In the previous article, Nucleic Acids 3: Same DNA, Different Cells — The Functional Organization of the Genome, we explored how cells can develop completely different forms and functions even though they contain the same genetic information. In the end, what determines a cell’s identity is not simply the genetic information it carries, but which genes it uses, when, where, and how much. So how does a cell choose which genes to turn on among the enormous amount of DNA it contains? And more importantly, how can regulatory elements that are not located right next to a gene control its transcription from such a great distance? To understand this, we need to look at the genome not as a long, linear molecule, but as a three-dimensional structure that folds and bends inside the nucleus. In this article, we’ll start with the units used to describe the distance and position of DNA, and then follow the story through DNA loops, cohesin, CTCF, Mediator, and TADs to see how the three-dimensional genome controls transcription in space.
Measuring Distance and Position Within a Gene in bp
When we want to describe where regulatory elements such as promoters and enhancers are located within or around a gene, we often use a unit called bp, or base pair. DNA has two strands, with nitrogenous bases paired together in the middle. We count these base pairs to describe the length and distance along a DNA molecule. In other words, one base pair corresponds to 1 bp.
The promoter contains a core promoter, and within this region is the transcription start site (TSS). We usually define the TSS itself as +1 and describe the positions of nearby regulatory elements relative to it. For example, the TATA box, which we will discuss below, is typically located around 25–30 base pairs upstream of the TSS, so we might describe it as being around –25 to –30. The Initiator (Inr) overlaps the TSS and is located around –2 to +4, while the DPE (Downstream Promoter Element) is located downstream of the TSS, around +28 to +32.
When looking at a gene map, you will also often see the terms upstream and downstream. These are relative terms. They describe whether a particular DNA sequence lies before or after the TSS relative to the direction in which transcription proceeds. For example, an enhancer or silencer regulating a particular gene may be located either upstream or downstream of that gene's TSS. This is why we use the terms upstream and downstream rather than simply saying "before" or "after."
Pre-initiation complex (PIC) and general transcription factors
Let’s take a closer look at the moment when transcription begins. Before transcription starts, a preparation process first takes place at a specific location on the DNA. During this process, several proteins and RNA polymerase gather at one location and form a complex structure. This is called the transcription initiation complex.
The pre-initiation complex begins in the promoter region, which is called the starting point of transcription. More precisely, it begins in the core promoter region, the minimum functional DNA region within the promoter that is necessary for transcription to start properly. The core promoter refers to a narrow DNA region around the transcription start site (TSS), and its range and composition can vary depending on the gene and the type of promoter. This region includes elements such as the TATA box, Initiator, and DPE (Downstream Promoter Element).
The pre-initiation complex begins when the transcription factor TFIID binds to these core sequences. The name TATA box literally comes from the fact that the DNA sequence is rich in T (thymine) and A (adenine). If a gene does not have a TATA box, TFIID can recognize and bind to other core promoter elements such as the Initiator (Inr) or DPE. In fact, the TATA box is found in only about 5–7% of promoters in eukaryotes, including humans. [1]
TFIID generally binds to the core promoter sequence first. TFIID is a large complex made up of the TATA-binding protein (TBP) and several TBP-associated factors (TAFs), and it serves as a kind of “platform” where other transcription factors can gather and bind. In genes transcribed by RNA polymerase II, a series of transcription factors known as TFII are essential for transcription initiation. RNA polymerase II cannot bind stably to the promoter region on its own, so with the help of these general transcription factors, it forms the pre-initiation complex. Because this process is common to the initiation of transcription in almost all genes, these transcription factors are called general transcription factors (GTFs).
Several general transcription factors are recruited to the pre-initiation complex. After TFIID, TFIIA and TFIIB bind sequentially, followed by RNA polymerase II, which joins the complex together with TFIIF. TFIIE and TFIIH then bind as well, completing the pre-initiation complex. TFIIH uses the activity of its helicase enzyme to locally unwind the DNA double strand at the promoter, making RNA synthesis possible at the transcription start site. At the same time, it phosphorylates the C-terminal domain (CTD) of RNA polymerase II, mainly at Ser5, guiding the polymerase to leave the promoter and enter the early stage of transcription.
TATA box
Now let’s think about the TATA box. Why is the TATA sequence located in the core promoter? Among DNA base pairs, A–T has two hydrogen bonds, while G–C has three. In other words, A–T base pairs have relatively weaker bonding than G–C base pairs, so they are much easier to separate. When TBP, which is part of the transcription factor complex, binds to the TATA box, it strongly bends and twists this relatively weak region. As the normally rigid DNA double helix bends, a space is created, providing a platform where the other transcription factors and RNA polymerase can assemble. This also reduces the energy burden on the helicase enzyme of TFIIH, which is responsible for unwinding the DNA, allowing transcription to proceed more quickly and strongly.
Therefore, TATA boxes are relatively common in genes that require relatively strong transcriptional regulation or need to be expressed immediately and rapidly under certain conditions. They are often found in inducible genes, such as inflammatory genes, immune genes, and stress-response genes, which are normally switched off or have low activity but can be switched on rapidly when external stimuli or environmental changes occur.
On the other hand, housekeeping genes, which perform basic functions that are always needed in all cells, often do not have a TATA box, and other core promoter elements and transcription factors are responsible for transcription initiation. Housekeeping genes are genes that need to be expressed steadily and continuously at a consistent level. In this way, cells have evolved a variety of promoter structures depending on the characteristics of each gene.
Core promoter, proximal promoter, and enhancer
Basically, all genes begin transcription in the same way. Various general transcription factors (General Transcription Factor, GTF) bind to RNA polymerase II in the core promoter region and form the pre-initiation complex. The speed and strength of transcription that begins from this complex can be described as the basal level, and to add stability to this complex, specialized transcription factors bind to the proximal promoter, which can be located tens to hundreds of bp upstream of the core promoter. These factors increase the speed or strength of transcription to some extent and determine the rate of gene expression.
However, if a cell needs to transcribe a particular gene in a very precise, fine, and powerful way depending on the tissue or situation, additional specialized transcription factors are needed beyond this basic pre-initiation complex. These tissue-specific transcription factors bind to sequences such as enhancers, which can be located thousands to more than hundreds of thousands of bp away from the core promoter. The reason they are located separately and so far away is that enhancers can combine many specialized transcription factors at the same time, allowing transcription to be regulated in a complex but flexible and precise way.
A single enhancer can also interact with the promoters of several genes and may flexibly contact different promoters depending on the time or stage. The highly sophisticated transcriptional regulatory ability of enhancers can have a major effect on gene expression and may even determine the fate of a cell. Some genes can show a large increase in expression when an enhancer is activated. That is probably why they are called enhancers, or “boosters.” Well-known powerful transcription factors such as NF-κB, AP-1, STAT, SMAD, HIF-1α, Myc, p53, GATA, SOX, OCT4, and FOXP3 are specialized transcription factors that bind to enhancers.
There are also silencers, which are very similar to enhancers but work in the opposite way. If an enhancer is an “on” switch, a silencer is an “off” switch. So, if enhancers and silencers can both be located very far away from the promoter region, how can they have such a strong influence on transcription from such a distance?
Gene expression and transcriptional regulation
The phrase “increasing gene expression” means that a gene is transcribed more often, more quickly, or in greater amounts, resulting in increased production of mRNA. One RNA polymerase can transcribe the same gene several times in succession, or multiple RNA polymerases can attach one after another and carry out transcription at the same time.
Enhancers promote transcription in several ways. They bring together activator transcription factors, make chromatin more open or maintain it in a more open state so that the transcription machinery can access the DNA more easily, recruit transcription-related coactivators, and help form or stabilize the pre-initiation complex at the promoter. In contrast, silencer sequences are also located far away from the promoter, and they bind to transcription factors called repressors to suppress transcription.
DNA loop: cohesin and CTCF
How can a distant enhancer regulate transcription? First, because DNA actually exists in a form that bends and folds in three dimensions. DNA does not simply exist as a long linear molecule. It forms loops, and the process that explains how these loops are formed is called the loop extrusion model.
According to this model, when part of the DNA passes through a ring-shaped cohesin complex, cohesin acts as a molecular motor that uses ATP to continuously pull the DNA inward, creating a progressively larger loop. As the DNA is pulled inward, the round loop gradually grows. This process of gradually pulling DNA inward and expanding the loop is called loop extrusion. The size of a loop is determined by a molecule called CTCF on the DNA. When cohesin reaches a pair of CTCF binding sites arranged in a convergent orientation, loop extrusion tends to stop. It is like a sign saying, “Stop pulling here.”
Cohesin can create loops and the loops can also disappear, so DNA is not simply fixed in one position but continues to move as loops of different sizes form and disappear. When DNA bends into a loop, an enhancer and promoter that were originally far apart can become closer to each other in three-dimensional space, and the probability of interaction between them can also increase.
Mediator complex
DNA loop formation can bring an enhancer and promoter closer together in space, but the role of transmitting signals between them and activating the transcription machinery is carried out through a large protein complex called the Mediator complex. Because specialized transcription factors, or activators, bound to enhancers rarely bind directly to RNA polymerase II in the promoter region, a mediator that can serve as a bridge between them is needed. That is, quite literally, the role of the Mediator complex. Mediator receives regulatory signals generated at the enhancer and passes them to RNA polymerase II and the pre-initiation complex bound to the promoter. [2] In other words, Mediator is not simply a physical bridge. It is closer to a regulatory hub that integrates multiple regulatory signals and helps determine whether transcription begins and how strongly it proceeds.
Mediator is a complex made up of about 25–30 different subunits. Structurally, it forms a core complex consisting of three basic modules: the head, middle, and tail. The Tail module mainly interacts with transcription factors bound to enhancers, while the Head module directly contacts RNA polymerase II and helps stabilize the transcription machinery. The Middle module connects the two structurally. Because different transcription factors bind to different subunits of Mediator, its structure allows it to selectively integrate various signals. A fourth component, the CDK8 kinase module (CDK8), can also bind as a separate component. This module is a cyclin-dependent kinase complex containing CDK8 and regulates the function of Mediator depending on the situation.
In the past, complexes made up only of the core module were sometimes understood as the active form, while those with the CDK8 module attached were considered the repressive form. However, a more detailed model is now accepted. The CDK8 module phosphorylates the C-terminal domain (CTD) of RNA polymerase II or transcription-related proteins, affecting how they interact with the pre-initiation complex and influencing the transition between stages of transcription. Therefore, rather than viewing the CDK8 module simply as an “off switch,” it is more appropriate to understand it as a regulatory device that coordinates the transition between transcription initiation and elongation. Under general conditions, it can act to limit or regulate the formation of the pre-initiation complex, but in stress responses such as the hypoxia response or under certain signaling conditions, it can instead promote the elongation of transcription for specific genes. [3]
The Mediator complex functions as a structural mediator connecting enhancers and promoters, while also acting as a dynamic regulatory device that interprets various cellular signals and adjusts the activity of RNA polymerase II. Therefore, it is not simply a connection between two regions, but an important component that finely controls the stages of transcription and affects gene expression.
TAD (Topologically Associating Domain)
Now that we have looked at the actions of cohesin and CTCF proteins, let’s use the TADs formed through their actions to understand the structure of the three-dimensional genome. DNA is not simply a long, straight molecule. Inside the nucleus, it forms loops and folds and bends in three dimensions. Because of these loops, certain regions can be positioned closer to one another and interact more frequently, while contact with other regions can be relatively limited.
These interaction-based regions are called TADs (Topologically Associating Domains), which literally means “topologically associating domains.” They are regions defined by contact frequency in three-dimensional space, regardless of their distance along the two-dimensional DNA sequence. In other words, they refer to groups of DNA regions that are physically close to one another inside the nucleus and statistically come into contact more frequently.
When we look at the three-dimensional structure of the genome, functionally related genes and regulatory elements are organized so that they can work within a certain spatial context. Interactions between genes and regulatory sequences inside chromatin do not happen randomly. They occur more frequently within certain spatial compartments. Therefore, understanding the spatial relationships between genes, and between genes and regulatory elements, within a TAD can provide an important clue to understanding gene expression and transcriptional regulation. In general, genes located within the same TAD tend to interact with one another more frequently, and in some cases they may show similar patterns of gene expression. In contrast, the frequency of interactions with genes located across a TAD boundary may be relatively lower.
TAD regions are domains within chromatin where contact frequency is relatively high, and in many cases CTCF binding sites play an important role in forming their boundaries. As we saw earlier, cohesin forms loops through the loop extrusion mechanism. During this process, when it reaches a pair of CTCF binding sites with a convergent orientation, loop extrusion tends to stop. These positions are often observed to coincide with TAD boundaries. However, not all TAD boundaries depend on CTCF, and various factors can be involved depending on the cell type and genomic context. The loops created by cohesin are dynamic structures. At the level of individual cells, they are continuously formed and broken down. However, when contact frequencies are averaged across many cells, relatively stable interaction patterns can be observed, and TAD structures are defined based on these patterns.
DNA regions inside a TAD show statistically higher contact frequencies than regions outside the TAD. This spatial proximity affects the probability of interactions between enhancers and promoters. Therefore, in many cases, enhancer–promoter interactions tend to occur preferentially within the same TAD. In this sense, TADs can be viewed as providing a spatial framework in which transcriptional regulation takes place. However, TADs do not absolutely determine which enhancer corresponds to which promoter, and interactions can also occur across TAD boundaries.
The image below shows a Hi-C contact map that measures how frequently different DNA regions come into contact with one another in three-dimensional space. It shows the three-dimensional structure formed by chromatin inside the nucleus and the contact frequencies within that structure, presenting them step by step from higher-level structures to smaller-scale structures. Think of it as zooming in with Google Earth. The more frequently two regions come into contact, the darker they are shown.
The first image (i) shows that chromosomes are not completely mixed together inside the nucleus. Instead, each chromosome exists relatively independently within its own territory, called a Chromosome territory, which could be thought of as its own home base. Contacts occur frequently within the same chromosome, while interactions between chromosomes are relatively limited. At the second level (ii), even within a single chromosome, chromatin can be divided into two large functional compartments (compartment A/B). In other words, it can be divided into open chromatin, euchromatin, where transcription is active, and closed chromatin, heterochromatin, where transcriptional activity is low. Although these concepts are not exactly the same, these regions also tend not to be mixed randomly. Instead, regions with similar functional states tend to gather together spatially and come into contact with one another more frequently. In other words, active chromatin tends to interact more frequently with other active chromatin, while inactive chromatin tends to interact more frequently with other inactive chromatin. As we move down to an even smaller scale (iii), TADs (topologically associating domains) can be observed. When the contact frequency between two particular points within a TAD is especially high, we can infer that a loop or another relatively specific spatial interaction may exist at that location. Loops near TAD boundaries can be identified more clearly, and loops can also be observed in enhancer–promoter interactions within TADs.
In this way, TADs are domains where the contact frequency within a particular region is much higher than in surrounding regions, and they appear to function as a kind of functional interaction unit. The fact that long-distance regulation between enhancers and promoters often occurs within the same TAD is also related to this spatial organization. TADs do not directly initiate or suppress transcription, but they provide a structural framework that somewhat limits the range in which transcriptional regulation can take place and increases or decreases the possibility of interactions between specific regulatory elements.
[References]
[1] Core Promoters in Transcription: Old Problem, New Insights
https://pmc.ncbi.nlm.nih.gov/articles/PMC4340783/
[2] The Mediator complex and transcription regulation
doi: 10.3109/10409238.2013.840259
[3] The Mediator Complex: A Regulatory Hub for Transcriptional Activity of Nuclear Receptors
https://doi.org/10.3390/cells14171335
