Portal de Periódicos da CAPES

Multiple Overlapping Positions of Nucleosomes with Single in Vivo Rotational Setting in the Hansenula polymorpha RNA Polymerase II MOX Promoter

1995; Elsevier BV; Volume: 270; Issue: 19 Linguagem: Inglês

10.1074/jbc.270.19.11091

1083-351X

Giovanna Costanzo, Ernesto Di Mauro, Rodolfo Negri, Gonçaio Pereira, Cornells Hollenberg,

DNA Repair Mechanisms

In vivo nucleotide-level mapping of nucleosomes in the promoter of the methanol oxidase (MOX) gene in the yeast Hansenula polymorpha is reported. The 4 nucleosomes analyzed are organized in families; they localize in alternative positions along a unique rotational phase, and the linker regions can be occupied by alternative nucleosomes. This organization underscores a substantial freedom of choice by histone octamers when nucleating on a promoter region. In vivo nucleotide-level mapping of nucleosomes in the promoter of the methanol oxidase (MOX) gene in the yeast Hansenula polymorpha is reported. The 4 nucleosomes analyzed are organized in families; they localize in alternative positions along a unique rotational phase, and the linker regions can be occupied by alternative nucleosomes. This organization underscores a substantial freedom of choice by histone octamers when nucleating on a promoter region. Knowledge of the exact in vivo distribution of nucleosomes in a large chromosomal region is the basis for the assessment of the significance of their positions in the regulation of transcription (van Holde, 1993van Holde K.E. Nature. 1993; 362: 111-112Crossref PubMed Scopus (33) Google Scholar) and for an understanding of their fate and function during the transcription process (Wolffe et al., 1986Wolffe A.P. Jordan E. Brown D.D. Cell. 1986; 44: 381-389Abstract Full Text PDF PubMed Scopus (105) Google Scholar; Lorch et al., 1987Lorch Y. La Pointe J.W. Rornberg R.D. Cell. 1987; 49: 203-210Abstract Full Text PDF PubMed Scopus (387) Google Scholar; Losa and Brown, 1987Losa R. Brown D.D. Cell. 1987; 50: 801-808Abstract Full Text PDF PubMed Scopus (92) Google Scholar; Lee and Garrard, 1991Lee M.S. Garrard W.T. EMBO J. 1991; 10: 607-615Crossref PubMed Scopus (101) Google Scholar; van Holde et al., 1992van Holde K.E. Lohr D.E. Robert C. J. Biol. Chem. 1992; 267: 2837-2840Abstract Full Text PDF PubMed Google Scholar; Clark and Felsenfeld, 1992Clark D.J. Felsenfeld G. Cell. 1992; 71: 11-22Abstract Full Text PDF PubMed Scopus (122) Google Scholar; O'Neill et al., 1993O'Neill T.E. Smith J.G. Bradbury E.M. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 6203-6207Crossref PubMed Scopus (30) Google Scholar; Adams and Workman, 1993Adams C.C. Workman J.L. Cell. 1993; 72: 305-308Abstract Full Text PDF PubMed Scopus (139) Google Scholar; Studitsky et al., 1994Studitsky V.M. Clark D.J. Felsenfeld G. Cell. 1994; 76: 371-382Abstract Full Text PDF PubMed Scopus (206) Google Scholar. The analysis of the sequence periodicities in a large family of nucleosome core DNAs (Satchwell et al., 1986Satchwell S.C. Drew H.R. Travers A.A. J. Mol. Biol. 1986; 191: 659-675Crossref PubMed Scopus (790) Google Scholar has shown that nucleosome positioning depends on sequence preferences and not on absolute sequence requirements. In particular, the structural characteristics of the preferred DNA sequences and their preferred distributions have revealed that nucleosome localization is determined by the phased repetition of local anisotropically flexible sequences, as predicted in Trifonov and Sussmann, 1980Trifonov E.N. Sussmann J.C. Proc. Natl. Acad. Sci. U. S. A. 1980; 77: 3816-3820Crossref PubMed Scopus (514) Google Scholar. This observation solves the problem of how nucleosomes can localize defined positions (Almer and Hörz, 1986Almer A. Hörz W. EMBO J. 1986; 5: 2681-2687Crossref PubMed Scopus (189) Google Scholar; Richard-Foy and Hager, 1987Richard-Foy H. Hager G.L. EMBO J. 1987; 6: 2321-2328Crossref PubMed Scopus (450) Google Scholar; Piña et al., 1990aPiña B. Barettino D. Truss M. Beato M. J. Mol. Biol. 1990; 216: 975-990Crossref PubMed Scopus (58) Google Scholar, Piña et al., 1990bPiña B. Brüggemeier V. Beato M. Cell. 1990; 60: 719-731Abstract Full Text PDF PubMed Scopus (335) Google Scholar on an essentially unlimited number of potential sites in genomic DNAs. Irrespective of the quality of the involved sequences, this statistical analysis has established that one important parameter in nucleosome localization is the repetition of the flexibility signals and their phasing, which was shown to be equal to the helical periodicity. In DNA, this helically phased periodic repetition defines the "rotational information," as opposed to the continuous "translational" information, determined by the local sequence in its entirety. The localization of the histone octamer on DNA and the energetics of this process depend on the interaction with 14 helically phased DNA sites. Several properties of nucleosomes descend from this fact. In the case, for instance, of two alternative histone octamers/DNA interactions shifted by only one helical turn one relative to the other, 13 out of 14 DNA interaction sites remain the same, only a single lateral interaction site being different. Therefore, unless invoking the presence in the nucleosomes of DNA sequences with special properties or unusual structural features that could act as external boundaries or internal determinants, the rotationally based recognition mechanism has an intrinsic potential consequence; the repetition of signals may in principle favor multiplicity of alternative localizations that could differ by one or a few helical turns. In order to verify this theoretical model and its in vivo effects, we have studied the in vivo distribution of nucleosomes on a RNA polymerase II promoter: the Hansenula polymorpha methanol oxidase (MOX) 1The abbreviation used is: MOXmethanol oxidase gene promoter. methanol oxidase H. polymorpha is a methylotrophic yeast able to grow on methanol as a sole energy and carbon source. Growth on methanol leads to proliferation and enlargement of peroxisomes and synthesis of large amounts of peroxisomal enzymes involved in methanol metabolism (Veenhuis et al., 1983Veenhuis M. van Dÿken J.P. Harder W. Adv. Microbial Physiol. 1983; 24: 1-82Crossref PubMed Scopus (177) Google Scholar. MOX is the most abundant of these enzymes and is encoded by the MOX gene, the transcription of which is completely abolished by catabolite repression (glucose or ethanol) and strongly enhanced when the cells grow in methanol (Roggenkamp et al., 1984Roggenkamp R. Janowicz Z. Stanikowsky B. Hollenberg C.P. Mol. Mol. Genetics. 1984; 194: 489-493Crossref PubMed Scopus (71) Google Scholar. Under these conditions, MOX makes out up to 30% of the soluble cell protein. The MOX gene has been sequenced (Ledeboer et al., 1985Ledeboer A.M. Edens L. Moat J. Visser C. Bos J.W. Verrips C.T. Janowicz Z.A. Eckart M. Roggenkamp R. Hollenberg C.P. Nucleic Acids Res. 1985; 13: 3063-3083Crossref PubMed Scopus (149) Google Scholar, and three cis-regulatory regions that are occupied by DNA-binding proteins have been mapped and analyzed (Gödecke et al., 1994Gödecke S. Eckart M. Janowics Z.A. Hollenberg C.P. Gene. 1994; 139: 35-42Crossref PubMed Scopus (37) Google Scholar. In addition, a region involved in glucose repression and derepression was identified to which an Adr1p homologue (FB2) may bind. 2G. Pereira and C. Hollenberg, unpublished observations. Early during derepression, low levels of transcription starting close to UAS1 and upstream to the FB2 binding site have been observed, which do not lead to Mox protein. Moreover, this transcription is abolished when the proper MOX transcription starts at a point downstream to the FB2 binding site. Whether FB2 plays a role in this transcription start point selection is unclear since this putative factor is present at nearly the same concentration in both repressed and derepressed cells. To be able to further understand the mechanism controlling the complex regulation of MOX, we have studied the chromatin structure of its promoter. We describe here the nucleotide-level localizations of four nucleosomes (termed –3, –2, –1, and +1, relative to the ATG) and of other proteins in the H. polymorpha MOX promoter. The low resolution localization of seven additional nucleosomes in the surrounding sequences is also reported. Unexpectedly, we observed that in vivo the –3 to +1 nucleosomes localize on multiple alternative positions characterized by (i) unique rotational phase and (ii) extensive overlapping of the alternative extremities. Materials—T4 polynucleotide kinase was purchased from U. S. Biochemical Corp.; Klenow DNA polymerase, exonuclease III, DNase-I, restriction endonucleases, and micrococcal nuclease were purchased from Boehringer Mannheim or New England Biolabs. Taq DNA polymerase was from Promega; radiochemicals were purchased from DuPont NEN, and nystatin was from Sigma. Strains and DNAs—The DNA region analyzed is the H. polymorpha MOX promoter and part of its coding sequence (Ledeboer et al., 1985Ledeboer A.M. Edens L. Moat J. Visser C. Bos J.W. Verrips C.T. Janowicz Z.A. Eckart M. Roggenkamp R. Hollenberg C.P. Nucleic Acids Res. 1985; 13: 3063-3083Crossref PubMed Scopus (149) Google Scholar. Growth conditions were as in Gödecke et al., 1994Gödecke S. Eckart M. Janowics Z.A. Hollenberg C.P. Gene. 1994; 139: 35-42Crossref PubMed Scopus (37) Google Scholar. The reported chromatin analysis was performed in the repressed condition on mid-log cells grown on glucose. In Vivo Chromatin Analysis: Micrococcal Nuclease—Low resolution in vivo nucleosome mapping was performed by micrococcal nuclease cleavage of nystatin-treated spheroplasts (see below) by indirect end-labeling (Wu, 1980Wu C. Nature. 1980; 286: 854-860Crossref PubMed Scopus (753) Google Scholar) using the probes indicated (see Fig. 1). High resolution in vivo mapping was performed according to Buttinelli et al., 1993Buttinelli M. Di Mauro E. Negri R. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 9315-9319Crossref PubMed Scopus (56) Google Scholar as follows: spheroplasts (Almer and Hörz, 1986Almer A. Hörz W. EMBO J. 1986; 5: 2681-2687Crossref PubMed Scopus (189) Google Scholar) were washed with 1 m sorbitol and suspended (1–5 × 107/ml) in 1 m sorbitol, 20 mm Tris-HCl, pH 8, 50 mm NaCl, 1.5 mm CaCl2, with nystatin 100 μg/ml. These conditions allow penetration of micrococcal nuclease or DNase-I into the spheroplasts and digestion of chromatin in the absence of lysis (Venditti and Camilloni, 1994Venditti S. Camilloni G. Mol. & Gen. Genet. 1994; 242: 100-104Crossref PubMed Scopus (29) Google Scholar). Micrococcal nuclease was added at 20 units/ml, or as otherwise indicated; after 10 min at 37 °C, the reaction was stopped with 5 mm EGTA. Spheroplasts were digested with proteinase K in 0.2% SDS, extracted 3 times with phenol/chloroform/isoamyl alcohol, 24:24:1 (v/v/v). Resuspended samples were electrophoresed in 1.5% agarose gels to resolve the nucleosomal DNA ladder. Monomer DNA was eluted, in part treated with phosphatase, labeled at low specific activity with kinase, and run on denaturing Polyacrylamide to remove internally nicked monomeric molecules. Full-length denatured monomelic DNAs were recovered, purified, quantified by spectrophotometry, and used in multiple cycles of primer extension by Taq polymerase with a 5′-labeled oligonucleotide primer (as mapped in Fig. 3), thus locating the borders of the monomer-length micrococcal nuclease digestion products (the in vivo protected nucleosomal DNA), which contain the indicated oligo primers.FIG. 3Localization of nucleosomes in cellular chromatin. Map of the data reported in Fig. 2. Oligos are indicated by couples of arrows and large-sized numbers; labeled extremities are starred, and their position is indicated. The experimentally defined extremities of mononucleosomes are indicated; when several bands are present, only the most upstream value is reported. Nucleosome positions are indicated by thick bars and numbered in families. The dashed line for the upper border of nucleosome 5 of the family –2 refers to mapping uncertainty. The borders at 108–110 and 117 of probe 8 are not reported in the map. Vertical lines indicate the downstream border of potential alternative occupancies (for quantitative evaluation, see text).View Large Image Figure ViewerDownload Hi-res image Download (PPT) The following oligos were used respectively for mapping mononucleosomes –3, –2, –1, and +1: 1 (Watson strand, W) and 2 (Crick strand, C), –350 to –331, GGCTCTGTTTGCTGGCGTAGCCGAGACAAACGACCGCATC; 3 (W) and 4 (C), –230 to –211, GTGGGGTGTCGGACAGGCTGCACCCCACAGCCTGTCCGAC; 5 (W) and 6 (C), –50 to –31, TACTGCTGCCAGTGCACGGTATGACGACGGTCACGTGCCT; 7 (W) and 8 (C), +100 to +119, CAGTTGCCCTGATCGAGGGTGTCAACGGGACTAGCTCCCA. DNase-I—DNase-I was subject to permeabilization as above (see legend to Fig. 4). The oligo used in the DNase-I footprint experiment was 1. Fig. 1 describes the chromatin structure (in glucose, repressed) of the MOX promoter. The micrococcal nuclease analysis was performed from both sides: with probe 1 from downstream upward (panel a) and with probe 2 from upstream downward (panel b). The results, mapped in panel c, show that (i) probe 1 identifies nine protected areas (numbered as –11 to –3, relative to the ATG) the sizes of which correspond to that of a nucleosome. The positions from –3 to –9 are indicated in the map. (ii) Between nucleosomes –3 and –4, a well defined hypersensitive region with a short central interruption is evident, which maps around the region previously identified as UAS1 (Gödecke et al., 1994Gödecke S. Eckart M. Janowics Z.A. Hollenberg C.P. Gene. 1994; 139: 35-42Crossref PubMed Scopus (37) Google Scholar. (iii) Probe 2 identifies five nucleosomes (from –3 to +2). (iv) The map shows that in the overlapping areas, the results obtained from one side match with those obtained from the other. Alternative Positions and Multiple Overlapping Borders—The low resolution mapping reported above was used to program the nucleotide-level localization of nucleosomes. For this analysis (which essentially consists in the identification of the borders of mononucleosomes produced by extensive micrococcal nuclease treatment in vivo, performed as described in Buttinelli et al., 1993Buttinelli M. Di Mauro E. Negri R. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 9315-9319Crossref PubMed Scopus (56) Google Scholar and under "Experimental Procedures"), oligos were used that map in the center of the identified particles (see "Experimental Procedures" and FIG. 2, FIG. 3). The validity of this type of assay or of other related no-background assays (Fedor et al., 1988Fedor M.J. Lue N.F. Kornberg R.D. J. Mol. Biol. 1988; 204: 109-127Crossref PubMed Scopus (184) Google Scholar; Buttinelli et al., 1993Buttinelli M. Di Mauro E. Negri R. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 9315-9319Crossref PubMed Scopus (56) Google Scholar; Georgel et al., 1993Georgel P. Dretzen G. Jagla K. Bellard F. Dubrowsky E. Calco V. Bellard M. J. Mol. Biol. 1993; 234: 319-330Crossref PubMed Scopus (11) Google Scholar; Yenidunya et al., 1994Yenidunya A. Davey C. Clark D. Felsenfeld G. Allan J. J. Mol. Biol. 1994; 237: 401-414Crossref PubMed Scopus (28) Google Scholar for nucleosome mapping is established by the coherence of the results with those obtained with several other approaches such as mapping of exonuclease III-resistant borders, mapping by restriction cleavages, and analysis of the DNase-I protection and the hydroxyl-radical degradation patterns. FIG. 2, FIG. 3 show the results of high resolution mapping of in vivo nucleosomes –3 (Fig. 2a), –2 (panel b), –1 (panel c), and +1 (panel d). Each panel is composed of (i) a size marker (lane M) (the 147-base pair marker band is indicated in all panels); (ii) a full-length monomer 146 ± 2 base pairs long purified and labeled (lane N); (iii) the result of the primer extension from a centrally located oligo to the micrococcal nuclease-induced cleavage sites, which identify the nucleosomal borders (lane β). For each monosome, both rightward and leftward primer extensions are reported. The borders are identified by their size (values at the right side of each panel). The lengths and positions of the protected DNA segments are easily determined summing up couples of values from the two sides, subtracting 20 (the size of the overlapping oligos) and choosing the values closest to the size of the monomer on which the analysis was performed (146 ± 2). The resulting matches are schematically shown in Fig. 3. In the primer-extension conditions chosen, no significant background due to abortive elongation or pausing was produced (Fig. 2). The map in Fig. 3 shows the localization of the borders and reveals a peculiar behavior of nucleosomal particles in vivo. Three out of four nucleosomes in this promoter occupy several alternative positions giving rise to a family of related alternatives rather than to a single position; only nucleosome –3 has one major and one quite minor position. The regions located between two adjacent nucleosomes can be alternatively engaged in the interaction with the upstream- or with the downstream-located particle. The number of helical turns engaged in these alternative occupancies increases moving from the UAS1 downward toward the coding region. The numerical analysis of the borders (as reported on the side of each lane β in Fig. 2) shows that the position of the downstream border of the downstream nucleosome of the –3 family (–3·2) is engaged with sequences that could be occupied, in an alternative series of nucleosomes, by the upstream extremity of the upstream nucleosome of the –2 family. The potentially common positions span 1 helical turn for the region common to the –3/–2 families, 3 helical turns for the –2/–1 families and 3 turns for –1/+1. We shall term these areas "internucleosome overlaps," keeping in mind that the overlap is only potential, not actual, and refers to alternative occupancies. DNase-I—The existance of overlapping positions on the same rotational phases was confirmed by the DNase-I digestion profiles of the chromatin of nystatin-treated spheroplasts. The resulting pattern shows two types of variations relative to purified DNA (FIG. 4, FIG. 5). (i) The first are regions of markedly increased sensitivity corresponding to inter-nucleosome overlaps, spaced with areas of decreased sensitivity that correspond to nucleosomal central regions. These central regions, common to all the members of a family, span for less than 14 helical periods (see the distribution in Fig. 3) and encompass positions that are always occupied, in agreement with the micrococcal nuclease results. (ii) The second is a continuous series of helical-period-wise alternations of increased-decreased sensitivity to DNase-I spanning the whole analyzed region, including both the nucleosomal central regions and the internucleosomal overlaps. The very fact that periodicity of the DNase-I cleavage pattern is observed throughout indicates that the whole region is organized in nucleosomes, without rigidly defined linker regions. Fig. 4 shows the DNase-I pattern of in vivo chromatin (panel a, lanes 1–3) and of purified DNA (lanes 4–6). The helical-period-wise sensitivities are indicated by numbered dots on the gel image. The helical periodicity of the DNase-I cleavages is evident throughout, as clearly seen also on the scanning pattern (panel c). Fig. 5 shows the same type of analysis, extended to the most upstream nucleosome analyzed (-3), yielding the same results. If the linkers were completely devoid of nucleosomes, their cleavage pattern would in principle be similar to that of naked DNA. A close inspection of the internucleosomal overlap areas (i.e. in Fig. 4 the –1/–2 overlap) shows that the nucleosomal DNA is organized in helically phased DNase-I hypersensitive sites and that a few intermediate sensitive sites exist that are sensitive also on naked DNA (i.e. between –160 and –150 and between –110 and –102). This residual sensitivity could be due to incomplete occupancy (that is, in population terms, each linker site is unoccupied in a fraction of the molecules) or to local topological strain on the fraction of free DNA tracts. The footprint revealed by the DNase-I cleavage pattern between –203 and –160 is caused by an ADR1-like protein complex.2 The interaction takes place on DNA engaged on a nucleosome; the sequences involved are in the central region of the nucleosomal family –2 (FIG. 1, FIG. 2, FIG. 3). We have described the organization in families of nucleosomes in the H. polymorpha MOX promoter. The entire region is organized into nucleosomes. Sequences appearing in linker regions are not always nucleosome-free, but they can also be engaged in a nucleosome. Each DNA molecule contains one of several sets of possible nucleosomal patterns related to the helical periodicity of DNA. These results bear to the general problem of nucleosome localization and dynamics. In nucleosomes, DNA is wrapped around the histone octamer 1.8 times in a left-handed solenoid; topology-related aspects of DNA/nucleosome interaction are reviewed in Freeman and Garrard, 1992Freeman L.A. Garrard W.T. Crit. Rev. Eukaryotic Gene Expression. 1992; 2: 165-209PubMed Google Scholar and Negri et al., 1994Negri R. Costanzo G. Buttinelli M. Venditti S. Di Mauro E. Biophys. Chem. 1994; 50: 1-13Crossref PubMed Scopus (11) Google Scholar. The fact that nucleosomes bend DNA and the distribution of flexible sequences (Travers and Klug, 1987Travers A.A. Klug A. Phil. Trans. R Soc. London. 1987; 317: 537-561Crossref PubMed Scopus (163) Google Scholar; Turnell and Travers, 1992Turnell W.G. Travers A.A. Methods Enzymol. 1992; 212: 378-400Google Scholar) according to the helical periodicity imply that intrinsically and correctly curved sequences favor nucleosome formation relative to unstructured DNA. The left-handed direction of DNA curvature on nucleosomes predicts that linkage reduction is also a favoring factor. Both predictions have been verified (Germond et al., 1975Germond J.E. Hirt B. Oudet P. Gross Bellard M. Chambon P. Proc. Natl. Acad. Sci. U. S. A. 1975; 72: 1843-1847Crossref PubMed Scopus (490) Google Scholar; Hsieh and Griffith, 1988Hsieh C.H. Griffith J.D. Cell. 1988; 52: 535-544Abstract Full Text PDF PubMed Scopus (64) Google Scholar; Zivanovic et al., 1988Zivanovic Y. Goulet I. Revet B. Le Bret M. Prunell A. J. Mol. Biol. 1988; 200: 287-290Crossref Scopus (67) Google Scholar; Pennings et al., 1989Pennings S. Muyldermans G. Meersseman G. Wyns L. J. Mol. Biol. 1989; 207: 183-192Crossref PubMed Scopus (41) Google Scholar; Negri et al., 1989Negri R. Costanzo G. Venditti S. Di Mauro E. J. Mol. Biol. 1989; 207: 615-619Crossref PubMed Scopus (14) Google Scholar; Costanzo et al., 1990Costanzo G. Di Mauro E. Salina G. Negri R. J. Mol. Biol. 1990; 216: 363-374Crossref PubMed Scopus (49) Google Scholar; Buttinelli et al., 1993Buttinelli M. Di Mauro E. Negri R. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 9315-9319Crossref PubMed Scopus (56) Google Scholar; Patterton and Van Holt, 1993Patterton H.G. Van Holt K. J. Mol. Biol. 1993; 229: 637-655Crossref PubMed Scopus (11) Google Scholar). The fact that both topological properties (of closed DNA systems) and intrinsic curvature (in open DNA systems) affect nucleosomal localization points to the relevance of the rotational information of DNA (as defined in the Introduction) in this process. In agreement, it was observed that in the Crithidia fasciculata kinetoplast curved DNA (Costanzo et al., 1990Costanzo G. Di Mauro E. Salina G. Negri R. J. Mol. Biol. 1990; 216: 363-374Crossref PubMed Scopus (49) Google Scholar and in the Saccharomyces cerevisiae 5 S rRNA repeat gene multiple nucleosomes are formed in vivo and in vitro that occupy all of the available alternative positions differing by one helical turn constantly on the same rotational phase (Buttinelli et al., 1993Buttinelli M. Di Mauro E. Negri R. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 9315-9319Crossref PubMed Scopus (56) Google Scholar. Unique Positions—Multiple nucleosomes are not the rule and have never been shown to occur on RNA polymerase II promoters; most of the available in vivo data do actually point to the contrary. Among others, uniquely located nucleosomes have been reported over the promoters of yeast PH05 (Almer et al., 1986Almer A. Hinnen A. Hörz W. EMBO J. 1986; 5: 2689-2696Crossref PubMed Scopus (349) Google Scholar; Almer and Hörz, 1986Almer A. Hörz W. EMBO J. 1986; 5: 2681-2687Crossref PubMed Scopus (189) Google Scholar; Fascher et al., 1990Fascher K.D. Schmitz J. Hörz W. EMBO J. 1990; 9: 2523-2528Crossref PubMed Scopus (111) Google Scholar, of adenovirus-2 major late genes (Workman et al., 1990Workman J.L. Roeder R.G. Kingston R.E. EMBO J. 1990; 9: 1299-1308Crossref PubMed Scopus (98) Google Scholar, Workman et al., 1991Workman J.L. Taylor I.C.A. Kingston R.E. Cell. 1991; 64: 533-544Abstract Full Text PDF PubMed Scopus (148) Google Scholar, of the mouse mammary tumor virus (Cordingley et al., 1987Cordingley M.G. Riegel A.T. Hager G.L. Cell. 1987; 48: 261-270Abstract Full Text PDF PubMed Scopus (313) Google Scholar; Perlmann and Wrange, 1988Perlmann T. Wrange O. EMBO J. 1988; 7: 3073-3079Crossref PubMed Scopus (235) Google Scholar; Piña et al., 1990aPiña B. Barettino D. Truss M. Beato M. J. Mol. Biol. 1990; 216: 975-990Crossref PubMed Scopus (58) Google Scholar, Piña et al., 1990bPiña B. Brüggemeier V. Beato M. Cell. 1990; 60: 719-731Abstract Full Text PDF PubMed Scopus (335) Google Scholar, of the rat tyrosine aminotransferase gene (Carr and Richard-Foy, 1990Carr K.D. Richard-Foy H. Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 9300-9304Crossref PubMed Scopus (59) Google Scholar), on globin genes (Buckle et al., 1991Buckle R. Balmer M. Yenidunya A. Allan J. Nucleic Acids Res. 1991; 19: 1219-1226Crossref PubMed Scopus (12) Google Scholar; Yenidunya et al., 1994Yenidunya A. Davey C. Clark D. Felsenfeld G. Allan J. J. Mol. Biol. 1994; 237: 401-414Crossref PubMed Scopus (28) Google Scholar, on centromere regions (Bloom et al., 1984Bloom K.S. Amaya E. Carbon J. Clarke C. Hill A. Yeh E. J. Cell Biol. 1984; 99: 1559-1568Crossref PubMed Scopus (51) Google Scholar; Gottschling and Cech, 1984), and on the ARS yeast elements (Simpson, 1991Simpson R.T. Prog. Nucleic Acids Res. Mol. Biol. 1991; 40: 143-184Crossref PubMed Scopus (202) Google Scholar); for reviews, see, among others, van Holde, 1988van Holde K.E. Chromatin.1st Ed. Springer Verlag, Heidelberg, Germany1988Google Scholar and Thoma, 1992Thoma F. Biochim. Biophys. Acta. 1992; 1130: 1-19Crossref PubMed Scopus (111) Google Scholar. Limiting Multiple Positions—The fact that the structural code for rotational positioning is redundant (Travers, 1993Travers A.A. DNA-Protein Interactions.1st Ed. Chapman & Hall, London1993: 1-27Crossref Google Scholar) implies that multiple positioning is intrinsically possible, underscoring a substantial freedom of choice by histone octamers when nucleating on a defined DNA area and in selecting one of several permitted positions. Therefore, obtaining unique nucleosome positioning requires additional elements. These have been reviewed extensively (Thoma, 1992Thoma F. Biochim. Biophys. Acta. 1992; 1130: 1-19Crossref PubMed Scopus (111) Google Scholar) and include defined DNA sequences acting as boundaries, dynamic processes, and asymmetries in the recognition sequences (as specifically discussed in terms of translational information in Travers and Klug, 1987Travers A.A. Klug A. Phil. Trans. R Soc. London. 1987; 317: 537-561Crossref PubMed Scopus (163) Google Scholar, Satchwell and Travers, 1989Satchwell S.C. Travers A.A. EMBO J. 1989; 8: 229-238Crossref PubMed Scopus (74) Google Scholar, and Travers, 1993Travers A.A. DNA-Protein Interactions.1st Ed. Chapman & Hall, London1993: 1-27Crossref Google Scholar). The nature and the role of translational information of DNA in causing unique nucleosomal positioning are difficult to define. It was proposed (Negri et al., 1994Negri R. Costanzo G. Buttinelli M. Venditti S. Di Mauro E. Biophys. Chem. 1994; 50: 1-13Crossref PubMed Scopus (11) Google Scholar that the translational message consists mostly in the equilibrium among the rotational signals present on each side of the center of symmetry of the histone octamer, the frequency of formation of a nucleosome on a defined DNA tract being determined by the sum of flexibility potentialities relative to a dyad central position. In other words, the elusive (Thoma, 1992Thoma F. Biochim. Biophys. Acta. 1992; 1130: 1-19Crossref PubMed Scopus (111) Google Scholar) nature of the translational messages is explained by the fact that these could be a quantitative aspect of the rotational information. Evidence has been provided for a boundary role of systemspecific proteins (Fedor et al., 1988Fedor M.J. Lue N.F. Kornberg R.D. J. Mol. Biol. 1988; 204: 109-127Crossref PubMed Scopus (184) Google Scholar; Simpson, 1991Simpson R.T. Prog. Nucleic Acids Res. Mol. Biol. 1991; 40: 143-184Crossref PubMed Scopus (202) Google Scholar; McPherson et al., 1993McPherson C.E. Shim E.Y. Friedman D.S. Zaret K.S. Cell. 1993; 75: 387-398Abstract Full Text PDF PubMed Scopus (284) Google Scholar. In particular, clear examples of proteins acting in the localization process have been obtained in yeast, where it was observed that α2 repressor positions nucleosomes in TPR1/ARS1 chromatin (Roth et al., 1990Roth S.Y. Dean A. Simpson R.T. Mol. Cell. Biol. 1990; 10: 2247-2260Crossref PubMed Google Scholar. In this same system, it was also shown that ABF1 contributes in reducing the number of possible alternative positions for the nucleosome located upstream of the α-domain, from the several positions observed in vitro to the largely dominant positions occupied in vivo on the chromosome (Venditti et al., 1994Venditti P. Costanzo G. Negri R. Camilloni G. Biochim. Biophys. Acta. 1994; 1219: 677-689Crossref PubMed Scopus (54) Google Scholar. The MOX Promoter—On this sequence, nucleosomes distribute themselves in families with limited multiplicity; their localization is not completely continuous (as observed in the C. fasciculata kinetoplast or in the S. cerevisiae 5 S gene systems mentioned above) and is not made of a series of unique positions. The average number of the components of each family is four to five, with the exception of nucleosome –3, whose family is composed by a major species (-3·2) and a quite minor one (-3·1). The unicity of this occupancy is predicted by the boundary model (Thoma, 1992Thoma F. Biochim. Biophys. Acta. 1992; 1130: 1-19Crossref PubMed Scopus (111) Google Scholar). In this case the putative boundary is the UAS1 with its binding protein(s) (Gödecke et al., 1994Gödecke S. Eckart M. Janowics Z.A. Hollenberg C.P. Gene. 1994; 139: 35-42Crossref PubMed Scopus (37) Google Scholar. Thus, in the H. polymorpha MOX promoter, each "linker region" appears to be engaged in nucleosomal complexes. The different organization of this area relative to the central part of the nucleosomal body consists in the frequency of its occupancy. The DNA tract that corresponds to the central part of a nucleosome is always occupied, whereas the internucleosomal tracts interact in a more complex way: a few DNA helical periods may interact with the terminal or paraterminal part of the upstream histone octamer or with the terminal or paraterminal extremity of the downstream one. Protection/exposure of DNA in these tracts is more variable. The increased sensitivity to DNase-I of the borders of DNA-protein complexes (FIG. 4, FIG. 5) is a well known phenomenon. It is noteworthy that the binding of the Adr1-like protein factor on nucleosome –2, as revealed by the DNase-I footprint (FIG. 4, FIG. 5), is not prevented by the multiple localizations of this particle. This indicates that the organization of nucleosomes in families allows the correct exposure of the relevant DNA sequences as long as they show a unique rotational setting. In conclusion, the analysis of the multiple nucleosome positions in the MOX promoter and of their distribution show that the rotational information appears to be the major determinant of nucleosome positioning. Interference with the rotational information, as described in Negri et al., 1994Negri R. Costanzo G. Buttinelli M. Venditti S. Di Mauro E. Biophys. Chem. 1994; 50: 1-13Crossref PubMed Scopus (11) Google Scholar, does actually alter nucleosome distributions in vivo and should provide an experimental basis to study nucleosome dynamics (a process that was shown to be an essential component of the topology of transcription in eukaryotes) (Wolffe et al., 1986Wolffe A.P. Jordan E. Brown D.D. Cell. 1986; 44: 381-389Abstract Full Text PDF PubMed Scopus (105) Google Scholar; Lorch et al., 1987Lorch Y. La Pointe J.W. Rornberg R.D. Cell. 1987; 49: 203-210Abstract Full Text PDF PubMed Scopus (387) Google Scholar; Losa and Brown, 1987Losa R. Brown D.D. Cell. 1987; 50: 801-808Abstract Full Text PDF PubMed Scopus (92) Google Scholar; Lee and Garrard, 1991Lee M.S. Garrard W.T. EMBO J. 1991; 10: 607-615Crossref PubMed Scopus (101) Google Scholar; van Holde et al., 1992van Holde K.E. Lohr D.E. Robert C. J. Biol. Chem. 1992; 267: 2837-2840Abstract Full Text PDF PubMed Google Scholar; Clark and Felsenfeld, 1992Clark D.J. Felsenfeld G. Cell. 1992; 71: 11-22Abstract Full Text PDF PubMed Scopus (122) Google Scholar; Meersseman et al., 1992Meersseman G. Pennings S. Bradbury E.M. EMBO J. 1992; 11: 2951-2959Crossref PubMed Scopus (262) Google Scholar; O'Neill et al., 1993O'Neill T.E. Smith J.G. Bradbury E.M. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 6203-6207Crossref PubMed Scopus (30) Google Scholar; Adams and Workman, 1993Adams C.C. Workman J.L. Cell. 1993; 72: 305-308Abstract Full Text PDF PubMed Scopus (139) Google Scholar; Studitsky et al., 1994Studitsky V.M. Clark D.J. Felsenfeld G. Cell. 1994; 76: 371-382Abstract Full Text PDF PubMed Scopus (206) Google Scholar; O'Donohue et al., 1994O'Donohue M.F. Duband-Goulet I. Hamiche A. Prunell A. Nucleic Acids Res. 1994; 22: 937-945Crossref PubMed Scopus (42) Google Scholar and to understand the regulatory properties of nucleosomes.