Chromatin and Gene Expression

Chromatin is the physiological form of eukaryotic DNA: a repeating complex of DNA wound around histone proteins, folded further into higher-order fibers, and organized within the nucleus into domains of differing accessibility. The basic repeating unit is the nucleosome core particle, in which 147 base pairs of DNA wrap 1.65 times in a left-handed superhelix around a disc-shaped histone octamer — two copies each of the four core histones H2A, H2B, H3, and H4. Nucleosome core particles are joined by stretches of linker DNA, typically 20-90 bp depending on cell type and species, giving nucleosome repeat lengths of roughly 165-240 bp. A fifth histone, linker histone H1, binds at the nucleosome entry/exit point and the linker DNA, helping to compact the fiber further. On electron micrographs of chromatin spread under low salt, this arrangement gives the famous "beads on a string" appearance — each bead a nucleosome, each string a stretch of linker DNA. Because the nucleosome bends DNA sharply and occludes roughly two turns of the double helix from free access, it is intrinsically repressive to any process that requires DNA-protein contact from the outside — transcription factor binding, RNA polymerase transit, replication origin firing, and repair enzyme access all must contend with it. Cells classify chromatin descriptively into euchromatin, which is less condensed, more accessible, and enriched for actively transcribed or transcribable genes, and heterochromatin, which is more condensed and largely transcriptionally silent. Heterochromatin is further divided into constitutive heterochromatin (permanently silent regions such as pericentromeric and telomeric repeats) and facultative heterochromatin (regions silenced in some cell types or developmental states but not others, such as the inactive X chromosome). This distinction, first made cytologically by Emil Heitz in the 1920s from differential staining of Drosophila chromosomes, remains the organizing vocabulary of chromatin biology. The crucial conceptual move, which took decades of genetics and biochemistry to establish, is that chromatin structure is not a fixed archival state but a regulated substrate. Cells possess dedicated enzyme complexes that reposition nucleosomes, evict them, exchange standard histones for variant histones, and covalently modify histone tails — and these activities are targeted to specific genes by the same sequence-specific transcription factors that "recruit" general transcription machinery. Chromatin regulation and transcription factor regulation are therefore not two separate systems but one integrated circuit, in which opening chromatin is frequently the rate-limiting first step of gene activation.

Step 1 — Nucleosome assembly and structure. The histone octamer core assembles as a (H3-H4)₂ tetramer flanked by two H2A-H2B dimers. DNA wraps around this core making roughly 14 contacts between the phosphate backbone and histone residues, mostly through hydrogen bonds and electrostatic interactions with the basic (arginine- and lysine-rich) histone surface — not sequence-specific base readout, which is why nucleosomes can form on almost any DNA sequence, though some sequences (curved, AT/GC periodic every ~10 bp) favor wrapping more strongly and create intrinsically preferred nucleosome positions. The N-terminal tails of all four core histones, roughly 20-35 residues each, protrude from the nucleosome surface between the DNA gyres and are the principal sites of post-translational modification (foreshadowing Histone Modifications and the Histone Code).

Step 2 — Higher-order folding. Strings of nucleosomes fold further into more compact fibers. The classical "30 nm fiber" model (solenoid or zigzag arrangements stabilized by linker histone H1) has been the textbook picture for decades, but modern imaging (electron microscopy of cryo-fixed nuclei, and super-resolution/ChromEM tomography) shows that in vivo chromatin more often exists as an irregularly folded, polymer-melt-like arrangement of nucleosome clutches rather than a uniform 30 nm helix — a point of active revision that should be flagged rather than glossed over. At the largest scale, chromosomes occupy discrete territories within the nucleus and fold into topologically associating domains (TADs), self-interacting neighborhoods bounded by architectural proteins (CTCF and cohesin), within which genes and their regulatory elements preferentially contact one another.

Step 3 — The barrier problem. A nucleosome positioned directly over a promoter's TATA box or a transcription factor's binding site sterically blocks that factor from binding — this has been measured directly, in vitro, as orders-of-magnitude reductions in binding affinity when a site is nucleosomal versus naked DNA. RNA polymerase II itself can transcribe through nucleosomes, but only slowly and with transient pausing, and efficient elongation in vivo depends on nucleosomes being destabilized ahead of the polymerase and reassembled behind it (via histone chaperones such as FACT, which transiently displaces an H2A-H2B dimer to let the polymerase pass, then helps redeposit it).

Step 4 — ATP-dependent chromatin remodeling. To overcome the nucleosome barrier at promoters, cells deploy four conserved families of ATP-dependent chromatin remodeling complexes, each built around a Snf2-family ATPase motor:

These enzymes use the energy of ATP hydrolysis to translocate DNA relative to the histone octamer, creating a transient DNA loop that propagates around the nucleosome and results in the octamer sliding to a new position, being pushed off the DNA entirely (eviction), or having one of its H2A-H2B dimers exchanged for a variant. Because this activity is targeted — remodelers are recruited by direct contact with sequence-specific activators, or by pre-existing histone marks read through bromodomains and chromodomains — the result is gene-specific, regulated accessibility rather than global unwrapping.

Step 5 — Reading the state: nucleosome positioning at a real promoter. At a typical inducible gene, the promoter region in the "off" state carries a positioned nucleosome directly over the transcription start site or over key activator sites, while flanking regions carry well-phased arrays. Upon induction, sequence-specific activators bind first to sites that remain exposed (often in linker DNA or on the nucleosome surface facing outward), then recruit SWI/SNF-type remodelers and histone-modifying enzymes, which together clear or reposition the promoter nucleosome — exposing the TATA box and initiator element for TFIID and Pol II loading. This "pioneer factor first, remodeler and general machinery second" logic is the standard model for how a repressed gene becomes transcribable, and it directly extends the Eukaryotic Transcriptional Regulation framework in which activators recruit coactivators: here, one class of coactivator recruited is the chromatin remodeling machinery itself.

Step 6 — Consequences for accessibility genome-wide. Because nucleosome occupancy is so consequential, cells and researchers alike use it as a readout of regulatory state. Regions of low nucleosome occupancy — nucleosome-depleted regions (NDRs) — flank active promoters and enhancers and are directly measured by nuclease sensitivity or by DNase/ATAC-seq (chromatin accessibility assays), giving genome-wide maps of "open" versus "closed" chromatin that correlate tightly with active transcription.

Establishing the nucleosome itself. In the early 1970s, chromatin's basic architecture was unknown; DNA was known to associate with histones, but not in what repeating structure. Two lines of evidence converged rapidly. Roger Kornberg (1974), synthesizing X-ray diffraction data on chromatin's repeating pattern with the known stoichiometry of core histones (roughly equal to DNA by mass, in defined pairs) and cross-linking data showing histones form defined oligomers, proposed the nucleosome model: DNA organized into repeating particles, each containing two copies of each core histone. Independently, Dean Hewish and Leigh Burgoyne (1973) had shown that a nuclease endogenous to nuclei cleaved chromatin's DNA into fragments that were multiples of a fixed unit size — approximately 200 base pairs — rather than randomly, implying a periodic protection pattern. Michael Noll (1974) then applied purified micrococcal nuclease (MNase) to chromatin in a titration series and, running the digestion products on gel electrophoresis, saw the now-iconic "nucleosome ladder": bands at ~200 bp, 400 bp, 600 bp, and so on. Partial digestion protects a "core" fragment (later refined to 147 bp) while linker DNA between particles is nuclease-accessible and cut first — direct biochemical proof that DNA in chromatin is organized into discrete, evenly spaced, protein-protected units, exactly as Kornberg's model predicted. This MNase-ladder assay remains the standard method for measuring nucleosome positioning and spacing today, and its higher-resolution descendant (MNase-seq) is used to map nucleosome occupancy genome-wide. The atomic structure was resolved definitively by Timothy Richmond and colleagues (1997), who solved the crystal structure of the nucleosome core particle at 2.8 Å resolution, visualizing the histone octamer, the precise path of the DNA superhelix, and the histone tails emerging between DNA gyres — confirming and extending every element of the Kornberg model at atomic detail.

Establishing that chromatin is a regulated barrier, not passive packaging. The decisive evidence came from yeast genetics in the 1980s. Studies of the yeast mating-type switching and sucrose fermentation pathways identified genes required for transcriptional activation of a subset of genes that were not general transcription factors themselves. Screens for mutants defective in mating-type SWitching (SWI genes, from the Winston/Herskowitz labs) and defective in Sucrose Non-Fermenting growth (SNF genes, from the Carlson lab, notably SNF2) converged on an overlapping set of genes. Genetic and biochemical work through the late 1980s and into 1992-1994 (Peterson and Herskowitz; Cairns, Kornberg and colleagues; Winston and colleagues) showed that these SWI/SNF gene products assemble into a large multi-subunit complex, and — critically — that this complex could alter nucleosome structure in an ATP-dependent manner in vitro, making previously nucleosome-occluded DNA sites accessible to restriction enzymes and transcription factors. This was the direct proof that a dedicated, ATP-powered machine exists whose job is to reposition nucleosomes to permit transcription factor access — establishing chromatin remodeling as a real, biochemically defined regulatory mechanism rather than an inference. Parallel work identified histone mutations and deletions (notably H4 tail truncations, and mutations in the histone chaperone/regulator genes) that altered transcription of specific genes without altering the coding sequence of any transcription factor, reinforcing that the histones and their packaging are themselves regulatory objects.

Establishing nucleosomes as occlusive in vitro. Direct biochemical demonstrations (e.g., studies by Workman and Kingston in the late 1980s-1990s) reconstituted promoters on nucleosomal templates in vitro and showed that transcription factor binding and transcription initiation were strongly repressed compared to naked DNA templates of identical sequence, and that this repression could be relieved by adding purified remodeling complexes or by pre-incubating with activators before nucleosome assembly ("assisted loading"). These experiments isolated the nucleosome itself, independent of any other cellular feature, as sufficient to block gene expression — completing the causal chain from nucleosome structure to transcriptional consequence.

Example 1 — Reading an MNase digestion ladder to infer nucleosome spacing and promoter state. A researcher digests chromatin from two cell types with a fixed concentration of MNase and runs the fragments on a gel. In cell type A (differentiated, gene X silent), the ladder shows sharp bands at 195, 390, 585, and 780 bp — an even, regular repeat of ~195 bp. In cell type B (progenitor, gene X active), the ladder shows the same regular array over the gene body, but the region 300 bp upstream of the transcription start site shows no distinct band — instead a smear, with a sharp additional cut site appearing exactly at the position expected for the promoter. Interpretation: the regular 195 bp ladder over the gene body in both cell types indicates a canonical nucleosome repeat length (147 bp core + ~48 bp linker) that is a general packaging feature independent of expression state. But the loss of a discrete nucleosome-sized protection band specifically at the promoter in cell type B, replaced by a site fully accessible to nuclease, indicates a nucleosome-depleted region (NDR) has formed there — consistent with a remodeling complex having evicted or displaced the promoter nucleosome to permit transcription factor and Pol II access in the active cell type. This is exactly the logic used in real genome-wide MNase-seq/ATAC-seq experiments: NDR width and position, not average nucleosome density, are the diagnostic feature of an active promoter.

Example 2 — Quantifying compaction: does the nucleosome explain the packing problem? Human diploid nuclear DNA totals about 6.4 billion base pairs (2 × 3.2 Gb per haploid genome). At 0.34 nm rise per base pair for B-form DNA, the total contour length is: 6.4 × 10⁹ bp × 0.34 nm/bp ≈ 2.18 × 10⁹ nm ≈ 2.18 m. A typical human nucleus is about 6 µm in diameter, so the DNA must be compacted from a straight-line length of ~2.18 m into a spherical volume of diameter 6 × 10⁻⁶ m — a linear compaction ratio of roughly: 2.18 m ÷ 6 × 10⁻⁶ m ≈ 360,000-fold. Nucleosome wrapping alone (147 bp of ~50 nm of extended contour length compacted into a particle roughly 11 nm in diameter and 5.5 nm tall) accounts for about a 6- to 7-fold linear compaction of the DNA into the "beads on a string" fiber. The remaining ~360,000/7 ≈ 50,000-fold compaction required to fit into the nucleus is achieved by higher-order folding (chromatin fiber formation, loop domains, TADs, and mitotic-level condensation during division) — illustrating quantitatively why the nucleosome, while essential, is only the first of several nested levels of packaging, and why higher-order structure (not just nucleosome wrapping) is itself a further site of regulation.

Roger Kornberg trained as a chemist and worked in the early 1970s at the MRC Laboratory of Molecular Biology in Cambridge, then in Aaron Klug's group, at a moment when chromatin structure was essentially an open problem — histones were known, DNA was known, but how the two combined into a physiological, repeating structure was not. Kornberg's insight was integrative rather than purely experimental: he combined X-ray diffraction patterns showing a repeating ~100 Å periodicity in chromatin fibers, the known near-equal mass ratio of histone to DNA, cross-linking data (from Klug's lab and others) showing histones form defined tetramer and octamer complexes, and micrococcal nuclease digestion patterns, into a single coherent model — the nucleosome — published in 1974. The model was audacious for its specificity: it predicted an exact histone stoichiometry (two each of H2A, H2B, H3, H4), an exact approximate DNA length per particle, and a specific octameric architecture, all before any atomic structure existed to confirm it. Every one of these predictions was subsequently vindicated, culminating in the 1997 crystal structure from Timothy Richmond's laboratory. Kornberg later moved into RNA polymerase II structural biology, for which he received the Nobel Prize in Chemistry in 2006 — but the nucleosome model remains, independently, one of the foundational structural discoveries of molecular biology.

Chromatin accessibility is now a direct clinical and biotechnological readout. ATAC-seq (Assay for Transposase-Accessible Chromatin using sequencing), developed in the 2010s, maps genome-wide open chromatin in a single afternoon from as few as thousands of cells, and is used routinely to characterize cell states in cancer biopsies, immune cell subsets, and developmental atlases — essentially an industrialized, sequencing-based descendant of the original MNase ladder experiment. In oncology, mutations in chromatin remodeling complex subunits are recurrent cancer drivers: the SWI/SNF complex subunit SMARCB1 (INI1) is inactivated in nearly all malignant rhabdoid tumors, and ARID1A, another SWI/SNF subunit, is among the most frequently mutated genes across human cancers overall, disrupting normal chromatin accessibility control at tumor suppressor and oncogene loci. Drugs targeting chromatin-modifying enzymes (histone deacetylase inhibitors such as vorinostat, and EZH2 inhibitors such as tazemetostat) are approved cancer therapies that work by shifting chromatin state toward or away from repression at specific loci. In regenerative medicine, reprogramming somatic cells into induced pluripotent stem cells requires overcoming heterochromatin barriers at pluripotency genes, and pioneer transcription factors among the Yamanaka factors (notably Oct4 and Sox2) function precisely because they can engage nucleosome-occluded DNA directly, then recruit remodelers to open the locus. In agriculture, chromatin accessibility profiling is used to identify regulatory elements controlling crop traits (drought tolerance, yield genes) for precision breeding and CRISPR-based editing of regulatory rather than coding sequence. And in basic developmental biology, the fact that chromatin state can be inherited through cell division — a nucleosome-level memory of "on" or "off" — underlies the maintenance of cell identity throughout an organism's life, the conceptual bridge into Epigenetic Inheritance.

Ask the tutor: How do histone chaperones like FACT let RNA polymerase II pass through a nucleosome without permanently destroying it?