The 4D Nucleome

The completion of the human genome sequence marked a watershed moment in biology, providing a linear blueprint of life. For years, the central paradigm focused on understanding this one-dimensional (1D) code—the sequence of nucleotides—and the epigenetic marks layered upon it.1 However, it has become profoundly clear that this linear perspective is insufficient. Within the confines of the eukaryotic cell nucleus, a structure with a diameter of less than 10 µm, over two meters of DNA must be packaged in a manner that is not only compact but also exquisitely functional.3 The genome is not a random tangle of thread but a highly organized, non-random, and profoundly dynamic three-dimensional (3D) structure.2 This 3D architecture represents a critical and distinct layer of biological information, operating in concert with the genetic sequence and the epigenome to orchestrate cellular life.1

The recognition of this spatial dimension has given rise to the concept of the 4D Nucleome, which extends the static 3D model to incorporate the fourth dimension: time.6 The 4D Nucleome framework seeks to understand the principles governing the structure and dynamics of the genome in space and time, and to decipher how this spatio-temporal organization influences nuclear function.6 This ambitious goal is the focus of major international consortia, such as the 4D Nucleome (4DN) Project, which aims to develop and benchmark experimental and computational approaches to map these dynamics and their functional consequences in human and mouse cells.6 The genome's architecture is in constant flux, reconfiguring in response to developmental programs, progression through the cell cycle, and external stimuli, making its temporal dynamics a key area of investigation.1

The spatial organization of the eukaryotic genome is best understood as a multi-level hierarchy, a series of nested folding patterns that package the vast length of DNA while ensuring access for essential nuclear processes. This hierarchy spans from the nanometer scale of the nucleosome to the micron scale of entire chromosomes, with each level contributing uniquely to genome function.2 Understanding this structural paradigm, which has filled the conceptual void between the nucleosome and the chromosome territory, is fundamental to deciphering the principles of the 4D nucleome.14

The most basic repeating subunit of chromatin is the nucleosome.3 The canonical nucleosome core particle consists of approximately 147 base pairs of DNA wrapped in about 1.75 left-handed superhelical turns around a protein core.24 This core is an octamer composed of two copies each of the four core histone proteins: H2A, H2B, H3, and H4.24 The resulting structure is a disc-like particle approximately 11 nm in diameter and 5.5 nm in height, which provides an initial linear compaction of the DNA by a factor of about seven.24 The nucleosome is completed by a stretch of "linker DNA" connecting it to the next nucleosome, and in many cases, a linker histone (e.g., H1) binds where the DNA enters and exits the core, further stabilizing the structure and contributing to higher-order folding.3

The hierarchical architecture of the genome is not a static state but the result of continuous, dynamic processes driven by a suite of molecular machines and biophysical principles. Understanding these mechanisms is key to deciphering how genome structure is established, maintained, and altered in response to cellular needs. The field has converged on two primary, and potentially competing, forces: the active, ATP-dependent process of loop extrusion, which organizes local domains, and the passive, thermodynamically driven process of phase separation, which governs large-scale compartmentalization. These processes are orchestrated by a growing cast of architectural proteins that act as the sculptors of the 3D genome.

The loop extrusion model has emerged as the leading hypothesis to explain the formation of TADs and many chromatin loops.47 First formalized through polymer simulations that remarkably recapitulated Hi-C data 47, the model has since gained substantial experimental support.

The genome's 3D architecture is not a fixed, static entity. It is a dynamic structure that is constantly remodeled over multiple timescales, from the dramatic reorganization that occurs during cell division to the subtle, rapid fluctuations in response to external signals. This temporal aspect—the fourth dimension—is a central focus of the 4D Nucleome project and is essential for understanding how the genome adapts to perform its diverse functions throughout a cell's life.6

The cell cycle imposes the most dramatic and cyclical changes on genome organization.87 To ensure the faithful segregation of genetic material, cells must transition between a decondensed, highly organized interphase state that permits transcription and replication, and a highly condensed, compact mitotic state.89

The intricate architecture of the 4D nucleome is not merely a feat of biological engineering; it is fundamentally intertwined with the function and integrity of the genome. The spatial and temporal organization of chromatin directly impacts core cellular processes, from the precise regulation of gene expression and the timing of DNA replication to the efficient repair of DNA damage. Consequently, when this architecture is disrupted, the consequences can be severe, leading to a spectrum of human diseases, including congenital developmental disorders and cancer.

Two of the most fundamental processes governed by 3D genome architecture are transcription and replication.

Our understanding of the 4D nucleome has been driven by a technological revolution over the past two decades. A powerful and complementary suite of experimental and computational methods has been developed to map, visualize, and model the genome's architecture in space and time. These approaches can be broadly categorized into genome-wide contact mapping, microscopy-based imaging, and in silico modeling, each providing unique and essential insights.119 A truly robust understanding of genome organization requires the integration of these multidisciplinary methods, as each has distinct strengths and limitations.34

The development of Chromosome Conformation Capture (3C) technology and its high-throughput derivatives has been the primary engine of discovery in the field. These methods are all based on the principle of proximity ligation: chromatin is first cross-linked in vivo with formaldehyde to fix spatial interactions, then digested with a restriction enzyme. The resulting sticky ends of DNA fragments that are held in close 3D proximity are then ligated together, creating novel chimeric DNA molecules. The frequency of these ligation products, quantified by sequencing, serves as a proxy for the interaction frequency between the two loci in the original cell population.29

The study of the 3D and 4D organization of the genome has transformed our understanding of nuclear biology, revealing a breathtakingly complex and dynamic architecture that is integral to genome function. We have moved from a static, linear view of DNA to a dynamic, four-dimensional perspective where the genome actively folds, reorganizes, and responds to its environment. This review has synthesized the current understanding of this architecture, from the hierarchical layers of organization—nucleosomes, loops, TADs, compartments, and territories—to the molecular mechanisms that sculpt them, most notably the competing forces of loop extrusion and phase separation. We have explored the profound functional consequences of this organization for gene regulation, replication, and DNA repair, and have seen how its disruption leads directly to human diseases, including developmental disorders...

This rapid progress has been propelled by a synergistic convergence of methodologies. High-throughput sequencing techniques like Hi-C and Micro-C provide population-averaged interaction maps; advanced imaging methods like super-resolution FISH offer direct visualization of structure in single cells; and sophisticated computational models based on polymer physics and artificial intelligence allow for mechanistic testing and prediction. The integration of these approaches is no longer just a best practice but a necessity for driving the field forward.34