Nucleosome core particles correspond to the structural units of eukaryotic chromatin. They are charged colloids, 101 Å in diameter and 55 Å in length, formed by the coiling of a 146/147 bp DNA fragment (50 nm) around the histone protein octamer. Solutions of these particles can be concentrated, under osmotic pressure, up to the concentrations found in the nuclei of living cells. In the presence of monovalent cations (Na+), nucleosomes self-assemble into crystalline or liquid crystalline phases. A lamello-columnar phase is observed at ‘low salt’ concentrations, while a two-dimensional hexagonal phase and a three-dimensional quasi-hexagonal phase form at ‘high salt’ concentrations. We followed the formation of these phases from the dilute isotropic solutions to the ordered phases by combining cryoelectron microscopy and X-ray diffraction analyses. The phase diagram is presented as a function of the monovalent salt concentration and applied osmotic pressure. An alternative method to condense nucleosomes is to induce their aggregation upon addition of divalent or multivalent cations (Mg2+, spermidine3+ and spermine4+). Ordered phases are also found in the aggregates. We also discuss whether these condensed phases of nucleosomes may be relevant from a biological point of view.
In eukaryotic nuclei, each DNA molecule, typically 1–10 cm long, is highly compacted into a small volume, approximately 1 μm in diameter that corresponds to the interphase chromosome. Before cell division, chromatin is further condensed into individual mitotic chromosomes that can be viewed under the optical microscope. The DNA lies in a complex environment, surrounded by water, multiple types of ions of different valencies and proteins. Among these, basic proteins called histones play a special role in organizing the DNA into a periodic structure called the nucleosomic filament. This ‘beads-on-a-string’ filament is highly organized. It must also be able to change its organization in order to allow access to a given portion of the genome that has to be transcribed at a given moment of the cell cycle, whereas, the activities of other genes must be repressed. The structures of both DNA and nucleosomes are relatively well understood, whereas higher levels of chromatin and chromosome organization are not. Numerous studies indicate various degrees of coiling, looping and random coil-like variability. These features need not be exclusive, and most structural models employ a hierarchy of organization scales (Alberts et al. 2002). It is clear that the organization of DNA and chromatin must vary, reflecting the functional differentiation of the chromosome into distinct domains. Irrespective of the multiple possible states of organization, the genetic material is always in a condensed state, but the degree of condensation may vary, with local DNA concentrations ranging from about 50 to 250 mg ml−1 in eukaryotic nuclei (100<Cnucleosomes<500 mg ml−1). The present issue is to understand (i) how phase transitions between these different states can be induced and controlled, and (ii) the physical properties of DNA and chromatin on which these morphological changes rely. A simplified model system can be used to understand the possible organization of chromatin. We isolated nucleosomes that can be prepared in large amounts and placed under a wide range of precisely defined ionic conditions. Here, we describe how the particles self-organize and form ordered phases that are either crystalline or liquid crystalline. We also discuss whether these ordered phases may be found inside living cells. The ultimate goal would be to relate the local structures with the functional activities of the genome.
2. The nucleosome core particle
The crystallographic structure of the nucleosome was solved a few years ago (Luger et al. 1997a) and has been refined further (Davey et al. 2002; Tsunaka et al. 2005). One tetramer of histones (H3–H4) and two dimers (H2A–H2B) form the histone octamer, around which 146/147 bp of DNA are wrapped in 1.65 turns of a tight left-handed superhelix to form the nucleosome core particle (NCP). As shown in figure 1, the core particle has the shape of a cylinder of 101 Å diameter and 55 Å length. It is a complex, chiral object, since the DNA with a right-handed double helix is wrapped around the protein core in a left-handed helix (figure 1a). The orientation of the pseudo-dyad axis (δ) and also the front and back sides of the particle, with the two free ends of the DNA on the front side, are seen when viewed from the top (figure 1b). We also note how the C- and N-terminal parts of the histones, also called the tails, may extend outside the particle, where their length varies from tail to tail. As a whole, the tails represent about 30% of the histone octamer weight. The NCP is mostly stabilized by electrostatic interactions between the highly negatively charged DNA (two charges per base pair) and the globally positively charged histone core. It is stable under a defined range of ionic conditions (Yager et al. 1989). As a whole, the NCP can be considered as a negatively charged colloid (150 e−), with a heterogeneous distribution of charges: the peripheral DNA crown is highly negatively charged, and the histone tails carry numerous positive charges from lysine and arginine amino acids (figure 1c; Van Holde 1988). The top and bottom faces are quite flat, with only NH4 tails extending in this direction.
Therefore, NCP behaves as a complex, colloidal particle with chiral and electrolytic properties. In a range of ionic conditions where they are stable, with no dissociation of DNA from the protein core, subtle changes of their conformation can be tuned with the ionic strength of the solution. In dilute solution and in the presence of monovalent ions, histone tails can be either condensed onto DNA (for NaCl concentration below 50 mM) or extended for NaCl concentration above this value (Mangenot et al. 2002a). Interactions between NCP are changed accordingly; they are repulsive under low salt conditions and turn attractive as soon as the tails are extended (Mangenot et al. 2002a,b). The attractive interactions, which are not large enough to induce aggregation, are suppressed upon enzymatic digestion of the tails (Bertin et al. 2004). Nevertheless, we do not observe any macroscopic precipitation of NCP in the presence of monovalent ions.
The shape and charge characters confer the NCP with multiple possibilities to monitor their mutual interactions and their supramolecular organizations, as described subsequently.
3. Preparation of condensed phases of nucleosome core particles
NCPs can be prepared from calf thymus, as represented schematically in figure 2a. The population of NCPs was found to have a moderate polydispersity. Indeed, the 146 bp core DNA is protected from enzymatic digestion by its interactions with the core proteins, but extensive digestion also degrades the DNA. Natural polydispersity also arises from the DNA sequence and the distribution of charges associated with the histone tails, since enzymatic modifications (acetylation, methylation, etc.) can introduce slight changes in the charges defined by the amino acid sequence of the histones. NCPs can also be reconstituted from recombinant histones and proteins expressed separately in the bacterium Escherichia coli and assembled in solution under defined ionic conditions (figure 2b). The histone octamer is reconstituted in the first step and the DNA fragment is further assembled to form the NCP. This technology produces a perfectly monodisperse solution of NCPs. All particles are identical, with the same sequence of bases along the 146 bp DNA fragment and the same amino acid sequences of the proteins, with no modification of the charges carried by the amino acids. Only small amounts of NCPs can be prepared using this approach, but its interest lies in the possibility of preparing NCPs that can be designed with precise modifications, for example, the absence of one or several histone tails (Luger et al. 1997b; Dyer et al. 2004). So far, our experiments have been done with native NCPs. Further work with recombinant particles is in progress.
We used two different methods to prepare the condensed phases of NCPs. (i) In the presence of monovalent ions, interactions between NCPs are globally repulsive and it is necessary to force the NCPs to come closer and closer to each other. Following the method of Parsegian et al. (1986), we used polyethylene glycol (PEG) to monitor the applied osmotic pressure under precisely controlled ionic conditions. Concentrations of PEG (MW 20 000 g mol−1) ranging from 19 to 35% were used to apply pressures ranging from 1.6 to 23 atm (http://www.brocku.ca/researchers/peter_rand/osmotic/osfile.html#data). This pressure range allowed us to reach NCP concentration as high as 500 mg ml−1 and to explore NCP concentrations that cover concentrations of the living cell. (ii) In the presence of divalent or multivalent ions, aggregation of NCPs may occur in dilute solution. Aggregates can be examined in the electron microscope or collected in a pellet for X-ray diffraction analyses.
4. Nucleosome core particle surrounded by monovalent ions
(a) Phase diagram
The phase diagram shown in figure 3 represents the multiplicity of organizations observed in the presence of monovalent ions (NaCl).
(i) Columnar isotropic and columnar nematic phases
A remarkable self-assembly property of NCPs is to form columns by stacking, on top of each other, above a critical concentration of particles in the solution. These columns were observed over all the explored range of monovalent salt concentration (from 3.5 to 150 mM NaCl+Tris; Leforestier et al. 2001; Mangenot et al. 2003b). The length of these columns may be quite polydisperse, where the dispersity depends on the number of particles stacked in the columns. We measured columns ranging from 25 to 120 nm in length using electron microscopy. NCPs lie normal to the direction of the column, but the position of each particle is not strictly defined in the column, as seen from close examination of columns trapped in a thin film of vitrified water (figure 4). Along a given column, DNA coiled around the successive NCP gives V shapes, inverted V shapes and more complex motifs that come from various side view projections (figure 4a–c). Each NCP column is therefore disordered in the sense that NCPs are apparently free to rotate around the axis normal to their flattening plane(s) (figure 4d). We suspect that the polydispersity of the DNA length may introduce some disorders in the relative positioning of successive NCPs.
Both the dilute solution of NCPs and the columnar, isotropic phase are extinguished between crossed polars. We therefore observe a transition between two optically isotropic phases, the more concentrated one being highly ordered compared with the less concentrated one. The difference comes from the nature of the objects, either isolated particles (flat cylinders with a height/diameter ratio of 0.5) or elongated columns with a height/diameter ratio that is rather polydisperse, ranging from about 2 to 20. The X-ray diffractograms of the columnar isotropic phase (figure 5a,a′) show three large diffraction rings that can be understood only by the presence of NCP columns. The first peak would correspond to the average distance between the columns. The second and subsequent peaks come from the form factor of the columns and the stacking distance between NCPs in the columns. More details cannot be extracted from these diffractograms in the absence of column alignment.
The columns may also form a nematic birefringent phase above a critical own concentration in the high salt range of the phase diagram. Note that the columns are also formed in the presence of divalent or trivalent cations (Grau et al. (1982) and our own experiments, see subsequent sections). Under higher pressures, these columns align to form more ordered phases. The nature of these phases depends on the ionic strength of the solution. The lateral organization of the columns is therefore mainly driven by the electrostatic interactions, which is not surprising owing to the distribution of charges on the peripheral crown of the particle (see figure 1). Slight changes in the ionic conditions are enough to induce phase transitions, significant change in the NCP concentration and DNA accessibility.
(ii) Lamello-columnar phase
The lamello-columnar phase is observed in the low salt range, under pressures ranging from about 3 to 25 atm. The textures seen in the optical microscope depend on the applied pressure (and concentration of the sample), showing mainly tubes (figure 6a) and spherulites (figure 6b). CryoEM of vitreous sections and X-ray diffraction were used in combination to determine the structure of this phase (Leforestier et al. 2001; Mangenot et al. 2003a). Columns of NCPs are aligned to form bilayers that are kept separated by layers of solvent (figure 7a). The periodicity of the lamellar structure (dL) decreases with the applied pressure and varies from 358 to 376 Å, as measured by X-ray diffraction in our experimental conditions (figure 5b; Mangenot et al. 2003a). The rear sides of the NCPs are oriented inward with respect to the bilayer. The front sides of the particles, with the free DNA ends, face the solvent. Within each layer, the NCPs present a two-dimensional, monoclinic ordering that is revealed by the diffraction peaks seen at high q-values in the X-ray profile (figure 5b,b′). The question was to know whether NCPs were lying flat in the columns and slightly shifted from one column to the next, or whether they were tilted in the column. The answer was found in the tangential cryosections of the layers. Two series of striations can be seen, making an angle of about 105° with respect to the direction of the columns (figure 7b). From these DNA patterns, which are in good agreement with the X-ray data (Mangenot et al. 2003a), we can deduce that the NCPs are tilted in the columns by an angle of about 15° and the tilt is opposite in the two parts of the bilayer, as predicted by the theory (Lorman et al. 2005). The structure of the lamello-columnar phase is sketched in figure 7c,d.
(iii) Inverse hexagonal phase
In the low salt range and for applied pressures above 25 atm, the lamello-columnar phase is replaced by an inverse hexagonal phase. The hexagonal patterns are recognized on freeze-fracture replicas (figure 8a), and several lattice parameters were measured: a1≈380 Å; a2≈450 Å (figure 8b); and a3≈540 Å (not shown). Since the diameter of the NCPs cannot be neglected, compared to the lattice parameters, any change in the NCP concentration leads to discrete variations of the parameter a. From these different values of a and from the corresponding NCP organization (with 9, 12 or 15 columns, respectively, in each hexagonal cell; figure 8b), the NCP concentration was calculated to be equal to about 315, 370 and 420 mg ml−1, respectively. The coexistence of domains with different parameters, as in figure 8a, affords intermediate concentrations. Note that the bilayer structure is kept in the walls separating the solvent channels, although we cannot ascertain whether NCPs are tilted or not in the columns.
(iv) Two-dimensional or three-dimensional columnar hexagonal phases
Discotic columnar hexagonal phases (either two dimensional or three dimensional) were found in the high salt range (50–160 mM) under pressures ranging from 4.7 to 17 atm. They were characterized by X-ray diffraction (figure 5c,d). In the two-dimensional phase, NCPs are piled up into columns that form a hexagonal lattice. No correlation is observed between the longitudinal order along columns and the transverse order in the plane normal to the columns. The distance separating NCPs in a column ranges from 57.7 to 60 Å and the distance between columns from 110 to 116 Å. In the three-dimensional structure, a partial orientation of the sample under a magnetic field of 10 T (figure 5d) allowed us to (i) differentiate diffraction rings from the alignment in the columns—with columns parallel to the magnetic field—and from the lateral organization of the columns (remember that the diameter of a NCP is approximately twice its thickness), and (ii) determine the structure. The ordering is not perfectly hexagonal, but slightly distorted into an orthorhombic structure, with b≈a√3, and c=2h, with h being the distance separating two nucleosomes along a column. The columns are parallel to the c-axis of the unit cell, and the concentration of NCPs are normal to the column axes. Over the entire pressure range, the NCPs' concentration calculated from a, b and c parameters varies from 500 to 580 mg ml−1. Similar concentrations were calculated in the two-dimensional hexagonal phase. In addition to the peaks corresponding to the orthorhombic lattice, a series of reflections that most probably originate from an incommensurate superstructure along the c-axis were observed (figure 5d,d′). They probably originate from a supramolecular chiral organization of the NCP, already detected by analysing the textures by optical microscopy. The initial (quasi)hexagonal domains show the characteristic sixfold symmetries of columnar hexagonal discotic liquid crystals (figure 6d; Leforestier & Livolant 1997). They retain their hexagonal shape and form chiral three-dimensional structures while growing or evolving into more classical hexagonal textures depending on whether the phase possesses three-dimensional or two-dimensional order (Livolant & Leforestier 2000; Mangenot et al. 2003b).
Two-dimensional and three-dimensional orderings were found under identical final conditions, but they were not formed exactly in the same way. Adding the stressing polymer PEG to a solution of NCP at a concentration below or above the concentration at which the NCP columns are formed lead to a two-dimensional or a three-dimensional hexagonal final structure, respectively (Mangenot et al. 2003a). One hypothesis would be that top to bottom contacts between NCPs must be established before the lateral organization of columns is set, in order to reach the higher ordered three-dimensional structure.
(b) Metastability, organization times and polydispersity effects
The hexagonal phase coexists with the lamello-columnar phase in the intermediate salt range (25–50 mM). The X-ray diffraction profile of a given sample analysed either one month or seven months after its preparation is shown in figure 9 to illustrate the time required for the organization of the sample in the intermediate salt range. On a regular basis, all samples were left to organize for three to six months in order to obtain beautiful textures suitable for analyses in the optical microscope and samples were equilibrated under PEG osmotic pressure for X-ray diffraction experiments. Nevertheless, the question arises as to whether or not we are dealing with equilibrium states or if some of these phases are metastable. The question is not easy to answer. We checked with optical microscopy whether identical phases were obtained using alternative preparation methods (especially dehydration of the sample in the absence of PEG, with the understanding that the ionic conditions were not precisely controlled). The sample used to record the diagram shown in figure 5d was also analysed again after 5 years, and it was found to be identical (less than 1% change in the parameters). We assume that the phase diagram given in figure 3 corresponds to an equilibrium phase diagram, but there is no way to prove it since metastable states may be trapped for years. Nevertheless, all capillaries have been kept, and it will be interesting to check whether, over time, we can detect any change in the structural organization. Fortunately, under this range of concentration, NCPs are also stable for years at room temperature.
Another difficulty comes from the polydispersity of the NCP prepared from native chromatin. Identical phases (columnar isotropic, lamello-columnar and columnar hexagonal) were obtained using NCPs with DNA fragments of various lengths (155±10 and 146±3 bp). Interestingly, the phase transitions were slightly shifted, in good agreement with a change in the global charge density of the particle (40 additional negative charges; Mangenot et al. 2003b). Nevertheless, some disorders in the supramolecular organization of NCPs may arise from this polydispersity. For example, the lamello-columnar and the two-dimensional or three-dimensional columnar hexagonal phases were lost when the associated DNA fragments were too long and polydisperse (170±15 bp). In the same way, the apparent free rotation of particles observed along isolated columns (figure 4c) may originate from the polydispersity of the DNA length. The position of one NCP relative to the next one may be influenced by the charge of each particle. Another source of polydispersity originates from the protein part of the NCPs, especially from the histone tails that might carry slight differences in their amino acid sequences (and charges). We can now get rid of these two sources of polydispersity and prepare particles that are perfectly identical, using recombinant technologies and checking the role of each parameter involved. Ordered phases were also obtained with recombinant NCP, either intact or with some of their tails deleted. This work is in progress.
5. Nucleosome core particle in the presence of multivalent counterions
Contrary to monovalent ions, the addition of a small amount of divalent, trivalent or higher valency ions or proteins to a dilute solution of NCPs induces their aggregation and precipitation in a pellet upon centrifugation. A phase rich in nucleosomes separates from a dilute supernatant solution. To determine the conditions of stability of nucleosome solutions, a series of precipitation curves were built for multiple initial NCP concentrations. A typical example is given in figure 10a. Starting from a dilute solution of NCP, increasing amounts of Mg2+ cations are added, and the relative amount of NCP in the supernatant is quantified. Interestingly, 2 mM Mg2+ ions (C1) is enough to initiate the precipitation of the NCPs. The precipitation of NCPs is maximum at a higher concentration of Mg2+ (called C2); it further decreases up to a concentration of Mg2+ (called C3) called the resolubilization threshold. A phase diagram was elaborated from these curves (de Frutos et al. 2001) and completed recently (Bertin in preparation), which is presented in figure 10b. A similar diagram is obtained with the trivalent cation spermidine, although the precipitation occurs for 10 times smaller concentrations (Bertin in preparation). Such diagrams remind us of the diagrams obtained with DNA in the presence of trivalent and tetravalent cations (Raspaud et al. 1998). We were interested in determining the organization of nucleosomes in these aggregates and to compare these structures with the structures observed in the presence of monovalent ions. The small aggregates formed with spermidine (3+) as the condensing agent showed multiple organizations, either isotropic, columnar isotropic, columnar nematic or columnar hexagonal (figure 11; Leforestier et al. 1999). We did not find the lamello-columnar phase observed with monovalent counterions. A systematic structural analysis was performed recently to follow the nature of the phases and control the variation of the crystallographic parameters as a function of precise ionic and NCP concentrations. We intend to relate the structural parameters to precise locations in the phase diagram, as it was done recently with short DNA fragments in the presence of spermine 4+ (Raspaud et al. 2005). An interesting property of multivalent cations is their ability, upon condensation onto the NCP, to reverse the global charge of the particle from negative to positive (de Frutos et al. 2001), which opens new possibilities for interactions of these particles. Interestingly, three-dimensional crystals of recombinant NCPs, used to determine the high-resolution structure of the particles, were obtained in the presence of divalent cations (37.5 mM manganese; Luger et al. 1997a,b), and these ions occupy precise locations in the crystals. Under such conditions, NCPs also stack to form columns that may be of different types, which cannot be described in detail here (Luger et al. 1997a; Harp et al. 2000; Davey et al. 2002).
6. Do we expect similar organizations of nucleosome core particles in the living cell?
In chromosomes, NCPs are connected together by the continuous DNA link. Other components are also present, namely the fifth histone protein H1 and other proteins involved in the functional activity of the genetic material. The living system is therefore of a much higher complexity, with local variations in the concentrations of the multiples components of chromatin. These local heterogeneities offer the possibilities to monitor local changes of short-range interactions between NCPs, via histone tails interactions (the current hypothesis predicts that distinct covalent modifications of the histone tails, such as acetylation, methylation, etc., would determine a code (Jenuwein & Allis 2001). Interestingly, phase transitions can be triggered experimentally using living cells. The whole nucleus already engaged in a process of cell division can be swollen to several times its initial size by changing the ionic conditions, further condensed again by returning to initial ionic conditions to finally complete its cell division process (Aaronson & Woo 1981; Bojanowski & Ingber 1998). Such experiments demonstrate the reversibility of chromatin structural changes upon variations of ionic conditions. Similar reversible phase transitions of unique chromosomes were also obtained experimentally by changing ionic conditions and using micromanipulation techniques (Poirier et al. 2002). Unfortunately, structural data to understand how chromatin reorganizes during these processes are missing. The ordered phases formed by isolated NCPs in solution can be considered as possible models for the organization of chromatin in vivo. They provide a library of structures that may possibly exist in situ, since they were found under biologically relevant salt and NCP concentration ranges. Telomeric chromatin (a chromatin located at the extremity of chromosomes and playing a crucial role in the control of the lifetime of a given cell; Blackburn 2005) was proposed to be organized in a columnar phase for biochemical reasons, such as the lack of nucleosome positioning and a specific spacing along the DNA (Fajkus & Trifonov 2001). Lamellar structures with a periodicity of about 50 nm were also described in the fish Scyliorhinus sperm cell at a given step of the spermiogenesis process (Gusse & Chevaillier 1978). This organization may remind us of the lamello-columnar phase of NCPs. These and other examples lead us to suspect the possible regular organization of nucleosomes in special types of chromatin, but the structures remain to be analysed carefully, using new available developments of ultrastructural methodology. More generally, in the classical eukaryotic chromatin, we do not expect the organization of NCPs to extend over long distances, but to be most probably restricted to micro-domains. In the context of the living cell, not only physico-chemical processes but also the activity of numerous enzymes and cofactors would help chromatin to organize in a sophisticated way, following an unknown path in the extremely complex phase diagram of chromatin. We guess that this complexity may offer extremely high possibilities of adaptation of chromatin organization to multiple local constraints.
G. Jackson (Chemical Engineering, Imperial College London, UK). Could you tell us more about the chirality of the phases that you have been studying? Are the nematic phases chiral?
F. Livolant. Both the columnar hexagonal phase and the lamello-columnar phase show evident chiral properties that we analysed.
Small growing domains of the columnar hexagonal phase present sixfold symmetries as in classical discotic phases (see figure 6d). A double-twist effect exists, but it is small and all the columns of nucleosomes align more or less normal to the sixfold axis. When these domains enlarge, the chirality of the phase becomes more visible. Competition between hexagonal packing and double-twist configurations leads to the individualization of six branches from a single hexagonal domain. These branches often twist to form ‘diabolo’ structures with a left-handed helicity. From optical observations, we deduced that the twist was right handed between columns of nucleosomes in these domains. More details of these structural analyses can be found in Livolant & Leforestier (2000). Recently, we also observed, in X-ray patterns of aligned samples of the three-dimensional orthorhombic phase, additional peaks (see figure 5d,d′) that could originate from a chiral organization of NCPs (Mangenot et al. 2003a).
The textures of the lamello-columnar phases that I presented are also the consequence of chiral properties of the nucleosomes. The tubes (see figure 6a) are often helical at their extremities (not shown here). Their formation itself is probably arising from the twist occurring between the columns of nucleosomes aligned in parallel in the bilayer. This twist distorts the bilayer into a twisted ribbon that relaxes into a tubular structure. In our experiments, on increasing the applied pressure, we observe the nucleation of new tubes at the extremity of the first ones, and this growth also comes from the chiral interactions between stacked bilayers, which lead finally to the formation of the spherulites (see figure 6b). Finally, the opposite tilt of nucleosomes in the two parts of the bilayer is also probably a consequence of the twist between columns from the two parts of the bilayer. A manuscript describing and discussing these aspects is in preparation.
E. T. Samulski (Department of Chemistry, University of North Carolina, USA). You alluded to magnetic orientation. The diamagnetic anisotropy of DNA is negative (ΔΧ<0) but when you wrap the double helix around the limestone fragment, the small column has ΔΧ>0. The column orients along the field. Therefore, the bilayer should align unless the twist of the bilayer causes a supramolecular structure that now has ΔΧ<0. Do you envisage the use of magnetic fields to elucidate the higher order supramolecular structure in the chromosome?
F. Livolant. We used high magnetic fields (10 T) to align the samples. We did not succeed with the lamello-columnar phase, but we obtained a good alignment of the columnar hexagonal phase (see figure 5d,d′). Columns align parallel to the field (DNA lies normal to the field) and this alignment allowed us to understand the X-ray diffractograms and differentiate rings that correspond to an ordering in the direction of the columns from rings coming from the direction normal to the columns. Indeed, the diameter of the nucleosome core particle (11 nm) is twice its thickness (5.7 nm), and without this alignment it was difficult to separate the two sets of diffraction rings. As you mention, our main interest is to understand the structure of chromatin and chromosome. In this purpose, we may try to use magnetic field to improve the organization of nucleosomes. We have not started such experiments yet.
A. S. Matharu (Department of Chemistry, University of York, UK). Within an NCP what is more important, the histones or the DNA that wraps around it? As the concentration changes, what role does the hydrophobic effect play and can you relate your phase diagram to these changes?
F. Livolant. Both DNA and histones are important in the nucleosome core particle. We took care to work under conditions where the NCP is stable, with no dissociation of the DNA from the histone core. The formation of the columns is driven by interactions between top and bottom surfaces of the NCP (histone–histone contacts), but we have no evidence that these interactions are driven by hydrophobic effects. Besides, lateral interactions between NCPs and between columns of NCPs are dominated by electrostatics, since the nature of the observed phases is dependent on the salt concentration. These lateral interactions are complex. DNA that brings 2×150 negative charges (carried by the phosphate groups of the 150 bp fragments) overcharges the protein core and confers a global structural negative charge to the particle. In dilute solution, these particles repulse each other up to a certain salt concentration, where the conformation of the NCP slightly changes. Attractive interactions between the positively charged amino acids of the terminal extensions of the histones tails and the DNA corona are screened and the histone ‘tails’ extend outside the particles. As soon as they extend, attractive interactions appear between NCPs. Therefore, lateral interactions depend on a delicate balance of repulsive and attractive interactions coming either from the DNA or the histone tails (see Mangenot et al. 2002a,b; Bertin et al. 2004).
V. Percec (Department of Chemistry, University of Pennsylvania, USA). Is any amplification of chirality at the transition from DNA to chromatins due to the rotational helical organization of DNA?
F. Livolant. I have already answered part of the question, and explained how the structures are the consequence of the chiral properties of the nucleosome core particle. So, yes, there is an amplification of the chirality up to the macroscopic scale. In the case of pure DNA, it has been shown that the right-handed helix of DNA is responsible for the left-handed twist between helices in the so-called ‘ψ-DNA’ aggregates (Maniatis et al. 1974) and in the cholesteric chromosome of the Dinoflagellates (Livolant et al. 1978). Nevertheless, in the dense phases of NCPs, it is not easy to determine whether the observed chiral effects come from the right-handed helical structure of the DNA molecule itself, or whether they come from the helical structure of the nucleosome, since the DNA follows a left-handed helix at the surface of the protein core. These two helices are of opposite handedness and their pitch is different. We may also wonder whether both helical structures may come into play, and whether one or the other may become dominant under defined conditions. So, it is a very interesting question. Experimental and theoretical approaches are needed to have a clear view of this point.
One contribution of 18 to a Discussion Meeting Issue ‘New Directions in Liquid Crystals’.
- © 2006 The Royal Society