular localization of the human papillomavirus (HPV)-16 E7 gene product in the cell lines CaSki and SiHa has been determined by both biochemical and immunocytochemical methods. These measurements show E7 to be localized in the cell nucleus, specifically with the nonchromatin nuclear structure or nuclear matrix. This localization of E7 required an unambiguous fractionation of the nuclear constituents. This was achieved by using a gentle sequential fractionation procedure to prepare the scaffold consisting of the nuclear matrix and intermediate filaments (NM-IF). Chromatin was cleaved with nuclease and the resulting nucleosomes eluted with 0.25 M ammonium sulfate. Immunostaining of cells after this extraction procedure with monoclonal antibodies (mAbs) to E7 revealed a fine grained, punctate nuclear fluorescence in CaSki and SiHa, which was absent in normal cervical keratinocytes and the HPV-negative cell line C33.1. Western blots of cell fractions with these mAbs showed that E7 was localized in the NM-IF fraction in SiHa and CaSki but was not detected in HPV-negative cells. A second protein of slightly higher mobility is identified by these antisera in HPV-16-containing cells. The data suggest that the previous inability to directly visualize E7 by immunocytology is due to the masking of epitopes by cellular components and not to low levels of protein.

Localization of heterogeneous nuclear ribonucleoprotein in the interphase nuclear matrix core filaments and on perichromosomal filaments at mitosis

Although heterogeneous nuclear RNA (hnRNA) has been localized to the core filament substructure of the nuclear matrix, its precise location in the filament network has been unknown. The fA12 monoclonal antibody can localize, at high resolution, hn ribonucleoproteins (hnRNPs) and, presumably, hnRNA. Gold bead immunolabeling of resinless electron microscopy sections showed the fA12 antigens were in the fibrogranular material enmeshed in the filament network and not in the filaments themselves. At mitosis, hnRNP antigens became dispersed into a halo surrounding the chromosomes and spindle poles. Immunogold staining showed fA12 stained fibrogranular material associated with perichromosomal and pericentriolar filaments distinct from the mitotic spindle fibers. fA12 also labeled the midbody remaining after cytokinesis.

Bio Vision: microscopy in three dimensions

Conventional electron microscopy is inadequate for visualizing the three-dimensional networks supporting cell architecture: the cytoskeleton and nuclear matrix. Consequently, we have not appreciated the extent to which the cell, its biochemistry, and its molecular biology are structured. A new technology combining in situ cell fractionation and resinless section electron microscopy allows the visualization of cell structure in three dimensions and permits the localization of individual components. These techniques reveal a far richer cell architecture than had been assumed and will allow important problems of biology, which have not surrendered their secrets to a purely biochemical approach, to be addressed.

Differential association of membrane-bound and non-membrane-bound polysomes with the cytoskeleton

We report here a differential release of specific mRNAs from the cytoskeleton by cytochalasin D treatment. Non-membrane-bound polysomal mRNAs, such as histone mRNA and c-fos mRNA, are readily released from the cytoskeleton of HeLa cells during cytochalasin D treatment. Over 90% of H3 and H4 histone mRNA is associated with the cytoskeleton in control cells and only 25% in cells treated with cytochalasin D (40 micrograms/ml). In contrast, the membrane-bound polysomal mRNAs for HLA-B7 and chorionic gonadotropin-alpha are inefficiently released from the cytoskeletal framework by cytochalasin D alone; approximately 98% of the HLA-B7 mRNA in control cells is associated with the cytoskeleton, whereas approximately 65% of the HLA-B7 mRNA is retained on the cytoskeleton in cells treated with cytochalasin D (40 micrograms/ml). Disruption of polysome structure with puromycin during cytochalasin D treatment results in the efficient release of HLA-B7 mRNA from the cytoskeleton. Under these conditions, only 25% of the HLA-B7 mRNA remains associated with the cytoskeletal framework. Thus, membrane-bound polysomes appear to be attached to the cytoskeleton through a cytochalasin D-sensitive site as well as through association with the nascent polypeptide and/or ribosome. These results demonstrate a complex association of polysomes with the cytoskeleton and elements of the endoplasmic reticulum.

Imaging cytoskeleton--mitochondrial membrane attachments by embedment-free electron microscopy of saponin-extracted cells

Embedment-free electron microscopy images the cytoskeleton and nuclear matrix, which are very difficult to visualize in conventional electron micrographs. However, to be effective, cell structures must be depleted of soluble proteins, which otherwise shroud cell architecture. Nonionic detergents effect this extraction, releasing soluble proteins but also destroying all membranes. Saponin can permeabilize plasma membranes, releasing soluble proteins while preserving many cytoplasmic membranes. Stereoscopic electron microscopy of resinless sections shows the many connections of the cytoskeleton to mitochondrial membranes.

Multiple types of mRNA-cytoskeleton interactions

Nearly all actively translated mRNAs are associated with the cytoskeleton in HeLa cells and the nature of this association is poorly understood. To gain insight into this association, we have examined and compared the cytoskeleton-mRNA interactions of a signal peptide-histone fusion mRNA (membrane-bound polysomal mRNA) to those of endogenous histone mRNA (nonmembrane-bound polysomal mRNA). We report here the detection of a cytoskeleton attachment site within the signal peptide-histone fusion mRNP/mRNA nucleotide sequence that is not present in wild-type histone mRNA or in HLA-B7 and chorionic gonadotropin-alpha membrane-bound polysomal mRNAs. These results support the possibility that there are multiple mechanisms for the attachment of specific classes of mRNAs to the cytoskeleton.

Progressive changes in the protein composition of the nuclear matrix during rat osteoblast differentiation

Primary cultures of fetal rat calvarial osteoblasts undergo a developmental sequence with respect to the temporal expression of genes encoding osteoblast phenotypic markers. Based on previous suggestions that gene-nuclear matrix associations are involved in regulating cell- and tissue-specific gene expression, we investigated the protein composition of the nuclear matrix during this developmental sequence by using high-resolution two-dimensional gel electrophoresis. The nuclear matrix was isolated at times during a 4-week culture period that represent the three principal osteoblast phenotypic stages: proliferation, extracellular matrix (ECM) maturation, and mineralization. The most dramatic changes in the nuclear matrix protein patterns occurred during transitions from the proliferation to the ECM maturation stage and from ECM maturation to the mineralization period, with only minor variations in the profiles within each period. These stage-specific changes, corresponding to the major transition points in gene expression, indicate that the nuclear matrix proteins reflect the progressive differentiation of the bone cell phenotype. Subcultivation of primary cells delays mineralization, and a corresponding delay was observed for the nuclear matrix protein patterns. Thus, the sequential changes in protein composition of the nuclear matrix that occur during osteoblast differentiation represent distinct stage-specific markers for maturation of the osteoblast to an osteocytic cell in a bone-like mineralized ECM. These changes are consistent with a functional involvement of the nuclear matrix in mediating modifications of developmental gene expression.

Core filaments of the nuclear matrix

The nuclear matrix is concealed by a much larger mass of chromatin, which can be removed selectively by digesting nuclei with DNase I followed by elution of chromatin with 0.25 M ammonium sulfate. This mild procedure removes chromatin almost completely and preserves nuclear matrix morphology. The complete nuclear matrix consists of a nuclear lamina with an interior matrix composed of thick, polymorphic fibers and large masses that resemble remnant nucleoli. Further extraction of the nuclear matrices of HeLa or MCF-7 cells with 2 M sodium chloride uncovered a network of core filaments. A few dark masses remained enmeshed in the filament network and may be remnants of the nuclear matrix thick fibers and nucleoli. The highly branched core filaments had diameters of 9 and 13 nm measured relative to the intermediate filaments. They may serve as the core structure around which the matrix is constructed. The core filaments retained 70% of nuclear RNA. This RNA consisted both of ribosomal RNA precursors and of very high molecular weight hnRNA with a modal size of 20 kb. Treatment with RNase A removed the core filaments. When 2 M sodium chloride was used directly to remove chromatin after DNase I digestion without a preceding 0.25 M ammonium sulfate extraction, the core filaments were not revealed. Instead, the nuclear interior was filled with amorphous masses that may cover the filaments. This reflected a requirement for a stepwise increase in ionic strength because gradual addition of sodium chloride to a final concentration of 2 M without an 0.25 M ammonium sulfate extraction uncovered core filaments.

Prompt heat-shock and heat-shifted proteins associated with the nuclear matrix-intermediate filament scaffold in Drosophila melanogaster cells

Elevated temperatures induced the synthesis of several new proteins in Drosophila melanogaster cells. Besides the conventional heat shock (HS) proteins, another set of temperature-induced proteins has been found. These latter resemble the prompt HS proteins of mammalian cells. The prompt HS proteins of Drosophila differ from the well-known conventional HS proteins in the following properties: (1) synthesis of the prompt HS proteins is insensitive to the transcription inhibitor actinomycin D, which blocks the appearance of conventional HS proteins; (2) induction of the prompt HS proteins requires a significantly higher temperature than conventional HS proteins; (3) prompt HS proteins associate strictly with the nuclear matrix-intermediate filament complex (NM-IF), while the conventional HS proteins are found in all subcellular fractions; (4) prompt HS proteins of Drosophila are induced by high temperature alone while the conventional HS proteins are also produced by a variety of stress conditions. Resinless-section electron micrographs show an altered nuclear matrix morphology in heat-shocked cells. The nuclear matrix fibers are altered in spatial distribution and have much additional electron-dense material. This added material probably reflects the soluble proteins shifted into the nuclear matrix at high temperature. The prompt HS proteins can be distinguished clearly from heat-shifted proteins by several criteria. Also, the prompt HS proteins are distinct from the heat-insensitive viral proteins of a persistent virus (HPS-1).

Immunolocalization in three dimensions: immunogold staining of cytoskeletal and nuclear matrix proteins in resinless electron microscopy sections

We describe two methods for staining resinless thin sections with antibodies and gold-conjugated second antibodies. Immunolocalization of specific proteins is a powerful tool for cell structure studies but current techniques do not develop its full potential. Immunofluorescence provides only low-resolution localization, whereas conventional thin-section electron microscopy images and immunostains only the section surface. Resinless sections of extracted cell structures offer a simple and effective means of immuno-electron microscopy. Without embedding plastic or soluble proteins, the cell cytostructure produces high-contrast, three-dimensional images. Resinless sections of detergent-extracted cells are prepared by embedding in diethylene glycol distearate, sectioning, and removing diethylene glycol distearate before microscopy. In the first method of immunostaining, extracted cells were fixed and stained with antibodies before embedment, sectioning, removal of the embedding resin, and critical point drying. In the postembedment method, the sample was embedded and sectioned, the diethylene glycol distearate was removed, and the sample was rehydrated before antibody staining. With these techniques, specific proteins were localized with high resolution throughout the entire section. Stereoscopic micrographs of resinless sections revealed the precise localization of specific cytoskeleton and nuclear matrix proteins in three dimensions with unprecedented clarity.

Association of adenovirus DNA transcribed activity with nuclear matrix of host cells

With gentle cell extraction techniques, various DNA components in the HeLa cells after 6 h of adenovirus infection have been obtained. Adenovirus, early transcribed regions (E2a, E1b) and a late transcribed region (L2) were used as probes in Southern hybridization, respectively. The experiment showed that only actively transcribed adenovirus DNA fragments would tightly bind to the nuclear matrix of host cells. We inferred that the nuclear matrix of host cells plays an important role in viral DNA transcription.

Chromatin architecture and nuclear RNA

The maintenance of normal chromatin morphology requires ongoing RNA synthesis. We have examined the role of RNA in chromatin organization, using selective detergent extraction of cells, RNA synthesis inhibitors, and enzymatic digestion of nuclear RNA. Comparison of extracted and unextracted cells showed that the important features of chromatin architecture were largely unchanged by the extraction procedure. Normally, chromatin was distributed in small heterochromatic regions and dispersed euchromatic strands. Ribonucleoprotein granules were dispersed throughout the euchromatic regions. Exposure to actinomycin led to the redistribution of chromatin into large clumps, leaving large empty spaces and a dense clustering of the remaining ribonucleoprotein granules. When the nuclei of extracted cells were digested with RNase A, there was a rearrangement of chromatin similar to but more pronounced than that seen in cells exposed to actinomycin. The inhibitor 5,6-dichloro-1-beta-D-ribofuranosylbenzimidizole also inhibits RNA synthesis but by a different mechanism that leaves no nascent RNA chains. The drug had little effect on chromatin after brief exposure but resembled actinomycin in its effect at longer times. We also examined the structure of the nuclear matrix to which most heteronuclear RNA remains associated. Pretreatment of cells with actinomycin or digestion of the nuclear matrix with RNase A caused the matrix fibers to collapse and aggregate. The experiments show a parallel decay of chromatin and of nuclear matrix organization with the depletion of nuclear RNA and suggest that RNA is a structural component of the nuclear matrix, which in turn may organize the higher order structure of chromatin.

Nuclear matrix proteins reflect cell type of origin in cultured human cells

The low abundance proteins of the nuclear matrix (NM) were separated from the intermediate filament (IF) proteins and analyzed by two-dimensional gel electrophoresis. Three human breast carcinoma lines had virtually identical patterns of 37 NM proteins. In contrast, cell lines derived from diverse tissues had qualitatively different NM protein patterns. Together, the five cell types examined here had a total of 205 distinguishable NM proteins with 125 of these proteins unique to a single cell type. The remaining NM proteins were shared among cell types to different degrees. Polyclonal antisera, obtained by immunization with total NM proteins as antigens, preferentially stained the nuclear interior and not the exterior IF. These observations suggest that the NM proteins, localized to the interior of the nucleus, vary in a cell-type-specific manner.

Alterations in nuclear matrix structure after adenovirus infection

Infection of HeLa cells with adenovirus serotype 2 causes rearrangements in nuclear matrix morphology which can best be seen by gentle cell extraction and embedment-free section electron microscopy. We used these techniques to examine the nuclear matrices and cytoskeletons of cells at 6, 13, 28, and 44 h after infection. As infection progressed, chromatin condensed onto the nucleoli and the nuclear lamina. Virus-related inclusions appeared in the nucleus, where they partitioned with the nuclear matrix. These virus centers consisted of at least three distinguishable areas: amorphously dense regions, granular regions whose granulations appeared to be viral capsids, and filaments connecting these regions to each other and to the nuclear lamina. The filaments became decorated with viral capsids of two different densities, which may be empty capsid shells and capsids with DNA-protein cores. The interaction of some capsids with the filaments persisted even after lysis of the cell. We propose that granulated virus-related structures are sites of capsid assembly and storage and that the filaments may be involved in the transport of capsids and capsid intermediates. The nuclear lamina became increasingly crenated after infection, with some extensions appearing to bud off and form blebs of nuclear material in the cytoplasm. The perinuclear cytoskeleton became rearranged after infection, forming a corona of decreased filament number around the nucleus. In summary, we propose that adenovirus rearranges the nuclear matrix and cytoskeleton to support its own replication.