Protein codes promote selective subcellular compartmentalization.

Cells have evolved mechanisms to distribute ~10 billion protein molecules to subcellular compartments where diverse proteins involved in shared functions must efficiently assemble. Here, we demonstrate that proteins with shared functions share amino acid sequence codes that guide them to compartment destinations. A protein language model, ProtGPS, was developed that predicts with high performance the compartment localization of human proteins excluded from the training set. ProtGPS successfully guided generation of novel protein sequences that selectively assemble in targeted subcellular compartments. ProtGPS also identified pathological mutations that change this code and lead to altered subcellular localization of proteins. Our results indicate that protein sequences contain not only a folding code, but also a previously unrecognized code governing their distribution in specific cellular compartments.

Distinct chemical environments in biomolecular condensates.

Diverse mechanisms have been described for selective enrichment of biomolecules in membrane-bound organelles, but less is known about mechanisms by which molecules are selectively incorporated into biomolecular assemblies such as condensates that lack surrounding membranes. The chemical environments within condensates may differ from those outside these bodies, and if these differed among various types of condensate, then the different solvation environments would provide a mechanism for selective distribution among these intracellular bodies. Here we use small molecule probes to show that different condensates have distinct chemical solvating properties and that selective partitioning of probes in condensates can be predicted with deep learning approaches. Our results demonstrate that different condensates harbor distinct chemical environments that influence the distribution of molecules, show that clues to condensate chemical grammar can be ascertained by machine learning and suggest approaches to facilitate development of small molecule therapeutics with optimal subcellular distribution and therapeutic benefit.

Genetic variation associated with condensate dysregulation in disease.

A multitude of cellular processes involve biomolecular condensates, which has led to the suggestion that diverse pathogenic mutations may dysregulate condensates. Although proof-of-concept studies have identified specific mutations that cause condensate dysregulation, the full scope of the pathological genetic variation that affects condensates is not yet known. Here, we comprehensively map pathogenic mutations to condensate-promoting protein features in putative condensate-forming proteins and find over 36,000 pathogenic mutations that plausibly contribute to condensate dysregulation in over 1,200 Mendelian diseases and 550 cancers. This resource captures mutations presently known to dysregulate condensates, and experimental tests confirm that additional pathological mutations do indeed affect condensate properties in cells. These findings suggest that condensate dysregulation may be a pervasive pathogenic mechanism underlying a broad spectrum of human diseases, provide a strategy to identify proteins and mutations involved in pathologically altered condensates, and serve as a foundation for mechanistic insights into disease and therapeutic hypotheses.

The nuclear receptor THRB facilitates differentiation of human PSCs into more mature hepatocytes.

To understand the mechanisms regulating the in vitro maturation of hPSC-derived hepatocytes, we developed a 3D differentiation system and compared gene regulatory elements in human primary hepatocytes with those in hPSC-hepatocytes that were differentiated in 2D or 3D conditions by RNA-seq, ATAC-seq, and H3K27Ac ChIP-seq. Regulome comparisons showed a reduced enrichment of thyroid receptor THRB motifs in accessible chromatin and active enhancers without a reduced transcription of THRB. The addition of thyroid hormone T3 increased the binding of THRB to the CYP3A4 proximal enhancer, restored the super-enhancer status and gene expression of NFIC, and reduced the expression of AFP. The resultant hPSC-hepatocytes showed gene expression, epigenetic status, and super-enhancer landscape closer to primary hepatocytes and activated regulatory regions including non-coding SNPs associated with liver-related diseases. Transplanting the hPSC-hepatocytes resulted in the engraftment of human hepatocytes into the mouse liver without disrupting normal liver histology. This work implicates the environmental factor-nuclear receptor axis in regulating the maturation of hPSC-hepatocytes.

RNA-Mediated Feedback Control of Transcriptional Condensates.

Regulation of biological processes typically incorporates mechanisms that initiate and terminate the process and, where understood, these mechanisms often involve feedback control. Regulation of transcription is a fundamental cellular process where the mechanisms involved in initiation have been studied extensively, but those involved in arresting the process are poorly understood. Modeling of the potential roles of RNA in transcriptional control suggested a non-equilibrium feedback control mechanism where low levels of RNA promote condensates formed by electrostatic interactions whereas relatively high levels promote dissolution of these condensates. Evidence from in vitro and in vivo experiments support a model where RNAs produced during early steps in transcription initiation stimulate condensate formation, whereas the burst of RNAs produced during elongation stimulate condensate dissolution. We propose that transcriptional regulation incorporates a feedback mechanism whereby transcribed RNAs initially stimulate but then ultimately arrest the process.

MeCP2 links heterochromatin condensates and neurodevelopmental disease.

Methyl CpG binding protein 2 (MeCP2) is a key component of constitutive heterochromatin, which is crucial for chromosome maintenance and transcriptional silencing1-3. Mutations in the MECP2 gene cause the progressive neurodevelopmental disorder Rett syndrome3-5, which is associated with severe mental disability and autism-like symptoms that affect girls during early childhood. Although previously thought to be a dense and relatively static structure1,2, heterochromatin is now understood to exhibit properties consistent with a liquid-like condensate6,7. Here we show that MeCP2 is a dynamic component of heterochromatin condensates in cells, and is stimulated by DNA to form liquid-like condensates. MeCP2 contains several domains that contribute to the formation of condensates, and mutations in MECP2 that lead to Rett syndrome disrupt the ability of MeCP2 to form condensates. Condensates formed by MeCP2 selectively incorporate and concentrate heterochromatin cofactors rather than components of euchromatic transcriptionally active condensates. We propose that MeCP2 enhances the separation of heterochromatin and euchromatin through its condensate partitioning properties, and that disruption of condensates may be a common consequence of mutations in MeCP2 that cause Rett syndrome.

Partitioning of cancer therapeutics in nuclear condensates.

The nucleus contains diverse phase-separated condensates that compartmentalize and concentrate biomolecules with distinct physicochemical properties. Here, we investigated whether condensates concentrate small-molecule cancer therapeutics such that their pharmacodynamic properties are altered. We found that antineoplastic drugs become concentrated in specific protein condensates in vitro and that this occurs through physicochemical properties independent of the drug target. This behavior was also observed in tumor cells, where drug partitioning influenced drug activity. Altering the properties of the condensate was found to affect the concentration and activity of drugs. These results suggest that selective partitioning and concentration of small molecules within condensates contributes to drug pharmacodynamics and that further understanding of this phenomenon may facilitate advances in disease therapy.

Pol II phosphorylation regulates a switch between transcriptional and splicing condensates.

The synthesis of pre-mRNA by RNA polymerase II (Pol II) involves the formation of a transcription initiation complex, and a transition to an elongation complex1-4. The large subunit of Pol II contains an intrinsically disordered C-terminal domain that is phosphorylated by cyclin-dependent kinases during the transition from initiation to elongation, thus influencing the interaction of the C-terminal domain with different components of the initiation or the RNA-splicing apparatus5,6. Recent observations suggest that this model provides only a partial picture of the effects of phosphorylation of the C-terminal domain7-12. Both the transcription-initiation machinery and the splicing machinery can form phase-separated condensates that contain large numbers of component molecules: hundreds of molecules of Pol II and mediator are concentrated in condensates at super-enhancers7,8, and large numbers of splicing factors are concentrated in nuclear speckles, some of which occur at highly active transcription sites9-12. Here we investigate whether the phosphorylation of the Pol II C-terminal domain regulates the incorporation of Pol II into phase-separated condensates that are associated with transcription initiation and splicing. We find that the hypophosphorylated C-terminal domain of Pol II is incorporated into mediator condensates and that phosphorylation by regulatory cyclin-dependent kinases reduces this incorporation. We also find that the hyperphosphorylated C-terminal domain is preferentially incorporated into condensates that are formed by splicing factors. These results suggest that phosphorylation of the Pol II C-terminal domain drives an exchange from condensates that are involved in transcription initiation to those that are involved in RNA processing, and implicates phosphorylation as a mechanism that regulates condensate preference.

Transcription Factors Activate Genes through the Phase-Separation Capacity of Their Activation Domains.

Gene expression is controlled by transcription factors (TFs) that consist of DNA-binding domains (DBDs) and activation domains (ADs). The DBDs have been well characterized, but little is known about the mechanisms by which ADs effect gene activation. Here, we report that diverse ADs form phase-separated condensates with the Mediator coactivator. For the OCT4 and GCN4 TFs, we show that the ability to form phase-separated droplets with Mediator in vitro and the ability to activate genes in vivo are dependent on the same amino acid residues. For the estrogen receptor (ER), a ligand-dependent activator, we show that estrogen enhances phase separation with Mediator, again linking phase separation with gene activation. These results suggest that diverse TFs can interact with Mediator through the phase-separating capacity of their ADs and that formation of condensates with Mediator is involved in gene activation.

Coactivator condensation at super-enhancers links phase separation and gene control.

Super-enhancers (SEs) are clusters of enhancers that cooperatively assemble a high density of the transcriptional apparatus to drive robust expression of genes with prominent roles in cell identity. Here we demonstrate that the SE-enriched transcriptional coactivators BRD4 and MED1 form nuclear puncta at SEs that exhibit properties of liquid-like condensates and are disrupted by chemicals that perturb condensates. The intrinsically disordered regions (IDRs) of BRD4 and MED1 can form phase-separated droplets, and MED1-IDR droplets can compartmentalize and concentrate the transcription apparatus from nuclear extracts. These results support the idea that coactivators form phase-separated condensates at SEs that compartmentalize and concentrate the transcription apparatus, suggest a role for coactivator IDRs in this process, and offer insights into mechanisms involved in the control of key cell-identity genes.

Transcriptional Dysregulation of MYC Reveals Common Enhancer-Docking Mechanism.

Transcriptional dysregulation of the MYC oncogene is among the most frequent events in aggressive tumor cells, and this is generally accomplished by acquisition of a super-enhancer somewhere within the 2.8 Mb TAD where MYC resides. We find that these diverse cancer-specific super-enhancers, differing in size and location, interact with the MYC gene through a common and conserved CTCF binding site located 2 kb upstream of the MYC promoter. Genetic perturbation of this enhancer-docking site in tumor cells reduces CTCF binding, super-enhancer interaction, MYC gene expression, and cell proliferation. CTCF binding is highly sensitive to DNA methylation, and this enhancer-docking site, which is hypomethylated in diverse cancers, can be inactivated through epigenetic editing with dCas9-DNMT. Similar enhancer-docking sites occur at other genes, including genes with prominent roles in multiple cancers, suggesting a mechanism by which tumor cell oncogenes can generally hijack enhancers. These results provide insights into mechanisms that allow a single target gene to be regulated by diverse enhancer elements in different cell types.

Integrated genomic analyses of de novo pathways underlying atypical meningiomas.

This corrects the article DOI: 10.1038/ncomms14433.

Corrigendum: Small genomic insertions form enhancers that misregulate oncogenes.

This corrects the article DOI: 10.1038/ncomms14385.

Integrated genomic analyses of de novo pathways underlying atypical meningiomas.

Meningiomas are mostly benign brain tumours, with a potential for becoming atypical or malignant. On the basis of comprehensive genomic, transcriptomic and epigenomic analyses, we compared benign meningiomas to atypical ones. Here, we show that the majority of primary (de novo) atypical meningiomas display loss of NF2, which co-occurs either with genomic instability or recurrent SMARCB1 mutations. These tumours harbour increased H3K27me3 signal and a hypermethylated phenotype, mainly occupying the polycomb repressive complex 2 (PRC2) binding sites in human embryonic stem cells, thereby phenocopying a more primitive cellular state. Consistent with this observation, atypical meningiomas exhibit upregulation of EZH2, the catalytic subunit of the PRC2 complex, as well as the E2F2 and FOXM1 transcriptional networks. Importantly, these primary atypical meningiomas do not harbour TERT promoter mutations, which have been reported in atypical tumours that progressed from benign ones. Our results establish the genomic landscape of primary atypical meningiomas and potential therapeutic targets.

Small genomic insertions form enhancers that misregulate oncogenes.

The non-coding regions of tumour cell genomes harbour a considerable fraction of total DNA sequence variation, but the functional contribution of these variants to tumorigenesis is ill-defined. Among these non-coding variants, somatic insertions are among the least well characterized due to challenges with interpreting short-read DNA sequences. Here, using a combination of Chip-seq to enrich enhancer DNA and a computational approach with multiple DNA alignment procedures, we identify enhancer-associated small insertion variants. Among the 102 tumour cell genomes we analyse, small insertions are frequently observed in enhancer DNA sequences near known oncogenes. Further study of one insertion, somatically acquired in primary leukaemia tumour genomes, reveals that it nucleates formation of an active enhancer that drives expression of the LMO2 oncogene. The approach described here to identify enhancer-associated small insertion variants provides a foundation for further study of these abnormalities across human cancers.

Recurrent somatic mutations in POLR2A define a distinct subset of meningiomas.

RNA polymerase II mediates the transcription of all protein-coding genes in eukaryotic cells, a process that is fundamental to life. Genomic mutations altering this enzyme have not previously been linked to any pathology in humans, which is a testament to its indispensable role in cell biology. On the basis of a combination of next-generation genomic analyses of 775 meningiomas, we report that recurrent somatic p.Gln403Lys or p.Leu438_His439del mutations in POLR2A, which encodes the catalytic subunit of RNA polymerase II (ref. 1), hijack this essential enzyme and drive neoplasia. POLR2A mutant tumors show dysregulation of key meningeal identity genes, including WNT6 and ZIC1/ZIC4. In addition to mutations in POLR2A, NF2, SMARCB1, TRAF7, KLF4, AKT1, PIK3CA, and SMO, we also report somatic mutations in AKT3, PIK3R1, PRKAR1A, and SUFU in meningiomas. Our results identify a role for essential transcriptional machinery in driving tumorigenesis and define mutually exclusive meningioma subgroups with distinct clinical and pathological features.

Activation of proto-oncogenes by disruption of chromosome neighborhoods.

Oncogenes are activated through well-known chromosomal alterations such as gene fusion, translocation, and focal amplification. In light of recent evidence that the control of key genes depends on chromosome structures called insulated neighborhoods, we investigated whether proto-oncogenes occur within these structures and whether oncogene activation can occur via disruption of insulated neighborhood boundaries in cancer cells. We mapped insulated neighborhoods in T cell acute lymphoblastic leukemia (T-ALL) and found that tumor cell genomes contain recurrent microdeletions that eliminate the boundary sites of insulated neighborhoods containing prominent T-ALL proto-oncogenes. Perturbation of such boundaries in nonmalignant cells was sufficient to activate proto-oncogenes. Mutations affecting chromosome neighborhood boundaries were found in many types of cancer. Thus, oncogene activation can occur via genetic alterations that disrupt insulated neighborhoods in malignant cells.

Models of human core transcriptional regulatory circuitries.

A small set of core transcription factors (TFs) dominates control of the gene expression program in embryonic stem cells and other well-studied cellular models. These core TFs collectively regulate their own gene expression, thus forming an interconnected auto-regulatory loop that can be considered the core transcriptional regulatory circuitry (CRC) for that cell type. There is limited knowledge of core TFs, and thus models of core regulatory circuitry, for most cell types. We recently discovered that genes encoding known core TFs forming CRCs are driven by super-enhancers, which provides an opportunity to systematically predict CRCs in poorly studied cell types through super-enhancer mapping. Here, we use super-enhancer maps to generate CRC models for 75 human cell and tissue types. These core circuitry models should prove valuable for further investigating cell-type-specific transcriptional regulation in healthy and diseased cells.