Major Determinants of Nucleosome Positioning

Abstract

The compact structure of the nucleosome limits DNA accessibility and inhibits the binding of most sequence-specific proteins. Nucleosomes are not randomly located on the DNA but positioned with respect to the DNA sequence, suggesting models in which critical binding sites are either exposed in the linker, resulting in activation, or buried inside a nucleosome, resulting in repression. The mechanisms determining nucleosome positioning are therefore of paramount importance for understanding gene regulation and other events that occur in chromatin, such as transcription, replication, and repair. Here, we review our current understanding of the major determinants of nucleosome positioning: DNA sequence, nonhistone DNA-binding proteins, chromatin-remodeling enzymes, and transcription. We outline the major challenges for the future: elucidating the precise mechanisms of chromatin opening and promoter activation, identifying the complexes that occupy promoters, and understanding the multiscale problem of chromatin fiber organization.

Figure 1

Level of digestion influences the set of mononucleosomal fragments obtained in MNase-seq experiments. Digestion of chromatin by MNase produces DNA fragments of various lengths, which can be separated by gel electrophoresis. An example of a gel shows the abundance of the bands (mononucleosomes (1n), dinucleosomes (2n), trinucleosomes (3n), and so forth) obtained by increasing levels of digestion by MNase. More extensive digestion trims the linker DNA but also invades the core particles. Marker: 50 bp DNA ladder. To see this figure in color, go online.

Figure 2

Nucleosome occupancy versus nucleosome dyad distribution. (A) Different cells contain nucleosomes at different positions. Each nucleosome footprint (blue oval) covers 147 bp of DNA, whose center corresponds to a nucleosome dyad (red diamond). (B) By stacking all nucleosome sequences and dividing by the number of cells, we obtain the nucleosome occupancy, i.e., the probability of each basepair to be covered by any nucleosome. The probability that any nucleosome dyad is located at a given basepair represents the nucleosome dyad distribution, and this is computed by stacking the single-basepair footprints of the nucleosome centers and dividing by the number of cells. Although we always compute the above probabilities by analyzing the ensemble of nucleosome configurations within a population of cells (i.e., computing ensemble averages), assuming that the system is ergodic, we can also think of these probabilities as time averages in a single cell. For example, when we map the nucleosomes (i.e., at a specific time), we may detect that a given basepair was covered by a nucleosome in 80% of the cells so that the nucleosome occupancy of that position is 0.80. Assuming the cell nucleus is an ergodic system, we may also say that if we follow the dynamics of a given cell throughout time, the same position will be covered by a nucleosome 80% of the time. To see this figure in color, go online.

Figure 3

Nucleosome organization near the gene ends in S. cerevisiae. (A) Average nucleosome occupancy (top) and nucleosome dyad distributions (bottom) at the TSS (left) and TTS (right) are shown. Profiles obtained for three levels of MNase digestion are shown. (B) A heat map shows the nucleosome dyad distribution near all TTSs, sorted according to the distance to the promoter (NDR; thick blue stripe) of the downstream gene. Only about one-third of the TTSs are located in nucleosome-depleted regions. These NDRs correspond to the promoters of the downstream genes in the tandem gene pairs. Inset: a scheme of the gene alignments is shown. All yeast genes (red arrows) are aligned at their TTSs and sorted according to the distance to the promoter (NDR) of the downstream gene (gray arrows). The aligned NDRs corresponding to the downstream genes (gray arrows) generate a blue stripe in the heat map, which is flanked by phased nucleosome arrays on both sides. The promoters corresponding to the genes aligned at their TTSs (the first in each pair of genes; red arrows) are not in phase with each other (because of different gene lengths) and do not generate a blue stripe in the heat map. (C) The heat maps show the dyad distributions near both gene ends: (left) 5′ ends are aligned at the +1 nucleosome and are sorted according to the distance to the promoter (NDR) of the upstream gene, and (right) 3′ ends are aligned at the TTS and are sorted according to the distance to the promoter (NDR) of the downstream gene. Genes are split into two groups according to the relative orientation of the neighboring gene: divergent/tandem and convergent/tandem, respectively. Insets: schemes of the alignments are shown; all yeast genes (red arrows) are aligned either at their 5′ (left panel) or 3′ ends (right panel) and sorted according to the distance to the NDR of the upstream and downstream genes (gray arrows), respectively. When the 5′ ends and the NDRs of the genes are aligned, a blue stripe appears in the heat map, indicating the locations of the NDRs. Annotations for +1 nucleosome position, NDR center, TSS, and TTS were obtained from (23), and MNase-seq data were obtained from (7).

Figure 4

Multiple determinants of nucleosome positioning. (A) An Integrative Genomics Viewer browser snapshot shows the nucleosome distribution on yeast chromosome IV (550,400–558,160). Nonhistone barriers (e.g., GRFs Reb1, Rap1, and Abf1) bind to DNA (light blue rectangles) and compete with histones, resulting in nucleosome-depleted regions. Weakly transcribed genes (light pink rectangles) are characterized by well-positioned nucleosomes forming regular arrays, whereas highly transcribed genes (red rectangles) are characterized by poorly positioned nucleosomes, i.e., increased cell-to-cell variability. Pol II occupancy (Rpb3 subunit) shows the relative transcription level. MNase-seq data and ChIP-seq data (Reb1, Rap1, Abf1, and Rpb3) are shown in individual Integrative Genomics Viewer tracks. For illustrative purposes, to distinguish linkers between individual nucleosomes, MNase-seq reads were trimmed symmetrically to 101 bp. ChIP-seq data show Reb1 (114), Rap1 (115), Abf1 (116), and Rpb3 (16). MNase-seq data from (7) are shown. (B) DNA sequence alone is not able to properly position nucleosomes at the gene promoters. The left heat map shows nucleosome dyads reconstituted in vitro by salt gradient dialysis (data from (52)) and the lack of nucleosome-depleted regions and nucleosome phasing at promoters. The other four heat maps show the distribution of nucleosome dyads in vivo in different remodeler mutants (data from (27)). A triple mutant in which three nucleosome-spacing enzymes have been deleted (chd1Δ isw1Δ isw2Δ) and a double mutant in which the two most important spacing enzymes are deleted (chd1Δ isw1Δ) both show vastly disrupted nucleosome phasing compared to that of wild-type cells (rightmost heat map). CHD1 is active in the single mutant isw1Δ, which has regular nucleosome arrays, shorter spacing, and weaker phasing compared to that of wild-type cells (27). All heat maps contain yeast promoters aligned at the NDR and sorted according to NDR width from wild-type cells. (C) The most important determinants of nucleosome organization include the following. 1) Stable barrier complexes prevent nucleosome formation to create NDRs at multiple regions along the genome. Promoters may be occupied by transcription-preinitiation complexes, transcription factors (TFs), and/or chromatin remodelers. tRNA genes are occupied by a stable preinitiation complex containing Pol III TFs TFIIIB and TFIIIC; replication origins are occupied by the origin recognition complex (ORC). 2) In combination with barrier complexes, chromatin remodelers (spacing enzymes) form regular nucleosome arrays between NDRs. 3) Pol II disrupts nucleosome arrays from gene bodies, whereas remodelers restore the regular organization. 4) The DNA sequence affects nucleosome positions either directly through its DNA-bending properties (e.g., periodic WW/SS dinucleotide distributions favor specific rotational positions, whereas poly(dA:dT) tracts disfavor nucleosome formation) or indirectly through the binding of other proteins that prevent nucleosome formation (e.g., the motif shown in the promoter is recognized by the Gcn4 activator, which activates transcription from this locus).