The exit from naive pluripotency: a platform for the study of enhancer mechanistics

Abstract

Multicellular life depends on the ability to activate and repress genes in a highly context-specific manner. With each cell state transition, a new transcriptional profile is established. As non-coding DNA elements, enhancers mediate their regulatory potential through the effectors they recruit. While ultimately instructed by the underlying DNA sequence, enhancer activity depends on several factors, such as transcription factor availability, chromatin state, and promoter proximity, all of which are dynamically regulated within the cell. Even when we understand the regulation of one enhancer, its genomic impact is dependent on its integration within the regulatory landscape. Thus, a full picture of enhancer dynamics can only be painted through broad, but controlled, approaches that integrate investigations into multiple levels of gene regulatory mechanisms. In this review, we will present the exit of naive pluripotency as a prime setting to do just that and contextualize how its contemporary use has been, and could be, used to reveal the intricacies of enhancer mechanistics.

Introduction

Spatiotemporal control of gene expression is central to the development of complex organisms, and the mechanisms that allow it remain a source of intrigue within modern biology. Promoters are limited in how much information can be encoded to instruct when and where the associated gene should be expressed. Consequently, the origin of multicellular life might have been possible through the evolution of additional gene regulatory elements [1]. Genetic innovation gave rise to regulatory elements that could act upon target genes from a distance, complementing the local mode of gene regulation that is presently associated with the transcription of housekeeping genes [2]. A broadened regulatory landscape meant that gene expression could be adapted to increasingly specific stimuli, allowing for complex concerted efforts between cells. Development is an ancient testament to the success of distal cis-regulatory elements, but the mechanistics by which they drive it are still being unraveled.

Central among cis-regulatory elements are enhancers—non-coding DNA sequences that support distal promoters in tuning the transcriptional response to cellular needs. Active enhancers recruit context-specific cofactors that ultimately act upon the associated gene promoters to promote their appropriate regulation in a poorly understood fashion. Recruiting transcriptional regulators in a sequence-specific manner, enhancers provide a physical extension to the regulatory environment of genes, increasing both the flexibility and specificity with which promoters drive transcription. Enhancers have evolved to control their target genes in specific tissues and at specific times. Hence, they stand as key regulators of development.

Different combinations of enhancers are active in different cell types and states. Most genes are regulated by several enhancers, and their cis-regulatory interactions are important for developmental processes [3]. In humans, the median number of enhancers associated with a promoter is proposed to be three per gene [4]. Enhancer collaboration is emerging as a cornerstone of complex gene regulation. Despite being central drivers of development, enhancers are remarkably redundant, with the deletion of a single enhancer often having little-to-no effect on the outcome of mammalian development [5]. This phenomenon aligns with early discoveries in Drosophila, where the concept of ‘shadow enhancers’ was established and a network of functionally redundant enhancers that help ensure robustness in gene expression patterns was revealed [6–8].

Enhancers that exhibit astounding evolutionary conservation can also demonstrate redundancy [9]. In some cases, the deletion of an ultraconserved enhancer does cause deleterious, tissue-specific, developmental phenotypes [10], but it can also have paradoxically systemic effects rather than the expected organ-specific ones [11]. It is clear that the importance of enhancer collaboration stretches further than providing redundancy to deleterious mutations [12], but a lot is still left to learn about how enhancer-enhancer interactions promote precise gene expression.

How enhancers turn on and off, and how their subsequent activity affects gene expression, is not well understood. Multiple layers of regulatory mechanisms are integrated to decide enhancer activity, and while reductionist approaches start to tease apart their individual importance, enhancers can only be truly understood within their endogenous chromatin context. During development, cells traverse between sequential cell states, each characterized by a cell state-specific enhancer landscape. To establish novel cell types, these landscapes must be rewired. As such, cell state transitions are exemplary settings in which to study enhancer mechanistics. In the embryo, many transitions occur in parallel, resulting in a heterogeneous and complex environment that is difficult to probe using currently available experimental techniques. Studying a single cell state transition in vitro is, therefore, an excellent experimental system for studying the interplay of the different levels of enhancer regulation.

The exit from naive pluripotency

The exit from naive pluripotency is a developmentally important cell state transition that can easily be modeled through an efficient and tractable in vitro differentiation system (Figure 1). Mouse embryonic stem cells (mESCs) are capable of self-renewal while retaining their capacity to contribute to all tissues of the embryo proper. This is often referred to as naive pluripotency. In vivo, naive pluripotency is highly transient and attained only by cells in the pre-implantation epiblast [13]. Fortunately, two small-molecule inhibitors (2i) enforcing activation of Wnt-signaling and inhibition of FGF-signaling, together with leukemia inhibitory factor (LIF), are sufficient to capture and maintain the naive state indefinitely in in vitro cell culture. Despite being noncommitted, naive pluripotent cells have mounted the molecular machinery needed to exit pluripotency [14,15]. Through removal of 2i/LIF, cells can be released along the path of differentiation, establishing a new cellular identity resembling the epiblast of the early post-implantation embryo at E5.5.

Figure 1. The exit from naive pluripotency.

Mouse embryonic stem cells are derived from early embryos and resemble pre-implantation epiblast cells when cultured with two small-molecule inhibitors (2i) and LIF. 2i + LIF maintains a balance between key signaling pathways that promote a transcriptional state where specific transcription factors (TFs), such as NANOG and KLF4, work together with the core pluripotency components OCT4 and SOX2 to maintain naive pluripotency; a cell state of indefinite self-renewal. Upon removal of 2i + LIF, and with the addition of FGF2, naive cells are ‘released’ to differentiate in a process mimicking that of post-implantation epiblast cells, where the first step is called formative pluripotency. Here, the cells attain the ability to differentiate into specific lineages, with the interaction partners of the core components changing accordingly. The factors converge on developmental enhancers, bringing forth a major cell state transition through a myriad of molecular mechanisms. LIF, leukemia inhibitory factor, FGF2, Fibroblast Growth Factor 2

The exit from naive pluripotency is fast, highly reproducible, and irreversible. After 2i/LIF withdrawal, all cells within the starting population exit naive pluripotency within 24–48 hours. However, not all cells initiate differentiation simultaneously. Perhaps surprisingly, the cell cycle is not the major source for the heterochronic exit from naive pluripotency [16–19]. The observed heterogeneity could instead arise from factors such as variations in the cellular microenvironment [20,21], metabolic cellular state [22], stochastic gene expression [23], or differences in the initial epigenetic landscape of the cells [24]. Despite being metachronous [17], cells exiting naive pluripotency follow a unidirectional differentiation trajectory [16,25,26]. In other words, the differentiation protocol is highly reproducible and irreversible.

As naive exit is deterministic, the system allows us to repeatedly probe different individual factors involved in cell state transitions. Over time, a holistic understanding of how they interact will emerge from parallel findings. The study of enhancer mechanistics during development relies on the integration of several layers of regulation; therefore, the exit from naive pluripotency is an excellent model system for dissection of the contribution of individual factors to enhancer activity. In this review, we will briefly summarize our current understanding of how different gene regulatory mechanisms determine enhancer activity and how they act within the context of the exit of naive pluripotency.

Using naive pluripotency to study and contextualize enhancer mechanistics

Enhancer activity is orchestrated through multiple levels of regulation, each contributing to the precise control of gene expression. This complex interplay allows cells to dynamically respond to developmental cues and to transition between cellular states in the correct space and time. In this section, we examine the main regulatory layers that modulate enhancer function and explore how these interactions are essential for cell state transition by examining their role in the exit of naive pluripotency.

Transcription factors

Transcription factors (TFs) are key regulators of cell state and enhancer activity. While the direct molecular action of TFs is often not clear, they are known to promote or repress transcription either directly or through recruiting effector proteins. TFs possess DNA-binding domains through which they recognize and bind to specific DNA sequence motifs. Enhancers are composed of clusters of motifs and, akin to genetic billboards, can promote the local attraction of TFs. In turn, the recruited factors act synergistically to promote the assembly of productive transcriptional complexes on promoters associated with the enhancer. Consequently, our understanding of enhancer mechanistics centers on the dynamics by which TFs interact with genomic DNA.

During naive pluripotency, the expression of specific pluripotency-related TFs maintains a gene regulatory network centered on developmental potency and self-renewal. TFs OCT4 and SOX2 form the core of this network [27], and together, they bind and activate most enhancers specific to pluripotency [28]. The core factors remain important throughout pre- and post-implantation development, but a plethora of TFs co-ordinate naive exit [15,27].

Interactions between TFs can influence their function, which includes what genes they target. Depending on the presence of accessory TFs, the binding patterns of the core factors and the effects of their associated enhancers change drastically. An example of this is OTX2, which is engaged in the early stages of naive exit [29]. It acts as an interaction partner to OCT4 to redirect its binding from naive pluripotency enhancers to those specific to early differentiation [28], thereby rewiring the enhancer landscape. Throughout naive exit, the expression of TFs that support the naive pluripotency network, such as KLF2 and TFCP2L1 [30,31], declines as signaling pathways are no longer active. Notably, some TFs specific to pluripotency, such as ESRRB, have been suggested to guide the core factors to new binding sites during the initial phase of differentiation [25]. Redirection of binding is one of several mechanisms through which TFs dynamically reshape enhancer activity during cell state transitions.

TFs can have direct effects on enhancer activity. In the exit of naive pluripotency, some TFs, such as FOXD3 [32], OCT6 [33], and ZIC3 [34], repress naive-specific enhancers and help dismantle the pluripotent state. Conversely, TFs can ensure that genes remain active even as their regulatory input changes. For example, GRHL2 defines which newly active enhancers sustain the expression of an already active gene [35]. Additionally, the expression of new TFs during this period can establish novel enhancer-promoter interactions, further promoting lineage commitment [36]. Thus, TFs not only determine which genes are expressed but also actively reshape the enhancer landscape, either by silencing regulatory elements tied to a previous state or by stabilizing new enhancer-promoter interactions that support differentiation.

While changes in TF expression levels can enforce lasting alterations in cell identity, cell state transitions are often driven by modulating the activity of pre-existing TFs. This can occur through cellular localization, as is the case of TFE3, which translocates out of the nucleus to allow the exit from naive pluripotency [37]. Similarly, nuclear export of KLF4, a TF that maintains long-range interactions at the Oct4/Pou5F1 locus needed for pluripotency [38], triggers the exit of naive pluripotency [39]. Post-translational modifications also play a crucial role in regulating TF activity. OCT4 itself is the target of numerous post-translational modifications such as sumoylation [40], phosphorylation [41], and O-GlcnAc [42] that affect its stability, target binding, and transcriptional output. Ultimately, a TF’s ability to influence the genome depends not just on its presence or activity but also on whether it can access its target sites, a constraint that is itself tightly regulated at multiple levels.

Chromatin accessibility

Chromatin refers to the dynamic composition of DNA and histone proteins, and it plays an essential role in how enhancers can be activated. To organize the DNA within the cell, DNA is wrapped around octamers of histone proteins forming nucleosomes. Organized DNA is known as chromatin, and at a higher level of organization, nucleosomes flexibly organize into a polymer melt-like structure with different levels of local compaction [43]. Open chromatin permits transcription factor binding to DNA and enables transcription, while closed chromatin restricts access and activity. However, accessibility is not binary; some factors can engage with compacted chromatin by specifically recognizing nucleosome-associated DNA or through recruiting chromatin remodelers. As a result, chromatin organization is not just a passive feature but actively shapes transcription by influencing which factors can bind in a given context.

Active enhancers are typically associated with open chromatin; however, the mere presence of accessible chromatin does not necessarily correlate with enhancer activity. Enhancer activity is shaped by both chromatin accessibility and the transcriptional landscape. Some enhancers require specific chromatin environments to become active (‘chromatin-dependent enhancers’) [44], while others function within closed chromatin, remaining dormant until activated by environmental or developmental cues [45]. Notably, many open chromatin regions lack enhancer function altogether. Even within ‘super-enhancers’, not all constituent elements directly contribute to transcription. Instead, non-canonical regulatory elements act as ‘facilitators’, ensuring full enhancer function by enabling long-range regulatory interactions [46]. This is particularly relevant during development and cell state transitions, where dynamic changes in chromatin accessibility and transcription factor availability dictate which enhancers become active, ensuring precise spatiotemporal regulation of gene expression necessary for lineage commitment and differentiation.

Pluripotency is linked to a uniquely accessible and dynamic chromatin landscape. Naive pluripotent cells show lower levels of chromatin compaction [47] and greater diversity in nuclear condensate behavior [48] compared with differentiated cells. Notably, RNA polymerase II has been shown to form transient condensates that correlate with transcriptional bursting activity in embryonic stem cells [49]. During naive exit, these condensates disperse, coinciding with broader nuclear reorganization and enhancer inactivation [50]. Differentiation is marked by a progressive loss of chromatin accessibility at pluripotency genes and their enhancers [51]. In pluripotency, chromatin state is linked to the activity of pluripotency factors, with OCT4 depletion leading to chromatin condensation in epiblast cells [52]. OCT4 works together with chromatin remodelers such as BAF both in naive pluripotency and during its exit [28,53,54]. These interactions suggest that chromatin remodelers are recruited by the core pluripotency factors independent of their specific binding patterns.

High chromatin accessibility is critical for pluripotency, but certain regulatory elements require targeted repression to prevent uncontrolled activation. Highly compacted DNA, marked by the epigenetic histone modification trimethylation of lysine 9 of histone 3 (H3K9me3), restricts aberrant transcriptional programs and safeguards naive pluripotency from activation of transposable elements and cryptic enhancers [55]. Recent findings suggest that H3K9me3 is not uniformly stable across the genome but is maintained through distinct, region-specific dynamics, with some regions losing H3K9me3 rapidly upon perturbation, while others exhibit more stable repression, revealing a nuanced balance between accessibility and long-term silencing [56]. This highlights how chromatin accessibility is highly dynamic and tightly intertwined with different histone modifications.

Histone modifications

In addition to chromatin remodelers, chromatin modifiers are recruited to enhancers to modify their structure and function, often through establishing post-translational modifications on histones. Numerous amino acid residues within the tail ends of histones are commonly modified, but the contribution of individual modifications to enhancer regulation remains unclear. Specific histone modifications correlate with different states of enhancer activity (see Table 1for a summary) and are often erroneously equated as causative [68]. Among the modifications, acetylation and methylation of lysines have received the most attention to date.

Table 1. Key epigenetic modifications commonly found on enhancers and their role in pluripotency.

Modification	Type	Writers	Erasers	Role in mESCs	Role in differentiation	Phenotype when disrupted	Citations
H3K4me1	Active	KMT2C/D (MLL3/4)	KDM1A (LSD1)	✔️ Required for enhancer priming	✔️ Required for lineage specification	Impaired activation of developmental enhancers	[57,58]
H3K27ac	Active	CBP, p300	HDAC1/2/3	❌ Not necessary	❌ Not necessary	Not essential; presence increases enhancer activity	[59,60]
H3K27me3	Repressive	PRC2: EZH2, EED, SUZ12	KDM6A/B (UTX/JMJD3)	❌ Not needed in steady state	✔️ Important for differentiation	Premature differentiation or failure of lineage commitment	[61–63]
H3K9me3	Repressive	KMT1A/B, KMT1E (SUV39H1/2, SETDB1)	KDM4A/B/C (JMJD2 family)	✔️ Required to silence aberrant transcription	✔️ Required for lineage commitment	De-repression of transposons and cryptic enhancers	[55,56,64]
H2AK119ub	Repressive	PRC1: RING1A/B	BAP1	❓ Context-dependent	✔️ Required for differentiation	Disrupted differentiation cues	[65,66]
DNA methylation (5mC)	Repressive	DNMT1, DNMT3A/B	TET1/2/3	✔️ Helps silence transposons	✔️ Important for lineage specification	Impaired lineage commitment	[67]

H3K9me3, trimethylation of lysine 9 of histone 3. mESC, mouse embryonic stem cell. PRC1, Polycomb-repressive complex 1. PRC2, Polycomb-repressive complex 2.

Histone acetylation is a relatively unspecific modification that correlates with transcriptional activity. It is deposited by histone acetyltransferases (HATs), such as p300 or CBP, which play direct roles in gene activation [69]. Furthermore, acetylation affects chromatin accessibility as it neutralizes the positive charge of histone tail residues, reducing their interaction with nucleosomal DNA. Conversely, histone methylation is a more specific modification deposited by histone methyltransferases and, sometimes, has distinct regulatory roles. For example, H3K9me3 is associated with repressed chromatin [64], while the same modification on lysine 4 (H3K4me3) is associated with active promoters [70]. Histone modifications are recognized by different reader proteins, which in turn recruit cofactors that have activating or repressive functions.

The exit from naive pluripotency provides insights into how histone modifications might be involved in the co-ordination of cell fate transitions. Active enhancers can be identified by the combination of multiple specific epigenetic marks: H3K27ac (acetylation of the 27th lysine residue of histone H3 [71]) and H3K4me1 (monomethylation of the 4th lysine residue of histone H3). Recently, H2BNTac (multi-site N-terminal acetylation of histone H2B) has also seen use as a marker of enhancers [72]. While H3K27ac is the main marker of active enhancers, preventing H3K27ac deposition at enhancers has a negligible effect on transcription in mESCs [59,60]. De novo H3K27 acetylation is also not an absolute prerequisite up-regulation of genes during naive exit [59]. This suggests an inherent robustness to, and independence of, the regulatory networks that work alongside the de facto deposited mark.

Similar to chromatin accessibility, the presence of H3K27ac does not always reflect intrinsic enhancer activity. The enhancer cluster regulating Fgf5 during the exit from naive pluripotency illustrates this complexity. There, multiple enhancers contribute to gene activation at different time points [73]. Each of the enhancers shows similar levels of H3K27ac; however, none of the elements has strong intrinsic enhancer activity, and only in co-operation are these elements able to activate the expression of Fgf5 during the exit from naive pluripotency (discussed in further detail below).

H3K4me1 is primarily catalyzed by the methyltransferases KMT2C and KMT2D (previously named MLL3 and MLL4) [74]. Despite mutations in both KMT2C and KMT2D being embryonic lethal beyond the post-implantation stages [75], these mutations do not significantly reduce global levels of H3K4me1. In mESCs, the deposition of H3K4me1 by KMT2C/D is not necessary for maintaining naive pluripotency but might be crucial for sustaining differentiation, with one study suggesting that KMT2D is needed for H3K27ac deposition by p300/CBP through H3K4me1-mediated priming and that this is required for mESC differentiation in vitro [76]. Refuting this proposed link, most, if not all, enhancers can gain H3K27ac independent of KMT2C/D-mediated H3K4me1, and their knockout has no significant effect on gene expression [57]. KMT2C/D have also been shown to have functions independent of their catalytic activity, as cells expressing catalytically deficient KMT2C/D, as opposed to their complete knockout, show only slight effects on transcription [77]. KMT2D is needed for the exit of naive pluripotency, but this is independent of its catalytic activity [78].

While H3K4me1 on its own might be dispensable for enhancer activation, it might play a supportive role in cell state transitions, ensuring that the correct genes are activated [58]. H3K4me1 aids in enhancer–promoter interactions and enhancer-driven transcription during differentiation, primarily mediated by KMT2C/D and KMT2B [79], which could be crucial for long-range chromatin interactions [80].

Differentiation relies on the repression of genes outside of the intended lineage. This is ensured in part by Polycomb-repressive complex 2 (PRC2)-mediated deposition of H3K27me3. In naive pluripotency, H3K27me3 is widespread across the genome [81]. In the absence of the mark, naive pluripotency can be maintained, but differentiation is impaired [61–63]. This stands in contrast with the proposedly negligible effect of H3K27ac loss [59,60], suggesting a dominant role for H3K27me3 in regulating enhancers, corroborated by its absence on repressed enhancers leading to aberrant gene expression [55]. PRC2 catalytic subunit EZH2 deletion is embryonically lethal post-implantation [82], as are deletions of subunits EED [83] and SUZ12 [84], emphasizing that, while H3K27me3 is not a requirement for productive gene regulatory networks per se, it contributes to proper spatiotemporal control of genes.

PRC2 works together with Polycomb-repressive complex 1 (PRC1) to silence genes and to organize chromatin. PRC1 mediates H2AK119 monoubiquitylation, which contributes to transcriptional repression by restricting RNA polymerase II initiation [65]. Two PRC1 configurations exist, and while naive pluripotency can be maintained without either, their simultaneous loss leads to differentiation [66]. The two PRC complexes are also important for chromatin architecture [85], facilitating enhancer:promoter contacts prior to transcription [86], likely independent of their enzymatic activity [87–89].

Cell state transitions involve priming the genome for expression of new genes and simultaneously preventing their misexpression. Enhancers bivalently marked by H3K4me1 and H3K27me3 can be found during naive pluripotency and are linked to developmentally relevant genes [90]. Such enhancers are proposed to be ‘poised’, inactive but prepared for activation or repression in cell state transitions downstream of the state in which they were first established. Poised enhancers are associated with orphan CpG islands, which suggests that bivalency could have a structural function by acting as tethering elements between poised enhancers and their cognate promoters [91]. Put differently, when not part of deliberate repression, bivalency could serve to establish the topology of the chromatin landscape needed for somatic cell differentiation in a PRC-dependent manner, which is especially important in terms of species-specific developmental control [86,91]. Further studies of naive exit will help elucidate how bivalent enhancers help cells navigate between activation and repression during cell state transitions.

DNA methylation

DNA methylation as a layer of gene regulation has a debated function in enhancer activity, but the exit from naive pluripotency could help to elucidate some of its functions. DNA methylation is depleted from the zygotic genome as pluripotency is attained and is reassembled to poise cells for differentiation as naive pluripotency is established. The naive state is marked by low global levels of DNA methylation, which increases upon differentiation [92,93]. Early studies suggested that loss of DNA methylation impairs differentiation potential of mESCs cultured under serum conditions, where signaling pathways are highly heterogeneous [94]. However, more recent evidence obtained from defined 2i/LIF conditions suggests that DNA methylation is largely dispensable for the exit from naive pluripotency and does not significantly impair differentiation potential [95].

DNA methylation in mammals occurs primarily as 5′-methylcytosine within CpG dinucleotides and is associated with compaction of DNA, thereby potentially preventing the binding of transcriptional activators. Loss of DNA methylation has little effect on mESC self-renewal and gene expression patterns, and can only be implied to be relevant in the formation of primordial germ cells [96]. The lack of methylation induces epigenetic changes primarily to enhancers and simultaneously extends the temporal window for germ cell differentiation [95]. This might reflect an overarching lack in the ability to establish strong restrictive epigenetic states, as a genome without methylation could likely see more promiscuous activation. Comparatively, the loss of TET enzyme function leads to widespread hypermethylation, particularly at gene promoters, and impairs the differentiation capacity of mESCs, as seen in embryoid bodies and teratomas [97]. Methylation can directly prevent DNA binding of proteins that recognize specific DNA motifs or create binding sites for factors that preferentially recognize methylated DNA [98]. Only a very small subset of enhancers is regulated through DNA methylation [99], hinting that DNA methylation is not a common tool used by the cell to control gene expression, but that it might be important in specific regulatory and developmental contexts [67].

Enhancer-promoter proximity

Often located tens to hundreds of kilobases away from their target promoters, enhancers circumvent separation in two-dimensional space by acting in three dimensions. One way to overcome separation by large genomic distances is to displace the genome, keeping two gene regulatory elements apart. This can be achieved through cohesin-mediated loop extrusion [100]. Cohesin proteins form a ring-shaped complex that acts as a molecular motor. When loaded onto chromatin, the DNA is extruded through the ring structure and thereby distant elements are brought into proximity to each other. The process is dynamic, and the constant shuttling of DNA could facilitate enhancer–promoter interactions [101]. It is also conceivable that, rather than through creating order, cohesin instead primarily functions as a disruptor of chromatin interactions, preventing excessive compaction to ensure that regulatory elements remain free to interact [102].

The Mediator complex is another key regulator of enhancer–promoter interactions, acting as a molecular bridge between TFs at enhancers and RNA polymerase II at promoters. Rapid depletion of Mediator leads to reduced enhancer-promoter contacts, decreased gene expression, and lowered binding of cohesin at enhancers [103]. Possibly, Mediator acts to strengthen and sustain cohesin-induced loops to ensure robust gene regulation.

While many long-distance enhancers might depend on cohesin [104], acute loss of cohesin has only minor effects on enhancer-promoter interactions and gene expression in steady state [105–107]. It was argued that, during differentiation, cohesin-mediated loop extrusion might be important to establish enhancer-promoter communication, but once established, loop extrusion might no longer be needed. However, depletion of the cohesin processivity factor Nipbl during differentiation has little effect on gene expression despite loss of most cohesin-mediated loops [108].

Several cases challenge the model of cohesin-mediated looping being the primary mechanism for enhancer–promoter interactions. For instance, the zone of polarizing activity regulatory sequence (ZRS) enhancer, which regulates Shh expression during limb development, functions over a megabase away from its target promoter [109,110]. Although promoter-enhancer looping has been proposed as the main activating step [111], studies suggest that Shh expression can persist even when the physical distance between the enhancer and promoter increases, indicating that other structural or regulatory mechanisms are at play [112]. Similarly, Sox2 transcription can occur without detectable enhancer-promoter interactions, questioning the necessity of stable loops for gene activation [113]. Rather than acting as a strict requirement, looping may play an early role in establishing genome topology, but once regulatory hubs form, transcription might continue independently of physical enhancer–promoter proximity [114,115].

High-resolution studies of the three-dimensional genome architecture have revealed so-called microcompartments. Within these, enhancers and promoters might connect independently of cohesin, since their interactions are largely unaffected by both transcriptional inhibition and loss of loop extrusion [116]. At the microcompartmental scale, enhancers show widespread interactions with both, other enhancers and target promoters, suggesting gene regulation is, at least in part, a product of local networks of collaborating regulatory elements. Accordingly, the co-operative action of multiple enhancers has been suggested to play a direct role in supporting cohesin-independent gene activation [108,117].

Enhancer cooperativity

Throughout differentiation, the enhancer network is continually rewired to promote novel gene regulatory networks that sustain specific cell states. This process is far more complex than a simple ‘enhancer on = gene on’ or ‘enhancer off = gene off’ model. Instead, as cells transition between states, enhancers are selectively activated, repressed, or repurposed, resulting in the up-regulation of some genes while others are silenced.

Crucially, enhancers do not act in isolation; they function within cooperative regulatory landscapes, where each element affects the function of the others. This coordinated enhancer remodeling ensures precise transcriptional control, allowing cells to establish and maintain distinct identities.

Enhancers that collaborate can carry out different roles. In the regulation of Klf4 during pluripotency, several elements interact in a hierarchical manner to sustain its expression. While disrupting a single enhancer only has a mild effect, the removal of multiple enhancers leads to a significant reduction in Klf4 expression, highlighting functional redundancy within the regulatory network [118]. Even super-enhancers are often modular and consist of multiple individual enhancer elements spread out over multiple kilobases. Five enhancer elements collaboratively activate Fgf5 expression during the exit from naive pluripotency, with four intergenic elements forming a super-enhancer that induces Fgf5 at distinct times. The fifth intronic enhancer consistently amplifies the expression of Fgf5, resulting in a strong, super-additive induction [73]. Enhancer cooperativity can significantly mitigate the reduction in gene activation caused by increased genomic distances, particularly when a weaker enhancer is positioned between a strong enhancer and its target promoter [117]. Therefore, weak enhancers can take on a facilitator role to help strong enhancers communicate with their target promoter efficiently [46]. The extent to which facilitator elements play a role in all enhancer-promoter communications has yet to be established.

Not all regulatory elements associated with a gene act to promote its expression: Enhancers also function together with silencers, and a developmentally relevant example of this is Cdx2, where a single TF motif turns from silencing to enhancing [119]. Enhancer interplay might establish the above-mentioned microcompartments and thereby increase the interaction of distal enhancers with the target promoter independent of cohesin [108]. This regulatory coordination is especially critical during cell state transitions, where chromatin structure is rapidly reorganized to establish new transcriptional programs. Recent findings reveal that as cells exit naive pluripotency, enhancerpromoter interactions undergo large-scale reconfiguration through the formation of multiway chromatin hubs, which bring together distant enhancers and promoters into shared regulatory environments [36]. These hubs provide a structural framework for genes associated with pluripotency exit to establish long-range interactions with emerging enhancers, suggesting that spatial genome architecture plays a key role in facilitating transcriptional transitions. Taken together, enhancers don’t work in isolation, and their impact can only be examined as part of a larger regulatory framework.

The exit from naive pluripotency presents an ideal model to study enhancer mechanisms, as it involves widespread chromatin reorganization, enhancer activation, and long-range genomic interactions necessary for lineage specification. By examining how enhancers integrate into multiway chromatin hubs during this transition, we can gain critical insights into how regulatory elements shape gene expression dynamics in development.

Discussion

The exit from naive pluripotency offers an invaluable opportunity for the holistic understanding of enhancer mechanistics. mESCs are easily maintained in vitro, highly amenable to perturbations, and their intrinsic differentiation trajectory recapitulates the rewiring of the enhancer landscape that allows for cell fate transitions during development. Most other developmental cell fate transitions are multi-pronged, and while successes have been made in probing regulatory elements in multicellular settings [120], the multi-omic approaches needed to reliably probe the complexity of enhancer functionality in heterogeneous systems are currently underdeveloped. mESCs circumvent the drawbacks of more complex systems, as they provide a near limitless amount of cells undergoing a single, developmentally relevant, cell state transition. Therefore, the exit of naive pluripotency remains the prime choice for studies that tease apart how enhancers work in determining cell fates in mammals.

Many aspects of gene regulation in development are still poorly understood, and it is not quite clear how individual factors studied in one cellular context contribute to cell fate transitions across multiple states. Despite its role in redirecting OCT4 binding to differentiation-specific enhancers, OTX2 is not vital for the exit of naive pluripotency. mESCs null for OTX2 display deficiencies in differentiating but will readily integrate into blastocysts and give rise to chimeric embryos [121]. Highly chimeric embryos show a phenotype consistent with the effect of its in vivo deletion, where the loss of OTX2 leads to the failure in the development of head structures [122]. This highlights that the effects of the lack of OTX2 are not caused by the lack of a single-cell fate decision, but through the compound effects that the absence of OTX2 has on anterior neural differentiation during development. Thus, pinpointing a distinct causal effect in development is difficult, as between one stage and another, a cell state can be largely indistinguishable from the wild type and a deleterious phenotype would only become apparent through the compound effects of subsequent transitions.

The robustness of cell state transitions and the transcription factor networks that drive them are a central theme in studies utilizing the exit of naive pluripotency as a model [123]. No single factor solely responsible for this cell state transition has been identified so far, despite genome-wide screens having been performed to saturation [15]. Transitioning between the naive and formative pluripotent state involves the modulation of multiple signaling pathways, including FGF/ERK, WNT/β-catenin, LIF, Notch, and mTORC1 [15]. Accordingly, the lack of activators and repressors of these pathways reproducibly correlates to defects in pluripotency and differentiation.

TCF7L1 recurrently appears as a target in loss of function screens of naive exit [124]. TCF7L1 is a TF that acts as a repressor of WNT target genes and its ablation replaces WNT activation in culture of mESCs [125]. Since WNT signaling is needed to uphold pluripotency and stemness in culture, which is mediated through repression of TCF7L1 [126], it is no surprise that TCF7L1 promotes naive exit. In the absence of TCF7L1, naive exit is delayed but not halted. This is true also for other central signaling mediators [15]. Only through the concurrent knockout of components belonging to different signaling pathways can ESCs be arrested in pluripotency. This is demonstrated in the triple knockout of TFs Etv5, Rbpj, and Tcf7l1 (facilitating FGF/ERK, Notch, and WNT signaling, respectively) arresting mESCs in naive pluripotency [127]. This highlights that the concerted efforts of different pathways are necessary to drive naive exit and that the enhancer landscape can be rewired even in the case of perturbations.

As our understanding of enhancer function deepens, future research must address how enhancer networks integrate multiple levels of regulatory input to generate stable, but flexible, transcriptional states. Advances in synthetic biology and genome engineering now offer the potential to build regulatory environments from the bottom up. By leveraging mESC models, enhancers can be introduced into synthetic loci and functionally probed in a developmental context, allowing researchers to systematically define enhancer logic and regulatory architectures in ways previously impossible. Understanding how enhancer landscapes can be manipulated in vivo will not only advance developmental biology but may also pave the way for precise engineering of cell fate transitions in regenerative medicine and biotechnological innovations.

Perspectives.

• Importance for the field: Enhancers are the main gene regulatory elements that convey spatiotemporal control of gene expression, specifically in development. The exit of naive pluripotency provides an accessible and versatile platform to study enhancer mechanistics during cell state transitions.

• Current thinking in the field: The embryo and systems modeling embryonic development are not ideal to understand how transcriptional changes are regulated. In larger gene regulatory networks and in enhancer clusters, deficiencies can often be overcome by other redundant factors.

• Future directions: Novel tools including developments in synthetic biology, targeted protein degradation, and dual screening strategies in combination with the exit from naive pluripotency have the potential to deepen our understanding of transcriptional regulation during cell state transitions.

Abbreviations

H2BNTac

multi-site N-terminal acetylation of histone H2B

H3K27ac

acetylation of lysine 27 on histone H3

H3K4me1

methylation of lysine 4 on histone H3

H3K9me3

trimethylation of lysine 9 of histone 3

H3K27me3

trimethylation of lysine 27 on histone H3

LIF

leukemia inhibitory factor

PRC1

Polycomb-repressive complex 1

PRC2

Polycomb-repressive complex 2

TF

transcription factor

2i

two small-molecule inhibitors

mESC

mouse embryonic stem cell

Competing Interests

The authors declare that there are no competing interests associated with the manuscript.

Funding

Research in the Buecker lab is supported by the European Reserach Council (ERC-CoG-2024, 101171485) and the Austrian Science Fund (FWF, PAT9017923).