A complex epigenome-splicing crosstalk governs epithelial-to-mesenchymal transition in metastasis and brain development.

Epithelial-to-mesenchymal transition (EMT) renders epithelial cells migratory properties. While epigenetic and splicing changes have been implicated in EMT, the mechanisms governing their crosstalk remain poorly understood. Here we discovered that a C2H2 zinc finger protein, ZNF827, is strongly induced during various contexts of EMT, including in brain development and breast cancer metastasis, and is required for the molecular and phenotypic changes underlying EMT in these processes. Mechanistically, ZNF827 mediated these responses by orchestrating a large-scale remodelling of the splicing landscape by recruiting HDAC1 for epigenetic modulation of distinct genomic loci, thereby slowing RNA polymerase II progression and altering the splicing of genes encoding key EMT regulators in cis. Our findings reveal an unprecedented complexity of crosstalk between epigenetic landscape and splicing programme in governing EMT and identify ZNF827 as a master regulator coupling these processes during EMT in brain development and breast cancer metastasis.

Histone marks regulate the epithelial-to-mesenchymal transition via alternative splicing.

Histone modifications impact final splicing decisions. However, there is little evidence of the driving role of these marks in inducing cell-specific splicing changes. Using CRISPR epigenome editing tools, we show in an epithelial-to-mesenchymal cell reprogramming system (epithelial-to-mesenchymal transition [EMT]) that a single change in H3K27ac or H3K27me3 levels right at the alternatively spliced exon is necessary and sufficient to induce a splicing change capable of recapitulating important aspects of EMT, such as cell motility and invasiveness. This histone-mark-dependent splicing effect is highly dynamic and mediated by direct recruitment of the splicing regulator PTB to its RNA binding sites. These results support a role for H3K27 marks in inducing a change in the cell's phenotype via regulation of alternative splicing. We propose the dynamic nature of chromatin as a rapid and reversible mechanism to coordinate the splicing response to cell-extrinsic cues, such as induction of EMT.

IRFinder-S: a comprehensive suite to discover and explore intron retention.

Accurate quantification and detection of intron retention levels require specialized software. Building on our previous software, we create a suite of tools called IRFinder-S, to analyze and explore intron retention events in multiple samples. Specifically, IRFinder-S allows a better identification of true intron retention events using a convolutional neural network, allows the sharing of intron retention results between labs, integrates a dynamic database to explore and contrast available samples, and provides a tested method to detect differential levels of intron retention.

A cell-to-patient machine learning transfer approach uncovers novel basal-like breast cancer prognostic markers amongst alternative splice variants.

BACKGROUND: Breast cancer is amongst the 10 first causes of death in women worldwide. Around 20% of patients are misdiagnosed leading to early metastasis, resistance to treatment and relapse. Many clinical and gene expression profiles have been successfully used to classify breast tumours into 5 major types with different prognosis and sensitivity to specific treatments. Unfortunately, these profiles have failed to subclassify breast tumours into more subtypes to improve diagnostics and survival rate. Alternative splicing is emerging as a new source of highly specific biomarkers to classify tumours in different grades. Taking advantage of extensive public transcriptomics datasets in breast cancer cell lines (CCLE) and breast cancer tumours (TCGA), we have addressed the capacity of alternative splice variants to subclassify highly aggressive breast cancers. RESULTS: Transcriptomics analysis of alternative splicing events between luminal, basal A and basal B breast cancer cell lines identified a unique splicing signature for a subtype of tumours, the basal B, whose classification is not in use in the clinic yet. Basal B cell lines, in contrast with luminal and basal A, are highly metastatic and express epithelial-to-mesenchymal (EMT) markers, which are hallmarks of cell invasion and resistance to drugs. By developing a semi-supervised machine learning approach, we transferred the molecular knowledge gained from these cell lines into patients to subclassify basal-like triple negative tumours into basal A- and basal B-like categories. Changes in splicing of 25 alternative exons, intimately related to EMT and cell invasion such as ENAH, CD44 and CTNND1, were sufficient to identify the basal-like patients with the worst prognosis. Moreover, patients expressing this basal B-specific splicing signature also expressed newly identified biomarkers of metastasis-initiating cells, like CD36, supporting a more invasive phenotype for this basal B-like breast cancer subtype. CONCLUSIONS: Using a novel machine learning approach, we have identified an EMT-related splicing signature capable of subclassifying the most aggressive type of breast cancer, which are basal-like triple negative tumours. This proof-of-concept demonstrates that the biological knowledge acquired from cell lines can be transferred to patients data for further clinical investigation. More studies, particularly in 3D culture and organoids, will increase the accuracy of this transfer of knowledge, which will open new perspectives into the development of novel therapeutic strategies and the further identification of specific biomarkers for drug resistance and cancer relapse.

SETDB1-dependent heterochromatin stimulates alternative lengthening of telomeres.

Alternative lengthening of telomeres, or ALT, is a recombination-based process that maintains telomeres to render some cancer cells immortal. The prevailing view is that ALT is inhibited by heterochromatin because heterochromatin prevents recombination. To test this model, we used telomere-specific quantitative proteomics on cells with heterochromatin deficiencies. In contrast to expectations, we found that ALT does not result from a lack of heterochromatin; rather, ALT is a consequence of heterochromatin formation at telomeres, which is seeded by the histone methyltransferase SETDB1. Heterochromatin stimulates transcriptional elongation at telomeres together with the recruitment of recombination factors, while disrupting heterochromatin had the opposite effect. Consistently, loss of SETDB1, disrupts telomeric heterochromatin and abrogates ALT. Thus, inhibiting telomeric heterochromatin formation in ALT cells might offer a new therapeutic approach to cancer treatment.

A lncRNA regulates alternative splicing via establishment of a splicing-specific chromatin signature.

Alternative pre-mRNA splicing is a highly cell type-specific process essential to generating protein diversity. However, the mechanisms responsible for the establishment and maintenance of heritable cell-specific alternative-splicing programs are poorly understood. Recent observations point to a role of histone modifications in the regulation of alternative splicing. Here we report a new mechanism of chromatin-mediated splicing control involving a long noncoding RNA (lncRNA). We have identified an evolutionarily conserved nuclear antisense lncRNA, generated from within the human FGFR2 locus, that promotes epithelial-specific alternative splicing of FGFR2. The lncRNA acts through recruitment of Polycomb-group proteins and the histone demethylase KDM2a to create a chromatin environment that impairs binding of a repressive chromatin-splicing adaptor complex important for mesenchymal-specific splicing. Our results uncover a new function for lncRNAs in the establishment and maintenance of cell-specific alternative splicing via modulation of chromatin signatures.

The non-coding genome: a universe in expansion for fine-tuning the coding world.

A report on the EMBO/EMBL Symposium on The Non-Coding Genome, held in Heidelberg, Germany, 9-12 October, 2013.

More than a splicing code: integrating the role of RNA, chromatin and non-coding RNA in alternative splicing regulation.

Large portions of the genome undergo alternative pre-mRNA splicing in often intricate patterns. Alternative splicing regulation requires extensive control mechanisms since errors can have deleterious consequences and may lead to developmental defects and disease. Recent work has identified a complex network of regulatory RNA elements which guide splicing decisions. In addition, the discovery that transcription and splicing are intimately coupled has opened up new directions into alternative splicing regulation. Work at the interface of chromatin and RNA biology has revealed unexpected molecular links between histone modifications, the transcription machinery, and non-coding RNAs (ncRNAs) in the determination of alternative splicing patterns.

Epigenetics in alternative pre-mRNA splicing.

Alternative splicing plays critical roles in differentiation, development, and disease and is a major source for protein diversity in higher eukaryotes. Analysis of alternative splicing regulation has traditionally focused on RNA sequence elements and their associated splicing factors, but recent provocative studies point to a key function of chromatin structure and histone modifications in alternative splicing regulation. These insights suggest that epigenetic regulation determines not only what parts of the genome are expressed but also how they are spliced.

Regulation of alternative splicing by histone modifications.

Alternative splicing of pre-mRNA is a prominent mechanism to generate protein diversity, yet its regulation is poorly understood. We demonstrated a direct role for histone modifications in alternative splicing. We found distinctive histone modification signatures that correlate with the splicing outcome in a set of human genes, and modulation of histone modifications causes splice site switching. Histone marks affect splicing outcome by influencing the recruitment of splicing regulators via a chromatin-binding protein. These results outline an adaptor system for the reading of histone marks by the pre-mRNA splicing machinery.

Targeted deficiency of the transcriptional activator Hnf1alpha alters subnuclear positioning of its genomic targets.

DNA binding transcriptional activators play a central role in gene-selective regulation. In part, this is mediated by targeting local covalent modifications of histone tails. Transcriptional regulation has also been associated with the positioning of genes within the nucleus. We have now examined the role of a transcriptional activator in regulating the positioning of target genes. This was carried out with primary beta-cells and hepatocytes freshly isolated from mice lacking Hnf1alpha, an activator encoded by the most frequently mutated gene in human monogenic diabetes (MODY3). We show that in Hnf1a-/- cells inactive endogenous Hnf1alpha-target genes exhibit increased trimethylated histone H3-Lys27 and reduced methylated H3-Lys4. Inactive Hnf1alpha-targets in Hnf1a-/- cells are also preferentially located in peripheral subnuclear domains enriched in trimethylated H3-Lys27, whereas active targets in wild-type cells are positioned in more central domains enriched in methylated H3-Lys4 and RNA polymerase II. We demonstrate that this differential positioning involves the decondensation of target chromatin, and show that it is spatially restricted rather than a reflection of non-specific changes in the nuclear organization of Hnf1a-deficient cells. This study, therefore, provides genetic evidence that a single transcriptional activator can influence the subnuclear location of its endogenous genomic targets in primary cells, and links activator-dependent changes in local chromatin structure to the spatial organization of the genome. We have also revealed a defect in subnuclear gene positioning in a model of a human transcription factor disease.

Distinct roles of HNF1beta, HNF1alpha, and HNF4alpha in regulating pancreas development, beta-cell function and growth.

Mutations in the genes encoding transcriptional regulators HNF1beta (TCF2), HNF1alpha (TCF1), and HNF4alpha cause autosomal dominant diabetes (also known as maturity-onset diabetes of the young). Herein, we review what we have learnt during recent years concerning the functions of these regulators in the developing and adult pancreas. Mouse studies have revealed that HNF1beta is a critical regulator of a transcriptional network that controls the specification, growth, and differentiation of the embryonic pancreas. HNF1beta mutations in humans accordingly often cause pancreas hypoplasia. By contrast, HNF1alpha and HNF4alpha have been shown to regulate the function of differentiated beta-cells. HNF1alpha and HNF4alpha mutations in patients thus cause decreased glucose-induced insulin secretion that leads to a progressive form of diabetes. HNF4alpha mutations paradoxically also cause in utero and neonatal hyperinsulinism, which later evolves to decreased glucose-induced secretion. Recent studies show that Hnf4alpha deficiency in mice causes not only abnormal insulin secretion, but also an impairment of the expansion of beta-cell mass that normally occurs during pregnancy. In line with this finding, we present data that Hnf1alpha-/- beta-cells expressing SV40 large T antigen show a severe impairment of proliferation and failure to form tumours. Collectively, these findings implicate HNF1beta as a regulator of pancreas organogenesis and differentiation, whereas HNF1alpha and HNF4alpha primarily regulate both growth and function of islet beta-cells.

A conditional model reveals that induction of hepatocyte nuclear factor-1alpha in Hnf1alpha-null mutant beta-cells can activate silenced genes postnatally, whereas overexpression is deleterious.

Humans with heterozygous loss-of-function mutations in the hepatocyte nuclear factor-1alpha (HNF1alpha) gene develop beta-cell-deficient diabetes (maturity-onset diabetes of the young type 3), indicating that HNF1alpha gene dosage is critical in beta-cells. However, whether increased HNF1alpha expression might be beneficial or deleterious for beta-cells is unknown. Furthermore, although it is clear that HNF1alpha is required for beta-cell function, it is not known whether this role is cell autonomous or whether there is a restricted developmental time frame for HNF1alpha to elicit gene activation in beta-cells. To address this, we generated a tetracycline-inducible mouse model that transcribes HNF1alpha selectively in beta-cells in either wild-type or Hnf1alpha-null backgrounds. Short-term induction of HNF1alpha in islets from adult Hnf1alpha(-/-) mice that did not express HNF1alpha throughout development resulted in the activation of target genes, indicating that HNF1alpha has beta-cell-autonomous functions that can be rescued postnatally. However, transgenic induction throughout development, which inevitably resulted in supraphysiological levels of HNF1alpha, strikingly caused a severe reduction of cellular proliferation, increased apoptosis, and consequently beta-cell depletion and diabetes. Thus, HNF1alpha is sensitive to both reduced and excessive concentrations in beta-cells. This finding illustrates the paramount importance of using the correct concentration of a beta-cell transcription factor in both gene therapy and artificial differentiation strategies.

IA1 is NGN3-dependent and essential for differentiation of the endocrine pancreas.

Neurogenin 3 (Ngn3) is key for endocrine cell specification in the embryonic pancreas and induction of a neuroendocrine cell differentiation program by misexpression in adult pancreatic duct cells. We identify the gene encoding IA1, a zinc-finger transcription factor, as a direct target of Ngn3 and show that it forms a novel branch in the Ngn3-dependent endocrinogenic transcription factor network. During embryonic development of the pancreas, IA1 and Ngn3 exhibit nearly identical spatio-temporal expression patterns. However, embryos lacking Ngn3 fail to express IA1 in the pancreas. Upon ectopic expression in adult pancreatic duct cells Ngn3 binds to chromatin in the IA1 promoter region and activates transcription. Consistent with this direct effect, IA1 expression is normal in embryos mutant for NeuroD1, Arx, Pax4 and Pax6, regulators operating downstream of Ngn3. IA1 is an effector of Ngn3 function as inhibition of IA1 expression in embryonic pancreas decreases the formation of insulin- and glucagon-positive cells by 40%, while its ectopic expression amplifies neuroendocrine cell differentiation by Ngn3 in adult duct cells. IA1 is therefore a novel Ngn3-regulated factor required for normal differentiation of pancreatic endocrine cells.

Hnf6 and Tcf2 (MODY5) are linked in a gene network operating in a precursor cell domain of the embryonic pancreas.

During pancreatic organogenesis endocrine cells arise from non self-renewing progenitors that express Ngn3. The precursors that give rise to Ngn3+ cells are presumably located within duct-like structures. However, the nature of such precursors is poorly understood. We show that, at E13-E18, the embryonic stage during which the major burst of beta-cell neogenesis takes place, pancreatic duct cells express Hnf1beta, the product of the maturity-onset diabetes of the young type 5 (MODY5) gene. Ngn3+ cells at this stage invariably cluster with mitotically competent Hnf1beta+ cells, and are often intercalated with these cells in the epithelium that lines the lumen of primitive ducts. We present several observations that collectively indicate that Hnf1beta+ cells are the immediate precursors of Ngn3+ cells. We furthermore show that Hnf1beta expression is markedly reduced in early pancreatic epithelial cells of Hnf6-deficient mice, in which formation of Ngn3+ cells is defective. These findings define a precursor cellular stage of the embryonic pancreas and place Hnf1beta in a genetic hierarchy that regulates the generation of pancreatic endocrine cells.