Physical constraints in the condensation of eukaryotic chromosomes. Local concentration of DNA versus linear packing ratio in higher order chromatin structures

The local concentration of DNA in metaphase chromosomes of different organisms has been determined in several laboratories. The average of these measurements is 0.17 g/mL. In the first level of chromosome condensation, DNA is wrapped around histones forming nucleosomes. This organization limits the DNA concentration in nucleosomes to 0. 3-0.4 g/mL. Furthermore, in the structural models suggested in different laboratories for the 30-40 nm chromatin fiber, the estimated DNA concentration is significantly reduced; it ranges from 0.04 to 0.27 g/mL. The DNA concentration is further reduced when the fiber is folded into the successive higher order structures suggested in different models for metaphase chromosomes; the estimated minimum decrease of DNA concentration represents an additional 40%. These observations suggest that most of the models proposed for the 30-40 nm chromatin fiber are not dense enough for the construction of metaphase chromosomes. In contrast, it is well-known that the linear packing ratio increases dramatically in each level of DNA folding in chromosomes. Thus, the consideration of the linear packing ratio is not enough for the study of chromatin condensation; the constraint resulting from the actual DNA concentration in metaphase chromosomes must be considered for the construction of models for condensed chromatin.     

THE HISTONES ASSOCIATED WITH CONDENSED AND EXTENDED CHROMATIN OF MOUSE LIVER

Histones were extracted from isolated mouse liver nuclei, and from mouse liver condensed and extended chromatin. Mouse liver histones were found to be very similar to those of calf thymus in their solubility properties, relative electrophoretic mobilities, and molecular weights as determined on SDS-polyacrylamide gels. Quantitative analysis by high-resolution gel electrophoresis demonstrated a remarkable similarity between the histones of condensed chromatin and those of extended chromatin. However, minor differences were found. A unique subspecies was found only in condensed chromatin histone and the relative amounts of fractions F2A1 and F2A2 differed in the two types of chromatin. The ratio of the parental to the acetylated form of F2A1 was identical in the two chromatin samples. Since DNA extracted from the condensed chromatin fraction consisted of approximately 50% satellite DNA, the general similarities between the histones of condensed and extended chromatin make it likely that even this simple, highly repetitive DNA is complexed with a number of histone subfractions.     

Higher-order structure of chromatin and chromosomes

The linear array of nucleosomes that comprises the primary structure of chromatin is folded and condensed to varying degrees in nuclei and chromosomes forming 'higher order structures'. We discuss the recent findings from novel experimental approaches that have yielded significant new information on the different hierarchical levels of chromatin folding and their functional significance.     

Generation of a third-order folded structure for chromatin

A model for the initiation of the diffuse-condensed transition of chromatin induced by a change in the conformation of lysine-rich histones is proposed. Three levels of folded structures are discussed. The first-order folded structure refers to the structure of the repeat unit of chromatin, which is called the nucleosome. The nucleosome contains a nuclease resistant region in which 140 base pairs of DNA are wrapped around the surface of a histone aggregated of two copies each of the histones H2A, H2B, H3 and H4. This DNA-histone aggregate is called a core particle. The nuclease accessible region of the nucleosome is approximately 60 base pairs of DNA which link the core particle, hence the terminology “linker DNA.” The lysine-rich histones, (Hl, H5), which are more loosely bound than the core histones, are associated with the linker DNA. The second-order folded structure refers to the conformation of a polynucleosome. Based on neutron scattering and quasielastic light scattering studies the second-order folded structure is assumed to be an extended helix in solution with 5–7 nucleosome units per turn. The third-order folded structure is defined as that structure resulting from the first stage in the condensation process induced by a conformational change in the lysine-rich histones. Generation of the third-order folded structure in the proposed model is effected by an increased affinity of the lysine-rich histones for super-helical DNA in the core particles in adjacent turns of the second-order folded structure. Since the lysine-rich histones preferentially bind to A-T rich regions in DNA, the distribution of these regions would determine the third-order folded structure. The net effect of a non-random distribution of A-T rich regions as in the proposed model is the generation of a helix for the third-order folded structure. The assumption of a non-random distribution of A-T rich regions is indirectly supported by proflavine binding studies reported herein and by the existence of repetitive and non-repetitive DNA regions inferred from renaturation studies. One consequence of the proposed mechanism is that the majority of the A-T rich regions are in the interior of the third-order folded structure. Promoter sites of high A-T content would then be inaccessible to polymerases. The proposed model also suggests a role for spacer DNA in the genome. Higher order folded structures must also be present in the final state of condensed chromatin since the three orders of folded structures considered in this communication accounts for only 2% of that required in the diffuse-condensed transition.     

Gene functioning and storage within a folded genome

In mammals, genomic DNA that is roughly 2 m long is folded to fit the size of the cell nucleus that has a diameter of about 10 μm. The folding of genomic DNA is mediated via assembly of DNA-protein complex, chromatin. In addition to the reduction of genomic DNA linear dimensions, the assembly of chromatin allows to discriminate and to mark active (transcribed) and repressed (non-transcribed) genes. Consequently, epigenetic regulation of gene expression occurs at the level of DNA packaging in chromatin. Taking into account the increasing attention of scientific community toward epigenetic systems of gene regulation, it is very important to understand how DNA folding in chromatin is related to gene activity. For many years the hierarchical model of DNA folding was the most popular. It was assumed that nucleosome fiber (10-nm fiber) is folded into 30-nm fiber and further on into chromatin loops attached to a nuclear/chromosome scaffold. Recent studies have demonstrated that there is much less regularity in chromatin folding within the cell nucleus. The very existence of 30-nm chromatin fibers in living cells was questioned. On the other hand, it was found that chromosomes are partitioned into self-interacting spatial domains that restrict the area of enhancers action. Thus, TADs can be considered as structural-functional domains of the chromosomes. Here we discuss the modern view of DNA packaging within the cell nucleus in relation to the regulation of gene expression. Special attention is paid to the possible mechanisms of the chromatin fiber self-assembly into TADs. We discuss the model postulating that partitioning of the chromosome into TADs is determined by the distribution of active and inactive chromatin segments along the chromosome.This article was specially invited by the editors and represents work by leading researchers.     

Differential accessibility of DNA in extended and condensed chromatin to pancreatic DNase I

Pancreatic DNase I has been used to study the interaction between DNA and chromosomal proteins in extended and condensed chromatin fractions isolated from mouse and Chinese hamster livers. It was found that DNase digests extended chromatin at a faster rate than condensed chromatin, and the evidence suggests that the chromosomal proteins are more tightly complexed to the DNA in condensed than in extended chromatin. This difference in DNA-protein interaction in extended and condensed chromatin may be related to the functional difference which characterizes these fractions, and might be one of the factors underlying the production of bands on metaphase chromosomes.     

Determination of the elastic properties of short ssDNA molecules by mechanically folding and unfolding DNA hairpins

The characterization of elastic properties of biopolymers is crucial to understand many molecular reactions determined by conformational bending fluctuations of the polymer. Direct measurement of such elastic properties using single‐molecule methods is usually hindered by the intrinsic tendency of such biopolymers to form high‐order molecular structures. For example, single‐stranded deoxyribonucleic acids (ssDNA) tend to form secondary structures such as local double helices that prevent the direct measurement of the ideal elastic response of the ssDNA. In this work, we show how to extract the ideal elastic response in the entropic regime of short ssDNA molecules by mechanically pulling two‐state DNA hairpins of different contour lengths. This is achieved by measuring the force dependence of the molecular extension and stiffness on mechanically folding and unfolding the DNA hairpin. Both quantities are fit to the worm‐like chain elastic model giving values for the persistence length and the interphosphate distance. This method can be used to unravel the elastic properties of short ssDNA and RNA sequences and, more generally, any biopolymer that can exhibit a cooperative two‐state transition between mechanically folded and unfolded states (such as proteins). © 2014 Wiley Periodicals, Inc. Biopolymers 101: 1193–1199, 2014.     

DNA-protein interactions and chromosome banding

Pancreatic DNase I was used as a probe to study DNA-protein interactions in condensed and extended chromatin fractions isolated from Chinese hamster liver, and in human lymphocyte and mouse L cell metaphase chromosomes in situ. By studying the rate of digestion of chromatin DNA by DNase, we have previously shown that DNA in extended chromatin is more sensitive to DNase digestion than that in condensed chromatin. In the current investigation, we have examined whether this differential sensitivity of the chromatin fractions to DNase is due to differences in protein binding to DNA or differences in the degree of chromatin condensation. By “decondensing” the condensed chromatin and comparing its rate of digestion to that of untreated condensed and extended chromatin, it was found that differences in the degree of binding of proteins to DNA rather than the degree of condensation of the chromatin primarily determines the sensitivity of each fraction to DNase. Extraction of the various classes of chromosomal proteins, followed by DNase digestion of the residual chromatin revealed that both the histone and non-histone proteins protect the DNA in the chromatin fractions from DNase attack; however, the more tightly associated non-histones appear to be specifically responsible for the differential sensitivity of the chromatin fractions to DNase digestion. These non-histones may be more tightly associated with the DNA in condensed than in extended chromatin, thereby protecting the DNA in condensed chromatin against DNase attack to a greater extent than that in extended chromatin. When metaphase chromosomes were briefly digested with DNase in situ and subsequently stained with Feulgen reagent, incontrovertible C-banding and some G-banding was obtained. This DNaseinduced banding demonstrates that the DNA in C-band and possibly G-band regions is less accessible to DNase than that in the interband regions, and our biochemical data suggest that this differential accessibility is caused by differential DNA-protein binding such that the non-histones are more tightly coupled to the DNA in the G- and C-band regions than they are in the interbands. Differences in the binding of non-histones to DNA in different segments of the metaphase chromosome may be involved in the mechanism of G- and C-banding.     

High concentration of DNA in condensed chromatin.

The lengths of the DNA molecules of eukaryotic genomes are much greater than the dimensions of the metaphase chromosomes in which they are contained during mitosis. From this observation it has been generally assumed that the linear packing ratio of DNA is an adequate measure of the degree of DNA compaction. This review summarizes the evidence suggesting that the local concentration of DNA is more appropriate than the linear packing ratio for the study of chromatin condensation. The DNA concentrations corresponding to most of the models proposed for the 30-40 nm chromatin fiber are not high enough for the construction of metaphase chromosomes. The interdigitated solenoid model has a higher density because of the stacking of nucleosomes in secondary helices and, after further folding into chromatids, it yields a final concentration of DNA that approaches the experimental value found for condensed chromosomes. Since recent results have shown that metaphase chromosomes contain high concentrations of the chromatin packing ions Mg2+ and Ca2+, it is discussed that dynamic rather than rigid models are required to explain the condensation of the extended fibers observed in the absence of these cations. Finally, considering the different lines of evidence demonstrating the stacking of nucleosomes in different chromatin complexes, it is suggested that the face-to-face interactions between nucleosomes may be the driving force for the formation of higher order structures with a high local concentration of DNA.     

Different structural states in oligonucleosomes are required for early versus late steps of base excision repair

Chromatin in eukaryotic cells is folded into higher order structures of folded nucleosome filaments, and DNA damage occurs at all levels of this structural hierarchy. However, little is known about the impact of higher order folding on DNA repair enzymes. We examined the catalytic activities of purified human base excision repair (BER) enzymes on uracil-containing oligonucleosome arrays, which are folded primarily into 30nm structures when incubated in repair reaction buffers. The catalytic activities of uracil DNA glycosylase (UDG) and apyrimidinic/apurinic endonuclease (APE) digest G:U mismatches to completion in the folded oligonucleosomes without requiring significant disruption. In contrast, DNA polymerase β (Pol β) synthesis is inhibited in a major fraction (~80%) of the oligonucleosome array, suggesting that single strand nicks in linker DNA are far more accessible to Pol β in highly folded oligonucleosomes. Importantly, this barrier in folded oligonucleosomes is removed by purified chromatin remodeling complexes. Both ISW1 and ISW2 from yeast significantly enhance Pol β accessibility to the refractory nicked sites in oligonucleosomes. These results indicate that the initial steps of BER (UDG and APE) act efficiently on highly folded oligonucleosome arrays, and chromatin remodeling may be required for the latter steps of BER in intact chromatin.     

H2A.Z alters the nucleosome surface to promote HP1alpha-mediated chromatin fiber folding

Controlling the degree of higher order chromatin folding is a key element in partitioning the metazoan genome into functionally distinct chromosomal domains. However, the mechanism of this fundamental process is poorly understood. Our recent studies suggested that the essential histone variant H2A.Z and the silencing protein HP1alpha may function together to establish a specialized conformation at constitutive heterochromatic domains. We demonstrate here that HP1alpha is a unique chromatin binding protein. It prefers to bind to condensed higher order chromatin structures and alters the chromatin-folding pathway in a novel way to locally compact individual chromatin fibers without crosslinking them. Strikingly, both of these features are enhanced by an altered nucleosomal surface created by H2A.Z (the acidic patch). This shows that the surface of the nucleosome can regulate the formation of distinct higher order chromatin structures mediated by an architectural chromatin binding protein.     

The arrangement of H5 molecules in extended and condensed chicken erythrocyte chromatin.

Chemical cross-linking with dithiobis(succinimidyl propionate) has been used to investigate the relative disposition of neighbouring H5 (H1) molecules in chicken erythrocyte chromatin in the extended (nucleosome filament) and condensed (300 A filament) states; in this chromatin H5 and H1 are interspersed along the nucleosome filament, rather than segregated into blocks, as shown by the nature of the cross-linked dimers and their relative amounts. Detailed analysis of the cross-linked H5 homopolymers from extended chromatin and condensed nuclear chromatin indicates which domains of H5 are in contact (or close proximity) in the two states. Two results suggest a polar, head-to-tail arrangement of H5 molecules along the nucleosome filament. This arrangement persists when chromatin adopts higher-order structure but in the folded state neighbouring basic C-terminal domains, in particular, are more closely juxtaposed than they are in extended chromatin.     

Chromatin condensation: does histone H1 dephosphorylation play a role?

In this article we describe three distinct biological systems where histone H1 phosphorylation is uncoupled from mitosis and highly condensed chromatin is enriched in dephosphorylated forms of H1: the amitotic macronucleus of Tetrahymena, terminally differentiated avian erythrocytes and sea urchin sperm. Each system offers informative contrasts to the idea that H1 hyperphosphorylation is causally related to mitotic chromosome condensation. Assuming that higher order chromatin folding is primarily driven by electrostatic interactions between H1 and DNA, an alternative model is presented for the role of H1 phosphorylation in chromatin condensation.     

Zinc binding modulates the entire folding free energy surface of human Cu,Zn superoxide dismutase

Over 100 amino acid replacements in human Cu,Zn superoxide dismutase (SOD) are known to cause amyotrophic lateral sclerosis, a gain-of-function neurodegenerative disease that destroys motor neurons. Supposing that aggregates of partially folded states are primarily responsible for toxicity, we determined the role of the structurally important zinc ion in defining the folding free energy surface of dimeric SOD by comparing the thermodynamic and kinetic folding properties of the zinc-free and zinc-bound forms of the protein. The presence of zinc was found to decrease the free energies of a peptide model of the unfolded monomer, a stable variant of the folded monomeric intermediate, and the folded dimeric species. The unfolded state binds zinc weakly with a micromolar dissociation constant, and the folded monomeric intermediate and the native dimeric form both bind zinc tightly, with subnanomolar dissociation constants. Coupled with the strong driving force for the subunit association reaction, the shift in the populations toward more well-folded states in the presence of zinc decreases the steady-state populations of higher-energy states in SOD under expected in vivo zinc concentrations (approximately nanomolar). The significant decrease in the population of partially folded states is expected to diminish their potential for aggregation and account for the known protective effect of zinc. The ∼ 100-fold increase in the rate of folding of SOD in the presence of micromolar concentrations of zinc demonstrates a significant role for a preorganized zinc-binding loop in the transition-state ensemble for the rate-limiting monomer folding reaction in this β-barrel protein.     

Chromatin fiber folding: requirement for the histone H4 N-terminal tail

We have developed a self-assembly system for nucleosome arrays in which recombinant, post-translationally unmodified histone proteins are combined with DNA of defined-sequence to form chromatin higher-order structure. The nucleosome arrays obtained are highly homogeneous and sediment at 53S when maximally folded in 1mM or 100mM MgCl(2). The folding properties are comparable to established systems. Analytical ultracentrifugation is used to determine the consequence of individual histone tail domain deletions on array folding. Fully compacted chromatin fibers are obtained with any one of the histone tails deleted with the exception of the H4 N terminus. The region of the H4 tail, which mediates compaction, resides in the stretch of amino acids 14-19.     

An elastic rod model to evaluate effects of ionic concentration on equilibrium configuration of DNA in salt solution

As a coarse-gained model, a super-thin elastic rod subjected to interfacial interactions is used to investigate the condensation of DNA in a multivalent salt solution. The interfacial traction between the rod and the solution environment is determined in terms of the Young–Laplace equation. Kirchhoff’s theory of elastic rod is used to analyze the equilibrium configuration of a DNA chain under the action of the interfacial traction. Two models are established to characterize the change of the interfacial traction and elastic modulus of DNA with the ionic concentration of the salt solution, respectively. The influences of the ionic concentration on the equilibrium configuration of DNA are discussed. The results show that the condensation of DNA is mainly determined by competition between the interfacial energy and elastic strain energy of the DNA itself, and the interfacial traction is one of forces that drive DNA condensation. With the change of concentration, the DNA segments will undergo a series of alteration from the original configuration to the condensed configuration, and the spiral-shape appearing in the condensed configuration of DNA is independent of the original configuration.