Structure of the chromatosome, a chromatin particle containing 160 base pairs of DNA and all the histones

Nucleotide sequence-directed mapping of the nucleosomes

The concept of sequence-dependent deformational anisotropy of DNA proposed earlier is further elaborated and a computational procedure is developed for the sequence-directed mapping of the nucleosomes along chromatin DNA nucleotide sequences. The deformational anisotropy is found to be nonuniform along the molecule of the nucleosomal DNA, suggesting that the DNA superhelix in the nucleosome is slightly oval rather than circular in projection. The number of superhelical turns in the nucleosome core particle is estimated to be 2.0 +/- 0.2. Preliminary mapping of the nucleosomes in various chromatin DNA sequences yields the distribution of linker lengths which shows several minima separated by about 10 base-pairs. This is explained by sterical exclusion effects due to overlapping of the nucleosomes in space when some specific linker lengths are chosen. The mapping procedure described is tested by comparing its results with all the most accurate experimental mapping data reported so far. The comparison demonstrates that the exact positions of all the nucleosomes appear to be determined exclusively by the nucleotide sequences.

Biochemical characterization of topoisomerase I purified from avian erythrocytes

A type I topoisomerase has been purified from avian erythrocyte nuclei. The most pure fraction contains one major polypeptide of Mr = 105,000 (80% of total) and several minor ones of lower molecular weight. Active forms of the topoisomerase were identified by covalently binding the enzyme to 32P-DNA, digesting with nuclease and detecting 32P labeled peptides by sodium dodecyl sulfate polyacrylamide gel electrophoresis. Topoisomerase activity, as measured by the ability to covalently bind DNA, is associated with the following peptides: Mr = 105, 83, 54 and 30,000. The similar chromatographic properties of the various forms of topoisomerase suggests a common structural identity as previously proposed for the HeLa topoisomerase I (Liu, L.F. and Miller, K.G. (1981) Proc. Natl. Acad. Sci. USA 78, 3487-3491). The avian enzyme is similar to other eucaryotic type I DNA topoisomerases in that it covalently binds double and single stranded DNA forming an enzyme linked to the 3'-phosphoryl end and after binding to single stranded DNA it can transfer the single stranded donor DNA to an acceptor DNA possessing 5'-OH end groups. The binding site size of topoisomerase on DNA has also been determined using micrococcal nuclease to digest unprotected DNA in the native enzyme/DNA complex. The enzyme blocks access to the helix over a span of 25 bp. These findings are discussed in light of the distribution and function of topoisomerase I in chromatin.

Spin-label study of histone H1-DNA interaction. Involvement of the central part of the molecule in reconstituted chromatin

As shown in a previous paper [J. J. Lawrence et al. (1980) Eur. J. Biochem. 107, 263-269], covalent spin labeling of basis residues in histone H1 allows the study of the interaction of this protein with DNA. Using a step gradient dialysis procedure to reconstitute chromatin from labeled H1 and stripped chromatin, it is shown that the process of interaction of the lysine residues and DNA is the same whether histone H1 is bound to linear purified DNA or to H1-depleted chromatin. In contrast, spin labeling of the unique tyrosine of histone H1 located in the globular part of the molecule shows that this part is more involved in the interaction with chromatin than it is with linear DNA (as judged from the lengthening of the rotational correlation time). These data are interpreted as reflecting different roles for the N and C termini of the molecule of H1 and the central globular part. A model, based on these observations together with examination of the primary structures of histones H1, is proposed which accounts for the H1 involvement in the chromatosome structure.

Histone H5 can correctly align randomly arranged nucleosomes in a defined in vitro system

In eukaryotic cells, DNA is packed into regularly spaced chromatin subunits called nucleosomes. The average distance between nucleosomes (the repeat length) varies in a tissue- and species-specific manner, with values ranging from about 160 to 240 DNA base pairs (bp). Thus, it has been recognized that the repeat length could be one of the factors underlying selective gene expression. In cells growing in culture, the characteristic repeat length for that type of cell seems to arise from an immature chromatin structure in which nucleosomes are initially irregularly spaced or are arranged in small closely packed clusters. At present no in vitro system has been described which is capable of reconstituting the mature physiological nucleosome spacing from purified chromatin components. Moreover, neither the factors necessary for spacing nor the reaction mechanism are known. We describe here an in vitro system that can restore the native subunit spacing in rearranged chromatin samples which have irregularly spaced nucleosomes similar to the situation apparent in newly replicated chromatin.

Isolation of a 167 basepair chromatosome containing a partially digested histone H5

A test has been made of the proposal that protection of the 167 basepair DNA length in the 'chromatosome' is due only to the central globular domain of the lysine-rich histones. Chicken erythrocyte chromatin was treated with trypsin to leave only the limit peptide from histones H1 and H5. Nucleosome monomers were then isolated on sucrose gradients following micrococcal nuclease digestion and were found to contain the 167 basepair DNA band as in intact chromatin. The presence of the limit peptide from H5 on the monomers was confirmed using an antibody to H5. It is concluded that the trypsin-susceptible domains of the lysine-rich histones are not involved in the protection of the 2-turn 167 basepair length of DNA in the nucleosome.

Structural features of a phased nucleosome core particle

Chicken erythrocyte inner histones associate with a cloned 260-base-pair (bp) segment of Lytechinus variegatus DNA in a unique location. The fragment contains a 120-bp segment encoding 5S rRNA, a 90-bp flanking sequence to the 5' side of the transcribed segment, and a 50-bp downstream flanking sequence. Association of DNA, uniquely labeled at one end or the other and at either the 3' or the 5' terminus of a given strand, with histones at 0.1 M ionic strength leads to formation of a compact complex which sediments at about 13 S. Analysis of cutting of the complex by DNase I shows that protection from the nuclease is confined to a region beginning 20 bp from the left end of the segment and extending to about 165 bp from the left end. Within the protected region, the two DNA strands differ in their susceptibilities to the nuclease, the precise location of nuclease cutting sites and the spacing between these sites, and the relative susceptibilities of specific cutting locations. It seems that information present in DNA and the histone octomer is sufficient to create a precisely phased nucleosome in which interactions of the two DNA strands with histones are not the same. The structure of this unique nucleosome is not predicted by the intellectual model based on studies of mixed populations of nucleosome core particles.

Thermal denaturation studies of acetylated nucleosomes and oligonucleosomes

The thermal melting behaviors of control and acetylated mononucleosomes, dinucleosomes and trinucleosomes have been studied. Along each series of oligonucleosomes, the melting profiles change in a manner consistent with the increasing number of nucleosomes. For the control mononucleosome, the melting profile exhibits a premelting region at about 61-64 degrees C and a major cooperative transition at 75-77 degrees C. The melting profiles of the control dinucleosomes and trinucleosomes show a premelt at 61-62 degrees C (similar to that of the nucleosome core); an intermediate transition at 73-74 degrees C for the dinucleosome and at 76-77 degrees C for the trinucleosome and a major cooperative transition at 79-80 degrees C for the dinucleosome and at 81-82 degrees C for the trinucleosome. The major cooperative transition at the highest melting temperatures in the melting profiles of the mononucleosome, dinucleosome and trinucleosome comes from the melting of the central region of DNA in the nucleosome which complexed with the core histones; the premelt region is attributed to two DNA segments per nucleosome which flank this central DNA region and are free or weakly complexed with histones. The origin of the intermediate transition found for the dinucleosomes and trinucleosomes is not fully understood but probably results from the melting of DNA at the entry to and exit from the nucleosome and the linker DNA which are complexed with histones. A very similar pattern of behavior is observed for the acetylated oligonucleosomes. Direct comparison of the melting profiles of acetylated and control mononucleosomes, dinucleosomes and trinucleosomes show that the premelt region is unaffected by histone acetylation whereas the intermediate and major cooperative transitions for the acetylated oligonucleosomes are broader and occur consistently at lower temperatures than for the controls. These differences support proposals that the N-terminal regions of core histones interact within the nucleosome and on linker DNA.

Developmental changes in DNAse I digestibility and RNA template activity of neuronal nuclei relative to the postnatal appearance of a short DNA repeat length

Neurons of the rat cerebral hemispheres are known to undergo a postnatal shift to a short DNA repeat length. In the present study we report that rat neuronal nuclei are more sensitive to digestion with DNAse I when isolated at a developmental stage after the shift in neuronal DNA repeated length compared to nuclei isolated before the shift. This observation may suggest that a decondensation of neuronal chromatin accompanies the postnatal shift in DNA repeat length. We have also found that neuronal nuclei isolated after the shift to a short DNA repeat length demonstrate an increased ability to synthesize RNA in vitro.

Alignment of nucleosomes along DNA and organization of spacer DNA in Drosophila chromatin

A series of mono- and dinucleosomal DNAs characterized by an about ten-base periodicity in the size were revealed in the micrococcal nuclease digests of Drosophila chromatin which have 180 +/- 5 base pair (bp) nucleosomal repeat. 20, 30, and 40 bp spacers were found to be predominant in chromatin by trimming DNA in dinucleosomes to the core position. Among several identified mononucleosomes (MN), MN170, MN180 and MN190 were isolated from different sources (the figures indicate the DNA length in bp). The presence of the 10, 20, and 30 bp long spacers was shown in these mononucleosomes by crosslinking experiments. The interaction of histone H3 with the spacer in the Drosophila MN180 particle was also shown by the crosslinking /5/. We conclude from these results that the 10 n bp long intercore DNA (n = 2, 3 and 4) is organized by histone H3, in particular, and together with the core DNA forms a continuous superhelix. Taken together, these data suggest that Drosophila chromatin consists of the regularly aligned and tightly packed MN180, as a repeating unit, containing 10 and 20 bp spacers at the ends of 180 bp DNA. Within the asymmetric and randomly oriented in chromatin MN180, the cores occupy two alternative positions spaced by 10 bp.

Differential inhibition of transcription of DNA by melanoma chromosomal proteins

Histones and 4 nuclear nonhistone protein fractions (NHP1-4) were extracted from nuclei of a Cloudman mouse melanoma cell line (NCTC 3960, CCL 53) and tested for their ability to bind to DNA and influence transcription. The histones and NHP fractions showed different binding affinities for DNA, with the histones and NHP1 exhibiting the highest affinity. The NHP fractions differentially affected both the rate of RNA synthesis and the size of RNA transcribed. NHP1 which inhibited RNA synthesis to the greatest extent, inhibited synthesis of all sizes of RNA except for major peaks of 28S and 8S RNS and discrete minor peaks of 7S, 6S, 5S, and 4S RNS. Histones markedly enhanced the effect of NHP1 on RNA synthesis. These results suggest that there are nonhistone proteins in Cloudman melanoma nuclei which have a high affinity for DNA and which may be involved in the regulation of transcription.

Nucleosome spacing in rat liver chromatin. A study with exonuclease III

Exonuclease III was used to uniformly trim DNA ends of micrococcal nuclease-prepared chromatin fragments down to the first major impediment encountered by the enzyme, which arises from the interaction of H1 with the nucleosome. This trimming, when performed on nucleosome dimers, allowed one to quantitatively determine the center-to-center distance of nucleosomes. This distance, of mean 198 base pairs, was found to essentially vary between about 180 and 215 base pairs, with extremes of 165 and 230 base pairs. Trimming of trimers further revealed that the overall arrangement of nucleosome center-to-center distances along the chromatin fiber is that expected on a statistical basis.

Higher order chromatin structure determines double-nucleosome periodicity of DNA fragmentation

Double-nucleosome periodicity of DNA fragmentation with DNAse I in the nuclei of cells differing in size of the linker DNA length and lysine-rich histone composition was analyzed by means of nondenaturing agarose gel electrophoresis. DNAse I revealed this type of periodicity in rat thymus and CHO cell nuclei as well as in erythrocyte nuclei. It has been deduced that the so-called nucleodisome structure is also typical of cells possessing a usual DNA repeat length (200 bp or less) and lysine-rich histone H1. Two probably related events are important for establishing a clear double-nucleosome periodicity of DNA fragmentation: the replacement of H1 histone by a specific arginine-rich histone fraction (H5 histone in the case of erythrocyte) and the increase of the linker DNA length. The results are interpreted in terms of supranucleosomal organization of chromatin which may determine the dinucleosome periodicity of DNA fragmentation due to a specific packing of nucleosomes.

Conditions for sliding of nucleosomes along DNA: SV 40 minichromosomes

'Sliding' of nucleosomes along DNA under nearly physiological conditions was studied using treatment of SV 40 minichromosomes with the single-cut restriction endonucleases EcoRI and BamHI. Each enzyme can convert no more than 20-25% of the circular DNA molecules of minichromosomes into the linear form irrespective of the presence of histone H1. This suggests absence of the nucleosomes lateral migration (sliding) along DNa at least in the vicinity of the restriction endonucleases cleavage sites during several hours of incubation. The sites available for EcoRI and BamHI in minichromosomes seem to be located predominantly in the spacer DNA regions of nucleosomes. Introduction of only one double-strand (but not single-strand) break into the DNA of minichromosomes stripped of histone H1 is sufficient to induce redistribution of the nucleosome core particles due to their sliding along DNA. Thus, sliding of the nucleosome core particles can be induced under physiological conditions by rather low energy expenditures.

Excision of benzo[a]pyrene diol epoxide I adducts from nucleosomal DNA of confluent normal human fibroblasts

The formation and removal of covalent adducts of racemic 7 beta, 8 alpha-dihydroxy-9 alpha,10 alpha-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene (BPDE I) was studied in nucleosomal DNA of confluent cultures of normal human fibroblasts (NF). For this purpose NF were prelabeled in their DNA with [14C]-thymidine and treated with [3H]BPDE I. The adducts were composed of 77% (7R)-N2-(7 beta, 8 alpha, 9 alpha-trihydroxy-7,8,9,10-tetrahydrobenzo[a]pyrene-10-yl)deoxyguanosine, 12% of the corresponding 7S-enantiomer and of minor amounts of adducts to cytosine and adenine. The adduct composition did not change significantly in 24-h post treatment incubation. Bulk mononucleosomes were prepared from micrococcal nuclease digested nuclei and their DNA analyzed by gel electrophoresis. The adduct concentrations were determined in 145 base pair (b.p.) nucleosomal core-DNA, 165 b.p. chromatosomal DNA and in total nuclear DNA. From these data the concentration in nucleosomal linker-DNA was calculated. The initial adduct distribution was non-random and 6.3 times higher in 47 b.p. linker-DNA relative to 145 b.p. core-DNA and 9.2 times higher in 27 b.p. linker-DNA relative to 165 b.p. chromatosomal DNA. Adduct removal was very rapid during the first 8 h and more efficient from linker-DNA than from core-DNA. After this early phase the adducts located in 145 b.p. core-DNA became refractory to further excision and represent a major fraction of the adducts persisting in DNA of NF over a prolonged period. In contrast, further adduct removal was observed from nucleosomal linker-DNA.

[Structural arrangement of chromatin (author's transl)]

Recent observations and hypotheses on the structure of chromatin are reviewed. Elementary "subunit" for higher structural orders is the nucleosome, consisting of a histone octamer and doublehelical DNA wrapped around it. During the last years details of the nucleosomal structure could be deduced down to a resolution of 2 nm. In the chromatin fiber, built up by (mono-)nucleosomes, the superhelical DNA has a tertiary structure, from which structures of still higher order (quaternary structure of DNA) can be formed. The correlation of these structures to the terms euchromatin and heterochromatin are discussed. Finally, some functional aspects, especially transcription and replication, are discussed in view of the new knowledge of chromatin structure.

Tightly bound non-histone proteins in nucleosomes from pig-liver chromatin

Core particles prepared by micrococcal nuclease digestion of pig liver chromatin have been adsorbed on hydroxyapatite and dissociated by gradual increase in ionic strength and finally by urea and guanidine. By this method non-histone proteins have been found to be associated with the core particles. Proteins tightly bound to the core particle DNA (i.e. dissociated only by urea and guanidine) have also been found: these are proteins with a limited heterogeneity, with respect to their molecular weights, since only six components are present with molecular weights ranging from 71000 to 20000. They show, furthermore, a peculiar amino acid composition. Other tightly bound proteins have been shown to be present only in the spacer regions. The existence of two different classes of tightly bound proteins probably reflects different modes of binding to the DNA, which are compatible or incompatible, respectively, with the simultaneous binding of the histone octamer.

Histones H1 and H5: one or two molecules per nucleosome?

We have determined histone stoichiometries in nuclei from several sources by a direct chemical method, with the particular aim of quantitating histone H1 and, in chicken erythrocytes, H5, and of distinguishing between one and two molecules per nucleosome. The four histones H3, H4, H2A and H2B are found in equimolar amounts, as expected for the core histone octamer. The molar ratio of H1 in lymphocyte and glial nuclei is 1.0 per octamer, and in liver nuclei from three species 0.8 per octamer. These results suggest that each nucleosome has one H1 molecule; nucleosomes could acquire two molecules of H1 only at the expense of others containing none. The stoichiometry of H5 in chicken erythrocyte nuclei is similar to that of H1 in other nuclei, being about 0.9 molecules per nucleosome; the H1 also present in these nuclei amounts to 0.4 molecules per nucleosome.

Modulation of nucleosome structure by histone subtypes in sea urchin embryos

Switches of the types of histones synthesized and incorporated into chromatin occur during sea urchin embryogenesis. In an attempt to define the possible effects of these variant histones on chromatin structure, I have isolated and characterized nucleosome core particles from Strongylocentrotus purpuratus blastula (nearly 100% early histones) and pluteus (75% late histones). Both particles contain 146-base-pair lengths of DNA wrapped around an octamer of H2A, H2B, H3, and H4. Although sharing these similarities with the canonical core particle, the nucleosome structures have certain features that differ from those of typical adult tissues. Both the reversible and the irreversible conformational transitions occurring on heating core particles are destabilized in the embryonic particles vs. "typical" core particles. The blastula core particle unfolds more easily than pluteus (or other) nucleosomes under the stress of low ionic strength. The rate of DNase I digestion of pluteus core particles is about half that of particles from blastula; certain cutting sites differ in their susceptibility between the two embryonic particles and between these two and the canonical core particle. The data demonstrate that the variant histones synthesized during early embryogenesis have demonstrable effects on chromatin structure, even at this basic level.

Localization of DNA methyltransferase in the chromatin of Friend erythroleukemia cells

Chromatin fragments released from intact Friend erythroleukemia cell nuclei during limited incubation with micrococcal nuclease, DNase II or DNase I were analyzed to determine the distribution of DNA methyltransferase in chromatin. The enzyme was released in a free form when internucleosomal DNA was digested with micrococcal nuclease but was found associated with Mg++-precipitable polynucleosomes after DNase II digestion. Less than 25% of the enzyme was released from nuclei incubated with DNase I under conditions where transcriptionally active chromatin should have been completely digested. These results indicated that the bulk of DNA methyltransferase was bound to "linker" DNA in condensed regions of chromatin. Preferential rebinding of free enzyme to linker DNA was also demonstrated in vitro. The possibility that chromatin proteins play a role in regulating access of DNA methyltransferase to specific sites in DNA is discussed in light of these findings.

Studies on structure and function of chromatin

This article covers research work in this laboratory on the structure and function of chromatin. Early studies have led to discovery of skeletal fibrils (nucleonemas) within the nuclei and showed the specific role of histone H1 in chromatin condensation and in restriction of transcription. More recently with the aid of a novel DNP electrophoresis technique the relation of histone H1 and non-histone proteins to nucleosomes was studied. Three types of mononucleosomes and a number of subnucleosomes were identified in chromatin digests. The complexes of certain HMG proteins with short DNA segments were isolated and found to originate frm transcriptionally active chromatin. Different forms of SV40 minichromosome were characterized. A method for the analysis of nucleosome distribution along the DNA sequence was elaborated and used to show non-random (phased) location of nucleosomes on SV40 DNA. The attachment of DNA to skeletal elements of interphase nuclei and metaphase chromosomes was shown to be a non-random, probably sequence-specific process.

Iodination of nucleosomes at low ionic strength: conformational changes in H4 and stabilization by H1

Radioactive iodine has been used to probe the relative reactivities of nucleosomal H4 tyrosine residues under various conditions of subphysiological ionic strength. We observe that tyrosine 72 of H4, which is not reactive over the range 20-150 mM NaCl, becomes the predominant site of iodination within H4 when nucleosomes are subjected to conditions of very low ionic strength. Conversely, the other H4 tyrosine residues, which are reactive within nucleosomes in solutions of moderate ionic strength (20-150 mM NaCl), become nonreactive when the ionic strength is reduced. This "flip-flop" in the H4 iodination pattern is the manifestation of a reversible nucleosomal conformational change. A method is presented which enables the conformational status of H4 in nucleosomes to be determined by simply electrophoresing the histones on a Triton gel after probing nucleosomes with labeled iodine. Using this technique, we demonstrate that the presence of H1 on one side of the nucleosome stabilizes a histone core domain on the other side so that all four tyrosines of H4 are maintained in their physiological ionic strength conformation even under conditions of no added salt.

Protection of discrete DNA fragments by the complex H1-octamerhistones or H5-octamerhistones after micrococcal nuclease digestion

Several authors, including ourselves, have reported the existence of chromatosomes with DNA size larger than 166 bp in bird erythrocyte chromatin. It was tempting to correlate this increased DNA size with the presence of histone H5. In order to substantiate this hypothesis, we performed a micrococcal nuclease digestion kinetic on: chicken erythrocyte chromatin, either native, selectively depleted from H1, or from H1 and H5; and rat liver chromatin, either native or partially H1 depleted. The comparative analysis of the lengths of DNA in the chromatosome size region led to the following conclusions: - denaturing gels clearly reveal a first discrete pause at 178 nucleotides in H1 depleted chicken erythrocyte chromatin as well as in partially H1-depleted rat liver chromatin, before the material accumulates at the next intermediate 166 nucleotide chromatosome pause. - the generation of all discrete chromatosome bands is critically dependent on low ionic strength conditions and low Ca++ concentrations during the digestion, suggesting it may result from the protection of DNA cleavage sites by histone H5 or H1, C or N terminal domains.

Specific interaction of histone H1 with eukaryotic DNA

The interaction of calf thymus histone H1 with homologous and heterologous DNA has been studied at different ionic strengths. It has been found that about 0.5 M NaCl histone H1, and its fragments N-H1 (residues 1-72) and C-H1 (residues 73-C terminal), precipitate selectively a small fraction of calf thymus DNA. This selective precipitation is preserved up to very high values (less than 2.0) of the input histone H1/DNA ratio. The percentage of DNA insolubilized by histone H1 under these ionic conditions is dependent upon the molecular weight of the nucleic acid, diminishing from 18% fro a Mw equals 1.0 x 10(7) daltons to 5% for a Mw equals 8.0 x 10(4) daltons. The base composition of the precipitated DNA is similar to that of the bulk DNA. Calf thymus histone H1 also selectively precipitates a fraction of DNA from other eukaryotes (herring, trout), but not from some prokaryotes (E. coli, phage gamma. On the other hand, at 0.5 M NaCl, the whole calf thymus DNA (but not E. coli DNA) presents a limited number of binding sites for histone H1, the saturation ratio histone H1 bound/total DNA being similar to that found in chromatin. A similar behavior is observed from the histone H1 fragments, N-H1 and C-H1, which bind to DNA in complementary saturation ratios. It is suggested that in eukaryotic organisms histone H1 molecules maintain specific interactions with certain DNA sequences. A fraction of such specific complexes could act as nucleation points for the high-order levels of chromatin organization.

Distribution of linker DNA in relation to core DNA in H1 depleted nucleosomes

The mode of interaction of histone H1 with the nucleosome is governed by the relative distribution of the linker with respect to the core DNA. Preliminary experiments (Simpson, R.T. 1978, Biochemistry, 17, 5524-5531) and tentative models (Thoma, F. et al. (1979), J. Cell. Biol., 83, 403-427) suggest that part of the linker complete two full turns of DNA around the histone core, probably by adding 10 base pairs at each end of the core DNA. In the present study Exonuclease III has been utilized to digest the 3' ends of H1 depleted nucleosomes. (i.e. the 195 base pair particle). The analysis of the resulting DNA fragments under denaturing conditions shows that the whole linker is distributed symmetrically with respect to the core DNA.

Location of the ethidium binding sites of high affinity in chromatin

The influence of H1 and H5 histones proteins upon the accessibility of ethidium bromide into chromatin is studied by steady-state fluorescence anisotropy in the range of r-values ([Dye]/[Phosphate]) smaller than 0.01. This corresponds to the very strong binding process. When H1 and H5 are present, the DNA segment which contains the binding sites is 25-30 base pairs long, even if H1 and H5 are digested by trypsin or by natural proteolysis, but presumably still interacting with the DNA chromatin. On the contrary, when H1 or H5 are separated from chromatin by an increase of the ionic strength, ethidium binds to a segment of DNA about 55-60 base pairs long. We may explain the results by assuming that the ethidium sites are located on a continuous segment constituting about one half of the linker, the other half interacting with H1 and H5. When chromatin is depleted from these proteins, the high affinity sites are distributed all along the linker.