SAGA-Dependent Histone H2Bub1 Deubiquitination Is Essential for Cellular Ubiquitin Balance during Embryonic Development.

Ubiquitin (ub) is a small, highly conserved protein widely expressed in eukaryotic cells. Ubiquitination is a post-translational modification catalyzed by enzymes that activate, conjugate, and ligate ub to proteins. Substrates can be modified either by addition of a single ubiquitin molecule (monoubiquitination), or by conjugation of several ubs (polyubiquitination). Monoubiquitination acts as a signaling mark to control diverse biological processes. The cellular and spatial distribution of ub is determined by the opposing activities of ub ligase enzymes, and deubiquitinases (DUBs), which remove ub from proteins to generate free ub. In mammalian cells, 1-2% of total histone H2B is monoubiquitinated. The SAGA (Spt Ada Gcn5 Acetyl-transferase) is a transcriptional coactivator and its DUB module removes ub from H2Bub1. The mammalian SAGA DUB module has four subunits, ATXN7, ATXN7L3, USP22, and ENY2. Atxn7l3-/- mouse embryos, lacking DUB activity, have a five-fold increase in H2Bub1 retention, and die at mid-gestation. Interestingly, embryos lacking the ub encoding gene, Ubc, have a similar phenotype. Here we provide a current overview of data suggesting that H2Bub1 retention on the chromatin in Atxn7l3-/- embryos may lead to an imbalance in free ub distribution. Thus, we speculate that ATXN7L3-containing DUBs impact the free cellular ub pool during development.

Homozygous TAF8 mutation in a patient with intellectual disability results in undetectable TAF8 protein, but preserved RNA polymerase II transcription.

The human general transcription factor TFIID is composed of the TATA-binding protein (TBP) and 13 TBP-associated factors (TAFs). In eukaryotic cells, TFIID is thought to nucleate RNA polymerase II (Pol II) preinitiation complex formation on all protein coding gene promoters and thus, be crucial for Pol II transcription. In a child with intellectual disability, mild microcephaly, corpus callosum agenesis and poor growth, we identified a homozygous splice-site mutation in TAF8 (NM_138572.2: c.781-1G > A). Our data indicate that the patient's mutation generates a frame shift and an unstable TAF8 mutant protein with an unrelated C-terminus. The mutant TAF8 protein could not be detected in extracts from the patient's fibroblasts, indicating a loss of TAF8 function and that the mutation is most likely causative. Moreover, our immunoprecipitation and proteomic analyses show that in patient cells only partial TAF complexes exist and that the formation of the canonical TFIID is impaired. In contrast, loss of TAF8 in mouse embryonic stem cells and blastocysts leads to cell death and to a global decrease in Pol II transcription. Astonishingly however, in human TAF8 patient cells, we could not detect any cellular phenotype, significant changes in genome-wide Pol II occupancy and pre-mRNA transcription. Thus, the disorganization of the essential holo-TFIID complex did not affect global Pol II transcription in the patient's fibroblasts. Our observations further suggest that partial TAF complexes, and/or an altered TFIID containing a mutated TAF8, could support human development and thus, the absence of holo-TFIID is less deleterious for transcription than originally predicted.

Cortical Layer Inversion and Deregulation of Reelin Signaling in the Absence of SOCS6 and SOCS7.

Mutations of the reelin gene cause severe defects in cerebral cortex development and profound intellectual impairment. While many aspects of the reelin signaling pathway have been identified, the molecular and ultimate cellular consequences of reelin signaling remain unknown. Specifically, it is unclear if termination of reelin signaling is as important for normal cortical neuron migration as activation of reelin signaling. Using mice that are single or double deficient, we discovered that combined loss of the suppressors of cytokine signaling, SOCS6 and SOCS7, recapitulated the cortical layer inversion seen in mice lacking reelin and led to a dramatic increase in the reelin signaling molecule disabled (DAB1) in the cortex. The SRC homology domains of SOCS6 and SOCS7 bound DAB1 ex vivo. Mutation of DAB1 greatly diminished binding and protected from degradation by SOCS6. Phosphorylated DAB1 was elevated in cortical neurons in the absence of SOCS6 and SOCS7. Thus, constitutive activation of reelin signaling was observed to be equally detrimental as lack of activation. We hypothesize that, by terminating reelin signaling, SOCS6 and SOCS7 may allow new cycles of reelin signaling to occur and that these may be essential for cortical neuron migration.

MOZ regulates B-cell progenitors and, consequently, Moz haploinsufficiency dramatically retards MYC-induced lymphoma development.

The histone acetyltransferase MOZ (MYST3, KAT6A) is the target of recurrent chromosomal translocations fusing the MOZ gene to CBP, p300, NCOA3, or TIF2 in particularly aggressive cases of acute myeloid leukemia. In this study, we report the role of wild-type MOZ in regulating B-cell progenitor proliferation and hematopoietic malignancy. In the Eµ-Myc model of aggressive pre-B/B-cell lymphoma, the loss of just one allele of Moz increased the median survival of mice by 3.9-fold. MOZ was required to maintain the proliferative capacity of B-cell progenitors, even in the presence of c-MYC overexpression, by directly maintaining the transcriptional activity of genes required for normal B-cell development. Hence, B-cell progenitor numbers were significantly reduced in Moz haploinsufficient animals. Interestingly, we find a significant overlap in genes regulated by MOZ, mixed lineage leukemia 1, and mixed lineage leukemia 1 cofactor menin. This includes Meis1, a TALE class homeobox transcription factor required for B-cell development, characteristically upregulated as a result of MLL1 translocations in leukemia. We demonstrate that MOZ localizes to the Meis1 locus in pre-B-cells and maintains Meis1 expression. Our results suggest that even partial inhibition of MOZ may reduce the proliferative capacity of MEIS1, and HOX-driven lymphoma and leukemia cells.

Experimental and spontaneous metastasis assays can result in divergence in clonal architecture.

Intratumoural heterogeneity is associated with poor outcomes in breast cancer. To understand how malignant clones survive and grow in metastatic niches, in vivo models using cell lines and patient-derived xenografts (PDX) have become the gold standard. Injections of cancer cells in orthotopic sites (spontaneous metastasis assays) or into the vasculature (experimental metastasis assays) have been used interchangeably to study the metastatic cascade from early events or post-intravasation, respectively. However, less is known about how these different routes of injection impact heterogeneity. Herein we directly compared the clonality of spontaneous and experimental metastatic assays using the human cell line MDA-MB-231 and a PDX model. Genetic barcoding was used to study the fitness of the subclones in primary and metastatic sites. Using spontaneous assays, we found that intraductal injections resulted in less diverse tumours compared to other routes of injections. Using experimental metastasis assays via tail vein injection of barcoded MDA-MB-231 cells, we also observed an asymmetry in metastatic heterogeneity between lung and liver that was not observed using spontaneous metastasis assays. These results demonstrate that these assays can result in divergent clonal outputs in terms of metastatic heterogeneity and provide a better understanding of the biases inherent to each technique.

Loss of TAF8 causes TFIID dysfunction and p53-mediated apoptotic neuronal cell death.

Mutations in genes encoding general transcription factors cause neurological disorders. Despite clinical prominence, the consequences of defects in the basal transcription machinery during brain development are unclear. We found that loss of the TATA-box binding protein-associated factor TAF8, a component of the general transcription factor TFIID, in the developing central nervous system affected the expression of many, but notably not all genes. Taf8 deletion caused apoptosis, unexpectedly restricted to forebrain regions. Nuclear levels of the transcription factor p53 were elevated in the absence of TAF8, as were the mRNAs of the pro-apoptotic p53 target genes Noxa, Puma and Bax. The cell death in Taf8 forebrain regions was completely rescued by additional loss of p53, but Taf8 and p53 brains failed to initiate a neuronal expression program. Taf8 deletion caused aberrant transcription of promoter regions and splicing anomalies. We propose that TAF8 supports the directionality of transcription and co-transcriptional splicing, and that failure of these processes causes p53-induced apoptosis of neuronal cells in the developing mouse embryo.

Computational Screening of Anti-Cancer Drugs Identifies a New BRCA Independent Gene Expression Signature to Predict Breast Cancer Sensitivity to Cisplatin.

The development of therapies that target specific disease subtypes has dramatically improved outcomes for patients with breast cancer. However, survival gains have not been uniform across patients, even within a given molecular subtype. Large collections of publicly available drug screening data matched with transcriptomic measurements have facilitated the development of computational models that predict response to therapy. Here, we generated a series of predictive gene signatures to estimate the sensitivity of breast cancer samples to 90 drugs, comprising FDA-approved drugs or compounds in early development. To achieve this, we used a cell line-based drug screen with matched transcriptomic data to derive in silico models that we validated in large independent datasets obtained from cell lines and patient-derived xenograft (PDX) models. Robust computational signatures were obtained for 28 drugs and used to predict drug efficacy in a set of PDX models. We found that our signature for cisplatin can be used to identify tumors that are likely to respond to this drug, even in absence of the BRCA-1 mutation routinely used to select patients for platinum-based therapies. This clinically relevant observation was confirmed in multiple PDXs. Our study foreshadows an effective delivery approach for precision medicine.

Histone H2Bub1 deubiquitylation is essential for mouse development, but does not regulate global RNA polymerase II transcription.

Co-activator complexes dynamically deposit post-translational modifications (PTMs) on histones, or remove them, to regulate chromatin accessibility and/or to create/erase docking surfaces for proteins that recognize histone PTMs. SAGA (Spt-Ada-Gcn5 Acetyltransferase) is an evolutionary conserved multisubunit co-activator complex with modular organization. The deubiquitylation module (DUB) of mammalian SAGA complex is composed of the ubiquitin-specific protease 22 (USP22) and three adaptor proteins, ATXN7, ATXN7L3 and ENY2, which are all needed for the full activity of the USP22 enzyme to remove monoubiquitin (ub1) from histone H2B. Two additional USP22-related ubiquitin hydrolases (called USP27X or USP51) have been described to form alternative DUBs with ATXN7L3 and ENY2, which can also deubiquitylate H2Bub1. Here we report that USP22 and ATXN7L3 are essential for normal embryonic development of mice, however their requirements are not identical during this process, as Atxn7l3-/- embryos show developmental delay already at embryonic day (E) 7.5, while Usp22-/- embryos are normal at this stage, but die at E14.5. Global histone H2Bub1 levels were only slightly affected in Usp22 null embryos, in contrast H2Bub1 levels were strongly increased in Atxn7l3 null embryos and derived cell lines. Our transcriptomic analyses carried out from wild type and Atxn7l3-/- mouse embryonic stem cells (mESCs), or primary mouse embryonic fibroblasts (MEFs) suggest that the ATXN7L3-related DUB activity regulates only a subset of genes in both cell types. However, the gene sets and the extent of their deregulation were different in mESCs and MEFs. Interestingly, the strong increase of H2Bub1 levels observed in the Atxn7l3-/- mESCs, or Atxn7l3-/- MEFs, does not correlate with the modest changes in RNA Polymerase II (Pol II) occupancy and lack of changes in Pol II elongation observed in the two Atxn7l3-/- cellular systems. These observations together indicate that deubiquitylation of histone H2Bub1 does not directly regulate global Pol II transcription elongation.

The site of breast cancer metastases dictates their clonal composition and reversible transcriptomic profile.

Intratumoral heterogeneity is a driver of breast cancer progression, but the nature of the clonal interactive network involved in this process remains unclear. Here, we optimized the use of optical barcoding to visualize and characterize 31 cancer subclones in vivo. By mapping the clonal composition of thousands of metastases in two clinically relevant sites, the lungs and liver, we found that metastases were highly polyclonal in lungs but not in the liver. Furthermore, the transcriptome of the subclones varied according to their metastatic niche. We also identified a reversible niche-driven signature that was conserved in lung and liver metastases collected during patient autopsies. Among this signature, we found that the tumor necrosis factor-α pathway was up-regulated in lung compared to liver metastases, and inhibition of this pathway affected metastasis diversity. These results highlight that the cellular and molecular heterogeneity observed in metastases is largely dictated by the tumor microenvironment.

Co-translational assembly of mammalian nuclear multisubunit complexes.

Cells dedicate significant energy to build proteins often organized in multiprotein assemblies with tightly regulated stoichiometries. As genes encoding subunits assembling in a multisubunit complex are dispersed in the genome of eukaryotes, it is unclear how these protein complexes assemble. Here, we show that mammalian nuclear transcription complexes (TFIID, TREX-2 and SAGA) composed of a large number of subunits, but lacking precise architectural details are built co-translationally. We demonstrate that dimerization domains and their positions in the interacting subunits determine the co-translational assembly pathway (simultaneous or sequential). The lack of co-translational interaction can lead to degradation of the partner protein. Thus, protein synthesis and complex assembly are linked in building mammalian multisubunit complexes, suggesting that co-translational assembly is a general principle in mammalian cells to avoid non-specific interactions and protein aggregation. These findings will also advance structural biology by defining endogenous co-translational building blocks in the architecture of multisubunit complexes.

Loss of p53 Causes Stochastic Aberrant X-Chromosome Inactivation and Female-Specific Neural Tube Defects.

Neural tube defects (NTDs) are common birth defects in humans and show an unexplained female bias. Female mice lacking the tumor suppressor p53 display NTDs with incomplete penetrance. We found that the combined loss of pro-apoptotic BIM and p53 caused 100% penetrant, female-exclusive NTDs, which allowed us to investigate the female-specific functions of p53. We report that female p53-/- embryonic neural tube samples show fewer cells with inactive X chromosome markers Xist and H3K27me3 and a concomitant increase in biallelic expression of the X-linked genes, Huwe1 and Usp9x. Decreased Xist and increased X-linked gene expression was confirmed by RNA sequencing. Moreover, we found that p53 directly bound response elements in the X chromosome inactivation center (XIC). Together, these findings suggest p53 directly activates XIC genes, without which there is stochastic failure in X chromosome inactivation, and that X chromosome inactivation failure may underlie the female bias in neural tube closure defects.