(C) PLOS One
This story was originally published by PLOS One and is unaltered.
 . . . . . . . . . .



ERK1/2 signalling dynamics promote neural differentiation by regulating chromatin accessibility and the polycomb repressive complex [1]

['Claudia I. Semprich', 'Division Of Cell', 'Developmental Biology', 'School Of Life Sciences', 'University Of Dundee', 'Scotland', 'United Kingdom', 'Lindsay Davidson', 'Adriana Amorim Torres', 'Harshil Patel']

Date: 2022-12 

Fibroblast growth factor (FGF) is a neural inducer in many vertebrate embryos, but how it regulates chromatin organization to coordinate the activation of neural genes is unclear. Moreover, for differentiation to progress, FGF signalling must decline. Why these signalling dynamics are required has not been determined. Here, we show that dephosphorylation of the FGF effector kinase ERK1/2 rapidly increases chromatin accessibility at neural genes in mouse embryos, and, using ATAC-seq in human embryonic stem cell derived spinal cord precursors, we demonstrate that this occurs genome-wide across neural genes. Importantly, ERK1/2 inhibition induces precocious neural gene transcription, and this involves dissociation of the polycomb repressive complex from key gene loci. This takes place independently of subsequent loss of the repressive histone mark H3K27me3 and transcriptional onset. Transient ERK1/2 inhibition is sufficient for the dissociation of the repressive complex, and this is not reversed on resumption of ERK1/2 signalling. Moreover, genomic footprinting of sites identified by ATAC-seq together with ChIP-seq for polycomb protein Ring1B revealed that ERK1/2 inhibition promotes the occupancy of neural transcription factors (TFs) at non-polycomb as well as polycomb associated sites. Together, these findings indicate that ERK1/2 signalling decline promotes global changes in chromatin accessibility and TF binding at neural genes by directing polycomb and other regulators and appears to serve as a gating mechanism that provides directionality to the process of differentiation.

Funding: CIS was supported by a Biotechnology and Biological Sciences Research Council studentship (1252451) https://bbsrc.ukri.org/ and AAT was supported by a CAPES studentship from the Brazilian Government. KGS is a Wellcome Trust Investigator (WT102817AIA) https://wellcome.ac.uk/ . JB, VM and HP are supported by the Francis Crick Institute which receives its core funding from Cancer Research UK (FC001051), https://www.cancerresearchuk.org/ ; the UK Medical Research Council (FC001051), https://mrc.ukri.org/ and the Wellcome Trust (FC001051), https://wellcome.ac.uk/ . VM is supported by a Sir Henry Dale Fellowship jointly funded by the Wellcome Trust and the Royal Society (Grant Number 218536/Z/19/Z), https://wellcome.ac.uk/ . The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Here, we investigate the role of ERK1/2 signalling in the engagement of a spinal cord neural differentiation programme in the caudal mouse embryo and in human spinal cord neural progenitors generated in vitro. We find inhibition of ERK1/2 activity rapidly increases chromatin accessibility across an exemplar neural progenitor gene, Pax6, and establish that this gene is a polycomb target in vivo. We confirm that occupancy of PRC proteins at the PAX6 locus also takes place in human NMP-like cells and show that ERK1/2 inhibition promotes precocious PAX6 transcription. The transition to a PAX6-positive state coincides with loss of PRC occupancy, while ERK2 association with chromatin remains unchanged. Moreover, genome-wide analysis of chromatin accessibility (ATAC-seq) and polycomb occupancy (ChIP-seq for Ring1B) revealed that the decline in ERK1/2 activity promotes increased chromatin accessibility across thousands of neural genes, including a subset targeted by Ring1B in NMPs. Using genomic footprinting [ 85 ], we identify the most likely candidate factors engaged in the genome-wide response to ERK1/2 signalling decline, uncovering the succession of genome-wide, molecular steps that operate downstream of ERK1/2 signalling decline to facilitate and coordinate the onset of neural gene expression.

The PRC1 core protein Ring1B is required for gastrulation in the mouse embryo [ 74 ] (and see [ 75 ]), and conditional Ring1B loss in motor neuron precursors has demonstrated its further requirement for maintenance of neuronal subtype identities [ 76 ], while roles of H2AK119Ub-dependent and H2AK119Ub-independent PRC1 activity have been identified in the developing cortex at early and late stages, respectively [ 77 ]. Ring1B is also present in both canonical and noncanonical PRC1 complexes [ 78 ]. Proteomic analyses further show dynamic changes in PRC complexes and subcomplexes during mESC neural differentiation [ 46 ] that prefigure a further and later role for PRC1 in augmenting enhancer–promoter interactions during gene activation [ 66 , 79 , 80 ]. Intriguingly, ERK1/2 inhibition in mES cells, while driving cells into a naïve pluripotent cell state, leads to reduction in PRC2 at developmental genes [ 81 – 83 ]. This suggests a role for ERK1/2 signalling in the recruitment or maintenance of polycomb-mediated repression. There is also evidence that this may involve regulation by ERK1/2 protein association with PRC-occupied chromatin in mES cells [ 82 ], although this appears not to be conserved in human ES cells [ 84 ].

Potential targets for FGF signalling include the polycomb repressive complexes 1 and 2 (PRC1 and 2), which cooperate in the repression of differentiation genes in many contexts ([ 38 – 44 ]; reviewed in [ 45 ]). PRC2 is composed of core subunits Ezh1 or Ezh2, Eed, Suz12, and associated context specific accessory proteins Rbbp4 or Rbbp7, Mtf2, Jarid2, and polycomb-like proteins Pali1/2 (isoforms of Lcor/Lcorl), Epop, and Aebp2 [ 46 – 50 ] and, via Ezh2, catalyses the methylation of histone 3 lysine 27 (H3K27me2/3) [ 46 , 51 – 54 ], which in mES cells is augmented by Jarid2 [ 55 – 59 ]. PRC1 contains the E3 ubiquitin ligase, Ring1A or B that catalyses mono-ubiquitination on lysine 119 of histone 2A (H2AK119Ub) and, in its canonical form, includes (cPRC1), PCGF2 (MEL18) or PCGF4 (BMI1), and a PHC (PHC1,2 or 3) and CBX subunit (CBX2, 4, 6, 7, or 8); PRC1 also has varied noncanonical (ncPRC1) forms (reviewed in [ 60 ]). Early studies established that PRC2 and PRC1 cooperate to recruit the composite complex: methylation of H3K27 provides a docking site for PRC1 [ 61 , 62 ], and, in turn, PRC1-dependent H2A ubiquitylation provides a scaffold for PRC2 binding [ 59 , 63 , 64 ], while ncPRC1 includes PRC2-independent PRC1 recruitment and can be involved in context-specific gene repression or activation [ 60 , 65 , 66 ]. Cooperative PRC2 and PRC1 interactions create polycomb domains characterised by gene repressive histone marks H3K27me3 and H2AK119Ub. PRC1-generated chromatin compaction has been elucidated in mESCs [ 67 – 70 ], and such changes in chromatin organisation are mediated by a range of mechanisms that include nucleosome compaction [ 71 ] that in mammals involves CBX2 [ 72 ], formation of Polycomb Group (PcG) bodies and higher-order subnuclear clustering via PHC2 [ 60 , 73 ].

A clue to the mechanism by which FGF regulates neural differentiation was revealed by analysis of chromatin organisation at key neural progenitor genes in the mouse embryo CLE [ 37 ]. This showed using fluorescent in situ hybridisation (FISH) that inhibition of FGF signalling in embryos decompacts chromatin around neural genes such as Pax6. This increased chromatin accessibility was also found in Raldh2 mutant embryos exposed to FGFR inhibitor in which retinoid deficiency resulted in the failure of Pax6 expression [ 37 ]. This indicated that FGF regulation of chromatin organisation is molecularly distinct from the machinery that drives subsequent gene transcription but left open the question of how FGF signalling modifies chromatin organisation.

As NMP progeny leave the CLE and embark on neural differentiation, they experience loss of the FGF effector kinase ERK1/2/MAPK (hereafter referred to as ERK1/2) activity in both chick and mouse embryos [ 26 , 27 ]. This suggests that ERK1/2 signalling dynamics may regulate the onset of neural differentiation. The in vitro manipulation of embryonic stem (ES) cells has provided some insight into the involvement of ERK1/2 signalling in this process. Rather than promoting differentiation, however, inhibition of ERK1/2 activity in mouse ES cells supports pluripotency [ 28 , 29 ], and exposure to FGF/ERK1/2 signalling appears to be an initial step in differentiation. In agreement with this, FGF is required for neural differentiation of mES cells [ 30 – 32 ], although the timing of this requirement differs between assays [ 33 ]. Importantly, however, only a short critical period of ERK1/2 signalling is required in mES cells for subsequent expression of neural genes [ 31 ], after which FGF inhibition then accelerates neural differentiation [ 31 , 34 , 35 ]. Moreover, in mouse epiblast stem cells (mEpiSCs), which rely on FGF for self-renewal, prolonged FGF signalling abrogates neural differentiation [ 36 ]. Consistent with this in both mEpiSCs and human ES cells (which also depend on FGF for self-renewal), inhibition of FGF/ERK1/2 signalling promotes neural differentiation [ 34 , 36 ]. These findings indicate that temporal control over FGF/ERK1/2 signalling is instrumental in the establishment of neural identity from epiblast cell precursors.

( A ) E8.5 mouse embryo schematic, showing the caudal region and signalling pathways known to regulate neural differentiation. ( B ) Schematic of the hanging drop culture used to treat E8.5 mouse embryos with small molecule MEKi (PD184352) or vehicle control DMSO for 1 hour. ( C and C’ ) Representative western blot of embryo lysates probed with antibodies against total (panERK1/2) and phosphorylated ERK1/2 (dpERK1/2) and LiCOR quantification data (n = 3 independent experiments, error bar = SEM, * p = <0.05) ( S1 Data ). ( D ) Position of fosmid probes flanking the Pax6 locus (interprobe probe distance ca. 45 kb). ( D’ ) Representative examples of fosmid probes in individual NT and CLE nuclei (white dashed line) visualised with DAPI (blue) in each untreated condition, vehicle control (DMSO) and MEKi (in DMSO) treated. ( D” ) Interprobe distance measured in >50 nuclei in NT and CLE in each of three embryos in each condition (n = > 150 nuclei/condition/region, Mann–Whitney test/rank-sum test **** p ≤ 0.0001) ( S1 Data ). ( E ) Schematic showing caudal end explant (full tissue thickness taken) in E8.5 embryo. ( E’ ) Chromatin from three biological replicates, each consisting of 30 pooled explants, were interrogated for H3K27me3 levels at TSSs of Pax6, known PRC target HoxD11 and at control regions (see text) compared to IgG background. Three biological replicates (circle, triangle, and diamond) and average (bar) shown, error bar = SEM ( S1 Data ). CLE, caudal lateral epiblast; FGF, fibroblast growth factor; IgG, immunoglobulin G; MEKi, MEK inhibitor; NT, neural tube; PRC, polycomb repressive complex; RA, retinoic acid; Raldh2, retinaldehyde dehydrogenase 2; RAR, retinoic acid receptor; TSS, transcription start site.

The requirement for transient FGF signalling to promote neural differentiation is particularly evident in the elongating embryonic body axis in which the spinal cord is generated progressively. Here, there is a clear spatial separation of the temporal events of differentiation. FGF acts, along with Wnt signalling, in the caudal lateral epiblast (CLE)/node streak border (which later form the tailbud) to maintain a bipotent cell population known as neuromesodermal progenitors (NMPs), which progressively gives rise to the spinal cord and paraxial mesoderm ( Fig 1A ) [ 11 – 19 ]. Blocking FGF signalling in this cell population accelerates the onset of neural differentiation genes, and ectopic maintenance of FGF inhibits this step [ 20 – 22 ]. During normal development, differentiation onset is promoted by rising retinoid signalling, provided by adjacent paraxial mesoderm, which represses Fgf8, restricting it to the tail end ( Fig 1A ) ([ 21 – 24 ]; reviewed in [ 25 ]). These findings indicate that a decline in FGF signalling promotes neural differentiation; however, little is known about the mechanism(s) by which such signalling dynamics mediate the coordinated activation of neural differentiation genes.

The identities of signals that induce particular cell fates are now well established, but how such signalling regulates chromatin to coordinate the transcription of differentiation genes and engage a differentiation programme are not well understood. Fibroblast growth factor (FGF) signalling has been implicated in the acquisition of neural cell fate in many vertebrate embryos ([ 1 – 7 ]; reviewed in [ 8 ]) (although the timing of involvement varies between species and anteroposterior regions; [ 6 , 9 , 10 ]). Intriguingly, while FGF is required for neural induction in most of these contexts, its decline is also necessary for differentiation progression.

In chick and mouse embryos, decline in FGF and downstream effector kinase ERK1/2 signalling takes place as NMP cells commence neural differentiation, mediated by rising levels of retinoid (RA) signalling, which represses Fgf8 (reviewed in [ 25 ]). Mouse NMPs are characterised by phosphorylated ERK1/2 (dpERK1/2) and compact chromatin across the exemplar neural differentiation gene Pax6, which is decorated with the PRC2 gene silencing mark H3K27me3. In human NMP-L cells, PRC2 (Jarid2) and PRC1 (Ring1B) occupy the locus and H3K27me3 is accompanied by gene activation associated histone modification H3K4me3, identifying this as a bivalent locus poised for transcription. During differentiation, loss of ERK1/2 signalling leads to chromatin decompaction across PAX6 in mouse embryo NMPs and in human NMP-L cells, where it is accompanied by loss of PRCs (PRC2 and PRC1). There is also evidence that levels of PRC2 protein Jarid2 decline during neural differentiation. Loss of ERK1/2 activity results in genome-wide increase in chromatin accessibility across neural genes, involving regulation of PRC and other chromatin accessibility complexes, opening sites that become bound by neural TFs (SOX family), detected prior to PAX6 transcription. The gene silencing histone mark H3K27me3 persists after PRC2 is lost and is removed coincident with transcription onset: ERK1/2 regulation of chromatin accessibility and neural TF binding appears distinct from this later step, which may depend on retinoid signalling, known to be required for PAX6 transcription in mouse and chick embryos. As transient loss of ERK1/2 activity removes PRC2 and PRC1 and these complexes are not reinstated on resumption of ERK1/2 signalling, decline in ERK1/2 activity and so PRC occupancy (and other chromatin remodelling complexes) acts as a gating mechanism that confers differentiation directionality as well as synchronising accessibility for neural TF binding and so engagement of the neural differentiation programme. FGF, fibroblast growth factor; NMP, neuromesodermal progenitor; NMP-L, NMP-like; RA, retinoic acid; PRC, polycomb repressive complex; TF, transcription factor.

To determine the significance of this loss, we assessed whether 12-hour MEKi exposure also altered chromatin accessibility around the PAX6 locus using FISH. Significant decompaction of this region was found in MEKi-treated cells compared with both DMSO and untreated controls ( Fig 6E–6E’ and S10 Data ). These findings correlate PRC loss with a distinct increase in chromatin accessibility across this key neural progenitor gene. This loss of PRC occupancy appears to represent the gating mechanism for the initiation of neural progenitor gene expression since the resumption of ERK1/2 activity (by 12 hours) did not reinstate PRC binding. Moreover, these data demonstrate that removal of H3K27me3 is not required for such chromatin reorganisation; indeed, the regulation of this histone modification appears molecularly distinct from the initial effects of ERK1/2 dephosphorylation and correlates with later transcriptional onset ( Fig 6C–6C” compare Fig 2D–2D” ). These distinct steps in regulation of the exemplar gene PAX6 are summarised in Fig 7 .

( A ) Differentiation protocol used to generate NMP-L (D3) cells and additional treatment with vehicle control DMSO or MEKi for 12 hours. ( B - B’ ) Representative western blot of cell lysates probed with antibodies against total (panERK1/2) and dual-phosphorylated-ERK1/2 (dpERK1/2) and LiCOR quantification data (n = 3 independent experiments, error bar = SEM) ( S10 Data ). ( C - C” ) ChIP-qPCRs investigating the histone modification H3K27me3 and polycomb repressive complex occupancy in NMP-L (D3) cells treated with MEKi or DMSO for 12 hours (n = 3 individual experiments indicated with circles, bar = average, * = p ≤ 0.05, t test) ( S10 Data ). ( D - D” ) Transcription levels of PAX6, HOXD11, and JARID2 assessed by RT-qPCR in undifferentiated cell (hESCs), NMP-L cells untreated, vehicle control (DMSO) treated or MEKi treated (n = 3 individual experiments, no significant differences between samples, t test) ( S10 Data ). ( E - E’ ) FISH to assess chromatin compaction around the PAX6 locus in NMP-Ls untreated and treated with DMSO or MEKi for 12 hours. Two probes flanking the target locus (interprobe distance ca. 51 kb) were hybridised and visualised by differential labelling nuclei (outlined with white dashed line) visualised with DAPI (blue). Interprobe distance measurements in >50 nuclei in the three conditions in three individual experiments (n = > 150 nuclei/cell type, Mann–Whitney test/rank-sum test, *** p ≤ 0.001) ( S10 Data ), this decompaction correlated with a 2.5-fold decrease in number of base pairs per nm compared to both controls (untreated NMP-L (D3): 205 bp/nm, DMSO NMP-L (D3): 218 bp/nm and MEKi NMPL (D3): 87 bp/nm). ChIP-qPCR, chromatin immunoprecipitation quantitative PCR; reverse transcription quantitative PCR; MEKi, MEK inhibitor; NMP-L, NMP-like; RT-qPCR, reverse transcription quantitative PCR.

We next assessed the consequences of loss of ERK1/2 activity at a later time point, exposing NMP-L (D3) cells to MEKi for 12 hours and conducting the same panel of ChIP-qPCR experiments ( Fig 6A ). At the end of this period, ERK1/2 phosphorylation levels had now returned to control levels ( Fig 6B and 6B’ and S10 Data ), indicative of transient MEKi effectiveness in the presence of exogenous FGF, although other consequences of reduced ERK1/2 activity, such as increased PKB phosphorylation, were still apparent ( S6 Fig and S11 Data ): This regime therefore delivered a transient inhibition of ERK1/2 activity for at least 3 hours. Moreover, after 12 hours, reduced Jarid2 and Ring1B occupancy were clearly apparent at PAX6 and HOXD11 loci in exposed cells (while control genomic regions possessed only low level H3K27me3 and Ring1B and Jarid2, which did not change with MEKi treatment) ( Fig 6C–6C’ and S10 Data ). Importantly, none of the gene loci analysed exhibited a change in H3K27me3 levels ( Fig 6C” and S10 Data ) nor did the loss of both PRC1 and PRC2 proteins correlate with transcription of PAX6 or HOXD11 nor with reduced JARID2 transcripts ( Fig 6D–6D” and S10 Data ). These data indicate that a transient reduction in ERK1/2 signalling is sufficient to trigger loss of both PRC2 and PRC1 proteins.

In the mouse embryo, we found that exposure to MEKi led to decompaction of chromatin around the locus of our exemplar neural progenitor gene Pax6, within one hour ( Fig 1 ). However, it was not practical to dissect and collect sufficient caudal embryonic tissue to assess the impact of loss of ERK1/2 signalling on PRC protein occupancy or the timing of such events. We therefore used human NMP-L cells to dissect the dynamics of such changes at the PAX6 locus. Exposure to MEKi for 3 hours (in the presence of FGF, which maintains the NMP-L cell state) reduced dpERK1/2 levels ( S5A–S5B’ Fig and S9 Data ); however, ChIP-qPCR revealed no change in Ring1B nor H3K27me3 occupancy ( S5C–S5E and S9 Data ). Moreover, MEKi exposure did not alter transcription of PAX6, HOXD11, nor JARID2 (S5F-S5F” Fig and S9 Data ). This short exposure was therefore insufficient to disassemble the PRC1 (Ring1B) complex or remove the H3K27me3 mark in the context of NMP-L (D3) cells.

To test whether Ring1B occupancy in NMP-L cells correlates with the binding of specific TFs, we further performed genomic footprinting on the Ring1B-bound subset and compared this with all other neural sites but found no differences in the factors enriched. However, comparison of predicted TF occupancy in MEKi and control conditions revealed that loss of ERK1/2 activity promoted the engagement of SOX family TFs (D5) and TFAP2 family TFs (D6) regardless of association with Ring1B. These findings indicate that while a decline in ERK1/2 signalling promotes increased accessibility at neural gene loci and promotes engagement of neural TFs, only a subset of the newly accessible loci was formerly occupied by Ring1B. Moreover, these data suggest that the activity of individual TF proteins such as SOX2 and SOX3 is subject to multiple regulatory mechanisms at different genomic sites. Overall, these data show that decline in ERK signalling broadly impacts accessibility at neural genes, can regulate PRC occupancy, and selectively promotes neural TF engagement signatures at neural gene loci.

To assess the significance of Ring1B occupancy in NMP-L (D3) cells, we next compared all such sites (D3 consensus sites) to our previously (ATAC-seq defined) neural sites, which exhibit increased accessibility from D3 to D8. This revealed that a subset of neural sites (22%, 1,698/7,877) are occupied by Ring1B in the NMP-L cell population, while GO term analysis revealed enrichment for neural and anterior-posterior patterning associated with these Ring1B bound sites ( Fig 5D and 5D’ ), linking PRC occupancy in NMP-L (D3) cells to neural differentiation. On the other hand, Ring1B occupancy was not predictive of which neural sites would show increased accessibility following exposure to MEKi: Only a subset of neural sites opening in response to MEKi were bound by Ring1B on D5 (258/1,176, 15%, includes PAX6) and D6 (373/1,587, 21.9%, includes PAX6 as well). This suggests that MEKi increases chromatin accessibility at canonical PRC and non-PRC-occupied sites.

To uncover the global relationship between PRC-occupied genes and subsequent sites of increased accessibility, we next undertook ChIP-seq with Ring1B in hESCs and D3 NMP-L cells. Comparison of Ring1B genome-wide occupancy in these two cell states revealed a striking pattern of change, with 1,827 sites detected as differential. Most of these sites (1,148) were decreased in NMP-L cells in comparison with hESCs, with a smaller set (679) constituting newly occupied sites in NMP-L cells. Such change in occupancy pattern further supports the conclusion that reduced Ring1B at the PAX6 locus ( Fig 4D and S5 Data ) is not simply a consequence of overall decline in Ring1B protein levels during differentiation. Moreover, GO analysis of genes associated with these differentially Ring1B occupied sites included many terms associated with spinal cord development and neural differentiation ( Fig 5C ). This establishment of a new pattern of genome wide Ring1B occupancy in the NMP-L (D3) cells that is aligned with spinal cord development supports the idea that the NMP cell state represents competency for differentiation into this tissue and suggests that this is linked to regulation of polycomb occupancy.

To assess how MEKi treatment might advance neural differentiation, we next analysed its impact on the repertoire of transcription factors (TFs) found at open chromatin sites. For this, we carried out genomic footprinting, a computational approach that predicts TF occupancy from ATAC-seq data on the basis of the Tn5 transposase integration frequency detected at base-pair resolution, combined with motif enrichment [ 85 ]. We assessed all accessible regions detected by ATAC-seq at D5 and D6, comparing control versus MEKi-treated conditions. This revealed an enrichment in neural TFs in MEKi conditions, notably SOX factors [ 96 ] at D5, as well as TFAP2A by D6, indicative of a neural crest-like signature [ 97 ]. By contrast CDX/HOX factors, associated with the NMP-L state [ 98 ], were found in control conditions ( Fig 5A and 5B and S8 Data ). These data suggest that MEKi exposure promotes global changes in the binding of TFs that mediate the transition to a neural identity. Moreover, a MEKi-induced increase in SOX factor binding at D5 occurs prior to precocious PAX6 expression on D6, suggesting that neural TF binding is an earlier distinct step that precedes neural gene transcription.

To interpret the genome-wide chromatin accessibility changes induced by MEKi, we analysed the GO biological processes associated with the gene cohorts linked to regions of increased accessibility. Strikingly, MEKi treatment on D5 corresponded to an increased representation of genes associated with neural development terms ( Fig 4F and S5 Data ). Moreover, at D6 of MEKi treatment, genes associated with later neural development, such as synapse organisation, were specifically increased ( Fig 4F and S5 Data ). We therefore analysed the GO terms for genes associated with more accessible regions that were unique to the MEKi data sets on D5 and D6 and found a strong enrichment towards later neuronal differentiation (see S3 Table for list of genes associated with loci of increased accessibility due to MEKi treatment). This suggests that MEKi treatment leads to increased global accessibility at neural differentiation genes and an accelerated progression to a more advanced differentiation cell state, consistent with MEKi-induced precocious PAX6 transcription and loss of Ring1B across this locus shown by ChIP-PCR by D6 ( Fig 4C, 4D, and 4D’ and S5 Data ). Moreover, examining the peaks detected by ATAC-seq at individual genes (DCX, GABRB2, and PAX6) supported these conclusions, showing an increase over time and that this was advanced with MEKi treatment assessed on D5 (arrowheads in Fig 4G–4G” ).

To determine whether changes to ERK1/2 activity broadly impact chromatin accessibility in cells transitioning to a neural identity, we performed ATAC-seq on cells at D5 and D6 exposed to MEKi or DMSO conditions during differentiation from D3 NMP-L cells ( Fig 4A ). A comparison of all regions opening between NMP-L (D3) and D5 in control (2,653 regions) and MEKi conditions (3,977 regions, associated with 417 newly accessible genes, and 368 more accessible genes), identified 2,193 regions with additional chromatin accessibility following MEKi treatment ( Fig 4E and S3 Table ). Many of the MEKi-induced regions of increased accessibility belonged to the previously defined neural sites (1,176) ( Fig 3C’ ), while a further (1,436) regions were unique to MEKi treatment and not included in neural sites ( Fig 4E ). Analysis of D6 control and MEKi datasets revealed a similar pattern, with 3,508 additional regions of increased accessibility induced by MEKi (associated with 593 newly accessible genes and 437 genes acquiring additional accessibility) ( Fig 4E’ ), with many coinciding with neural sites (1,587), and a further 1,921 regions were detected outside of control D6 and D8 data sets ( Fig 4E’ and see below).

A more direct mechanism by which ERK1/2 signalling might regulate PRC occupancy has been proposed in mouse ESCs. Here, ChIP-seq experiments revealed that ERK1/2 activity promotes ERK2 occupancy at developmental genes, in addition to the recruitment of polycomb proteins at such genes in an ERK2-dependent manner [ 82 ]. Using ChIP-qPCR and the same ERK2 antibody used and validated in [ 82 ], we surveyed ERK2 occupancy during differentiation from NMP-L (D3) cells. ERK2 was detected at TSSs of PAX6, HOXD11 and the constitutively expressed control gene GAPDH ( S6A Fig ). Although ERK2 occupancy at PAX6 and HOXD11 declined during differentiation ( S4G Fig and S7 Data ), inhibition of ERK1/2 activity revealed no difference in ERK2 occupancy between treated and control samples ( S4H and S4H’ Fig and S7 Data ). These findings suggest that a decline in ERK1/2 activity, and not its association with chromatin, regulates PRC occupancy at the exemplar neural gene PAX6 in this cellular context.

These findings identify ERK1/2 signalling as a regulator of PRC occupancy and transcription of a key neural gene. This might involve mechanisms that influence the activity of PRC proteins but could be mediated by regulation of PRC protein expression. As Ring1B occupancy is significantly reduced by D6 ( Fig 4D’ and S5 Data ), we assessed levels of Ring1B by western blot during differentiation from hESCs. This revealed a sharp decline as hESCs become NMP-L cells and a further small decrease in Ring1B as NMP-L (D3) cells become NPs (D8) ( S4A and S4B Fig and S7 Data ). This raised the possibility that reduction in global levels of Ring1B contribute to reduced PRC occupancy and that loss of ERK signalling might accelerate differentiation by reducing Ring1B expression. To test whether ERK signalling regulates Ring1B levels, we repeated our initial experiment treating cells with MEKi for 3 days (D3-D6 as in Fig 4A ). This regime reduced pERK1/2 and induced precocious PAX6 expression ( S4C and S4D Fig and for comparison, also see Fig 4C and S7 Data ) but did not alter Ring1B protein levels. This indicates that ERK signalling decline does not impact PRC occupancy by regulating expression or turnover of Ring1B.

( A ) Differentiation protocol generating NMP-L (D3) cells and spinal cord NPs by D8 and additional treatment with vehicle control DMSO or MEKi (PD184352) every 24 hours from D3 for the duration of the differentiation protocol. ( B ) Western blot confirming that exposure to MEKi lead to reduced ERK1/2 phosphorylation after 24 hours in differentiation conditions (D3-D4). ( C ) PAX6 transcript levels in NMP-L (D3), in control conditions (WT and DMSO) or in MEKi-treated cells on D5 and D6 and NP(D8) determined by RT-qPCR (n = 3 independent experiments indicated by circles, triangles, and diamonds, bar = average, error bars = SEM) ( S5 Data ). ( D - D’ ) ChIP-qPCR for Ring1B and H3K27me3 on D6 note H3K27me3 below detection in MEKi conditions (n = 3 independent experiments indicated with circles, bar = average, error bars = SEM, * = p ≤ 0.05, t test, S5 Data ). ( E - E’ ) Venn diagrams representing distribution of genomic regions that become more accessible on D5 (E) and D6 (E’) in control condition and MEKi-treated cells compared to NMP-L (not restricted to regions accessible in NPs (D8)) as well as more accessible regions on D8. ( F - F’ ) GO term analysis of genes associated with more accessible chromatin in D5 (F) and D6 (F’) in control condition and MEKi-treated cells compared to NMP-L ( S5 Data ). ( G - G” ) ATAC-seq peak tracks in NMP-L, D5 DMSO and MEKi, D6 DMSO and MEKi, and NP (D8) cells for the genomic regions of DCX (G), GABRB2 (G’), and PAX6 (G”), grey boxes outline peaks within the gene body, black boxes peaks outside of the gene body and red box genomic region of a known PAX6 enhancer, example comparisons between peaks in MEKi and DMSO conditions indicated with black arrowhead pairs. Note MEKi exposure does not alter levels of Ring1B protein while it promotes neural differentiation ( S4 Fig ). ChIP-qPCR, chromatin immunoprecipitation quantitative PCR; GO, gene ontology; MEKi, MEK inhibitor; NMP-L, NMP-like; NP, xxxx; RT-qPCR, reverse transcription quantitative PCR; WT, wild type.

Gene ontology (GO) term analysis of the genes associated with regions of increased accessibility on D5, D6, and D8 revealed terms related to neural development and showed a progression with more genes associated with these terms over time ( Fig 3E and S3 Data , gene list S2 Table ). Focussing on the GO term Neurogenesis, we selected 2 key neural progenitor genes, in addition to PAX6, and examined the ATAC-seq signal in detail around their gene bodies. Chromatin accessibility around RARB and GLI3 showed an increase similar to that associated with PAX6 as early as D5 ( Fig 3F and 3F’ ). Overall, these findings indicate that coordinated regulation of chromatin accessibility is a general mechanism that mediates the onset of neural gene transcription, while analysis of PRC occupancy dynamics at the exemplar PAX6 locus identifies this repression complex as a target of such regulation.

Genome-wide comparison of accessibility patterns across TSSs and gene bodies in D5, D6, and D8 cells revealed that although overall accessibility levels remained relatively constant over time, a clear peak in accessibility appeared at D6 of differentiation ( Fig 3C ). Detailed analysis comparing D3 and D8 identified 7,877 regions with increased, and 11,603 regions with decreased accessibility (analysis performed with diffReps with thresholds FDR ≤0.01 and log2FC ≥1). The regions of increased accessibility correlated with known active enhancer and promoter marks such as H3K4me1 and H3K27ac detected in human embryonic spinal cord and brain (data from the ENCODE regulatory element database [ 95 ]; see S2 Fig and S4 Data ). We termed these regions “neural sites” and further analysis revealed that these were located in introns (7,003), exons (333), promoter/TSSs (199), TTS (175), and intergenic regions (167), while additional comparison with ENCODE datasets for human neurons and hESCs identified overlaps between neural sites and peaks for the transcriptional coactivator p300, indicating that recruitment of this histone acetyltransferase is a potential mechanism for these changes in chromatin accessibility. Overall, we found that neural sites were associated with 4,001 genes ( S1 Table ) and that accessibility increased across the TSS and body of these genes over time, with accessibility peaking at D6 ( Fig 3C’ and S3 Data ). The changes included additional chromatin opening across genes already accessible at D3 as well as new genes with increased accessibility by D8 ( Fig 3D ).

( A ) ATAC-seq peak tracks in NMP-L (D3), D5, D6, and NP (D8) cells for the genomic regions of PAX6, grey boxes are peaks within the gene body, black boxes, peaks outside of the gene body and red box genomic region of a known PAX6 enhancer. ( B - B” ) ChIP-qPCR for Jarid2, Ring1B, and H3K27me3 on D5, D6, and D7, note by D7 Jarid2 signal falls below detection (n = 3 independent experiments, each indicated by circles), bar = average, error bars = SEM, although there is a declining trend for Jarid2, there are no significant differences between samples, t test ( S3 Data ). ( C - C’ ) Metaprofiles comparing chromatin accessibility in NMP-L (D3) and NP (D8) cells along the gene body genome-wide (C) and focussed on genes associated with increased accessibility in NPs (D8) (neural sites, C’) compared to NMP-L ( S3 Data ). ( D ) Venn diagram representing distribution of genomic regions more accessible in NPs (D8) (neural sites) compared to NMP-L (D3) and their appearance over time (D5 and D6). Comparison of accessible regions and associated genes on D5 and after D5 revealed approximately 19% of genes had additional accessible regions, while the majority were found in newly accessible genes, similarly 40% of D6 genes with additional regions became more accessible in NP (D8). ( E ) GO term analysis of genes associated with neural sites compared to NMP-L (D3) and genes associated with more accessible chromatin regions on D5 and D6 out of the neural sites ( S3 Data ). ( F - F’ ) ATAC-seq peak tracks in NMP-L (D3), D5, D6, and NP (D8) cells for the genomic regions of RARB (E) and GLI3 (E’), grey boxes are peaks within the gene body. ChIP-qPCR, chromatin immunoprecipitation quantitative PCR; GO, gene ontology; NMP-L, NMP-like; NP, xxxx.

While these data indicate that PRC loss precedes neural gene transcription, it remained unclear whether the regulation of chromatin accessibility is specific for PAX6 or reflects a general mechanism that mediates onset of neural gene transcription. To address this, we performed a global analysis of the chromatin accessibility landscape using ATAC-seq [ 94 ], as cells differentiated into NPs (D8) from the NMP-L (D3) state. This involved differentiation of NMP-L (D3) cells (as in Fig 2A ) and sampling cell populations at intervals (D5, D6, and D8). Analysis of the chromatin configuration across the PAX6 locus revealed increased accessibility at multiple sites, some as early as D5 and most prominent by D6 ( Fig 3A ). These early increases in chromatin accessibility prefigure reduction in PRC protein occupancy at the PAX6 locus detected by ChIP-qPCR, which only becomes significant by D8 ( Fig 3B and 3B’ and S3 Data ; and also see Fig 2D and 2D’ ). In contrast to the declining trend in levels of Jarid2 occupancy during differentiation, H3K27me3 levels at the PAX6 TSS and along the gene body remained unchanged between D5 and D7 ( Fig 3B’ ), only dropping later at D8 ( Fig 2D” ). Thus, a reducing trend in PRC occupancy, likely reflecting differentiation progression of some cells, is observed concurrent with increased chromatin accessibility, while loss of H3K27me3 appears to be a later step that coincides with robust PAX6 transcription.

To elucidate how chromatin compaction is regulated at the PAX6 locus, we interrogated PRC occupancy over developmental time using ChIP-qPCR in NMP-L (D3) cells versus NPs (D8). We assessed the occupancy of the core PRC1 protein Ring1B (which is required in the embryonic mouse epiblast that contains NMPs, while Ring1A loss affects only later development) [ 74 ] and for PRC2, Jarid2 (as the latter is highly expressed in the NMP containing caudal lateral epiblast and newly forming neural progenitors in chick [ 93 ] and mouse embryos ( S1 Fig )). Together with H3K27me3, this analysis established that PRC proteins are present at the TSS and along the gene body of PAX6 and at the TSS of known PRC target HOXD11 in NMP-L (D3) cells. By contrast, reduced occupancy was detected at PAX6 and HOXD11 in NPs (D8) (while baseline detection in the gene desert region (Gw10) and the housekeeping gene GAPDH remained unchanged) ( Fig 2D-D ” and S2 Data ). Strikingly, analysis of the active histone modification H3K4me3 revealed no change at the HOXD11 or PAX6 TSSs during neural differentiation ( Fig 2D” and S2 Data ). This suggests that polycomb-mediated repression is a major activity restricting PAX6 gene expression in the NMP-L cells, and not the low levels of H3K4me3. Together, these findings demonstrate that site-specific deposition of PRC at PAX6 is transient and that its removal correlates with chromatin decompaction at this locus, characteristic of the neural progenitor cell state.

( A ) Schematic of differentiation regime used to generate first NMP-like cells (NMP-L (D3)) and then spinal cord progenitors (NPs (D8)) from human ESCs (D1 = Day 1. ( B - B”‘ ) Transcription levels of BRA, SOX2, HOXD11, and PAX6 assessed by RT-qPCR in undifferentiated cells (hESCs), NMP-Ls (D3), and NPs (D8) for PAX6 additionally on day 5 (D5) and day 6 (D6) of differentiation (n = 3 independent experiments, indicated with circles, triangles, and diamonds, error bars = SEM) ( S2 Data ). ( C - C” ) FISH to assess chromatin compaction around the PAX6 locus in NMP-L (D3) and NP (D8). Two probes flanking the target locus (interprobe distance ca. 51 kb) were hybridised and labelling visualised in DAPI-stained nuclei (blue, outlined with white dashed line). Interprobe distance measurements in >50 nuclei in NMP-Ls and NPs in 3 individual experiments (n = > 150 nuclei/cell type, Mann–Whitney test/rank-sum test, *** p ≤ 0.001) ( S2 Data ). ( D - D”‘ ) ChIP-qPCRs investigating polycomb repressive complex occupancy (Ring1B/PRC1 and Jarid2/PRC2) and the histone modifications H3K27me3 and H3K4me3 in NMP-L and NP cells (note, D8 Jarid2 is below IgG control) (n = 3 independent experiments, each indicated by circles), bar = average, error bar = SEM, * p ≤ 0.01, t test ( S2 Data ). ChIP-qPCR, chromatin immunoprecipitation quantitative PCR; ESC, embryonic stem cell; FISH, fluorescent in situ hybridisation; IgG, immunoglobulin G; NMP-L, NMP-like; NP, xxxx; RT-qPCR, reverse transcription quantitative PCR.

To test this, we developed an in vitro system to study neural differentiation during the generation of the spinal cord. We reasoned that the use of human ESCs, with the slower progression to neural progenitor cell identity, would provide a better temporal resolution of mechanisms mediating neural differentiation. To this end, we directed the differentiation of human ESCs towards spinal cord using a protocol established previously [ 91 , 92 ] ( Fig 2A and see Methods ). This involved the generation of NMP-like cells on day 3 (NMP-L D3) (characterised by the coexpression of BRA and SOX2, which are analogous to NMP cells in the mouse CLE) and their differentiation into PAX6 expressing neural progenitors by day 8 (NP D8), which also transcribed the known PRC target HOXD11 ( Fig 2B–2B” and S2 Data ). Detailed analysis of this differentiation revealed low level onset of PAX6 on D6, anticipating the robust expression detected on D8 ( Fig 2B”’ ). This neural differentiation was further slowed by not including exogenous retinoic acid, which acts in part by inhibiting FGF/ERK signalling; this modification allowed us to test the ability of MEK inhibition/loss of ERK1/2 activity to promote neural differentiation ( Fig 2A and see Methods ). Having established this assay, we next asked whether changes in chromatin compaction that we observe around the Pax6 locus in mouse embryo also occur at PAX6 during human neural differentiation in vitro ( Fig 1D ). FISH was carried out in NMP-L (D3) and NPs (D8) cells, and the distances between the PAX6-flanking probes were measured ( Fig 2C–2C” and S2 Data ). This revealed a significant increase in interprobe distance in NPs (D8) compared to NMP-Ls (D3) ( Fig 2C’ and 2C” and S2 Data ). This indicates that the regulation of chromatin compaction is a conserved mechanism operating during neural differentiation that can be recapitulated and examined in vitro.

To determine whether PRCs regulate chromatin compaction around the Pax6 locus in vivo, we carried out chromatin immunoprecipitation (ChIP) for the PRC-mediated histone modification H3K27me3. Mouse embryo microdissection was performed to enrich for the caudal region at E8.5 for ChIP-qPCR ( Fig 1E and see Methods ). The Pax6 transcription start site (TSS) was marked by high levels of H3K27me3 indicative of polycomb activity and consistent with repression of this gene in the CLE ( Fig 1E’ and S1 Data ). Similar levels were also detected at the known PRC target HoxD11 [ 86 – 88 ], which is also not expressed in the mouse CLE at E8.5 [ 89 ]. By contrast, known PRC targets transcribed in the CLE, HoxA7, and HoxB1 [ 90 ] exhibited minimal levels of H3K27me3 at their TSSs. This was similar to levels detected at a control intergenic region and the house-keeping gene GAPDH indicative of a lack of polycomb activity at these loci ( Fig 1E’ and S1 Data ). Together, these data demonstrate that polycomb activity at target genes is selective in vivo and identify Pax6 as a polycomb target in caudal tissue. This raises the question of whether ERK1/2 activity directs polycomb occupancy and chromatin compaction and thus determines the onset of neural gene transcription.

Discussion

In this study, we uncover a molecular mechanism by which the neural inducing signal FGF regulates higher-order chromatin organisation to engage the neural differentiation programme. We demonstrate that loss of the activity of FGF-effector kinase ERK1/2 leads to rapid chromatin decompaction at the neural gene PAX6 in the caudal lateral epiblast of mouse embryos and in analogous hESC-derived NMP-L cells in vitro. Using ATAC-seq, we show that this reflects global remodelling involving increased chromatin accessibility across thousands of neural genes. Focusing on PAX6 as an exemplar neural progenitor gene, we find that ERK1/2 dephosphorylation results in reduced occupancy by polycomb proteins and that this is not explained by changes in PRC protein levels, although PRC2 protein Jarid2 declines during differentiation. Moreover, ERK1/2 inhibition did not alter ERK2 association with chromatin at this locus. We further show that while PAX6 is a bivalent gene poised for expression in NMP-L cells, PRC protein loss following ERK1/2 inhibition is distinct from later actions that mediate removal of the gene silencing mark H3K27me3 and transcriptional onset. Importantly, transient ERK1/2 inhibition triggered increased chromatin accessibility and PRC loss, which was not reinstated on resumption of ERK1/2 signalling, and this suggests that a decrease in ERK1/2 activity serves as a gating mechanism that prompts spinal cord differentiation progression. Furthermore, genomic footprinting of sites identified by ATAC-seq together with ChIP-seq for polycomb protein Ring1B revealed that ERK1/2 inhibition increased neural TF occupancy at non-polycomb as well as polycomb associated sites. Together, these findings suggest a model in which FGF/ERK1/2 signalling promotes or maintains PRCs at a subset of neural gene loci while its decline initiates the neural programme by synchronising removal of this and other chromatin remodelling complexes, globally increasing accessibility for neural TFs, which drive differentiation (summarised for exemplar gene PAX6 in Fig 7).

The ability of ERK1/2 inhibition to elicit chromatin decompaction at the PAX6 locus in both the mouse embryo and human ESC derived NMP-L cells indicates that the molecular mechanism by which ERK1/2 regulates chromatin accessibility is conserved across species. Establishing PAX6 as a polycomb target in both contexts further identified this repressive complex as the likely mediator of chromatin compaction. PRC1-generated compaction has been elucidated in mESCs [67–70], and Pax6 has been identified as a PRC target in this context too [99]. The rapid decompaction of chromatin (within just one hour) following ERK1/2 inhibition in the mouse embryo uncovered here further suggests that ERK1/2 activity may directly maintain PRC occupancy. Exposure of human NMP-L cells to MEKi led to loss of the PRC1/Ring1B and PRC2/Jarid2 by 12 hours at the PAX6 locus, accompanied by local chromatin decompaction, while in MEKi-treated differentiating NMP-L cells Ring1B was lost, Jarid2 was now almost below detection in all conditions. These findings raise the possibility that PRC1 is the primary target of ERK1/2 activity, while low level Jarid2 at chromatin in differentiating NMP-L cells may reflect global decline in PRC2 proteins during neural differentiation, as documented during mouse ESC differentiation into forebrain neural progenitors [46] and suggested by decreasing levels of Jarid2 detected here by western blot. This is consistent with declining chromatin-associated Jarid2 evident in wild-type conditions, where it is enriched on D3 and D5 and falls below immunoglobulin G (IgG) control levels on D7 and D8 (Figs 2D’ and 3B). Further supporting this, Jarid2 in mouse embryos is highly transcribed in the CLE and later tailbud and is down-regulated as cells form the neural tube (S1 Fig). However, the finding that in NMP-L cells MEKi exposure does not alter Jarid2 transcript levels after 12 hours while reducing Jarid2 occupancy across PAX6 and at HOXD11 (Fig 6D”) indicates that ERK1/2 activity loss can acutely impact Jarid2 protein availability/association with chromatin. This suggests that decline in ERK1/2 activity may act at multiple levels to attenuate PRC2 at differentiation genes. Indeed, while Jarid2 expression is characteristic of NMPs, it would be interesting to monitor regulation of other PRC2 proteins by this signalling pathway. Moreover, although we found Ring1B protein levels reduced slightly during neural differentiation of NMP-L cells (confirming similar mild down-regulation in mESC differentiation; [46]), ERK1/2 inhibition did not alter Ring1B levels (S3 Fig), indicating that ERK1/2 signalling dynamics do not impact PRC1 occupancy by altering availability of this core protein. Furthermore, we note that while it would be interesting to assess the impact of declining ERK1/2 activity on further PRC1 proteins, we found that MEKi exposure led to global increase in chromatin accessibility at neural genes many of which were Ring1B occupied in NMPs, indicating that even if a Ring1A containing PRC1 complex is present in this context this is not operating to maintain chromatin compaction following loss of Ring1B occupancy. Together, these data using representative PRC 1 and PRC2 proteins suggest that loss of ERK1/2 activity reduces these complexes at differentiation genes by multiple distinct mechanisms; indeed, this may initially impact PRC1, while PRC2 gene/protein levels also decline during differentiation.

A potentially direct mechanism by which ERK1/2 regulates PRC2 has been reported in mESCs, where ERK2 protein associates with PRC2/Jarid2 occupied chromatin [82]. In this study, abrogating ERK1/2 signalling reduced Jarid2 and H3K27me3 at developmental genes, raising the possibility that ERK1/2 protein is directly involved in PRC recruitment and/or maintenance [82]. We found evidence for decline in ERK2 association with chromatin at PAX6 and HOXD11 during neural differentiation. However, we also observed ERK2 at the GAPDH TSS indicating that in this context, its association with DNA is not PRC target gene specific. Furthermore, inhibition of ERK1/2 activity did not alter ERK2 association with chromatin at these PRC target loci. Our finding of change in PRC occupancy over time may be explained by a decline in ERK1/2 levels during neural differentiation, while experiments involving acute ERK1/2 inhibition suggest that a decrease in ERK1/2 activity and not ERK2 occupancy correlates with reduced PRC in the context of differentiating NMP-L cells. The differences between our findings and those in mESCs [82] might reflect operation of distinct regulatory mechanisms in these very different cellular contexts. Indeed, ERK1/2 inhibition promotes self-renewal in the mouse ES cell state, while loss of FGF and so ERK1/2 promotes differentiation progression of human ESC and NMP-L cells. Moreover, the observation that NMP-L cells exposed to MEKi lose PRC occupancy and decompact the PAX6 locus, despite resumption of ERK1/2 signalling (in the continued presence of exogenous FGF) indicated that the effects of ERK1/2 inhibition in differentiation conditions are not easily reversed. This contrasts with the findings in mES cells, where rescue of ERK1/2 mutant ESCs with ERK2 restored PRC mark H3K27me3 [82]. Loss of ERK1/2 signalling in NMP-L cells may therefore serve as a gating mechanism that promotes differentiation progression and appears an initial step in the process of neural commitment.

Importantly, in the experiments in which NMP-L cells were exposed to MEKi for 12 hours, chromatin decompaction and loss of both Jarid2 and Ring1B at the PAX6 locus was not accompanied by change in H3K27me3 levels, and PAX6 was not transcribed. These findings and those in differentiation conditions at D6 show that the mechanism(s) by which loss of ERK1/2 activity removes PRC are distinct from those that direct subsequent gene transcription. Such distinct regulation of chromatin accessibility is consistent with observations in Ring1B-deficient mES cells, where HoxB and D loci decompact while their gene promoters remain decorated with H3K27me3 [67]. These findings are also consistent with the retention of H3K27me3 after PRC2 protein binding is lost during mESC neural differentiation [46] and in mouse mutant embryos in which PRC2 genes have been conditionally removed [76,100] Although there are examples of transcription from loci carrying H3K27me3 [40,101], our findings here indicate that H3K27me3 removal is closely temporally associated with the transcriptional derepression of PAX6 (this mark is not significantly lost until D8) (Figs 2D and 3B). These findings emphasise the value of the temporal resolution afforded by this slow in vitro differentiation assay in which signalling can be manipulated and the molecular mechanisms regulating chromatin accessibility decoupled from those directing transcription initiation.

Using ATAC-seq, we identified genes with increasing accessibility during neural differentiation of NMP-L cells and identified thousands of genes with GO terms associated with neural development. These included PAX6, but also other neural progenitor genes (e.g., SOX1, GlI3, GlI2, RARB, NKX6.1, DBX1, NEUROG2), which commence expression in the mouse and chick embryonic body axis, like Pax6, after FGF/ERK1/2 signalling decline [20,22,35]. We compared our “neural sites” (i.e., those that show increased accessibility during differentiation of NMP-L cells (D3) into NP cells (D8)) with the data for the H3K27me3 mark in hESCs available on ENCODE, and this showed only a small overlap (S2 Fig). This suggested that a subset of “neural sites” are PRC targets and that increased accessibility at other sites is regulated either as a downstream consequence of expression of PRC target genes such as PAX6, which is known to regulate expression of other neural differentiation TFs, including the master neuronal differentiation gene Neurog2 [102] or by a PRC-independent mechanism. The extent to which these genes are direct PRC targets was then tested by intersection of ATAC-seq-defined D5 and D6 neural sites uniquely induced by MEKi with global sites of Ring1B occupancy in NMP-L cells. An important conclusion from this work is that the deposition of Ring1B alone does not globally define all sites that later change in accessibility upon ERK1/2 signalling decline. Instead, our analysis identified a subset of sites/associated genes that are likely direct PRC targets. This suggests that declining ERK1/2 activity increases chromatin accessibility at neural genes by removing PRC but also affecting other chromatin regulatory machinery. Indeed, genomic footprinting of these MEKi-induced neural sites provided evidence that this event is likely driven by increased SOX family TF engagement at both PRC and non-PRC-associated sites. Moreover, our analysis indicated that SOX family TFs are enriched in neural sites on D5 in MEKi conditions, prior to the time when such treatment leads to precocious PAX6 transcription on D6, indicating that increased chromatin accessibility and neural TF binding are ERK1/2-regulated steps distinct from the machinery of transcription (Fig 7). A further possibility for the action of declining ERK1/2 activity could be through the regulation of enhancers that in turn activate gene expression and lead to PcG eviction. This further mechanistic possibility could be explored in the future once enhancer sequences and their target promoters have been functionally validated and may reveal not only SOX motifs, but also those for TFs whose activity is directly controlled by ERK1/2, such as ETS factors.

Further ways in which FGF/ERK1/2 signalling decline may direct chromatin reorganisation include alteration of BAF (or SWI/SNF) complex composition, leading to its recruitment and so eviction of PRC proteins from gene loci, as reported in mouse fibroblasts and ESCs [103]. This is an attractive possibility as it might also provide specificity for PRC removal at neural genes, encoded in BAF complex composition [104]. This complex alters as neural progenitor cells become neurons [104,105], while in differentiating NMP-L cells, we would expect loss of ERK1/2 activity to lead to initial formation of the BAF complex characteristic of neural progenitors. In support of such a step, transcription of key BAF complex component the ATP-dependent helicase SMARCA2 (Brahma homolog) is sharply down-regulated as FGF/ERK1/2 signalling declines in the elongating embryonic body axis [93]. Interesting, this coincides with down-regulation of Jarid2 and also the histone deacetylase HDAC1 in the embryo: Moreover, here, HDAC1 expression depends on FGFR signalling [93]. These observations point to alignment of FGF/ERK1/2 signalling loss and a coordinated change in the expression of genes that contribute to multiple chromatin regulatory complexes as differentiation commences.

The extent to which FGF/ERK1/2 signalling directs progression of differentiation by regulating PRC occupancy in contexts other than caudal epiblast/NMP-L cells requires further investigation. In mouse ESCs maintained in serum and LIF, PRC and H3K27me3 are detected at differentiation gene loci and this constitutes primed pluripotency, characterised by heterogenous cell states in a dynamic flux between pluripotency and differentiation [81,106]. This contrasts with the ground-state or naïve pluripotent cell state achieved by treatment with both ERK1/2 and GSK3beta inhibitors (2i) [29,107], in which differentiation genes lack both PRC and H3K27me3 [81]. Importantly, differentiation of such 2i ES cells involves progression through the primed pluripotent cell state, suggesting that this involves recruitment of PRC and establishment of bivalent loci. Progression towards differentiation involves FGF/ERK1/2 signalling [32,108], but only a transient period is required to initiate this process [31]. This last step is consistent with the differentiation of mEpiSCs and hESCs (for which FGF is a self-renewal factor) when FGF signalling is blocked [34,36]. These observations, together with the requirement for PRC1 and 2 for mESC differentiation [88], indicate a good correlation between exposure to a sufficient period of FGF/ERK1/2 signalling, PRC presence at differentiation genes, and subsequent differentiation from the ES cell state, following ERK1/2 signalling decline. These findings, along with our observation that ERK1/2 activity is required for PRC occupancy at PAX6 in NMP-L cells and that its inhibition triggers precocious PAX6 transcription in differentiation conditions, suggest that promotion of PRC at neural genes by ERK1/2 signalling may function as a rite of passage for subsequent differentiation.

In the future, it will be important to investigate whether ERK1/2 signalling dynamics operate to coordinate chromatin accessibility and expression of differentiation genes in other FGF-expressing progenitor cell populations in the developing embryo, including in limb, lung, and tooth buds [109–112], as well as in distinct regions of the nervous system [113]. Intriguingly, differentiation during epidermal homeostasis also requires ERK1/2/MAPK phosphatase DUSP6 in this adult tissue context [114]. This underscores the idea supported here that a transient ERK1/2 signalling dynamic acting on PRC and other chromatin complexes constitutes core and conserved molecular machinery that is a prelude to cellular differentiation.

[END]
---
[1] Url: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3000221

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/