Nuclear matrix-bound DNA primase. Elucidation of an RNA priming system in nuclear matrix isolated from regenerating rat liver

Recent findings in purified systems demonstrate the universality of DNA polymerase-primase complexes which may function in the priming and continuation of eucaryotic DNA replication. In this report we characterize an in vitro, nuclear matrix-associated, priming and continuation system that can utilize either endogenous matrix-bound DNA or exogenous single-stranded DNA as template. 30-40% of total nuclear DNA primase activity was recovered in association with the isolated nuclear matrix fraction from regenerating rat liver. Matrix-bound primase catalyzed the alpha-amanitin, actinomycin D-resistant synthesis of oligonucleotide chains of 8-50 nucleotides on the endogenous template. At least a portion of the RNA primers were continued by DNA polymerase alpha with deoxynucleoside triphosphate incorporation up to 300-600 nucleotides. Nearest neighbor analysis revealed ribodeoxynucleotide covalent linkages in these RNA-DNA chains. The matrix-bound primase preferred single-stranded fd DNA as exogenous template over synthetic homopolymers and was strictly dependent on the presence of ribonucleoside triphosphates. Appropriate subfractionation revealed that the matrix-bound primase activity is exclusively localized in the nuclear matrix interior. The ability of primase and DNA polymerase to synthesize covalently linked RNA-DNA products demonstrates the potentially useful role of the nuclear matrix in vitro system for elucidating the organizational and functional properties of the eucaryotic replication apparatus in the cell nucleus.     

The effect of in vitro heat exposure on the recovery of nuclear matrix-bound DNA polymerase alpha activity during the different phases of the cell cycle in synchronized HeLa S3 cells.

HeLa S3 cells were synchronized by a double thymidine block or aphidicolin treatment and the levels of nuclear matrix-bound DNA polymerase alpha activity were then measured using activated calf thymus DNA as template. The nuclear matrix was obtained by 2 M NaCl extraction and DNase I digestion of isolated nuclei incubated at 37 degrees C for 45 min prior to subfractionation. In all phases of the cell cycle 25-30% of nuclear DNA polymerase alpha activity remained matrix-bound, even when cells were in the G1 phase. No dynamic association of DNA polymerase alpha activity with the matrix was seen, at variance with previous results obtained in regenerating rat liver. The variations measured in matrix-bound activity closely followed those detected in isolated nuclei throughout the cell cycle. If nuclei were not heat-stabilized very low levels of DNA polymerase alpha activity were measured in the matrix (1-2% of total nuclear activity). Heat incubation of nuclei failed to produce any enrichment in matrix-associated newly replicated DNA, whereas the sulfhydryl cross-linking chemical sodium tetrathionate did. Therefore the results obtained after the heat stabilization procedure do not completely fit with the model that envisions the nuclear matrix as the active site where eucaryotic DNA replication takes place.     

Pre-replicative association of multiple replicative enzyme activities with the nuclear matrix during rat liver regeneration

As a step toward the molecular elucidation of the putative replicational apparatus associated with the nuclear matrix, we have investigated the possible matrix association of several replicational related enzymes. In addition to the previously identified DNA polymerase alpha, DNA primase, 3'-5' exonuclease, RNase H, and DNA methylase were all recovered at significant levels (20-30% of total nuclear activity) in nuclear matrix isolated from regenerating rat liver during maximal in vivo replication (22 h post-hepatectomy). In contrast, DNA ligase was not detected on the nuclear matrix even though significant activity was present in isolated nuclei. Examination of the replicative dependency of these enzyme activities following partial hepatectomy revealed pre-replicative elevations which were distinct for each matrix-bound enzyme. A second late-replicative peak in DNA methylase is consistent with a role of this matrix-bound enzyme in the maintenance of the inheritable methylation pattern. Mild sonication resulted in a significant release of all of these activities except RNase H. A major portion of the matrix-solubilized DNA polymerase alpha, DNA primase, 3'-5' exonuclease, and DNA methylase activities cosedimented on sucrose gradients between approximately 8-12 S. Our results are consistent with the organization of at least a portion of these replicative enzymes into nuclear matrix-bound replicational complexes. We also propose a novel pre-replicative assembly model of the matrix-bound replicational apparatus in which DNA primase plays an initial and critical role.     

Cytochemical localization of DNA loop attachment sites to the nuclear lamina and to the inner nuclear matrix

Summary The rat liver nuclear matrix, obtained by endogenous nuclease digestion and extraction with low and high lonic strength media, contains residual DNA fragments that are considered to represent the attachment sites of the chromatin domains to the nucleoskeleton. These sites, protected against nuclease digestion by their binding with the nucleoskeleton proteins, should be either mainly linked to the peripheral lamina or to the inner nuclear matrix. The DNA fragment distribution at the level of the different components of the nuclear matrix has been evaluated in samples embedded in Epon and in hydrophilic resins by means of the DNase-gold technique. The labeling obtained suggests that the chromatin loops are prevailingly associated with the interior of the matrix; in fact about twice of the label is present in the inner matrix with respect to the peripheral lamina area.These results confirm the hypothesis that in interphase the chromatin maintains an organization similar to that of chromosomes, with loops radiating from a central scaffold, instead of being mainly attached to the lamina as otherwise suggested.     

Preferential binding of DNA primase to the nuclear matrix in HeLa cells

Studies of the spatial organization of DNA replication have provided increasing evidence of the importance of the nuclear matrix. We have previously reported a relationship between rates of DNA synthesis and the differential binding of DNA polymerase alpha to the nuclear matrix over the S-phase. We now report the detection of DNA primase bound to the HeLa nuclear matrix. Matrix-bound primase was measured both indirectly, by the incorporation of [32P]dAMP into an unprimed single-stranded template, poly(dT), and directly, by the incorporation of [3H]AMP into matrix DNA. Characteristics of this system include a requirement for ATP, inhibition by adenosine 5'-O-(thiotriphosphate), a primase inhibitor, and insensitivity to aphidicolin and alpha-amanitine, inhibitors of polymerase alpha and RNA polymerase, respectively. Subcellular quantification of primase and polymerase alpha activity revealed that while most (approximately 72%) primase activity is bound to the matrix, only a minority (approximately 32%) of polymerase alpha activity is matrix-bound. Treatment of the nuclear matrix with beta-D-octylglucoside allowed the solubilization of approximately 54% of primase activity and approximately 39% of the polymerase alpha activity. This data provides further evidence of a structural and functional role for the nuclear matrix in DNA replication. The ability to solubilize matrix-bound replicative enzymes may prove to be an important tool in the elucidation of the spatial organization of DNA replication.     

Intranuclear dynamics of DNA polymerase alpha differs between the transplanted R3230AC mammary adenocarcinomas and the host mammary gland depending on lactation cycle

DNA polymerase alpha activity was markedly higher in all nuclear subfractions, including nuclear matrix, from transplanted R3230AC mammary adenocarcinomas than in the analogous fractions from mammary gland of same tumor-bearing pregnant or lactating rats. Changes in host lactational status had no significant effect on subnuclear distribution of tumor DNA polymerase alpha activity, with the majority (60-75%) localized in soluble nucleoplasm and a significant amount (13-20%) retained in the nuclear matrix. In the host mammary gland, nuclear matrix-bound DNA polymerase alpha was highest, accounting for 48% of total nuclear activity, during late pregnancy when mammary cells undergo rapid raplication. During lactation, when cells in mammary gland cease to divide, only 8% of enzyme activity was in the nuclear matrix, while the majority (60-80%) of DNA polymerase alpha activity was localized in nucleoplasm. In both R3230AC tumor and mammary gland regardless of host's lactational status, the majority (60-80%) of DNA polymerase beta activity was localized in the high salt-soluble chromatin. These present data thus suggest that, regardless of host lactational status, R3230AC tumor has many cycling cells, each with a large pool of DNA polymerase alpha molecules maintaining maximal and constant replicative activity, while normal mammary gland cells have a smaller pool of DNA polymerase alpha which become primarily matrix-bound only during active cell replication during late pregnancy. A constant localization of nuclear DNA polymerase beta in chromatin in both mammary gland and the tumor suggest it is not important in mammary cell proliferation.     

Is the nuclear matrix the site of DNA replication in eukaryotic cells?

Four types of experiment were carried out to test the recently proposed model of matrix-bound replication in eukaryotic cells. In experiments with pulse-labelling we found preferential association of newly replicated DNA with the matrix only when the procedure for isolation includes first high-salt treatment of isolated nuclei and then digestion with nucleases, or when prior to digestion the nuclei have been stored for a prolonged time. In both cases, however, evidence was found that this preferential association is due to a secondary, artifactual binding of the newly replicated chromatin region to the matrix elements. Pulse-chase experiments and experiments with continuous labelling were carried out to answer the question whether during replication the DNA is reeled through the replication complexes, i.e., whether newly replicated DNA is temporarily or permanently associated with the matrix. The results showed that at that time the matrix DNA does not move from its site of attachment. Since, according to the model of matrix-bound replication, the forks are assumed to be firmly anchored to high-salt resistant proteinaceous matrix structures, the chromatin fragments isolated with endonuclease not recognizing newly replicated DNA and purified by sucrose gradient centrifugation should be free of replication intermediates. The electronmicroscopic analysis of such fragments revealed the existence of intact replication micro-bubbles. Moreover, the fragments with replication configurations appeared as smooth chromatin fibres not attached to elements characteristic for the matrix. All these experiments suggest that the nuclear skeleton is not a native site of DNA replication in eukaryotic cells.     

The interrelation between DNA synthesis rates and DNA polymerases bound to the nuclear matrix in synchronized HeLa cells

Quantitative rates of DNA synthesis can be determined by DNA:propidium fluorescence measurements of synchronized cells progressing through S-phase. We have previously reported that HeLa cells have discontinuous rates with values of about 2.9, 1.6, and 4.4 pg of DNA/h for early, middle, and late S-phase, respectively. In attempts to understand why two peaks of DNA synthesis rates are observed, we have examined the nuclear DNA polymerases alpha and beta over the S-phase. Nuclear matrices isolated from HeLa cells contained 2% of the alpha polymerase and 12% of the beta polymerase that was present in cell lysates, and about 2% of the original DNA. The amounts of endogenous DNA synthesis in isolated nuclear matrices were comparable to the amounts observed when exogenous DNA was added. DNase treatment abolished the endogenous DNA synthesis but not the exogenous DNA synthesis, suggesting that polymerase alpha binding does not depend on matrix-bound DNA. As synchronized cells progressed through the S-phase, there appeared two peaks of enzymatic activity of alpha polymerase bound to the nuclear matrix which correlated with in vitro DNA synthesis in these nuclear matrices and with the two peaks of quantitative DNA synthesis rates. Two peaks of alpha polymerase activity were also observed with isolated nuclei, but not with cell lysates or cytosol. Our results suggest that, over the S-phase, the differential binding of polymerase alpha to the nuclear matrix determines the differential rates of DNA synthesis.     

DNA polymerase alpha from the nuclear matrix of cells infected with simian virus 40.

The nuclear matrix prepared from normal, simian virus 40 (SV40)-infected, and SV40-transformed cells contained DNA polymerase activities. Approximately 12% of the total DNA polymerase activities in isolated nuclei remained with the nuclear matrix. alpha-polymerase was the major matrix DNA polymerase activity as judged by sensitivity to various inhibitors: aphidicolin, dideoxy-TTP, and N-ethylmaleimide. Approximately 2-4 fold higher DNA polymerase activity was detected in matrices obtained from lytically infected and virus-transformed cells than that found in normal cells. In lytically infected cells, 30-50% of the matrix-bound DNA polymerase activity solubilized by sonication co-sedimented with majority of the matrix T-antigen, and was co-precipitated with anti-T sera. The results suggest that alpha-polymerase and viral T-antigen may form a functional complex in the matrix.     

Enhanced processivity of nuclear matrix bound DNA polymerase alpha from regenerating rat liver

Translocation of DNA during in vitro DNA synthesis on nuclear matrix bound replicational assemblies from regenerating rat liver was determined by measuring the processivity (average number of nucleotides added following one productive binding event of the polymerase to the DNA template) of nuclear matrix bound DNA polymerase alpha with poly(dT).oligo(A)10 as template primer. The matrix-bound polymerase had an average processivity (28.4 nucleotides) that was severalfold higher than the bulk nuclear DNA polymerase alpha activity extracted during nuclear matrix preparation (8.9 nucleotides). ATP at 1 mM markedly enhanced the activity and processivity of the matrix-bound polymerase but not the corresponding salt-soluble enzyme. The majority of the ATP-dependent activity and processivity enhancement was completed by 100 microM ATP and included products ranging up to full template length (1000-1200 nucleotides). Average processivity of the net ATP-stimulated polymerase activity exceeded 80 nucleotides with virtually all the DNA products greater than 50 nucleotides. Release of nuclear matrix bound DNA polymerase alpha by sonication resulted in a loss of ATP stimulation of activity and a corresponding decrease in processivity to a level similar to that of the salt-soluble polymerase (6.8 nucleotides). All nucleoside di- and triphosphates were as effective as ATP. Stimulation of both activity and processivity by the nonhydrolyzable ATP analogues adenosine 5'-O-(3-thiotriphosphate), 5'-adenylyl imidodiphosphate, and adenosine 5'-O-(1-thiotriphosphate) further suggested that the hydrolysis of ATP is not required for enhancement to occur.(ABSTRACT TRUNCATED AT 250 WORDS)     

The nuclear matrix continues DNA synthesis at in vivo replicational forks

Alkaline cesium chloride gradient analysis of in vivo [3H]bromodeoxyuridine-labeled and in vitro [alpha-32P]dCTP-labeled DNA was used to determine whether in vitro DNA synthesis in regenerating rat liver nuclei and nuclear matrices continued from sites of replication initiated in vivo. At least 70 and 50% of the products of total nuclear and matrix-bound in vitro DNA synthesis, respectively, were continuations of in vivo initiated replicational forks. The relationship of the in vitro DNA synthetic sites in total nuclei versus the nuclear matrix was examined by using [3H]bromodeoxyuridine triphosphate to density label in vitro synthesized DNA in isolated nuclei and [alpha-32P]dCTP to label DNA synthesized in isolated nuclear matrix. A minimum of about 40% of matrix-bound DNA synthesis continued from sites being used in vitro by isolated nuclei. Furthermore, nuclear matrices prepared from in vitro labeled nuclei were 5-fold enriched in DNA synthesized by the nuclei and were several-fold enriched, compared to total nuclear DNA, in a particularly high density labeled population of DNA molecules.     

Differential compartmentalization of plasmid DNA microinjected into Xenopus laevis embryos relates to replication efficiency.

Circular plasmid DNA molecules and linear concatemers formed from the same plasmid exhibit strikingly different fates following microinjection into Xenopus laevis embryos. In this report, we prove quantitatively that only a minority of small, circular DNA molecules were replicated (mean = 14%) from fertilization through the blastula stage of development. At all concentrations tested, very few molecules (approximately 1%) underwent more than one round of DNA synthesis within these multiple cell cycles. In addition, unlike endogenous chromatin, the majority of circular templates became resistant to cleavage by micrococcal nuclease. The extent of nuclease resistance was similar for both replicated and unreplicated templates. Sequestration of circular molecules within a membranous compartment (pseudonucleus), rather than the formation of nucleosomes with abnormal size or spacing, apparently conferred the nuclease resistance. In contrast, most linearly concatenated DNA molecules (derived from end-to-end joining of microinjected monomeric plasmid DNA) underwent at least two rounds of DNA replication during this same period. Linear concatemers also exhibited micrococcal nuclease digestion patterns similar to those seen for endogenous chromatin yet, as judged by their failure to persist in later stages of embryogenesis, were likely to be replicated and maintained extrachromosomally. We propose, therefore, that template size and conformation determine the efficiency of replication of microinjected plasmid DNA by directing DNA to a particular compartment within the cell following injection. Template-dependent compartmentalization may result from differential localization within endogenous nuclei versus extranuclear compartments or from supramolecular assembly processes that depend on template configuration (e.g., association with nuclear matrix or nuclear envelope).     

No discrete complexes containing DNA polymerase α activity can be solubilized from the heat-stabilized nuclear matrix prepared from HeLa S3 cells

Most of the DNA polymerase α activity, bound to the heat-stabilized nuclear matrix prepared from HeLa S3 cells, was released as a matrix extract by sonication. When the extract was centrifuged in a 5–20 per cent linear sucrose gradient no definite peaks of activity could be identified. Most of the activity sedimented to the bottom of the tube under all the conditions tested, whilst the remaining activity was associated with matrix fragments of various and irregular size. No 10 S complexes, containing polymerase activity, were seen after incubation of the extract for 16 h before centrifugation. Other solubilization procedures (i.e. treatment of the matrix with chelating agents, high pH associated with reducing agents, ionic and nonionic detergents) failed to produce release of matrix-bound DNA polymerase α activity. In contrast, we released 10 S complexes, containing polymerase activity, from the matrix prepared from nuclei not exposed to heat. We conclude that a 37°C incubation of isolated nuclei before extraction with 2 M NaCl and DNase I digestion causes DNA polymerase α to bind to the nuclear matrix in a form that cannot subsequently be released as discrete components, at variance with previous results obtained with the matrix prepared from regenerating rat liver.     

Evidence for attachment of interphase chromatin to the nuclear matrix via matrix-bound nucleosomes

Chromatin structure has been studied in the sites of attachment to the nuclear matrix in interphase mouse liver and spleen nuclei. The patterns of fragmentation of the DNA belonging to these sites (0.3-2% of total DNA in spleen and liver, respectively) with staphylococcal nuclease and DNAase I were very close to those of usual nucleosomal chains. Moreover, the nuclear matrix preparations contained all five major histones, including H1, in almost stoichiometric amounts. The histone/DNA ratios for the matrix were also similar to those found in nuclei. These findings and the size of the matrix-protected DNA indicated that interphase chromatin was attached to the nuclear matrix via matrix-bound nucleosomes and, to a much lesser extent, oligonucleosomes up to 5-6 units long. Two-dimensional electrophoretic separation of the matrix-bound histones revealed that modifications of histone H1 and, probably, of other histones were distinguished from those in bulk chromatin. Study of binding of exogenously added labeled histone octamers or mononucleosomal size DNA to nuclear matrix excluded the possibility of their artifactual trapping during the isolation procedure.     

Screening of isolated DNA for sequences released from anchorage sites in nuclear matrix

Isolated chromosomal DNA is associated with polypeptides that are not released from DNA by several methods designed to purify DNA, e.g. treatment with sodium dodecyl sulphate. DNA fragments associated with these very tight DNA/protein complexes show high affinity to nitrocellulose filters in the presence of salt concentrations of 500 mM or greater. Consequently, a fraction of AluI-fragmented native DNA comprising the complexes and 0.2 to 0.3 micron of vicinal DNA can be isolated by one filtration step. This fraction of DNA shows characteristics of residual DNA sequences retained in nuclei after extraction with nucleases and high salt (nuclear matrix). The DNA fragments retained on filters are highly enriched in replicative DNA; and their degree of hybridization with poly(A)+ RNA points to enrichment in actively transcribed sequences. The results support previous work indicating that the very tight DNA/polypeptide complexes co-isolating with DNA under conditions that release other peptide materials from DNA may be anchorage sites of DNA in the nuclear matrix. Moreover, the method described here allows isolation of replicating and actively transcribed DNA sequences directly from isolated total genomic DNA by skipping artefact-prone isolations of the nuclear matrix.     

Isolation of a fraction from cauliflower mosaic virus-infected protoplasts which is active in the synthesis of (+) and (-) strand viral DNA and reverse transcription of primed RNA templates

Sub-cellular fractions, isolated from cauliflower mosaic virus (CaMV)-infected turnip protoplasts, are capable of synthesising CaMV DNA in vitro on an endogenous template and of reverse transcribing oligo dT-primed cowpea mosaic virus RNA. The activity was not detected in mock-inoculated protoplasts. In vitro-labelled DNA hybridized to single-stranded M13 clones complementary to the putative origins of (-) and (+) strand CaMV DNA synthesis and to restriction endonuclease fragments encompassing more than 90% of the CaMV genome. The synthesis of (-) and (+) strand DNA appeared asymmetric. The template(s) for in vitro CaMV DNA synthesis are in a partially nuclease-resistant form. Fractions capable of in vitro CaMV DNA synthesis contained CaMV RNA both heterogeneous and as discrete species; they also contained a range of different sizes of CaMV DNA. Several lines of evidence indicate that this range of in vitro-labelled CaMV DNA, extending from 0.6kb to 8.0kb in length, represents elongating (-) strand DNA. These are discussed in relation to their role as possible replicative intermediates.     

Binding of the DNA polymerase alpha-DNA primase complex to the nuclear matrix in HeLa cells

It is well-known that there are multiple forms of DNA polymerase alpha. In order to determine which form(s) is (are) tightly bound, the activities were dissociated from DNA-poor nuclear matrices, with octyl beta-D-glucoside. Sucrose gradient sedimentation analysis revealed three bands with s values of 7.5, 10.5, and 13. The 7.5S form was free of DNA primase and represented only 10% of the total DNA polymerase alpha bound to the nuclear matrix. The 13S and the 10.5S forms each contained DNA primase activity. The 10.5S form comprised 85% of the DNA polymerase alpha activity and 95% of the DNA primase activity, dissociated from the nuclear matrix. Neither temperature of nuclease digestion nor various salt treatments of nuclei had significant effects on the proportions of DNA polymerase alpha and DNA primase activities bound to, or subsequently dissociated from, nuclear matrices. In a comparison of primase activity bound to the nuclear matrix, dissociated from the nuclear matrix, and in the soluble fraction, it was found that the bound activity had a lower ATP dependence, had less KCl inhibition, and was less sensitive to heat, compared to the dissociated and soluble activities. No differences in Mg2+ or pH dependence were noted. The amounts of DNA polymerase alpha and DNA primase activities bound to the nuclear matrix varied over the cell cycle of synchronized cells. Over the S phase, there were two peaks of matrix-bound DNA primase and two peaks of subsequently dissociated DNA polymerase alpha-DNA primase complex.(ABSTRACT TRUNCATED AT 250 WORDS)