Histone H1 compacts DNA under force and during chromatin assembly

Histone H1 binds to linker DNA between nucleosomes, but the dynamics and biological ramifications of this interaction remain poorly understood. We performed single-molecule experiments using magnetic tweezers to determine the effects of H1 on naked DNA in buffer or during chromatin assembly in Xenopus egg extracts. In buffer, nanomolar concentrations of H1 induce bending and looping of naked DNA at stretching forces below 0.6 pN, effects that can be reversed with 2.7-pN force or in 200 mM monovalent salt concentrations. Consecutive tens-of-nanometer bending events suggest that H1 binds to naked DNA in buffer at high stoichiometries. In egg extracts, single DNA molecules assemble into nucleosomes and undergo rapid compaction. Histone H1 at endogenous physiological concentrations increases the DNA compaction rate during chromatin assembly under 2-pN force and decreases it during disassembly under 5-pN force. In egg cytoplasm, histone H1 protects sperm nuclei undergoing genome-wide decondensation and chromatin assembly from becoming abnormally stretched or fragmented due to astral microtubule pulling forces. These results reveal functional ramifications of H1 binding to DNA at the single-molecule level and suggest an important physiological role for H1 in compacting DNA under force and during chromatin assembly.     

Xenopus egg extracts: a model system for chromatin replication

A cell-free system derived from Xenopus eggs enables in vitro reproduction of the steps occurring during eukaryotic DNA replication. With a circular single-stranded DNA template, extracts obtained from high-speed centrifugation perform complementary DNA strand synthesis coupled to chromatin assembly. Nucleosomes are formed on the newly replicated DNA and the overall reaction mimics the events occurring during chromosomal replication on the lagging strand at the replication fork. ATP is necessary at all steps examined individually, including RNA priming, elongation of DNA strands and chromatin assembly. Although not required for nucleosome formation, ATP is involved in the correct spacing of nucleosomes and the stability of the assembled chromatin. Replication of double-stranded DNA was observed only with extracts obtained from low-speed centrifugation using demembraned sperm nuclei as substrate. Nuclei are reconstituted around the DNA and then undergo a series of events characteristic of a cell cycle. In contrast, neither DNA elongation or chromatin assembly require formation of the nucleus, and both are independent of the cell cycle.     

Xenopus egg extracts: A model system for chromatin replication

A cell-free system derived from Xenopus eggs enables in vitro reproduction of the steps occurring during eukaryotic DNA replication. With a circular single-stranded DNA template, extracts obtained from high-speed centrifugation perform complementary DNA strand synthesis coupled to chromatin assembly. Nucleosomes are formed on the newly replicated DNA and the overall reaction mimics the events occuring during chromosomal replication on the lagging strand at the replication fork. ATP is necessary at all steps examined individually, including RNA priming, elongation of DNA strands and chromatin assembly. Although not required for nucleosome formation, ATP is involved in the correct spacing of nucleosomes and the stability of the assembled chromatin. Replication of double-stranded DNA was observed only with extracts obtained from low-speed centrifugation using demembraned sperm nuclei as substrate. Nuclei are reconstituted around the DNA and then undergo a series of events characteristic of a cell cycle. In contrast, neither DNA elongation or chromatin assembly require formation of the nucleus, and both are independent of the cell cycle.     

Requirement of DNA topoisomerases for in vitro chromatin assembly by 3T6 mouse cell extracts

Nuclear extracts of 3T6 mouse cells were able to assemble in vitro minichromosomes which displayed a 150-bp periodicity. Activities of both DNA topoisomerases I and II were detected in these extracts. When a supercoiled pUC DNA was added, it was first relaxed in less than 3 min, then slowly supercoiled again in 1-4 h. Both reactions occurred either in the absence or the presence of added Mg2+ and/or ATP, they were not blocked by DNA topoisomerase II inhibitors and they were inhibited by an antiserum against DNA topoisomerase I and by camptothecin. These findings led us to propose that, under our in vitro assay conditions, chromatin assembly is mainly carried out by a DNA topoisomerase I.     

Assembly of spaced chromatin involvement of ATP and DNA topoisomerase activity.

Undiluted extracts from eggs or oocytes of Xenopus laevis support the assembly of chromatin with physiologically spaced nucleosomes. Micrococcal nuclease and DNase I digestion experiments show that nucleosome formation as well as supercoiling of circular DNA concomitant to assembly do not require ATP or Mg2+. However these factors are essential for the stability and the physiological spacing of the assembled chromatin. gamma-S-ATP can substitute for ATP in this process. With topoisomers of defined linking number topological interconversions proceed by steps of unity, both in vitro as well as in vivo, indicating that topoisomerase I is dominantly acting in this process. Novobiocin sensitivity occurred only with diluted extracts and was unrelated to an inhibition of topoisomerase II. Finally, nucleosome assembly occurs efficiently on linear DNA although the assembled DNA is less stable than with circular DNA. From these results we propose that mature chromatin is formed in a two-step reaction. In the first step, nucleosome deposition occurs independently of ATP and Mg2+. Thus, nucleosome formation can be uncoupled from their spacing. In this step, topoisomerase activity is involved in the relaxation of the topological constraints generated by chromatin assembly rather than in the process of assembly itself. The second step, requiring ATP and Mg2+, generates properly spaced chromatin.     

Novobiocin inhibits passive chromatin assembly in vitro.

Novobiocin, an inhibitor of prokaryotic DNA gyrase and eukaryotic type II topoisomerase enzymes, interferes with in vitro chromatin assembly using purified histones, DNA and nucleoplasmin. The target of inhibition is not topoisomerase II; this energy-independent assembly system lacks any ATP and Mg2+-dependent type II topoisomerase or gyrase activities. Rather, novobiocin interacts with histones, disrupting histone-histone associations required for octamer formation, and causing histones to precipitate from both nucleoplasmin-histone and histone-DNA complexes. Thus, novobiocin is able to generate 'dynamic' chromatin in vitro in the absence of ATP and Mg2+ by removing histones from previously assembled static chromatin, so that the DNA supercoils, previously constrained by conventional nucleosomes, become susceptible to removal by topoisomerase I.     

The chloroplast HSP70B-CDJ2-CGE1 chaperones catalyse assembly and disassembly of VIPP1 oligomers in Chlamydomonas

The vesicle-inducing protein in plastids (VIPP1) is essential for the biogenesis of thylakoid membranes in cyanobacteria and plants. VIPP1 and its bacterial ancestor PspA form large homo-oligomeric rings of >1 MDa. We recently demonstrated that VIPP1 interacts with the chloroplast J-domain co-chaperone CDJ2 and its chaperone partner HSP70B, and hypothesized that the chaperones might be involved in the assembly and/or disassembly of VIPP1 oligomers. To test this hypothesis, we analysed the composition of VIPP1/chaperone complexes in Chlamydomonas reinhardtii cell extracts and monitored effects of the chaperones on VIPP1 assembly states in vitro. We found that CGE1, the chloroplast GrpE homologue, is also part of complexes with HSP70B, CDJ2 and VIPP1. We observed that CDJ2-VIPP1 accumulated as low- and high-molecular-weight complexes in ATP-depleted cell extracts, but as intermediate-size complexes in extracts supplemented with ATP. This was consistent with a role for the chaperones in VIPP1 assembly and disassembly. Using purified proteins, we could demonstrate that the chaperones indeed facilitated both the assembly and disassembly of VIPP1 oligomers. Electron microscopy studies revealed that, in contrast to PspA, VIPP1 rings assembled into rod-shaped supercomplexes that were morphologically similar to microtubule-like structures observed earlier in various plastid types. VIPP1 rods, too, were disassembled by the chaperones, and chaperone-mediated rod disassembly also occurred when VIPP1 lacked an approximately 30-aa C-terminal extension present in VIPP1 homologues but absent in PspA. By regulating the assembly state of VIPP1, the chloroplast HSP70 chaperone system may play an important role in the maintenance/biogenesis of thylakoid membranes.     

Comparison of replicative and non-replicative chromatin assembly pathways in HeLa cell extracts.

It has been reported that chromatin assembly in mammalian cell extracts depends exclusively or preferentially on ongoing DNA replication (Stillman, B. (1986) Cell 45, 555-565). More recently, this view has been challenged demonstrating that, in the same extracts, chromatin can also be formed efficiently in the absence of DNA replication (Gruss et al. (1990) EMBO J. 9, 2911-2922). The experiments, described in this communication, were performed to resolve this apparent contradiction. We found that there are at least two distinct in vitro pathways for chromatin assembly in HeLa cell extracts. The replicative pathway requires a nuclear protein, most likely identical with the chromatin assembly factor, described by Stillman (1986, Cell 45, 555-565), and the free soluble histones present in the cytosol of S phase cells. In contrast, a non-replicative pathway was identified that depends on isolated nuclear histones. As one component of the non-replicative assembly pathway we identified a cytosolic factor that was purified to apparent homogeneity and shown to be an acidic 50 kDa polypeptide. The isolated cytosolic 50 kDa protein efficiently promoted nucleosome assembly as demonstrated by one- and two-dimensional gel electrophoresis of in vitro packaged plasmid DNA.     

The histone chaperone Asf1 is dispensable for direct de novo histone deposition in <Emphasis Type="Italic">Xenopus</Emphasis> egg extracts

Histone chaperones that escort histones during their overall lifetime from synthesis to sites of usage can participate in various tasks. Their requirement culminates in the dynamic processes of nucleosome assembly and disassembly. In this context, it is important to define the exact role of the histone chaperone Asf1. In mammals, Asf1 interacts with two other chaperones, CAF-1 and HIRA, which are critical in DNA synthesis-coupled and synthesis-uncoupled nucleosome assembly pathways, respectively. A key issue is whether Asf1 is able or not to deposit histones onto DNA by itself in both pathways. Here, to delineate the precise role of Asf1 in chromatin assembly, we used Xenopus egg extracts as a powerful system to assay de novo chromatin assembly pathways in vitro. Following characterization of both Xenopus Asf1 and p60 (CAF-1), we used immunodepletion strategies targeting Asf1, HIRA, or CAF-1. Strikingly, the depletion of Asf1 led to the simultaneous depletion of HIRA and consequently impaired the DNA synthesis-independent nucleosome assembly pathway. The rescue of nucleosome assembly capacity in such extracts was effective when adding HIRA along with H3/H4 histones, yet addition of Asf1 along with H3/H4 histones did not work. Moreover, nucleosome assembly coupled to DNA repair was not affected in these Asf1/HIRA-depleted extracts, a pathway impaired by CAF-1 depletion. Thus, these data show that Asf1 is not directly involved in de novo histone deposition during DNA synthesis-independent and synthesis-dependent pathways in egg extracts. Based on our results, it becomes important to consider the implications for Asf1 function during early development in Xenopus.     

Cell cycle control of higher-order chromatin assembly around naked DNA in vitro.

We have developed an in vitro system in which higher-order chromatin structures are assembled around naked DNAs in a cell cycle-dependent manner. Membrane-free soluble extracts specific to interphase and mitotic states were prepared from Xenopus eggs. When high molecular weight DNA is incubated with interphase extracts, fluffy chromatin-like structures are assembled. In contrast, mitotic extracts produce highly condensed chromosome-like structures. Immunofluorescence studies show that a monoclonal antibody MPM-2, which recognizes a class of mitosis-specific phosphoproteins, stains the "core" or "axis" of condensed mitotic chromatin but not interphase chromatin. By adding mitotic extracts, interphase chromatin structures are synchronously converted into the condensed state. The increasingly condensed state of chromatin correlates with the appearance and structural rearrangements of the MPM-2-stained structures. These results suggest that mitosis-specific phosphoproteins recognized by MPM-2 may be directly involved in the assembly of the chromosome scaffold-like structures and chromatin condensation. Although both extracts promote nucleosome assembly at the same rate, topoisomerase II (topo II) activity is four to five times higher in mitotic extracts compared with interphase extracts. The addition of a topo II inhibitor VM-26 into mitotic assembly mixtures disturbs the organization of the MPM-2-stained structures and affects the final stage of chromatin condensation. This in vitro system should be useful for identifying cis- and trans-acting elements responsible for higher-order chromatin assembly and its structural changes in the cell cycle.     

Postreplicative chromatin assembly by Drosophila and human chromatin assembly factor 1.

To study the relationship between DNA replication and chromatin assembly, we have purified a factor termed Drosophila chromatin assembly factor 1 (dCAF-1) to approximately 50% homogeneity from a nuclear extract derived from embryos. dCAF-1 appears to consist of four polypeptides with molecular masses of 180, 105, 75, and 55 kDa. dCAF-1 preferentially mediates chromatin assembly of newly replicated DNA relative to unreplicated DNA during T-antigen-dependent simian virus 40 DNA replication in vitro, as seen with human CAF-1. Analysis of the mechanism of DNA replication-coupled chromatin assembly revealed that both dCAF-1 and human CAF-1 mediate chromatin assembly preferentially with previously yet newly replicated DNA relative to unreplicated DNA. Moreover, the preferential assembly of the postreplicative DNA was observed at 30 min after inhibition of DNA replication by aphidicolin, but this effect slowly diminished until it was no longer apparent at 120 min after inhibition of replication. These findings suggest that the coupling between DNA replication and chromatin assembly may not necessarily involve a direct interaction between the replication and assembly factors at a replication fork.     

Nuclear assembly with lambda DNA in fractionated Xenopus egg extracts: an unexpected role for glycogen in formation of a higher order chromatin intermediate

Crude extracts of Xenopus eggs are capable of nuclear assembly around chromatin templates or even around protein-free, naked DNA templates. Here the requirements for nuclear assembly around a naked DNA template were investigated. Extracts were separated by ultracentrifugation into cytosol, membrane, and gelatinous pellet fractions. It was found that, in addition to the cytosolic and membrane fractions, a component of the gelatinous pellet fraction was required for the assembly of functional nuclei around a naked DNA template. In the absence of this component, membrane-bound but functionally inert spheres of lambda DNA were formed. Purification of the active pellet factor unexpectedly demonstrated the component to be glycogen. The assembly of functionally active nuclei, as assayed by DNA replication and nuclear transport, required that glycogen be pre-incubated with the lambda DNA and cytosol during the period of chromatin and higher order intermediate formation, before the addition of membranes. Hydrolysis of glycogen with alpha- amylase in the extract blocked nuclear formation. Upon analysis, chromatin formed in the presence of cytosol and glycogen alone appeared highly condensed, reminiscent of the nuclear assembly intermediate described by Newport in crude extracts (Newport, J. 1987. Cell. 48:205- 217). In contrast, chromatin formed from phage lambda DNA in cytosol lacking glycogen formed "fluffy chromatin-like" structures. Using sucrose gradient centrifugation, the highly condensed intermediates formed in the presence of glycogen could be isolated and were now able to serve as nuclear assembly templates in extracts lacking glycogen, arguing that the requirement for glycogen is temporally restricted to the time of intermediate formation and function. Glycogen does not act simply by inducing condensation of the chromatin, since similarly isolated mitotically condensed chromatin intermediates do not form functional nuclei. However, both mitotic and fluffy interphase chromatin intermediates formed in the absence of glycogen can be rescued to form functional nuclei when added to a second extract which contains glycogen. This study presents a novel role for a carbohydrate in nuclear assembly, a role which involves the formation of a particular chromatin intermediate. Potential models for the role of glycogen are discussed.     

Topoisomerase function during replication-independent chromatin assembly in yeast.

DNA topoisomerases I and II are the two major nuclear enzymes capable of relieving torsional strain in DNA. Of these enzymes, topoisomerase I plays the dominant role in relieving torsional strain during chromatin assembly in cell extracts from oocytes, eggs, and early embryos. We tested if the topoisomerases are used differentially during chromatin assembly in Saccharomyces cerevisiae by a combined biochemical and pharmacological approach. As measured by plasmid supercoiling, nucleosome deposition is severely impaired in assembly extracts from a yeast mutant with no topoisomerase I and a temperature-sensitive form of topoisomerase II (strain top1-top2). Expression of wild-type topoisomerase II in strain top1-top2 fully restored assembly-driven supercoiling, and assembly was equally efficient in extracts from strains expressing either topoisomerase I or II alone. Supercoiling in top1-top2 extract was rescued by adding back either purified topoisomerase I or II. Using the topoisomerase II poison VP-16, we show that topoisomerase II activity during chromatin assembly is the same in the presence and absence of topoisomerase I. We conclude that both topoisomerases I and II can provide the DNA relaxation activity required for efficient chromatin assembly in mitotically cycling yeast cells.     

Nucleosome assembly of simian virus 40 DNA in a mammalian cell extract.

We report here a mammalian cell-free system that can support chromatin assembly. Effective nucleosome assembly in HeLa cell extracts occurred at 125 to 200 mM KCl or potassium glutamate. At this physiological K+ ion concentration, two types of chromatin assembly were observed. The first was interfered with by Mg2+. Other cations such as Mn2+, Ca2+, Fe3+, and spermidine also inhibited this type of nucleosome assembly. The second type of assembly occurred in the presence of Mg2+ and at least equimolar ATP. However, even in the presence of ATP, excess Mg2+ inhibited assembly and promoted catenation of DNA; these effects could be circumvented by excess ATP, GTP, EDTA, or polyglutamic acid. The critical DNA concentration for optimum assembly in both pathways suggested a stoichiometric association of histones with DNA. The spacing of nucleosomes formed by both types of assembly on linear and circular DNA was reasonably regular, but chromatin assembled in the presence of ATP and Mg2+ was more stable.     

In vitro chromatin assembly promoted by the Xenopus laevis S-150 cell-free extract is enhanced by treatment with RNase A.

Cell-free extracts employed as chromatin assembly systems contain a myriad of proteins, polyanions and nucleic acids. The roles of ATP, MgCl2 and other cofactors in the catalysis of nucleosome formation by the Xenopus laevis oocyte S-150 have yet to be established unequivocally. In this study we examine the influence of RNA in the assembly process. Under reaction conditions that inhibit nucleosome formation (+ EDTA), pretreatment of the extract with RNase A revives the chromatin assembly machinery while the rate of DNA supercoiling is stimulated significantly. Addition of purified RNA blocks DNA supercoiling. Taken together, these data suggest that the parameters surrounding in vitro chromatin assembly are variable and subject to modulation by endogenous factors.