Article ID: pjab.101.020
The organization and dynamics of chromatin are critical for genome functions such as transcription and DNA replication/repair. Historically, chromatin was assumed to fold into the 30-nm fiber and progressively arrange into larger helical structures, as described in the textbook model. However, over the past 15 years, extensive evidence including our studies has dramatically transformed the view of chromatin from a static, regular structure to one that is more variable and dynamic. In higher eukaryotic cells, chromatin forms condensed yet liquid-like domains, which appear to be the basic unit of chromatin structure, replacing the 30-nm fiber. These domains maintain proper accessibility, ensuring the regulation of DNA reaction processes. During mitosis, these domains assemble to form more gel-like mitotic chromosomes, which are further constrained by condensins and other factors. Based on the available evidence, I discuss the physical properties of chromatin in live cells, emphasizing its viscoelastic nature—balancing local fluidity with global stability to support genome functions.
The human body consists of approximately 40 trillion cells.1) Each cell houses 2 meters of deoxyribonucleic acid (DNA), the blue print of life. In the 19th century, W. Flemming described a nuclear substance that became visible after staining with a basic dye using primitive light microscopes, and named it “chromatin”,2) which later turned out to be a complex of DNA, protein, RNA, and other factors. In 1928, Emil Heitz reported a differential staining of chromosomal regions and identified euchromatin and heterochromatin using improved cytological staining techniques.3) Even before the DNA structure was discovered,4) chromatin had already garnered significant interest among biologists. In this review, I explore how chromatin is structured/organized in live cells, emphasizing the physical properties of chromatin and their functional relevance.
DNA is a negatively charged polymer that produces electrostatic repulsion between adjacent regions (Fig. 1A).4) To fold into a small space like the cell,5),6) the long, negatively charged polymer is wrapped around a basic core histone octamer to form a nucleosome (Fig. 1B).7)–9) The detailed nucleosome structure was elucidated: 147-base pairs (bp) of DNA are wrapped around the histone octamer composed of H2A, H2B, H3, and H4 histones, whose surface is positively charged (right, Fig. 1B).10),11) Each nucleosome particle is connected by linker DNA (∼50 bp) to form repetitive motifs of ∼200 bp. This string of nucleosomes was described initially as the “beads on a string” structure12) and is also known as the 10-nm fiber (Fig. 1C). Because only about half of DNA’s negative charges are neutralized in the 10-nm fiber, its remaining charge must be neutralized by other factors, including linker histone H1, cations, and other positively charged proteins or small molecules for further folding. This implies that the electrostatic state of the surrounding environment can greatly influence folding and dynamics of the 10-nm fiber.13)–20)
(Color online) DNA, histones, nucleosomes, and chromatin. (A) Negatively charged DNA (red). (B, left) The positively charged core histone octamers (yellow). (B, right) Nucleosome structure at 1.9 Å resolution10); totally 10 histone tail domains (intrinsically disordered regions [IDRs]) are extended away from the nucleosome core. The dyad position was indicated with an arrowhead. Basic residues (lysines and arginines) in the tail domains are colored in orange. Asterisks indicate sites of lysine acetylation (illustration was modified from Ref. 168 with permission). (C) 10-nm fiber. Note that the fiber remains negatively charged. (D) A classical view of 30-nm chromatin fibers folded into a hierarchical structure. (E) A current view of irregularly folded 10-nm fibers in the chromatin domain. (F) The nucleus has chromosomes in different colors based on their territories. This figure was reproduced from Ref. 122 with modifications.
How is the 10-nm fiber folded? In 1976, Finch and Klug first observed in vitro that the purified 10-nm fiber easily folded into a regular fiber with a diameter of 30 nm, known as the “30-nm chromatin fiber (30-nm fiber)” (Fig. 1D; center, Fig. 2A).21) The in vitro condition used the linker histone H1 and low concentrations of cations (e.g., <1 mM Mg2+ or <∼50 mM Na+). A possible formation mechanism is that low concentrations of cations cause nucleosomal fibers to gently repel each other due to insufficient screening of negative charges, and each nucleosome binds selectively to close neighbor nucleosomes on the DNA strand (Fig. 2A and B).20),22)–24) Two major structural models of the 30-nm fibers were reported. One model is called “solenoids” or the “one-start helix”, where consecutive nucleosomes are located adjacent to one another in the fiber (center, Fig. 2A).21),25) The second model is termed “zigzag” or the “two-start helix”, which assumes that each nucleosome in the fiber is bound to its second neighbor (center, Fig. 2A).26)–30) Since its discovery,21) the 30-nm fiber had long been assumed to be the basic structural unit of chromatin in eukaryotic cells and used as the textbook model.31),32) Building on this assumption, it was further proposed and widely believed that the 30-nm fiber folded progressively into larger fibers, the so-called “hierarchical helical folding model” (Fig. 1D).33)
(Color online) Structural variations of chromatin in vitro. (A) Left, 10-nm fibers form in no cation conditions. Center, three types of 30-nm fibers form with a low concentration of cations/Mg2+: 1) a solenoid (one-start), 2) a two-start zigzag, and 3) a zigzag tetranucleosomal.27) Right, large chromatin condensates with interdigitated 10-nm fibers that form in more physiological cations or higher cations. Note, this condensate lacks the 30-nm structures.20) Recently, under a certain specific condition, the condensates were shown to be liquid droplets formed by liquid-liquid phase separation.92),93) (B) Liquid-like nucleosome fibers. Under low cation and/or molecular crowding conditions, 10-nm fibers could form 30-nm chromatin fibers via intra-fiber nucleosome associations. An increase in the cation concentration and/or depletion attraction (also known as macromolecular crowding) results in inter-fiber nucleosomal contacts that interfere with intra-fiber nucleosomal associations, leading to a liquid-like state. Note that we show a highly simplified two-dimensional nucleosome model in these illustrations. Panels A and B were modified from Refs. 168 and 180 with permissions.
Although the 30-nm fiber had long been assumed to be the basic structure of chromatin31),32) (Fig. 1D; center, Fig. 2A), a pioneering cryo-electron microscopy (EM) study by the Dubochet group in 1986,34) our studies in collaboration with Eltsov et al.,35)–39) and subsequent research have demonstrated that chromatin consists of irregular and variable structures in most of the cells examined (Fig. 1E). The 30-nm fibers were observed in only a few exceptions, such as starfish spermatozoa and chicken erythrocyte with complete genome silencing.38),40)–42) Techniques have contributed to this new concept, including cryo-EM,34),36),43),44) X-ray scattering,38),39) electron spectroscopic imaging,18),41),45) conventional EM tomography,46) super-resolution imaging STORM (stochastic optical reconstruction microscopy),47) chromosome conformation capture (3C),48) Hi-C,49)–51) and computational modeling.52)–54) Recent cryo-EM studies have visualized chromatin fibers in near-native states of frozen-hydrated cell sections,55)–58) revealing no regular 30-nm fibers in human and other cells.
Considering these findings, it is intriguing to explore why the regular 30-nm fibers are absent in most cells. One major reason is the abundance of cations (∼100 mM K+ and ∼1 mM Mg2+) and other positively charged molecules inside of cells. While chromatin can form 30-nm fibers at low ionic strength in vitro, higher concentrations of cations converted chromatin into large globular condensates, which lack the regular 30-nm structure (right, Fig. 2A and B).20),59) Under such conditions, electrostatic repulsion between adjacent nucleosomes is almost completely diminished, allowing nucleosomes to interact freely with more distal nucleosome partners.37),60),61) We demonstrated this scenario in vitro20) using a model chromatin system, the 12-mer nucleosome array59) (Fig. 2A and B). In these large condensates, interdigitated chromatin folding prevents 30-nm fiber formation by sequestering the H4 tail, which mediates both the 30-nm fiber15) and large condensate formation.61),62) Frequent nucleosome loss or irregular spacing in native chromatin also disrupts regular 30-nm fiber formation.52),59),63) Furthermore, low levels of linker histones,63),64) acetylation of histone tails,65) histone variants, and non-histone protein binding53) all destabilize 30-nm fibers. These in vitro and cellular evidence indicate that the 30-nm fiber is not a basic structure of chromatin in the cell.
What is the basic structure/organization of chromatin in the cell? We consider that the irregular clusters of nucleosomes termed chromatin domains (or globules) are basic chromatin structures in higher eukaryotic cells (Fig. 1E; Fig. 3A),45),66),67) for the following two reasons.
(Color online) Schemes for hierarchical chromatin organization inside the cell nucleus and for depletion attraction. (A) Negatively charged DNA is wrapped around a positively charged core histone octamer to make a nucleosome. A string of nucleosomes, the 10-nm chromatin fiber, can form a loop structure held by cohesin, CTCF, and other proteins. The 10-nm fiber is compacted into chromatin domains (or globules) (e.g., topologically associating domain (TAD)/contact domain/loop domain), which interact over long distances to form chromatin A/B compartments. These compartments represent a transcriptionally active/open chromatin state (orange, A compartment) and an inactive/closed chromatin (blue, B compartment). A single interphase chromosome consisting of several compartments occupies a chromosome territory. Note that this scheme is highly simplified, and a more complex organization is inside the cell. Panel was reproduced from Refs. 168 and 181 with modifications. (B) Schematic representation of the depletion attraction (also known as the macromolecular crowding effect). Chromatin domains (colored spheres) are depicted within a square. The excluded volume is shown in gray below. When chromatin domains associate with each other (e.g., spheres outlined with dotted lines), the accessible space for soluble macromolecules (red) increases, leading to a reduction in excluded volume. This reduction is entropically favorable, effectively generating an attractive force between the chromatin domains. This panel was reproduced from Ref. 97 with modifications.
First, a large-scale chromatin structure with a diameter around 100–300 nm has long been observed by light and electron microscopic imaging in higher eukaryotic cells.68)–72) Consistently, recent combination studies of 3D super-resolution and scanning electron microscopy revealed “chromatin domains”, which were observed as irregularly sized clusters of nucleosomes (∼100–300 nm) (Fig. 3A).66),73) Interestingly, although a typical textbook model has long described euchromatin as open and heterochromatin as closed and condensed,31),32) new studies including ours demonstrated that the chromatin domains in euchromatic regions maintain condensed states with active epigenetic marks, but their sizes (orange in Fig. 3A) were smaller than those of heterochromatin (blue in Fig. 3A).66),67) Euchromatin domains seem to be not fully open, but condensed except for enhancers and active transcription start sites,66),67),74)–76) which are marked with histone modification of H3K27ac and range only from 200 to 1500 bp.77)
Second, Hi-C technology,78) which can generate a fine contact probability map of genome DNA, has also revealed that the higher eukaryotic genome is partitioned into chromatin domains (Fig. 3A) called topologically associating domains (TADs) (several hundreds of kb79)–81)) or contact/loop domains (mean size of ∼185 kb).82) These structures were confirmed visually by super resolution imaging using fluorescence in situ hybridization (FISH).83)–85) The loop domains seem to be held by cohesin (Fig. 3A), in cooperation with CTCF.83),86)–90) These domains are often clustered as two distinct, mega-base scale A/B compartments,78) which likely serve transcriptionally active chromatin and inactive chromatin, respectively (Fig. 3A). These compartments occupy the “chromosome territory” in the nucleus.91) Taken together, the chromatin domains seem to be fundamental chromatin structures (Fig. 1F; Fig. 3A), irrespective of euchromatin or heterochromatin.45),66),67)
How does the chromatin domain form in the cell? In addition to the cohesin complex that can crosslink chromatin, nucleosome-nucleosome interactions seem critical to forming condensed domains.45),70),75) The 10 long histone tails of each nucleosome are highly positively charged (right, Fig. 1B), which bind to neighboring nucleosomes and makes chromatin very “sticky”. Cations like Mg2+ can facilitate these interactions, as shown in the formation of Mg2+-dependent chromatin condensates in vitro (right, Fig. 2A and B).20),45),59),92),93) Furthermore, high concentrations of proteins, RNAs, and other macromolecules surrounding chromatin likely contribute to the formation of condensed chromatin domains through depletion attraction (also known as the macromolecular crowding effect) (Figs. 2B and 3B). Depletion attraction is an effective, attractive force that arises between large structures in crowded nuclear environments,94)–97) where the total molecular density, including DNA, RNA, and proteins, has been optically measured at 136 mg/ml97),98) (for more details on the force, see Fig. 3B legend).
The condensed organization (or globules) can modulate the accessibility of larger protein complexes to target sites (Fig. 4).66),67),99)–101) The condensed chromatin domains likely hinder the access of large protein complexes, such as transcription factors and replication initiation complexes, from their inner cores (Fig. 4).66),67),99)–101) Decompaction of such domains with histone modifications or the action of other proteins can increase accessibility to the complexes to turn on gene transcription (Fig. 4).
(Color online) Scheme for the eukaryotic chromatin domain. Euchromatic condensed domains (or globules) provide higher-order regulation of transcription by physically excluding large transcription complexes (green spheres) from the inner core of the domains. Attachment of large transcription complexes (green spheres) can retain binding regions on the domain surface for transcriptional activation, resembling a buoy (for details, see Ref. 67). An enhancer may also act as a buoy, interacting with the gene promoter (open green nucleosomes marked with ‘P’) and causing it to float up to the domain surface to be transcribed. These regions can be retained at the surface by enhancer–promoter interactions with a transcription condensate. The two “buoy” parts are marked. Scheme was modified from Ref. 67 with permission.
It should be emphasized that condensed chromatin domains have additional functional roles to protect the genome. First, the condensed chromatin organization can generate a spring-like restoring force that resists nuclear deformation by mechanical stress and plays an important role in maintaining genomic integrity. Nuclei with condensed chromatin possess significant elastic rigidity, while those with decondensed chromatin are considerably softer.102)–105) Second, condensed chromatin seems more resistant to radiation and chemical damage than the decondensed form, probably because condensed chromatin has lower reactive radical generation and is less prone to chemical attack.106)–109)
The view of irregular and variable structures of chromatin discussed above implies that chromatin is less physically constrained and more flexible and dynamic than expected in the regular structure model with helical coiling (left, Fig. 1D).37) Consistent with this view, live-cell imaging studies have long revealed dynamic chromatin movements using lacO/LacI-GFP (Fig. 5A),110)–114) as well as related ANCHOR/ParB approaches,115) and the CRISPR/dCas9-based technique (Fig. 5B).116)–118) The lacO/LacI-GFP method targets regions spanning 20–50 nucleosomes, while the ANCHOR [ParB/INT (or parS)] and CRISPR/dCas9-based techniques track <1 kb regions containing several nucleosomes.
(Color online) Schemes for various imaging and viscoelastic properties of chromatin. Schemes for lacO/LacI-GFP (A) and CRISPR-based chromatin labeling (B). (C) Left, scheme for single-nucleosome imaging. A small number of nucleosomes are labeled with a fluorescent tag and imaged using oblique illumination microscopy (center). Right, a typical single-nucleosome image. (D) Left, schematic for calculating mean squared displacement (MSD). Right, MSD plot (±SD among cells) of the H2B-HaloTag-TMR labeled nucleosome motion in untreated control (black) and formaldehyde (FA)-fixed (red) HeLa cells from 0.05 to 0.5 s. ***P < 0.0001 by Kolmogorov–Smirnov test. (E) Left, schematic for dual color (green and red) imaging of two neighboring nucleosomes in the chromatin domain. Right, typical trajectory sets of two neighboring nucleosomes. rc is the correlation coefficient of the trajectory set. (F) Two-point MSD plot (±SD among ≥3 clusters) in living siRAD21 (cohesin depletion)-treated HeLa cells (red), untreated cells (black), and TSA-treated cells (blue). Panels A–D were reproduced from Refs. 128 and 180 with modifications. Panels D–F were reproduced from Ref. 75 with modifications.
How does a nucleosome behave within the chromatin domain? To address this question, we developed single-nucleosome imaging119) using sparse fluorescent labeling and oblique illumination microscopy (Fig. 5C).120) This imaging sensitively and accurately measures individual nucleosome motion, providing useful structural information on chromatin in live cells.121)–127) We observed that nucleosomes could move ∼80 nm within 100 ms in interphase (Supplementary Movie S1),128) which is approximately 8 times larger than the size of a nucleosome—an unexpectedly large movement within such a short time. A mean squared displacement (MSD) quantitates how molecules spatially move during a certain time (left, Fig. 5D). MSD plots of nucleosomes appeared sub-diffusive (right, Fig. 5D), suggesting the nucleosomes are somehow constrained.
However, simple single-nucleosome imaging does not tell whether individual nucleosomes fluctuate within the domain, the domain itself moves (making nucleosomes appear to “move”), or both occur simultaneously. To clarify these possibilities, we focused on the distances between two neighboring nucleosomes in the chromatin domain (<150 nm distance) (left, Fig. 5E) and analyzed the MSD between them (two-point MSD)89) because the distance between nucleosomes is much less affected by either translational or rotational movements of the domain during imaging. While motions of the neighboring nucleosomes exhibited a weak correlation (right, Fig. 5E; Supplementary Movie S2),75) the two-point MSD between neighboring nucleosomes increased over time (Fig. 5F), suggesting that nucleosomes fluctuate within the domain in a liquid-like manner. Here, we define a liquid-like state as one characterized by diffusive movement without crystal-like long-range order. Notably, the two-point MSD increased with cohesin depletion or hyperacetylation of histone H3 and H4 tails by HDAC inhibitor trichostatin A (TSA) treatment, indicating that both cohesin and local nucleosome-nucleosome interactions are involved in constraining nucleosomes within a domain.75) These findings suggest that chromatin behaves like a liquid on a size scale of the domain (approximately 200 nm) (Fig. 6).75)
(Color online) Model of euchromatin organization in living cells. Nucleosomes in chromatin regions form condensed domains with local nucleosome contacts and cohesin (rings), where nucleosomes fluctuate like a liquid. A single interphase chromosome is stably occupied in a chromosome territory (highlighted as different colors). Illustration was reproduced from Ref. 75 with modifications.
Consistent with these findings, 12-mer nucleosome arrays59) with cations formed large chromatin condensates in vitro, which exhibited liquid-like behavior through liquid-liquid phase separation (right, Fig. 2A),92),93) where some macromolecules condense while others are excluded, creating functional compartments in the cell.129),130) While the 12-mer arrays used had uniform properties in length, spacing, size, and modifications,45),92),93) native chromatin with heterogeneous properties also transitioned into liquid droplets in the presence of cations and macromolecular crowders such as PEG, BSA, and dextran.97) Consistently, when a restriction enzyme was introduced into mitotic cells to fragment chromatin, chromosomes transformed into round structures resembling liquid droplets.131) Since intrinsically disordered regions (IDRs) in proteins are known to mediate the multivalent interactions that drive droplet formation,132) the 10 IDRs in the tail regions of the core histones (right, Fig. 1B)133),134) may contribute to the liquid droplet formation of chromatin.71),92),135),136) Notably, histone tail modifications that neutralize positive charges, such as acetylation, play a critical role in regulating the formation of chromatin condensates and droplets.92),137) The evidence discussed here and in the previous section strongly suggests that the liquid-like chromatin domain is the fundamental structural unit of chromatin in higher eukaryotic cells, replacing the 30-nm fiber (Fig. 6).
As discussed above, chromatin behaves like a viscous liquid within chromatin domains (∼200 nm/∼0.5 s spatiotemporal scale),75) allowing smaller proteins, such as transcription factors, to access their target sequences even within the condensed regions (right, Fig. 6; Fig. 4). This enhances chromatin accessibility and facilitates key DNA processes, such as RNA transcription and DNA replication/repair.67),100),119) Interestingly, local nucleosome motion, on average, remains constant throughout interphase.128) This steady motion enables cells to perform their routines like housekeeping functions (e.g., RNA transcription and DNA replication) under similar conditions throughout interphase.
How does chromatin behave at the chromosome scale (up to several micrometers)? On this scale, chromatin behaves more like a solid, likely due to numerous local constraints.45),75) This property supports that each chromosome is quite stably occupied in the territory (left, Fig. 6) without excess intermingling with other chromosomes, which contributes to keeping genome integrity. With loss of this property, more mixing of chromosomes and subsequent chromosome breaks/fusions may happen, leading to various genetic disorders, including tumorigenesis.
Additionally, the elastic characteristics of chromatin play a crucial role in safeguarding the genome by counteracting mechanical stress-induced nuclear deformation. As discussed previously, nuclei with condensed chromatin exhibit significant elastic rigidity, whereas those with decondensed chromatin—such as that induced by histone acetylation—are considerably softer.102)–104) These findings reinforce the viscoelastic principle of chromatin:138)–143) chromatin is locally dynamic and reactive but globally stable (Fig. 6).
Given that both euchromatin and heterochromatin form condensed domains in the cell, what distinguishes them? They exhibit distinct histone modifications (e.g., active and inactive marks) and histone variants, which play unique roles in genomic functions.144)–147) However, the physical differences between euchromatin and heterochromatin remain unclear, particularly in living cells, due to the limited availability of specific labeling methods for these chromatin regions.70),75),110),148)–152)
Recently, we have developed a new method called replication-dependent histone labeling (Repli-Histo labeling) to label euchromatin and heterochromatin specifically in live cells (Fig. 7A).100) This method was based on the eukaryotic DNA replication timing: euchromatin replicates in the early S phase, while heterochromatin replicates in the late S phase.153),154) Using Repli-Histo labeling, we found that more euchromatic regions exhibited greater nucleosome motion (Fig. 7B).100) Two-point MSD analysis between two neighboring nucleosomes further revealed that nucleosomes in euchromatin domains exhibited larger two-point MSD than those in heterochromatin domains.100) These findings suggest that nucleosomes in euchromatin fluctuate like a liquid at domain-size dimension scale and second time-scale, while those in heterochromatin are more constrained, resembling a more gel-like state, likely due to additional crosslinks such as with heterochromatin protein 1 (HP1).155)
(Color online) Dynamics of euchromatin, heterochromatin, and mitotic chromosomes. (A) Based on DNA replication timing, nucleosomes in euchromatin (left) and heterochromatin (right) were specifically labeled by Repli-Histo labeling.100) Nucleosome motion in euchromatin is more dynamic than in heterochromatin. (B) Schematic of an MSD plot of nucleosome motions in euchromatin and heterochromatin. For details, see Ref. 100. (C) Upper, sparse labeling of nucleosomes in mitotic chromosomes. Lower, MSD plot (±SD among cells) of TMR-labeled nucleosomes in live interphase (black), metaphase (magenta), and FA-fixed (gray) HeLa cells from 0.05 to 0.5 s. (D) Upper, condensins act as molecular crosslinkers to make loops. Lower left, condensins (red) locate around the chromosome center. Right, nucleosome–nucleosomes interactions are weakened when chromosomes are treated with the HDAC inhibitor Trichostatin A. Nucleosomes around the periphery (those mostly free from condensins) are less constrained and have higher mobility than those around the axis. (E) During mitosis, the chromatin domains may be assembled by condensin (and topoisomerase IIα, TopoII), transient increases in Mg2+,159) and depletion attraction97) to obtain a rod-like shape. Panels A and B and panels C and D were reproduced from Refs. 156 and 181 with modifications, respectively.
It is intriguing to discuss how interphase chromatin might reorganize into mitotic chromosomes during cell division. Our single-nucleosome imaging (upper, Fig. 7C) revealed that the nucleosomes in mitotic chromosomes fluctuate but are more constrained than those of interphase chromatin (lower, Fig. 7C; Supplementary Movie S3).156) Condensins are key protein complexes for mitotic chromosome assembly,157) and act as molecular crosslinkers enriched along the chromosome axis, which locally constrain nucleosomes and organize chromosomes (left, Fig. 7D).156) Nucleosome-nucleosome interactions mediated by histone tails also constrain and compact whole chromosomes. Treatment with HDAC inhibitors such as TSA, which disrupt nucleosome-nucleosome interactions, results in chromosome decondensation131),156) and increased nucleosome motion (right, Fig. 7D).156) Together, these findings suggest that chromosomes exhibit more gel-like properties in live cells.
Since condensed chromatin domains are the basic structure of chromatin in interphase, it is tempting to speculate that they act as building blocks for mitotic chromosomes (Fig. 7E). Upon mitotic entry, extensive histone tail deacetylation occurs,158) potentially promoting the fusion of interphase chromatin domains. Furthermore, increased levels of free Mg2+ and depletion attraction were observed during mitosis97),159) and may facilitate these fusion events. While Hi-C studies on mitotic chromosomes have not identified structures such as TADs or A/B compartments,160),161) electron microscopy, PALM (photo-activated localization microscopy), and STORM (stochastic optical reconstruction microscopy) have consistently revealed large-scale structures or chromatin domains within mitotic chromosomes that are approximately 200 nm in diameter.70)–72),162),163) This discrepancy may arise from the nature of Hi-C, which captures chromatin interactions based on cross-linking and ligation rather than directly visualizing spatial structures. It is possible that the cross-linking efficiency of formaldehyde in mitotic chromosomes is lower than in interphase chromatin due to the reduced abundance of nonhistone proteins bound to mitotic chromosomes compared to interphase chromatin.164) Alternatively, formaldehyde may induce higher levels of ‘nonspecific’ cross-linking in the highly condensed mitotic chromosomes, potentially masking domain structures in Hi-C analysis.
Condensed chromatin domains as building blocks in mitotic chromosomes likely facilitate the smooth assembly and disassembly of chromosomes. Epigenetic markers within these building blocks can also be reliably preserved throughout the cell cycle. How condensins, topoisomerase IIα, and other factors157),165),166) function in this process remains unclear and warrants further investigation.
Finally, what drives nucleosome fluctuations in interphase chromatin domains or mitotic chromosomes? We consider that local nucleosome fluctuation is primarily driven by thermal fluctuations because the motion of a compact polymer model167) with thermal fluctuations closely recapitulated the profile of nucleosome motion in live cells.128) Local nucleosome motions are increasingly constrained from euchromatin to heterochromatin (Fig. 7B),100) and are most constrained in mitotic chromosomes (lower, Fig. 7C).156) What creates these motion differences? Numerous protein and environmental factors contribute to restricting the thermal-fluctuating motion of nucleosomes within the cell,168) determining the overall nucleosome motion in euchromatin, heterochromatin, and mitotic chromosomes. For example, transcription machinery151),169),170) and cohesin70),89),90),171) are likely to impose constraints on nucleosomes in euchromatin regions, while HP1 and the nuclear lamina may play key roles in restricting nucleosome motion in heterochromatin. In mitotic chromosomes, condensins serve as the primary constraint.156)
We ponder that ATP energy is not a major driving force of nucleosome fluctuations in the cell. Although many studies have concluded that chromatin motion across different timescales is ATP-dependent, based on ATP-reduction experiments using inhibitors of respiration (e.g., NaN3) and glycolysis (e.g., 2-deoxyglucose [2-DG]),112),172)–176) this conclusion warrants reconsideration for several reasons. While ATP-reduction treatments do lead to decreased chromatin motion (Fig. 8A and B),128) they also induce chromatin condensation (Fig. 8C),18),70),128),159),177) likely due to a rapid increase in free Mg2+ levels (Fig. 8D). Most intracellular Mg2+ (10–20 mM) exists as a complex with ATP and other molecules, such as proteins.178) ATP hydrolysis releases Mg2+ bound to ATP,159) which induces chromatin condensation and subsequently reduces chromatin motion within the cell.70),128),159) These findings highlight the need for careful reevaluation of the ATP-dependency of chromatin motion. Systematic and rapid knockdown analyses of ATP-dependent chromatin proteins, such as remodelers, in the cell would provide new insight into this issue.
(Color online) ATP-reduction effects on chromatin organization and dynamics in live cells and scheme for possible chemical fixation effects on the cell. (A) Measurements of intracellular ATP levels using a Luciferase-based assay for HeLa cells: untreated cells (left), cells treated with 2-DG (ATP-reduction) and sodium azide (center), and without cells (right). (B) MSD plot (±SD among cells) of nucleosome motions in MQ water treated control cells (blue) and cells treated with 2-DG and sodium azide (pink) from 0.05 to 0.5 s. (C) Chromatin condensation by ATP-reduction in live cells (right). (D) Levels of intranuclear free Mg2+ determined by their Mg-green fluorescent signal. Data in panels A, B, and D were reproduced from Ref. 128. (E) In living cells, most molecules (circles) move and diffuse freely. Following chemical cross-linking, such as with FA fixation or methanol treatment, molecules adhere together or to cellular structures (green lines) until they form solid, continuous artificial structures. In addition, permeabilization with certain detergents for immunostaining or alcohol dehydration for EM can extract unfixed materials such as small molecules and proteins from the cells. Panel E was reproduced from Ref. 180.
Over the past 15 years, extensive evidence has dramatically transformed the view of chromatin from a static, regular structure to one that is more irregular and dynamic. Chromatin locally behaves like a liquid in living cells. Liquid-like chromatin domains represent the fundamental structural unit of chromatin, replacing the 30-nm chromatin fiber.
In this context, live-cell imaging is becoming increasingly crucial for understanding chromatin. As Dubochet noted,179) many molecules including proteins, DNAs, RNAs, and liquids freely diffuse in the aqueous medium of the nucleoplasm or cytoplasm in living cells (left, Fig. 8E). However, chemical fixation, such as formaldehyde, glutaraldehyde, or methanol, cannot fully preserve the dynamic states of intracellular molecules. Instead they form solid, continuous artificial structures, while other molecules are lost during sample preparation (right, Fig. 8E).179) With the careful evaluation of each research method’s strengths and limitations, a comprehensive approach that integrates imaging, genomics, proteomics, and computational modeling is essential for elucidating the true nature of genome chromatin organization and dynamics within the cell, which govern cell proliferation, differentiation, and developmental processes.
I am grateful to Ms. Sachiko Tamura for the preparation of figures and providing some results, and Dr. K. M. Marshall for critical reading and editing of this manuscript. I am also grateful to Dr. Mikhail Eltsov for performing our early work together and stimulating discussions. I thank Dr. K. Asakawa, Dr. S. Shinkai, Ms. K. Nakazato, Mr. M.A. Shimazoe, and Mr. K. Minami for constructive comments on this manuscript, all our collaborators who contributed to our studies, and Maeshima laboratory members for helpful discussions and support. I must also apologize to our colleagues in the chromatin field for citing a limited number of their papers due to space constraints. This work was supported by the Japan Society for the Promotion of Science (JSPS) and MEXT KAKENHI grants (JP20H05936, JP23K17398, JP22H04925 (PAGS), and JP24H00061) and the Takeda Science Foundation.
The authors declare no competing interests.
Edited by Yoshinori OHSUMI, M.J.A.
Correspondence should be addressed to: K. Maeshima, National Institute of Genetics, 1111 Yata, Mishima, Shizuoka 411-8540, Japan (e-mail: kmaeshim@nig.ac.jp).
Kazuhiro Maeshima was born in 1969. He graduated from the College of Biological Sciences at the University of Tsukuba in 1993. He earned his Ph.D. from Osaka University in 1999, where he studied protein-DNA interactions involving the DNA recombination protein RAD51. This research sparked his interest in chromatin and chromosome structure, leading him to work as a postdoctoral fellow in the lab of Prof. Ulrich K. Laemmli at the University of Geneva, Switzerland. In 2004, he moved to RIKEN in Wako to work in Dr. Naoko Imamoto’s lab, where he became a Research Scientist in 2006. In 2007, he proposed a novel chromatin structure model. Two years later, he accepted a position as a full professor at the National Institute of Genetics in Mishima. His research group pioneered single-nucleosome imaging, providing groundbreaking insights into chromatin organization and behavior in live cells. For his achievements, he was honored with the Kihara Memorial Foundation Academic Award in 2016. He also serves as an Associate Editor or Editorial Board Member for several scientific journals, including Epigenetics and Chromatin, Nucleus, Chromosoma, Cell Structure and Function, and Scientific Reports.
