Cell-to-cell variation is a universal feature of life that impacts a wide range of biological phenomena, from developmental plasticity1,2 to tumor heterogeneity3. While recent advances have improved our ability to document cellular phenotypic variation4–8 the fundamental mechanisms that generate variability from identical DNA sequences remain elusive. Here we reveal the landscape and principles of cellular DNA regulatory variation by developing a robust method for mapping the accessible genome of individual cells via assay for transposase-accessible chromatin using sequencing (ATAC-seq). Single-cell ATAC-seq (scATAC-seq) maps from hundreds of single-cells in aggregate closely resemble accessibility profiles from tens of millions of cells and provides insights into cell-to-cell variation. Accessibility variance is systematically associated with specific trans-factors and cis-elements, and we discover combinations of trans-factors associated with either induction or suppression of cell-to-cell variability. We further identify sets of trans-factors associated with cell-type specific accessibility variance across 8 cell types. Targeted perturbations of cell cycle or transcription factor signaling evoke stimulus-specific changes in this observed variability. The pattern of accessibility variation in cis across the genome recapitulates chromosome topological domains9 de novo, linking single-cell accessibility variation to three-dimensional genome organization. All together, single-cell analysis of DNA accessibility provides new insight into cellular variation of the “regulome.”

Phosphoinositides are a family of lipid signalling molecules that regulate many cellular functions in eukaryotes. Phosphatidylinositol-4,5-bisphosphate (PtdIns4,5P2), the central component in the phosphoinositide signalling circuitry, is generated primarily by type I phosphatidylinositol 4-phosphate 5-kinases (PIPKIalpha, PIPKIbeta and PIPKIgamma). In addition to functions in the cytosol, phosphoinositides are present in the nucleus, where they modulate several functions; however, the mechanism by which they directly regulate nuclear functions remains unknown. PIPKIs regulate cellular functions through interactions with protein partners, often PtdIns4,5P2 effectors, that target PIPKIs to discrete subcellular compartments, resulting in the spatial and temporal generation of PtdIns4,5P2 required for the regulation of specific signalling pathways. Therefore, to determine roles for nuclear PtdIns4,5P2 we set out to identify proteins that interacted with the nuclear PIPK, PIPKIalpha. Here we show that PIPKIalpha co-localizes at nuclear speckles and interacts with a newly identified non-canonical poly(A) polymerase, which we have termed Star-PAP (nuclear speckle targeted PIPKIalpha regulated-poly(A) polymerase) and that the activity of Star-PAP can be specifically regulated by PtdIns4,5P2. Star-PAP and PIPKIalpha function together in a complex to control the expression of select mRNAs, including the transcript encoding the key cytoprotective enzyme haem oxygenase-1 (refs 8, 9) and other oxidative stress response genes by regulating the 3'-end formation of their mRNAs. Taken together, the data demonstrate a model by which phosphoinositide signalling works in tandem with complement pathways to regulate the activity of Star-PAP and the subsequent biosynthesis of its target mRNA. The results reveal a mechanism for the integration of nuclear phosphoinositide signals and a method for regulating gene expression.

Prader-Willi syndrome (PWS) and Angelman syndrome (AS) are oppositely imprinted autism-spectrum disorders with known genetic bases, but complex epigenetic mechanisms underlie their pathogenesis. The PWS/AS locus on 15q11-q13 is regulated by an imprinting control region that is maternally methylated and silenced. The PWS imprinting control region is the promoter for a one megabase paternal transcript encoding the ubiquitous protein-coding Snrpn gene and multiple neuron-specific noncoding RNAs, including the PWS-related Snord116 repetitive locus of small nucleolar RNAs and host genes, and the antisense transcript to AScausing ubiquitin ligase encoding Ube3a (Ube3a-ATS). Neuronspecific transcriptional progression through Ube3a-ATS correlates with paternal Ube3a silencing and chromatin decondensation. Interestingly, topoisomerase inhibitors, including topotecan, were recently identified in an unbiased drug screen for compounds that could reverse the silent paternal allele of Ube3a in neurons, but the mechanism of topotecan action on the PWS/AS locus is unknown. Here, we demonstrate that topotecan treatment stabilizes the formation of RNA:DNA hybrids (R loops) at G-skewed repeat elements within paternal Snord116, corresponding to increased chromatin decondensation and inhibition of Ube3a-ATS expression. Neural precursor cells from paternal Snord116 deletion mice exhibit increased Ube3a-ATS levels in differentiated neurons and show a reduced effect of topotecan compared with wild-type neurons. These results demonstrate that the AS candidate drug topotecan acts predominantly through stabilizing R loops and chromatin decondensation at the paternally expressed PWS Snord116 locus. Our study holds promise for targeted therapies to the Snord116 locus for both AS and PWS.rader-Willi syndrome (PWS) and Angelman syndrome (AS) are imprinted neurodevelopmental disorders caused by oppositely inherited deficiencies of chromosome 15q11-q13. AS and PWS are both characterized by hypotonia at birth, disordered sleep, autistic features, and intellectual disabilities, but the diseases differentiate into phenotypically distinct syndromes in early childhood (1, 2). Seizures, ataxia, and inappropriate laughter characterize AS, whereas hyperphagia leading to obesity and obsessive-compulsive behaviors characterize PWS. Maternal mutations in UBE3A/Ube3a in humans and mice have identified the loss of function of this ubiquitin E3 ligase encoding gene as the cause of AS (3, 4). For PWS, small deletions of the HBII-85/ SNORD116 locus (5-7) and two mouse models of Snord116 deletions (8, 9) have identified the minimal causative deficiency to be the paternally expressed, highly repetitive, long noncoding RNA (lncRNA) that is processed into multiple small nucleolar RNAs (snoRNAs) and spliced nuclear retained host genes (116HG and 115HG) (10,11).A recent drug screen discovered that topoisomerase inhibitors, including topotecan, reduce Ube3a-ATS by an unknown mechanism to reverse the silencing of paternal Ube3a in mouse neurons and brain (12). Topotec...

Methyl CpG binding protein-2 (MeCP2) is an essential epigenetic regulator in human brain development. Rett syndrome, the primary disorder caused by mutations in the X-linked MECP2 gene, is characterized by a period of cognitive decline and development of hand stereotypies and seizures following an apparently normal early infancy. In addition, MECP2 mutations and duplications are observed in a spectrum of neurodevelopmental disorders, including severe neonatal encephalopathy, X-linked mental retardation, and autism, implicating MeCP2 as an essential regulator of postnatal brain development. In this review, we compare the mutation types and inheritance patterns of the human disorders associated with MECP2. In addition, we summarize the current understanding of MeCP2 as a central epigenetic regulator of activity-dependent synaptic maturation. As MeCP2 occupies a central role in the pathogenesis of multiple neurodevelopmental disorders, continued investigation into MeCP2 function and regulatory pathways may show promise for developing broad-spectrum therapies.

We have recently identified Star-PAP, a nuclear poly(A) polymerase that associates with phosphatidylinositol-4-phosphate 5-kinase I␣ (PIPKI␣) and is required for the expression of a specific subset of mRNAs. Star-PAP activity is directly modulated by the PIPKI␣ product phosphatidylinositol 4,5-bisphosphate (PI-4,5-P 2 ), linking nuclear phosphoinositide signaling to gene expression. Here, we show that PI-4,5-P 2 -dependent protein kinase activity is also a part of the Star-PAP protein complex. We identify the PI-4,5-P 2 -sensitive casein kinase I␣ (CKI␣) as a protein kinase responsible for this activity and further show that CKI␣ is capable of directly phosphorylating Star-PAP. Both CKI␣ and PIPKI␣ are required for the synthesis of some but not all Star-PAP target mRNA, and like Star-PAP, CKI␣ is associated with these messages in vivo. Taken together, these data indicate that CKI␣, PIPKI␣, and Star-PAP function together to modulate the production of specific Star-PAP messages. The Star-PAP complex therefore represents a location where multiple signaling pathways converge to regulate the expression of specific mRNAs.Polyadenylation of most mRNAs is required for their efficient transcription and export as well as regulating their stability and translational efficiency (1). Polyadenylation of mRNA is achieved through the activity of poly(A) polymerases (PAPs). 2There are multiple PAPs in mammalian cells, including PAP␣, which is thought to be primarily responsible for the polyadenylation of newly transcribed mRNAs in the nucleus (2, 3). Additionally, there are "non-canonical" PAPs, including Gld2 and Trf4, which regulate stability and degradation of specific RNAs through polyadenylation (4, 5).Star-PAP is a non-canonical poly(A) polymerase that is distinct from all other currently characterized PAP enzymes (6). Like the canonical PAP␣, Star-PAP has an RNA recognition motif and is a nuclear enzyme. However, similar to non-canonical PAPs, the Star-PAP protein complex and architecture differ significantly from PAP␣, and consequently, Star-PAP specifically targets a select subset of mRNAs. This suggests that Star-PAP is a hybrid PAP that is required for the 3Ј-end formation of newly transcribed pre-mRNAs but functions in a regulatory role to control mRNA expression levels. Star-PAP has a unique domain structure relative to all other known PAPs (6). One unique feature is a 205-amino acid proline-rich region (PRR) inserted into the catalytic PAP core. The PRR splits the catalytic PAP domain, and this region represents a potential site for regulation of Star-PAP function.Poly(A) polymerases operate as large multiprotein complexes responsible for the 3Ј-processing of RNAs (7, 8). The Star-PAP polyadenylation complex is similar to that of poly(A) polymerases that utilizes mRNA as a substrate and includes cleavage and polyadenylation specificity factor (CPSF) subunits, cleavage stimulatory factor (CstF) subunits, symplekin, and RNA polymerase II (6, 9). However, unlike other poly(A) polymerase complexes, the Star-PAP complex...

bMutations in the gene encoding methyl-CpG-binding protein 2 (MeCP2) lead to disrupted neuronal function and can cause the neurodevelopmental disorder Rett syndrome. MeCP2 is a transcriptional regulator that binds to methylated DNA and is most abundant in neuronal nuclei. The mechanisms by which MeCP2 regulates gene expression remain ambiguous, as it has been reported to function as a transcriptional silencer or activator and to execute these activities through both gene-specific and genome-wide mechanisms. We hypothesized that posttranslational modifications of MeCP2 may be important for reconciling these apparently contradictory functions. Our results demonstrate that MeCP2 contains multiple posttranslational modifications, including phosphorylation, acetylation, and ubiquitylation. Phosphorylation of MeCP2 at S229 or S80 influenced selective in vivo interactions with the chromatin factors HP1 and SMC3 and the cofactors Sin3A and YB-1. pS229 MeCP2 was specifically enriched at the RET promoter, and phosphorylation of MeCP2 was necessary for differentiation-induced activation and repression of the MeCP2 target genes RET and EGR2. These results demonstrate that phosphorylation is one of several factors that are important for interpreting the complexities of MeCP2 transcriptional modulation.

Mutations in MECP2 cause the neurodevelopmental disorder Rett syndrome (RTT OMIM 312750). Alternative inclusion of MECP2/Mecp2 exon 1 with exons 3 and 4 encodes MeCP2-e1 or MeCP2-e2 protein isoforms with unique amino termini. While most MECP2 mutations are located in exons 3 and 4 thus affecting both isoforms, MECP2 exon 1 mutations but not exon 2 mutations have been identified in RTT patients, suggesting that MeCP2-e1 deficiency is sufficient to cause RTT. As expected, genetic deletion of Mecp2 exons 3 and/or 4 recapitulates RTT-like neurologic defects in mice. However, Mecp2 exon 2 knockout mice have normal neurologic function. Here, a naturally occurring MECP2 exon 1 mutation is recapitulated in a mouse model by genetic engineering. A point mutation in the translational start codon of Mecp2 exon 1, transmitted through the germline, ablates MeCP2-e1 translation while preserving MeCP2-e2 production in mouse brain. The resulting MeCP2-e1 deficient mice developed forelimb stereotypy, hindlimb clasping, excessive grooming and hypo-activity prior to death between 7 and 31 weeks. MeCP2-e1 deficient mice also exhibited abnormal anxiety, sociability and ambulation. Despite MeCP2-e1 and MeCP2-e2 sharing, 96% amino acid identity, differences were identified. A fraction of phosphorylated MeCP2-e1 differed from the bulk of MeCP2 in subnuclear localization and co-factor interaction. Furthermore, MeCP2-e1 exhibited enhanced stability compared with MeCP2-e2 in neurons. Therefore, MeCP2-e1 deficient mice implicate MeCP2-e1 as the sole contributor to RTT with non-redundant functions.

Graphical Abstract Highlights d Multimodal analysis differentiates cells beyond transcriptomic classification d Single-cell analysis links stimulus-induced calcium elevations to transcriptomes d Cell-type-specific responses to neurotransmitters are associated with maturation d Serotonergic signaling in human radial glia promotes radial fiber formation