Regulatory architecture of cell identity genes and housekeeping genes

Abstract

Gene regulation and chromosome architecture are intimately linked. Genes with prominent roles in cell identity are often regulated by clusters of enhancer elements. In contrast, a recent study shows housekeeping genes are often regulated through clustering of promoters. We discuss here new regulatory insights for these two types of genes.

Proper regulation of gene expression is fundamental to cell identity and homeostasis. This regulation involves dynamic interactions among DNA, protein and RNA molecules that influence local genome architecture. Here we describe recent insights into regulatory mechanisms that feature prominently at cell identity genes and housekeeping genes and discuss questions that emerge from these insights.

Gene regulatory elements: promoters and enhancers

Promoters and enhancers are DNA elements that play key roles in gene regulation [1]. Promoters contain the transcriptional start sites of genes and binding sites for factors generally involved in early transcriptional events. Enhancers are cell-type specific elements that bind transcription factors (TFs) and cause increased expression of the genes with which they associate. Promoter and enhancer elements share the ability to bind TFs, recruit transcription apparatus and initiate transcription. The tendency of RNA polymerase II molecules to pause transcription soon after initiation [1] means that most active promoters and enhancers form assemblies of DNA, RNA and protein molecules (Figure 1A).

Figure 1:

Regulatory architecture of genes. (A) Enhancers and promoters are DNA regulatory elements that are bound by transcription factors, which recruit the transcription apparatus. RNA polymerase II can transcribe portions of these elements, so active promoters and enhancers form assemblies of DNA, RNA and protein molecules. (B) Regulatory elements can interact with each other due to tethering proteins that can bind these elements and dimerize or bind local RNA species. (C) Condensates formed at regulatory elements may fuse to form larger assemblies. (D) Cell identity and housekeeping genes share the features of clustered regulatory elements, but are distinguished by enhancer versus promoter clustering and the nature of regulatory and tethering transcription factors.

The exact DNA sequences of promoters and enhancers are not precisely defined, as promoters are typically inferred from the 5’ end sequences of RNA molecules and enhancers are generally identified using evidence for occupancy of histones with certain modifications. Nonetheless, we can make some estimates of their typical dimensions by considering what is known about their components [1,2]. Promoters assemble the basal transcription apparatus and RNA polymerase II, and so promoters might generally span ~100bp. A typical active enhancer might encompass 100-300bp of DNA and be bound by 10-40 TF molecules together with multiple coactivator complexes and RNA polymerase II molecules.

Promoter and enhancer elements, bound by various protein and RNA molecules, can interact in myriad ways: enhancer-enhancer, enhancer-promoter and promoter-promoter interactions are observed [3]. These interactions can be due to tethering proteins that bind sites within or adjacent both types of DNA elements and interact with one another [3-5] (Figure 1B), and to the propensity of protein/RNA complexes to assemble into larger complexes that encompass multiple DNA elements [6,7] (Figure 1C).

Cell identity genes and housekeeping genes

Genes can roughly be divided into two categories of function: cell identity and housekeeping. Cell identity genes tend to be expressed in a cell- or lineage-specific manner, encode machineries that contribute to cell-specific functions, and are regulated by cell- or lineage-specific enhancer elements. Thus, genes that encode transcription factors that control cell identity are generally under the control of enhancers that are active predominantly in those cells.

Housekeeping genes, in contrast, are expressed in most cells and encode critical biosynthetic functions such as the metabolic enzymes that are necessary for all cells. Although some housekeeping genes can be regulated by enhancers, they typically lack these elements. Despite constituting the majority of genes expressed in cells, their regulation has not garnered the same level of attention given to cell-specific genes, so there has been more limited knowledge about the mechanisms involved in their control. Some transcription factors are expressed broadly across cell types and bind promoters, and these have been assumed to play roles in expression of housekeeping genes.

Clusters of regulatory elements and their tethering factors

For genes that play prominent roles in cell identity, it is common to find clusters of enhancer elements, bound by cell-type specific transcription factors, that interact with each other and with distant promoters [8]. Tethering proteins such as CTCF and YY1 have been identified that can facilitate interactions among DNA regulatory elements, and these proteins may bind sites within the DNA elements or adjacent to them [4,9].

Recent studies reveal that housekeeping genes with essential roles in metabolism and cell maintenance, which tend to be clustered in the mammalian genome, often have promoters with a conserved sequence that is bound by the Ronin transcription factor [5]. Ronin has emerged as a tethering factor that facilitates interactions among these promoters.

Thus, cell identity and housekeeping genes share the features of clustered regulatory elements, but are distinguished by enhancer versus promoter clustering and the nature of regulatory and tethering transcription factors (Figure 1D).

Clustering and transcriptional condensates

The clusters of enhancers at cell identity genes assemble large numbers of transcription factor, cofactor, RNA polymerase II and RNA molecules [6,7]. These large assemblies, where hundreds of transcription factor molecules interact with large numbers of coactivator and RNA polymerase II molecules, are highly dynamic, typically forming and dissolving in seconds to minutes. These assemblies exhibit features expected for biomolecular condensates; their components are enriched in disordered regions and exhibit rapid molecular exchange kinetics, and the bodies they form show evidence of fusion in live cells. The transcriptional output of these condensates correlates with their size and lifetime.

The clusters of promoters at housekeeping genes also assemble a large amount of transcription apparatus and these assemblies likewise exhibit behaviors expected for biomolecular condensates [6]. The formation of these condensates is dependent on normal levels of Ronin protein and the DNA sequence that it binds in these promoters.

The physicochemical properties of condensates contribute to gene regulation in several important ways [6]. The hundreds of different types of protein molecules that must assemble at an individual gene can associate efficiently without each molecule requiring a specific high-affinity interaction with a partner in a complex with defined stoichiometry. Proteins with shared functions can be selectively concentrated in specific condensates by virtue of favorable chemical interactions that are not dependent on stable structures [10]. Condensates can be regulated by the charge balance afforded by their association with RNA; low levels of RNA produced initially in regulatory regions can facilitate formation of transcriptional condensates, whereas a burst of RNA produced when a gene is transcribed can dissolve them, providing a mechanism for RNA-mediated feedback regulation [11].

New questions emerge from these insights

The clustering of regulatory elements raises numerous questions about the potential advantages of clustering, the factors that contribute to this behavior, the potential functions of RNA molecules that are components of such clusters, and how differences in the factors present at cell identity and housekeeping genes may contribute to different regulatory behaviors (Figure 2).

Figure 2.

Opportunities for further study of regulatory architecture. (A) What opportunities for gene regulation emerge from clustering of regulatory elements? (B) What are the factors and forces that contribute to interactions among regulatory elements? (C) What are the functions of RNA molecules that are transcribed from regulatory elements? (D) What are the functions of different regulatory factors at clustered DNA elements for cell identity and housekeeping genes

Why might evolution favor clustering of key regulatory elements at genes (Figure 2A)? Clustering offers the opportunity for different enhancer elements to respond to different signals involved in the regulation of a single gene. Promoter clustering can facilitate coordinate regulation of multiple genes. Clustering of regulatory elements will also favor the formation of transcriptional condensates, whose physicochemical properties can contribute to gene regulation.

What protein factors contribute to tethering and clustering of regulatory elements (Figure 2B)? Tethering factors such as CTCF, YY1 and Ronin contribute to these interactions and there are likely to be others that both bind to DNA and form multimeric complexes. When crowded on DNA, proteins that are enriched in small modular repeats or intrinsically disordered regions, such as transcription factors and cofactors, will promote formation of condensates whose properties can be regulated by diverse catalytic enzymes and small molecules.

What are the functions of RNA molecules that are transcribed from regulatory elements (Figure 2C)? Many transcription factors and cofactors can bind RNA [12]. Differences in the affinity of protein regulators for different RNAs might influence gene regulation in a manner that is not currently appreciated. RNA molecules might act as tethers between proteins bound to different DNA elements [13]. The ability of RNA molecules to contribute to the charge balance in condensates provides yet another regulatory function.

Why might cells have evolved to use different regulatory factors at clustered DNA elements for cell identity and housekeeping genes (Figure 2D)? One reason may be their need to respond to different environmental signals. In response to developmental signals, signaling factors, often lacking DNA-binding domains, preferentially interact with the clustered regulatory regions of key cell identity genes [14]. In response to metabolic signals, a portion of insulin receptor molecules, which lack DNA binding domains, become associated with metabolic gene promoters [15]. Future studies will be necessary to more fully understand how these diverse signaling factors selectively associate with these clustered elements.