Pioneer factors: Roles and their regulation in development

Abstract

Pioneer factors are a subclass of transcription factors that can bind and initiate opening of silent chromatin regions. Pioneer factors subsequently regulate lineage-specific genes and enhancers and thus activate zygotic genome after fertilization, they guide cell fate transitions during development, and they can promote various forms of human cancers. As such, pioneer factors are useful in directed cell reprogramming. In this review, we define structural and functional characteristics of pioneer factors, how they bind and initiate opening of closed chromatin regions, and the consequences on chromatin dynamics and gene expression during cell differentiation. We also discuss emerging mechanisms that modulate pioneer factors during development.

Introduction: Pioneer factor definition and recent discoveries

The vast majority of DNA in the nucleus is wrapped into nucleosomes and assembled into chromatin fibers, making DNA inaccessible to many transcription factors and other gene regulatory proteins (Figure 1A). Active genes and enhancers typically harbor an open chromatin conformation and exist in euchromatin. Active promoters and enhancers can include nucleosome-depleted regions where DNA is accessible to transcription factors and protein machineries [1]. They are often flanked by nucleosomes harboring the histone modifications H3K27Ac, H3K4me3 and/or H3K4me1 (Figure 1B). However, euchromatin regions represent a small fraction of the nuclear genome; the largest fraction being nucleosome-dense and folded into compacted and silent chromatin. We now appreciate that silent chromatin can be in distinct molecular forms [2, 3], including naïve chromatin which is enriched for linker histone H1 isoforms that compact chromatin [2], H3K9me3-marked heterochromatin, which is bound by the chromatin compacting protein Heterochromatin protein 1 (HP1), and H3K27me3-marked heterochromatin, which is bound by the chromatin-compacting Polycomb repressive complex 1 (PRC1) and Polycomb repressive complex 2 (PRC2) (see Glossary and Figure 1B). These three silent chromatin states, naïve, H3K9me3-marked, and H3K27me3-marked, repress lineage-specific genes, enhancers, DNA regulatory elements, and repeat elements [3–5]. During development, new spatiotemporal, lineage-specific gene expression networks are required to induce cell differentiation, necessitating a transition from silent chromatin states to an open chromatin conformation at promoters and enhancers. The transition is often initiated by pioneer factors, which thus are among the master regulators of cell fate changes [6–12] (Figure 1A). Herein we primarily focus on the latest studies in the field.

Figure 1: Pioneer factors bind silent chromatin and drive cell differentiation.

(A) Left: Pioneer factors binds closed chromatin regions. Middle: Pioneer factor induces local DNA accessibility to other transcription factors and protein complexes. Right: Pioneer factor induces chromatin reorganization and thus DNA sequence activation. Protein machineries include histone modifying complexes, the RNA polymerase transcription complex, and additional protein complexes regulating cis-regulatory elements. (B) Chromatin state of cis-regulatory elements in euchromatin regions (top) and silent chromatin regions (bottom). Unlike canonical transcription factors (TFs), pioneer factors (PFs) are able to bind silent chromatin regions. (C) Pioneer factors guide cell differentiation; their dysregulation might lead to cancer or cell reprogramming. Pioneer factors drive zygotic gene activation.

Pioneer factors interact with silent chromatin regions and enable their remodeling into an active chromatin conformation (Figure 1A) or further compaction and silencing, as reviewed recently [13–15]. Pioneer factors’ opening chromatin can lead to activation of lineage-specific genes and enhancers [6, 16, 17] and guide cell differentiation [7, 9, 10, 16, 18, 19]. Indeed the factors’ ectopic expression can initiate cell fate transitions [8, 17, 20–24], whereas pioneer factor-depleted cells often fail to differentiate [6, 7, 9, 11, 19, 22, 24–30]. Furthermore, pioneer factor dysregulation can cause aberrant cell phenotypes, such as cancer [31, 32] or cell reprogramming [33–39], as recently reviewed [40] (Figure 1C). Indeed, chromosome translocations can form oncogenic-fusion pioneer factors, such as Paired box gene 3-Forkhead box protein O1 (PAX3-FOXO1) in rhabdomyosarcoma cancer cells, leading to cancer [41] by altering enhancer repertory usage and gene expression profile [42–44].

The fundamental role of pioneer factors in development has been affirmed by their ability to drive zygotic gene activation [6, 11, 25, 26, 30] (Figure 1C). First characterized in Drosophila, the pioneer factor Zelda initiates zygotic gene transcription in the early embryo [45, 46]. In mouse and human, Double homeobox (DUX) family factors, which can target closed chromatin [47], have been identified as being required for the zygotic genome activation, since their zygotic depletion leads to defective genome zygotic activation and impairs embryonic development [48]. In Zebrafish as well, the POU Domain, class 5, transcription factor 3 (Pou5f3), the SRY-box transcription factor 19b (Sox19b), together act as pioneer factors with homeobox protein NANOG (Nanog) to activate zygotic gene transcription. Triple zygotic mutants are deficient in zygotic gene expression and fail to induce embryonic development [26]. Upon fertilization, the zygotic genome harbors closed and inactive chromatin. Pioneer factors, such Nuclear receptor subfamily 5 group A (Nr5a2) in murine zygotic cells [6], target closed chromatin and initiating its opening, which results in zygotic gene expression during early developmental stage [45, 48].

While the distinction between canonical transcription factors and pioneer factors can be considered to be a continuum [49], pioneer factors are defined by specific properties that distinguish them from other transcription factors [50–52]. Whereas canonical transcription factors have weak affinity for closed chromatin regions and their binding on DNA target site is impeded by nucleosomes [51–53], pioneer factors present a higher affinity for closed chromatin [50, 51, 53–56] and weak dissociation from nucleosomal DNA [51, 53, 57]. Therefore, pioneer factors are defined by their ability to bind their target DNA site on nucleosomes in vitro [25, 36, 52, 53, 58–72] and in vivo [33, 73–75]. Moreover, recent biophysical and structural studies highlight details by which pioneer factors can differentially interact with nucleosome core particles [51, 59, 63, 65, 67, 72, 76], whereas canonical transcription factors do it poorly or undetectably [51]. Such structural studies are relatively new, so we can anticipate that additional mechanisms of nucleosome interactions will be revealed as more structural studies are performed [77]. Pioneer factors are also characterized by their ability, upon binding in vitro, to perturb the structure of an underlying nucleosome in an ATPase-independent manner [51, 58, 59, 63, 64, 67, 71, 73], though full opening of chromatin sites targeted by the factors in vivo involves ATP-dependent remodelers [32, 34, 78, 79]. In this review, we describe the latest studies on characteristics of pioneer factors and how they regulate cell fate, and then present mechanisms of regulation of pioneer factors during development.

Pioneer factors scan and bind locally closed chromatin

Investigations into the underlying mechanisms of action reveal that the act of “pioneering” occurs through multiple steps: chromatin scanning, nucleosomal DNA targeting, and local chromatin reorganization. First, pioneer factors need to identify their specific DNA binding sites by scanning the chromatin fiber. While they can freely diffuse in the nucleus, they also exhibit many associations with nucleosomal chromatin to reach their DNA binding site [80, 81] (Figure 2A). Single molecule tracking microscopy assays showed that pioneer factors have a short residence time on closed chromatin while non-pioneers seem to avoid it altogether [50, 54–56, 80]. Nonspecific electrostatic interactions with the phosphodiester backbone, as it is exposed on the nucleosome surface, allows the pioneer factors to scan closed chromatin loci across the nucleus [50, 55, 74]. While their on-off behavior and local proximity to DNA mass would allow them to scan nucleosomal DNA via one-dimensional diffusion [80], a recent biophysical study of the GAGA pioneer factor (GAF) pioneer factor indicates that nucleosome scanning is more of a three-dimensional process [56]. Detailed single-molecule-tracking studies comparing the Forkhead box A1 (FOXA1) and SRY-box transcription factor (SOX2) pioneer factors for chromatin scanning showed that FOXA1 diffuses slowly in closed chromatin domains, whereas SOX2 diffuses more rapidly in such domains; yet in ectopic expression experiments, both factors similarly target closed chromatin target sites with low-turnover nucleosomes [50]. By contrast, comparably expressed Hepatocyte nuclear factor-4-alpha (HNF4A), a transcription factor that appears not to be a pioneer factor, scans closed chromatin poorly and when it does so, it targets sites with high-turnover nucleosomes [50]. Observing differences for chromatin scanning among the few pioneer factors assessed in this fashion suggests that there will be a variety of mechanisms by which such factors scan chromatin.

Figure 2: Mechanism of action of Pioneer factors to induce silent chromatin opening.

(A) Unlike canonical transcription factor (TF), pioneer factors (PFs) can transiently interact with silent chromatin in order to identify their binding sites. (B) Pioneer factor interacts with nucleosomal DNA containing its DNA motif through its DNA-binding domain (DBD) and its non-DNA-binding domain. The non-DNA-binding domain establishes non-specific electrostatic interactions with nucleosomal DNA and direct contacts with nucleosome core particle, as described by Donovan et al. (2023) [51]. (C) Pioneer factor binding induces local silent chromatin opening by inducing local DNA accessibility and local chromatin decompaction.

While scanning chromatin, pioneer factors can transiently interact with target motifs on nucleosomal DNA [25, 51] (Figure 2B). The DNA-binding domain (DBD) itself is sufficient to bind nucleosomes, as observed for FOXA1 [82] and SOX2 [58], but remains to be confirmed for additional pioneer factors. DBD deletion of Octamer-binding transcription factor 4 (OCT4) abolishes binding [69], confirming its fundamental role in nucleosomal DNA binding. DNA-binding domains of pioneer factors typically fold into an α-helix or short helical twist structure, allowing nucleosome invasion and DNA anchoring [60]. Pioneer factors preferentially bind their target DNA motif when it localizes at nucleosome entry/exit site or at DNA groove exposed on nucleosome surface [32, 52, 59, 65, 67, 71, 72, 83].

Mutations of pioneer factors, outside the DNA-binding domain, can selectively impact nucleosome binding [10, 84–86] and the ability to drive cell fate [76]. A striking example is provided by the deletion of four amino acids within linker region between two DNA-binding domains of the OCT4 that ablates nucleosome binding and OCT4’s reprogramming function, without affecting how the factor binds free DNA [69]. Recent cross-linking studies have showed that non-DNA-binding domains can be in close proximity to the nucleosome surface and establish pioneer factor-nucleosome interactions [51, 59, 67, 69, 72, 76]. Different pioneer factors use different forms of interaction with nucleosome. A pioneer factor residue can interact with specific residues of the histone core or N-terminal extensions (tails) [59, 76], such as for Cbf1, where its E253 residue interacts with K75 residue of H2A histone [51] (Figure 2B). Pioneer factors can also establish electrostatic interactions with the nucleosome surface [67]. OCT4 has a small acidic patch with which it can interact with positively charged lysine and arginine residues of histones [67, 87]. The various emerging mechanisms illustrate how we are only beginning to understand how pioneer factors interact with nucleosomes.

Pioneer factors can open locally closed chromatin

After binding to closed chromatin, pioneer factors can induce local chromatin decompaction. Recent ChIP-nexus experiments on Drosophila embryos demonstrate that the Zelda pioneer factor initiates and drives chromatin opening, while other transcription factors tested only increase chromatin accessibility that was already initiated by pioneer factors [88]. How does this initiate? Pioneer factor binding to nucleosomes often leads to DNA accessibility, making the nucleosomal DNA sensitive to nucleases [58, 60, 65, 71, 89] or accessible to other proteins [58, 89–91]. Incubating FOXA1 [92], PU.1 [61], or OCT4 [71] with a linker histone (H1)-compacted nucleosome array induces DNAseI cleavage underlying its binding site, in the absence of an ATP-dependent nucleosome remodeler. Pioneer factors elicit the initial step of chromatin opening by inducing DNA distortions [51, 74, 87] and limited DNA unwrapping of the nucleosome [51, 63, 65, 72, 87] (Figure 2C). The binding of OCT4 on nucleosome unwraps 25 bp of nucleosomal DNA [71]. Pioneer factors alone can also induce nucleosome sliding [51, 74, 87] and histone tail reposition [58]. Upon its binding, OCT4 and H4/H2B histone tails are in close proximity, leading to conformation changes and relocalization of histone tails [67, 71]. This could decrease inter-nucleosome interactions and nucleosome stacking, mechanistically explaining local chromatin opening by the factors [74] (Figure 2C). Finally, due to their ability to bind DNA at nucleosome entry/exit site, or near the nucleosome dyad axis, pioneer factors can displace and compete with histone linker H1 [59, 71, 89] (Figure 2C). Indeed, Abbreviated LFY (Leafy) pioneer factor has structure similarity with H1 [73]. We could assume that this competition mechanism could be shared by several pioneer factors, since many bind the nucleosome entry/exit site [6, 32, 51, 59, 65, 67, 69, 71].

Pioneer factors enable ATP-dependent chromatin reorganization and cis-regulatory element activation

After local chromatin decompaction, additional protein machineries are required to promote complete opening of closed chromatin regions. Foundational in vitro studies using chromatin assembled with Drosophila embryo extracts showed that pioneer factor GAF’s nucleosome disruption depends on ATP hydrolysis via the Nucleosome remodeling factor (NURF) complex [93], indicating that complete chromatin remodeling is an energy dependent pathway. In a recent in vitro study using fully purified, H1-compacted nucleosome arrays, purified PU.1 pioneer factor and purified canonical BRG1/BRM-associated factor (cBAF) Switch/sucrose nonfermenting (SWI/SNF) nucleosome remodeling complex showed that PU.1 alone could initiate DNA hypersensitivity underlying its binding site, which was enabled by an unstructured domain of PU.1. The initially open domain was expanded by the cBAF complex only in presence of ATP [61] (Figure 3A). In addition, incubation of cBAF complex with a PU.1 mutant deficient in its recruitment of cBAF, but which still harbors the PU.1 unstructured domain and a robust local chromatin opening activity, did not allow cBAF action [61]. The result argues against a model whereby the cBAF remodeler scans chromatin for a partially open site and argues for a model where the pioneer factor must elicit local chromatin opening and simultaneously recruit or stabilize the remodeler at the site.

Figure 3: Pioneer factors drive closed chromatin reorganization and gene expression activation.

(A) In vitro, PU.1 is able to recruit the cBAF SWI/SNF nucleosome remodeling complex on H1-compacted nucleosome arrays to expand local DNA accessibility, as described by Frederick et al. (2023) [61]. (B) In vivo, a pioneer factor (PF) leads to chromatin reorganization and thus enhancer activation. Pioneer factor recruits SWI/SNF nucleosome remodeling complexes to promote nucleosome eviction and chromatin opening. Pioneer factors promote deposition of active chromatin on enhancer by recruiting MLL3/4 and p300 enzymes.

During cell differentiation, pioneer factors directly recruit SWI/SNF nucleosome remodeling complexes by interacting with different subunits of SWI/SNF complexes [94, 95], such as Brahma-related gene-1 (BRG1) [16, 34, 36, 78, 79, 96] or BAF [9]. Thus, pioneer factors guide SWI/SNF-dependent remodeler on specific silent loci which are important for cell differentiation. After their recruitment on chromatin, SWI/SNF nucleosome remodeling complexes evict nucleosomes [34, 78] and promote chromatin rearrangement, leading to an open chromatin conformation [9, 36, 78, 94, 96, 97] (Figure 3B). While pioneer factors recruit and stabilize SWI/SNF complexes on chromatin [34, 36, 61, 78, 94, 98], the open domains created by SWI/SNF complexes can subsequently stabilize pioneer factors on chromatin [34, 78, 79]. Indeed, inhibition of BRG1 led to a reduction of PU.1 binding [79]. The recent results suggest a positive feedback loop between pioneer factors and SWI/SNF nucleosome remodeling complexes that can maintain an open chromatin conformation on cis-regulatory elements.

After initial chromatin opening, additional transcription factors and other protein complexes bind the newly accessible region [97, 99]. Pioneer factor binding correlates with a reduction of cytosine methylation of 5-methylCpG residues [86, 100–103]; CpG methylation (5mC) normally represses enhancers [104]. Indeed pioneer factors can recruit Ten-eleven translocation (TET) enzymes to enhancers [101, 105], leading to conversion of 5mC into 5-hydroxymethylcytosine (5hmC) [105, 106]. Pioneer factors also establish an active chromatin state on enhancers by recruiting Mixed Lineage Leukemia-3 and -4 (MLL3/4) [107, 108] and p300 [16, 26, 47, 109] enzymes, which deposit H3K4me1 and H3K27ac on enhancers, respectively (Figure 3B). FOXA1 can recruit TET enzymes and reduce 5mC on enhancers [106, 110], while also recruiting MLL3 and promoting H3K4me1 deposition [108]; both of which increase enhancer activity and gene expression. Finally, pioneer factors mediate enhancer-promoter proximity [33, 94, 111] and thus stimulate target lineage-specific gene expression. T cell factor 1 (TCF1) regulates CCCTC-binding factor (CTCF) on enhancers [112]. A recent mechanism based on protein aggregation has been identified [113–115]. Indeed OCT4 [114] and Nanog [115] can form protein aggregates to favor enhancer-promoter communications and gene activation.

Pioneer factors can also result in further closing of chromatin and gene silencing

To direct cell differentiation, pioneer factors also elicit silencing of other lineage specific-genes [19, 21, 23, 33, 95, 100]. They can indirectly inactivate gene expression by inducing passive enhancer inactivation. That is, pioneer factor binding on chromatin leads to transcription factors and protein machinery relocalization on the enhancer repertoire; thus away from active enhancers [17, 116, 117]. For example, induction of SRY-box transcription factor 9 (SOX9), a pioneer factor driving the hair follicle fate, in epidermal stem cells indirectly silences epidermal related genes by interacting with SWI/SNF remodeling complexes and MLL4/3, and recruiting them away from active epidermal enhancers [95] (Figure 4A). Pioneer factors can also directly inactivate genes by establishing silent chromatin on enhancers and promoters, which can include chromatin closure. For example, FOXA1, which when it recruits the Groucho homolog (Grg) transcriptional repressor to naïve chromatin, can induce local chromatin compaction and impairs binding of other transcription factors [118] (Figure 4B). Pioneer factors can also drive chromatin closure by recruiting histone deacetylases, such as Nucleosome remodeling and deacetylase (NuRD) complex, which causes erasure of the H3K27ac active histone mark and loss of chromatin accessibility on enhancers [119, 120] (Figure 4C). A recent study showed that the GAF pioneer factor can elicit H3K9me3 deposition on DNA repeats and transcriptional silencing during Drosophila zygotic genome activation [121]. Pioneer factors interact with diverse repressive complexes [33, 120]. Some pioneer factors, such as PU.1, can recruit and stimulate the PRC2 complex, inducing H3K27me3-marked heterochromatin deposition on lineage-specific enhancers and promoters [23, 33, 100, 119] (Figure 4C). To summarize, pioneering occurs when the transcription factor targets a naïve domain of chromatin and, via partners, elicits a new functional capacity to the domain: either open and active or closed and further silenced.

Figure 4: Pioneer factors induce gene silencing.

(A) Pioneer factors induce passive enhancer inactivation by relocalizing protein complexes on chromatin. As described by Yang et al. (2023) [95], SOX9 recruits MLL3/4 and SWI/SNF nucleosome remodeling complex away from active epidermal enhancers during a hair follicle cell fate transition. (B) In vitro, FOXA1 increases local compaction of closed chromatin by recruiting Grg transcriptional corepressor, as described by Sekiya et al. (2007) [118]. (C) In vivo, pioneer factors directly induces enhancer inactivation. It erases active chromatin and induces chromatin closure on enhancers by recruiting histone deacetylase complex (NuRD). Pioneer factors also establish H3K27me3-marked heterochromatin by recruiting and stimulating the PRC2 complex.

Pioneer factors maintain cell identity

During mitosis, chromatin is massively condensed into mitotic chromosomes, many DNA-binding proteins are evicted from chromosomes, and transcription is diminished, but not completely silent [122]. After mitosis exit, cells reactivate lineage-specific gene expression to restore their identity. Live imaging, using fluorescent protein-fused pioneer factors, showed that at least a subset of pioneer factors, such as FOXA1, PAX3, SOX2 and GATA binding protein 2 (GATA2), stay associated with mitotic chromosomes [123–125]. Their mitotic chromosome binding ability is related to their affinity for closed chromatin, since it depends on their DNA-binding domains and their ability to establish non-specific electrostatic interactions with closed chromatin [123–127]. SOX2 degradation specifically in mitosis leads to a loss of pluripotency in mouse embryonic stem cells [127]. Thus, pioneer factors have an important role in the maintenance of cell identity.

Modulating pioneer factor activity

During tissue specification, pioneer factors have dynamic binding and action on silent chromatin to progressively guide lineage specific gene expression [7, 10, 17, 19–21, 24, 100, 117, 128, 129]. Pioneer factors do not generate “peaks” in binding assays at all of their potential DNA binding sites in a given cell [20, 79, 83]. Yet, examining subthreshold peaks reveals extensive binding at alternative cell sites, called “sampling” [20, 50]. Moreover, DNA motif enrichment analyses have shown that pioneer factors can bind chromatin regions weakly or not enriched for their DNA binding motif [9, 21, 130]. Also, pioneer factors bind closed chromatin regions without necessarily inducing DNA accessibility [36, 90, 107]. These observations indicate that pioneer factors are modulated for their stable binding events and chromatin opening activity.

What are ways that pioneer factors can be modulated during chromatin binding? Pioneer factor binding on nucleosomal DNA depends on nucleotide sequence and the position and orientation of its target DNA motif on the nucleosome surface [52, 64, 83, 87, 131, 132]. A target DNA motif to close or facing the nucleosome surface could hinder factor binding, due to steric clashes [64, 72, 87, 131]. A recent in-vitro study using a basic helix-loop-helix (bHLH) pioneer factor binding motif tiled at one base pair intervals across the DNA template shown that bHLH factors preferentially bind nucleosomes when binding motif is exposed at the +/−7 to +/−5 superhelical region on nucleosome surface [72], which is in agreement with an in vitro binding selection study for bHLH proteins [52]. To limit steric clashes, pioneer factors are able to bind partial or degenerate DNA motifs [65, 75], due to flexibility of their DNA binding domain [68]. Given the emerging structural information showing that pioneer factors interact with the core histones within a bound nucleosome, binding to only a partial motif may be compensated by histone interactions that stabilize binding.

During development, pioneer factors contribute to binding of other pioneer and/or transcription factors [88] leading to their cooperative binding on nucleosomal DNA [24, 38, 65, 131]. This may be elicited by a first bound pioneer factor to distort nucleosomal DNA leading to DNA motifs being accessible for other pioneer [58, 64, 72, 87, 131] or non-pioneer [107, 133] factor binding (Figure 5A). An increase in pioneer factor concentration correlates with an increase in nucleosome binding [25, 51, 80] and with binding of new genomic regions during development [10, 26, 134]. More recently, pioneer factors have been identified as able to form aggregates on chromatin [81, 115, 135–139]. The mechanism may ensure local increase of pioneer factor concentration and their binding events [81]. It may also promote cooperation between pioneer factors, since several pioneer factors may colocalize on a same aggregate [136]. However, how pioneer factors form aggregates is poorly understood, since deletion of their unstructured domains, known to promote protein aggregate formation, does not always destabilize pioneer factor aggregates [135, 136].

Figure 5: Pioneer binding and opening activity are modulated.

(A) Cooperative binding. A first pioneer factor binds and distorts nucleosomal DNA, leading to exposure of a second DNA motif and thus binding of a second factor. (B) Pioneer factor binding and opening activity depend on pioneer factor cofactors. As described by Zhang et al. (2019) [143], SOX2 interacts with OCT4 in embryonic stem cells, then loss its interaction with OCT4 to gain interaction with PAX6 during neural fate transition leading to SOX2 genome wide relocalization. (C) Poised pioneer factor binds closed chromatin regions but does not induce its local opening. Closed chromatin opening may occur later during development. (D) Histone modifications might favor or impede pioneer factor binding/local opening activity.

Pioneer factor function on silent chromatin can be regulated by pioneer factor’s partners. For example Tripartite motif containing 24 protein (Trim24), a p53’s cofactor, prevents p53 opening activity on closed chromatin regions [140]. In addition, during cell differentiation, pioneer factors loss of certain partners and acquisition of new partners can lead to relocalization of pioneer factor genomic occupancy and modulation of pioneer factor chromatin opening activity [129, 141]. Proteomic analysis of SOX2 between embryonic stem cells and neuronal cells shown cell-specific interactors [128, 142]. In embryonic stem cells, SOX2 interacts with OCT4. During embryonic stem cell differentiation into neural progenitor, SOX2 loss interaction with OCT4 and progressively gain interaction with a new pattern: Paired box 6 (PAX6) [143] (Figure 5B). Thus, SOX2 targets and regulates different set of lineage-specific genes during cell differentiation [143].

A ‘poised’ state of enhancers has been identified during cell differentiation, where the enhancers are bound by a pioneer factor but still harbor a closed conformation [19, 107, 144–146] (Figure 5C). They become open, and thus activated, later during development [7]; such enabling of developmental competence characterized the initial pioneer factor description for FOXA1 in undifferentiated mouse endoderm [147]. Interestingly, in cells, ectopically expressed pioneer factor can bind chromatin within 0.5–24 hours, but overt chromatin remodeling and opening may be delayed by one to five days [90]. The delay could be due to the time for subsequent recruitment of a nucleosome remodeler [61, 129], a requirement for additional developmental signals [24, 146], or loss of binding of a local inhibitor which then allows chromatin opening [140].

Histone post-translational modifications might favor [67, 148] or impede [21, 149] pioneer factor binding or local opening action by inducing steric clashes [64], altering electrostatic histone surface charges [149], de/stabilizing pioneer factor-histone interactions or DNA-histone interactions [67]. Cryo-EM structural studies using H3K27acetylated nucleosomes showed that the acetylation induces DNA sliding, increases DNA binding site exposition, and thus modulate pioneer factor binding [67] (Figure 5D). Moreover, peptide microarray assays have showed that some pioneer factors are unable to interact with specific modified histone tails [21, 149]. For example, H3K27ac histone modification prevents Myb proto-oncogene protein (c-Myb) binding on the histone tail and thus poorly colocalizes on chromatin [149] (Figure 5D). This suggests that H3K27ac might protect active enhancers from aberrant c-Myb binding or ensure c-Myb detachment and recycling from chromatin after enhancer activation. Thus, some histone modifications, and protein complexes that bind them, might repulse pioneer factors from chromatin to protect cis-regulatory elements from their action at specific stages of development.

Given all of the parameters that could affect pioneer factor binding, it may be no surprise that pioneer factors harbor diverse affinities and mechanism of actions on the three states of silent chromatins: naïve chromatin, H3K9me3-marked and H3K27me3-marked heterochromatin. First, pioneer factors seem to preferentially target naïve chromatin regions [38], since most of pioneer factor bound regions which have a closed conformation but contain low or no H3K9me3 and H3K27me3 repressive histone marks [17, 20, 107] (Figure 1B). Moreover, H3K9me3-heterochromatin had been characterized as a barrier to reprogramming [38, 150]. Yet, recent epigenomic studies found that certain pioneer factors, such as Paired box 7 (PAX7) or TCF1, target regions covered by H3K27me3 [22–24, 140] or H3K9me3 marks [21, 22, 90, 140] (Figure 1B) and then induce underlying DNA sequence activation [22–24, 140]. However, mechanisms of engagement of pioneer factors with these heterochromatin regions are poorly understood. During pituitary lineage differentiation, PAX7 has a weaker enrichment on heterochromatin regions than on naïve chromatin regions. Moreover, PAX7 induces activation of enhancers covered by H3K9me2-heterochromatin, but not by H3K9me3-heterochromatin [90], suggesting different mechanisms of regulation between the diverse types of silent chromatin. It may be that H3K9me3- and H3K27me3-heterochromatin have to be destabilized by pioneer factor-independent mechanisms, in order to facilitate pioneer factor binding and action on a more naïve-like chromatin structure during cell differentiation.

Finally recent studies revealed consequential post-translational modifications on the pioneer factors themselves. The modifications might modulate their scanning, binding, or local opening activities [60, 151–156]. For example, acetylation of PAX7 modulates its binding on chromatin and thus gene expression profile [154], while its methylation is required for its mitotic chromosome binding activity [156]. Identification of post-translational modification on pioneer factors, the associated modifying enzymes, and their consequences on pioneer factor function is a key point for the future. Such studies will allow a better understanding pioneer factor roles during development and diseases such as cancer, where pioneer factors might be mutated on potential post-translational modified residues or where modifying enzymes might be dysregulated or mutated.

Concluding remarks:

The nucleosome-targeting feature of pioneer factors enables them to drive development by targeting naïve and heterochromatic domains to elicit chromatin opening or further closing and promote subsequent cell fate transitions. RNA-binding domains have been recently identified on pioneer factors [157], which impact DNA binding; however, their role in scanning, binding and local opening of silent chromatin is unknown. In general, non-DNA-binding domains, including apparently unstructured domains that are critical for local chromatin opening by the factors [61], need to be characterized to understand their role in pioneer factor function (see Outstanding questions). Recent studies identified additional pioneer factor co-factors and post-translational modifications, but underlying mechanisms need to be better investigated to fully understand the dynamic role of pioneer factors in terms of mechanisms and roles in cell differentiation. Finally, pioneer factors are able to target diverse types of silent chromatin [21–24, 90]. However, it is unclear how pioneer factors recognize and reorganize H3K9me3- or H3K27me3-marked-heterochromatin. Addressing these questions in the future is key to fully understand pioneer factor function on silent chromatin during development and in disease (see Outstanding questions).

Outstanding Question Box:

• What are different modes of interaction with nucleosomes for pioneer factors? What are mechanistic consequences of pioneer factor-nucleosome interactions on nucleosome-wrapped DNA and local chromatin decompaction? How do histone variants and histone modifications modulate pioneer factor-nucleosome interactions and chromatin opening by pioneer factors?

• What is the role of non-DNA-binding domains, especially intrinsically disordered regions, in pioneer factor scanning, binding, and opening of chromatin? How do post-translational modifications of pioneer factors modulate the activities?

• What are the developmental signals and their molecular mechanisms that activate enhancers that are ‘poised’ by pioneer factors?

• How do pioneer factors distinguish diverse type of silent chromatin? How do pioneer factors recognize, destabilize, and reorganize H3K9me3- and H3K27me3-heterchromatin regions?

Highlights:

• Pioneer factors can be master regulators of development. They are responsible for zygotic genome activation after fertilization and drive cell fate transitions during development.

• Pioneer factors bind nucleosomal DNA containing their DNA binding motifs. The binding is mediated by the factor’s DNA-binding domain and via establishment of histone-pioneer factor interactions.

• Pioneer factors initiate local chromatin decompaction by partially unwrapping DNA from nucleosomes and reducing local nucleosome interactions Pioneer factors recruit nucleosome remodeling complexes and chromatin modifying enzymes to promote silent chromatin reorganization, deposition of active chromatin modifications, and thus activation of underlying promoter and enhancer DNA sequences.

• Pioneer factors can vary with regard to their ability to target diverse types of silent chromatin during cell differentiation.

Glossary:

Nucleosome

The nucleosome core particle is the basic repeating unit of chromatin. A nucleosome consists of about 150 bp of DNA sequence wrapped around an octamer of the four core histones: H2A, H2B, H3 and H4. Nucleosomes limit DNA sequence accessibility, and what distinguishes pioneer factors is their ability to target nucleosomal DNA.

Transcription factor

A transcription factor is a protein able to bind a specific DNA motif on free DNA, through its DNA-binding domain, in order to regulate gene expression.

Pioneer factor

A pioneer factor is a subclass of transcription factor able to bind a sequence motif on free DNA and when its motif DNA is exposed on a nucleosome, though the latter may be limited by the position of the motif on the nucleosome. Recent studies reveal that pioneer factors interact with DNA and histones within the nucleosome core. Pioneer factors induce local nucleosomal accessibility in chromatin to promote binding of other transcription factors or protein complexes that can either open the chromatin further or lead to its stronger compaction.

Euchromatin

Euchromatin is an active chromatin conformation of genome where genes are expressed and cis-regulatory elements are active. Active promoters and enhancers reside in euchromatin and typically have central regions that are depleted for nucleosomes, keeping the DNA accessible to transcription factors and gene regulatory protein complexes. They are also flanked by nucleosomes enriched for specific histone modifications, such as H3K27ac, H3K4me3 and H3K4me1.

Naive chromatin

Naive chromatin harbors a closed conformation and is transcriptionally silent. It enriches in histone linker H1 but does not harbor, or low enrichment in, specific histone modifications.

H3K9me3-heterochromatin

H3K9me3-heterochromatin is a silent chromatin enriched in H3K9me3 and Heterochromatin protein 1 (HP1). HP1 induces H3K9me3-heterochromatin compaction. H3K9me3-heterochromatin is established by Histone methyltransferases (HMTs) Suppressor of variegation 3–9 homologs-1/2 (Suv39h-1/2) and SET domain bifurcated 1 (SETDB1) on DNA repeat sequences and lineage specific-genes and enhancers.

H3K27me3-heterochromatin

H3K27me3-heterochromatin is a silent chromatin enriched in H3K27me3 and H2AK119ub mediated by Polycomb Repressive Complex 2 (PRC2) and Polycomb Repressive Complex 1 (PRC1) complexes, respectively. H3K27me3-heterochromatin represses lineage-specific genes to ensure cell differentiation.

Zygotic gene activation

Zygotic gene activation is the first transcriptional expression of embryonic genes after egg fertilization.