Insulators and domains of gene expression Tamer Ali 1 , Rainer Renkawitz and Marek Bartkuhn The genomic organization into active and inactive chromatin domains imposes specific requirements for having domain boundaries to prohibit interference between the opposing activities of neighbouring domains. These boundaries provide an insulator function by binding architectural proteins that mediate long-range interactions. Among these, CTCF plays a prominent role in establishing chromatin loops (between pairs of CTCF binding sites) through recruiting cohesin. CTCF- mediated long-range interactions are integral for a multitude of topological features of interphase chromatin, such as the formation of topologically associated domains, domain insulation, enhancer blocking and even enhancer function. Address Institute for Genetics, Justus-Liebig-University, Heinrich-Buff-Ring 58, D35392 Giessen, Germany Corresponding author: Renkawitz, Rainer (rainer.renkawitz@gen.bio.uni- giessen.de) 1 Permanent address: Faculty of Science, Benha University, Benha, Egypt. Current Opinion in Genetics & Development 2016, 37:17–26 This review comes from a themed issue on Genome architecture and expression Edited by Frederic Berger and Pamela Geyer For a complete overview see the Issue and the Editorial Available online 20th January 2016 http://dx.doi.org/10.1016/j.gde.2015.11.009 0959-437X/# 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creative- commons.org/licenses/by-nc-nd/4.0/). Introduction The concept of inactive and active chromatin domains was suggested quite early on as a way to interpret compact and less dense chromatin packaging in diploid interphase nu- clei or in polytene chromosomes. The existence of such domains specifically requires the presence of domain boundaries to insulate the opposite activities of neighbour- ing domains. Such shielding elements, known as insulators, have been functionally identified by a position-indepen- dent high-level expression of a transgene in mice and flies [1,2]. In contrast to this barrier effect of an insulator, another shielding activity was called enhancer blocking [3], since it interferes with the action of an enhancer on a specific promoter when the insulator is positioned between the two. Following the discovery of several Drosophila-specific insulator binding proteins (IBPs), such as BEAF32 [4], Su(Hw) [5,6] and Zw5 [7], the vertebrate factor CTCF [8,9] was shown to mediate insulation [10]. Later, the high conservation of chromatin insulation was demonstrated by the identification of CTCF in Drosophila (dCTCF) [11–13] and by comparing shared features (Table 1). Here, we summarize recent results on the genome-wide binding of these and more recently discovered insulator factors, and the projection of these binding sites onto the three-dimensional chromatin structure. These observa- tions and results from high throughput analyses and functional tests are discussed with respect to a unifying mechanism for insulator-mediated barrier function and enhancer blocking activity. CTCF: inhibitor and facilitator of enhancer function Enhancer blocking activity of an insulator depends on its arrangement, that is, it has to be situated between the enhancer and promoter. This fact alone implies that the enhancer blocking activity is achieved by interfering with the chromatin looping required for enhancer/promoter contact. Detailed analysis of three-dimensional looping and the role played by the insulator protein CTCF revealed that CTCF not only possesses interference (enhancer blocking) activity, but also additionally med- iates chromatin contacts or loops required for enhancer function. Examples for such bivalent consequences of loop formation are discussed below. Bioinformatics evaluation of genome-wide chromatin in- teraction data led to the construction of a genome-wide interaction map of regulatory elements, which indicated that enhancer–promoter interactions are highly cell-type specific. Key interacting components are CTCF and cohesin [14]. This is exemplified by the MHC-II locus, which is active in B cells and bound by CTCF at 15 sites. In plasmablasts, this locus is inactive and only one third of the CTCF sites are bound. This correlates with the finding that CTCF is required for the cell type specific three-dimensional architecture of the locus and for maxi- mal MHC-II gene expression in B cells [15  ]. Another example is where CTCF/cohesin organizes a loop pattern that includes the promoter of the PTGS2 gene such that the PTGS2 gene is activated. In cancer cells the CpG island at the PTGS2 promoter is methyl- ated and the gene is turned off. This silencing mechanism is in part caused by the methylation-induced loss of CTCF binding, which results in a change in chromatin looping and abrogation of gene activity [16]. Regulation of dCTCF binding in Drosophila development is seen at the homeotic gene Ultrabithorax (Ubx), which is activated by Ubx enhancer elements in the third thoracic leg imaginal disc. Here, a dCTCF site at the enhancer generates a loop with the gene promoter. In inactive tissues Available online at www.sciencedirect.com ScienceDirect www.sciencedirect.com Current Opinion in Genetics & Development 2016, 37:17–26