Intrinsic histone-DNA interactions are not the major determinant of nucleosome positions in vivo

Abstract

We assess the role of intrinsic histone-DNA interactions by mapping nucleosomes assembled in vitro on genomic DNA. Nucleosomes strongly prefer yeast DNA over Escherichia coli DNA, indicating that the yeast genome evolved to favor nucleosome formation. Many yeast promoter and terminator regions intrinsically disfavor nucleosome formation, and nucleosomes assembled in vitro show strong rotational positioning. Nucleosome arrays generated by the ACF assembly factor have fewer nucleosome-free regions, reduced rotational positioning and less translational positioning than obtained by intrinsic histone-DNA interactions. Notably, nucleosomes assembled in vitro have only a limited preference for specific translational positions and do not show the pattern observed in vivo. Our results argue against a genomic code for nucleosome positioning, and they suggest that the nucleosomal pattern in coding regions arises primarily from statistical positioning from a barrier near the promoter that involves some aspect of transcriptional initiation by RNA polymerase II.

Figure 1. Micrococcal nuclease digestion analysis of chromatin and generation of mononucleosomes

Chromatin assembled using ACF (a) or salt dialysis (b) was partially digested with micrococcal nuclease. (c) Mononucleosomes were generated by extensive digestion of chromatin with micrococcal nuclease. DNA size markers (M), 123-bp ladder (Invitrogen). (d) Number of sequence tags for yeast and E. coli from the indicated samples.

Figure 2. Nucleosome density profiles

(a-c) Nucleosome density profiles around transcriptional initiation sites (TSS) of 1752 genes with isolated promoters (defined as having 1 kb upstream regions that do not overlap with other genes) for chromatin assembled by salt dialysis, ACF, or in vivo in YPD medium. Each tag was extended to 146 bp, piled up, and then normalized by average sequencing coverage. Nucleosome density for each gene is represented by a horizontal line, and genes are ranked from top to bottom by expression level. (d-f) Nucleosome density profiles of the indicated chromatin preparations around transcriptional termination sites (TTS) of 1548 genes (ranked by expression level) with isolated terminator regions (defined as having 1 kb downstream regions that do not overlap with other genes. (g-h) Nucleosome density profiles in vivo around the binding sites (ranked by genomic location) of transcription factors Abf1 and Reb1.

Figure 3. Rotational positioning

(a) Fraction of AA/TT/TA dinucleotides of yeast nucleosomes aligned by their 5′ ends from the indicated samples. (b) Power spectrum analysis of di-nucleotide periodicity. Discrete Fourier transform was performed on signals in (+11 to +160) interval, and the signals were first subtracted the mean value to remove the DC component. (c) Fraction of AA/TT/TA dinucleotides of E. coli nucleosomes aligned by their 5′ ends. (d) Power spectrum of AA/TT/AT pattern for yeast and E. coli genomic DNA. Each genome was split into 1,024 nt fragment, discrete Fourier transform was performed on each fragment, and the results presented as the average of all fragments.

Figure 4. Intrinsic histone-DNA interactions do not account for the nucleosome positioning pattern in vivo

(a) Degree of translational positioning for each nucleotide position is determined by the number of nucleosome centers within a 20 bp window divided by the number of nucleosome centers within the corresponding 160 bp window. Then, for the indicated samples, the percentage of regions (Y-axis) above the maximal positioning degree within a 160 bp window (X-axis; a value of 1 indicates a perfectly positioned nucleosome) is plotted. (b-d) Relationship between nucleosomal positions in vivo and in vitro. Predefined 20 bp windows for the centers of +1 through +10 nucleosomes were defined on a gene-by-gene basis in chromatin from cells grown in YP-ethanol. For the indicated samples (in vivo represents cells grown in YPD and serves as a control for experimental variation), the percent of nucleosomal loci (Y-axis) is plotted as a function of the number of tags (defined by the central position of each nucleosome) within the predefined windows (different colors indicate the +1 through +10 nucleosomes). (e) Tag position relationship. Start-to-start distances of tags in the same strand are shown, with the right plots being a blow-up of the boxed areas in the left plots. (f) The distribution of distance between +1 nucleosome center location and transcriptional start site (TSS), binned by 10-bp.