A LINE1-Nucleolin Partnership Regulates Early Development and ESC Identity.

Transposable elements represent nearly half of mammalian genomes and are generally described as parasites, or "junk DNA." The LINE1 retrotransposon is the most abundant class and is thought to be deleterious for cells, yet it is paradoxically highly expressed during early development. Here, we report that LINE1 plays essential roles in mouse embryonic stem cells (ESCs) and pre-implantation embryos. In ESCs, LINE1 acts as a nuclear RNA scaffold that recruits Nucleolin and Kap1/Trim28 to repress Dux, the master activator of a transcriptional program specific to the 2-cell embryo. In parallel, LINE1 RNA mediates binding of Nucleolin and Kap1 to rDNA, promoting rRNA synthesis and ESC self-renewal. In embryos, LINE1 RNA is required for Dux silencing, synthesis of rRNA, and exit from the 2-cell stage. The results reveal an essential partnership between LINE1 RNA, Nucleolin, Kap1, and peri-nucleolar chromatin in the regulation of transcription, developmental potency, and ESC self-renewal.

The H3.3 chaperone Hira complex orchestrates oocyte developmental competence.

Successful reproduction requires an oocyte competent to sustain early embryo development. By the end of oogenesis, the oocyte has entered a transcriptionally silenced state, the mechanisms and significance of which remain poorly understood. Histone H3.3, a histone H3 variant, has unique cell cycle-independent functions in chromatin structure and gene expression. Here, we have characterised the H3.3 chaperone Hira/Cabin1/Ubn1 complex, showing that loss of function of any of these subunits causes early embryogenesis failure in mouse. Transcriptome and nascent RNA analyses revealed that transcription is aberrantly silenced in mutant oocytes. Histone marks, including H3K4me3 and H3K9me3, are reduced and chromatin accessibility is impaired in Hira/Cabin1 mutants. Misregulated genes in mutant oocytes include Zscan4d, a two-cell specific gene involved in zygote genome activation. Overexpression of Zscan4 in the oocyte partially recapitulates the phenotypes of Hira mutants and Zscan4 knockdown in Cabin1 mutant oocytes partially restored their developmental potential, illustrating that temporal and spatial expression of Zscan4 is fine-tuned at the oocyte-to-embryo transition. Thus, the H3.3 chaperone Hira complex has a maternal effect function in oocyte developmental competence and embryogenesis, through modulating chromatin condensation and transcriptional quiescence.

Extracellular vesicles secreted by cumulus cells contain microRNAs that are potential regulatory factors of mouse oocyte developmental competence.

The role of cumulus cells (CCs) in the acquisition of oocyte developmental competence is not yet fully understood. In a previous study, we matured cumulus-denuded fully-grown mouse oocytes to metaphase II (MII) on a feeder layer of CCs (FL-CCs) isolated from developmentally competent (FL-SN-CCs) or incompetent (FL-NSN-CCs) SN (surrounded nucleolus) or NSN (not surrounding nucleolus) oocytes, respectively. We observed that oocytes cultured on the former could develop into blastocysts, while those matured on the latter arrested at the 2-cell stage. To investigate the CC factors contributing to oocyte developmental competence, here we focused on the CCs' release into the medium of extracellular vesicles (EVs) and on their miRNA content. We found that, during the 15-h transition to MII, both FL-SN-CCs and FL-NSN-CCs release EVs that can be detected, by confocal microscopy, inside the zona pellucida (ZP) or the ooplasm. The majority of EVs are <200 nm in size, which is compatible with their ability to cross the ZP. Next-generation sequencing of the miRNome of FL-SN-CC versus FL-NSN-CC EVs highlighted 74 differentially expressed miRNAs, with 43 up- and 31 down-regulated. Although most of these miRNAs do not have known roles in the ovary, in silico functional analysis showed that seven of these miRNAs regulate 71 target genes with specific roles in meiosis resumption (N = 24), follicle growth (N = 23), fertilization (N = 1), and the acquisition of oocyte developmental competence (N = 23). Overall, our results indicate CC EVs as emerging candidates of the CC-to-oocyte communication axis and uncover a group of miRNAs as potential regulatory factors.

HIRA contributes to zygote formation in mice and is implicated in human 1PN zygote phenotype.

Elucidating the mechanisms underpinning fertilisation is essential to optimising IVF procedures. One of the critical steps involves paternal chromatin reprogramming, in which compacted sperm chromatin packed by protamines is removed by oocyte factors and new histones, including histone H3.3, are incorporated. HIRA is the main H3.3 chaperone governing this protamine-to-histone exchange. Failure of this step results in abnormally fertilised zygotes containing only one pronucleus (1PN), in contrast to normal two-pronuclei (2PN) zygotes. 1PN zygotes are frequently observed in IVF treatments, but the genotype-phenotype correlation remains elusive. We investigated the maternal functions of two other molecules of the HIRA complex, Cabin1 and Ubn1, in mouse. Loss-of-function Cabin1 and Ubn1 mouse models were developed: their zygotes displayed an abnormal 1PN zygote phenotype. We then studied human 1PN zygotes and found that the HIRA complex was absent in 1PN zygotes that lacked the male pronucleus. This shows that the role of the HIRA complex in male pronucleus formation potentially has coherence from mice to humans. Furthermore, rescue experiments in mouse showed that the abnormal 1PN phenotype derived from Hira mutants could be resolved by overexpression of HIRA. We have demonstrated that HIRA complex regulates male pronucleus formation in mice and is implicated in humans, that both CABIN1 and UBN1 components of the HIRA complex are equally essential for male pronucleus formation, and that rescue is feasible.

Ectopic pregnancy and epithelial to mesenchymal transition: is there a link?

Ectopic pregnancy (EP) is defined as the implantation of an embryo outside of the uterus and is a leading cause of first trimester maternal mortality and morbidity. This article discusses a possible role for epithelial to mesenchymal transition in the pathogenesis of EP, given the notable similarity of protein expression between the two processes.

Identifying the Translatome of Mouse NEBD-Stage Oocytes via SSP-Profiling; A Novel Polysome Fractionation Method.

Meiotic maturation of oocyte relies on pre-synthesised maternal mRNA, the translation of which is highly coordinated in space and time. Here, we provide a detailed polysome profiling protocol that demonstrates a combination of the sucrose gradient ultracentrifugation in small SW55Ti tubes with the qRT-PCR-based quantification of 18S and 28S rRNAs in fractionated polysome profile. This newly optimised method, named Scarce Sample Polysome Profiling (SSP-profiling), is suitable for both scarce and conventional sample sizes and is compatible with downstream RNA-seq to identify polysome associated transcripts. Utilising SSP-profiling we have assayed the translatome of mouse oocytes at the onset of nuclear envelope breakdown (NEBD)-a developmental point, the study of which is important for furthering our understanding of the molecular mechanisms leading to oocyte aneuploidy. Our analyses identified 1847 transcripts with moderate to strong polysome occupancy, including abundantly represented mRNAs encoding mitochondrial and ribosomal proteins, proteasomal components, glycolytic and amino acids synthetic enzymes, proteins involved in cytoskeleton organization plus RNA-binding and translation initiation factors. In addition to transcripts encoding known players of meiotic progression, we also identified several mRNAs encoding proteins of unknown function. Polysome profiles generated using SSP-profiling were more than comparable to those developed using existing conventional approaches, being demonstrably superior in their resolution, reproducibility, versatility, speed of derivation and downstream protocol applicability.

Treatment effectiveness of levamisole plus prednisolone on oral lichen planus patients with emphasis on levamisole-induced agranulocytosis or pancytopenia.

BACKGROUND/PURPOSE: Physicians' and dentists' knowledge of levamisole-induced agranulocytosis or pancytopenia remains incomplete. This study aimed to evaluate the treatment effectiveness of levamisole plus prednisolone on oral lichen planus (OLP) patients with emphasis on levamisole-induced hematological changes. METHODS: Ninety patients with erosive OLP were given 120 mg/day new levamisole (Levazol) and 15 mg/day prednisolone for three consecutive days each week. Three cases with levamisole-induced blood-cytopenias were assessed and treated within one year. RESULTS: Most patients reported significant pain relief and showed no evidence of erosive OLP after 4-8 weeks of treatment with few side effects; nevertheless, three female patients developed agranulocytosis or granulocytopenia with concomitant thrombocytopenia or pancytopenia within 2-6 weeks after levamisole (Levazol) treatment. One case with previously unknown double episodes of agranulocytosis revealed her first episode following interruption of levamisole (Decaris) treatment for 4 months. High fever and sore throat were the most common symptoms, but two agranulocytosis cases remained asymptomatic one week before diagnosis, and were treated with levamisole withdrawal and empiric antimicrobial initiation as well as utilization of granulocyte colony-stimulating factors. Neutrophil recovery took about 1 week, but over 4 weeks in one of the cases (an elderly patient) with septic shock. CONCLUSION: Agranulocytosis or pancytopenia usually developed within 2 months after levamisole treatment, but it might be delayed. Agranulocytosis was more likely to occur in females and onset was acute. Levamisole is an effective immunomodulator for OLP patients; however, it should be used with caution and administered with regular blood monitoring.

Chd1 is essential for the high transcriptional output and rapid growth of the mouse epiblast.

The pluripotent mammalian epiblast undergoes unusually fast cell proliferation. This rapid growth is expected to generate a high transcriptional demand, but the underlying mechanisms remain unknown. We show here that the chromatin remodeler Chd1 is required for transcriptional output and development of the mouse epiblast. Chd1(-/-) embryos exhibit proliferation defects and increased apoptosis, are smaller than controls by E5.5 and fail to grow, to become patterned or to gastrulate. Removal of p53 allows progression of Chd1(-/-) mutants only to E7.0-8.0, highlighting the crucial requirement for Chd1 during early post-implantation development. Chd1(-/-) embryonic stem cells (ESCs) have a self-renewal defect and a genome-wide reduction in transcriptional output at both known mRNAs and intergenic transcripts. These transcriptional defects were only uncovered when cell number-normalized approaches were used, and correlate with a lower engagement of RNAP II with transcribed genes in Chd1(-/-) ESCs. We further show that Chd1 directly binds to ribosomal DNA, and that both Chd1(-/-) epiblast cells in vivo and ESCs in vitro express significantly lower levels of ribosomal RNA. In agreement with these findings, mutant cells in vivo and in vitro exhibit smaller and more elongated nucleoli. Thus, the RNA output by both Pol I and II is reduced in Chd1(-/-) cells. Our data indicate that Chd1 promotes a globally elevated transcriptional output required to sustain the distinctly rapid growth of the mouse epiblast.

Distributed Newton Methods for Deep Neural Networks.

Deep learning involves a difficult nonconvex optimization problem with a large number of weights between any two adjacent layers of a deep structure. To handle large data sets or complicated networks, distributed training is needed, but the calculation of function, gradient, and Hessian is expensive. In particular, the communication and the synchronization cost may become a bottleneck. In this letter, we focus on situations where the model is distributedly stored and propose a novel distributed Newton method for training deep neural networks. By variable and feature-wise data partitions and some careful designs, we are able to explicitly use the Jacobian matrix for matrix-vector products in the Newton method. Some techniques are incorporated to reduce the running time as well as memory consumption. First, to reduce the communication cost, we propose a diagonalization method such that an approximate Newton direction can be obtained without communication between machines. Second, we consider subsampled Gauss-Newton matrices for reducing the running time as well as the communication cost. Third, to reduce the synchronization cost, we terminate the process of finding an approximate Newton direction even though some nodes have not finished their tasks. Details of some implementation issues in distributed environments are thoroughly investigated. Experiments demonstrate that the proposed method is effective for the distributed training of deep neural networks. Compared with stochastic gradient methods, it is more robust and may give better test accuracy.

Histone variant H3.3 maintains a decondensed chromatin state essential for mouse preimplantation development.

Histone variants can replace canonical histones in the nucleosome and modify chromatin structure and gene expression. The histone variant H3.3 preferentially associates with active chromatin and has been implicated in the regulation of a diverse range of developmental processes. However, the mechanisms by which H3.3 may regulate gene activity are unclear and gene duplication has hampered an analysis of H3.3 function in mouse. Here, we report that the specific knockdown of H3.3 in fertilized mouse zygotes leads to developmental arrest at the morula stage. This phenotype can be rescued by exogenous H3.3 but not by canonical H3.1 mRNA. Loss of H3.3 leads to over-condensation and mis-segregation of chromosomes as early as the two-cell stage, with corresponding high levels of aneuploidy, but does not appear to affect zygotic gene activation at the two-cell stage or lineage gene transcription at the morula stage. H3.3-deficient embryos have significantly reduced levels of markers of open chromatin, such as H3K36me2 and H4K16Ac. Importantly, a mutation in H3.3K36 that disrupts H3K36 methylation (H3.3K36R) does not rescue the H3.3 knockdown (KD) phenotype. In addition, H3.3 KD embryos have increased incorporation of linker H1. Knockdown of Mof (Kat8), an acetyltransferase specific for H4K16, similarly leads to excessive H1 incorporation. Remarkably, pan-H1 RNA interference (RNAi) partially rescues the chromosome condensation of H3.3 KD embryos and allows development to the blastocyst stage. These results reveal that H3.3 mediates a balance between open and condensed chromatin that is crucial for the fidelity of chromosome segregation during early mouse development.

Epigenetic reprogramming of the zygote in mice and men: on your marks, get set, go!

Gametogenesis (spermatogenesis and oogenesis) is accompanied by the acquisition of gender-specific epigenetic marks, such as DNA methylation, histone modifications and regulation by small RNAs, to form highly differentiated, but transcriptionally silent cell-types in preparation for fertilisation. Upon fertilisation, extensive global epigenetic reprogramming takes place to remove the previously acquired epigenetic marks and produce totipotent zygotic states. It is the aim of this review to delineate the cellular and molecular events involved in maternal, paternal and zygotic epigenetic reprogramming from the time of gametogenesis, through fertilisation, to the initiation of zygotic genome activation for preimplantation embryonic development. Recent studies have begun to uncover the indispensable functions of epigenetic players during gametogenesis, fertilisation and preimplantation embryo development, and a more comprehensive understanding of these early events will be informative for increasing pregnancy success rates, adding particular value to assisted fertility programmes.

Subsampled Hessian Newton Methods for Supervised Learning.

Newton methods can be applied in many supervised learning approaches. However, for large-scale data, the use of the whole Hessian matrix can be time-consuming. Recently, subsampled Newton methods have been proposed to reduce the computational time by using only a subset of data for calculating an approximation of the Hessian matrix. Unfortunately, we find that in some situations, the running speed is worse than the standard Newton method because cheaper but less accurate search directions are used. In this work, we propose some novel techniques to improve the existing subsampled Hessian Newton method. The main idea is to solve a two-dimensional subproblem per iteration to adjust the search direction to better minimize the second-order approximation of the function value. We prove the theoretical convergence of the proposed method. Experiments on logistic regression, linear SVM, maximum entropy, and deep networks indicate that our techniques significantly reduce the running time of the subsampled Hessian Newton method. The resulting algorithm becomes a compelling alternative to the standard Newton method for large-scale data classification.

Large-scale linear rankSVM.

Linear rankSVM is one of the widely used methods for learning to rank. Although its performance may be inferior to nonlinear methods such as kernel rankSVM and gradient boosting decision trees, linear rankSVM is useful to quickly produce a baseline model. Furthermore, following its recent development for classification, linear rankSVM may give competitive performance for large and sparse data. A great deal of works have studied linear rankSVM. The focus is on the computational efficiency when the number of preference pairs is large. In this letter, we systematically study existing works, discuss their advantages and disadvantages, and propose an efficient algorithm. We discuss different implementation issues and extensions with detailed experiments. Finally, we develop a robust linear rankSVM tool for public use.

One-Class SVM Probabilistic Outputs.

One-class support vector machine (SVM) is an extension of SVM to handle unlabeled data. As a mature technique for outlier detection, one-class SVM has been widely used in many applications. However, similar to standard two-class SVM, the design of one-class SVM does not give probabilistic outputs. For two-class SVM, some methods have been proposed to effectively obtain probabilistic outputs, but due to the difficulty of no-label information, less attention has been paid to one-class SVM. Our aim in this work is to propose some practically viable techniques to generate probabilistic outputs for one-class SVM. We investigate existing methods for two-class SVM and explain why they may not be suitable for one-class SVM. Due to the lack of label information, we think a feasible setting is to have probabilities mimic to the decision values of training data. Based on this principle, we propose several new methods. Detailed experiments on both artificial and real-world data demonstrate the effectiveness of the proposed methods.

Jak/Stat3 signaling promotes somatic cell reprogramming by epigenetic regulation.

Although leukemia inhibitory factor (LIF) maintains the ground state pluripotency of mouse embryonic stem cells and induced pluripotent stem cells (iPSCs) by activating the Janus kinase/signal transducer and activator of transcription 3 (Jak/Stat3) pathway, the mechanism remained unclear. Stat3 has only been shown to promote complete reprogramming of epiblast and neural stem cells and partially reprogrammed cells (pre-iPSCs). We investigated if and how Jak/Stat3 activation promotes reprogramming of terminally differentiated mouse embryonic fibroblasts (MEFs). We demonstrated that activated Stat3 not only promotes but also is essential for the pluripotency establishment of MEFs during reprogramming. We further demonstrated that during this process, inhibiting Jak/Stat3 activity blocks demethylation of Oct4 and Nanog regulatory elements in induced cells, which are marked by suppressed endogenous pluripotent gene expression. These are correlated with the significant upregulation of DNA methyltransferase (Dnmt) 1 and histone deacetylases (HDACs) expression as well as the increased expression of lysine-specific histone demethylase 2 and methyl CpG binding protein 2. Inhibiting Jak/Stat3 also blocks the expression of Dnmt3L, which is correlated with the failure of retroviral transgene silencing. Furthermore, Dnmt or HDAC inhibitor but not overexpression of Nanog significantly rescues the reprogramming arrested by Jak/Stat3 inhibition or LIF deprivation. Finally, we demonstrated that LIF/Stat3 signal also represents the prerequisite for complete reprogramming of pre-iPSCs. We conclude that Jak/Stat3 activity plays a fundamental role to promote pluripotency establishment at the epigenetic level, by facilitating DNA demethylation/de novo methylation, and open-chromatin formation during late-stage reprogramming.

A study on L2-loss (squared hinge-loss) multiclass SVM.

Crammer and Singer's method is one of the most popular multiclass support vector machines (SVMs). It considers L1 loss (hinge loss) in a complicated optimization problem. In SVM, squared hinge loss (L2 loss) is a common alternative to L1 loss, but surprisingly we have not seen any paper studying the details of Crammer and Singer's method using L2 loss. In this letter, we conduct a thorough investigation. We show that the derivation is not trivial and has some subtle differences from the L1 case. Details provided in this work can be a useful reference for those who intend to use Crammer and Singer's method with L2 loss. They do not need a tedious process to derive everything by themselves. Furthermore, we present some new results on and discussion of both L1- and L2-loss formulations.

Recent advances in the understanding of tubal ectopic pregnancy.

Ectopic pregnancy (EP) is described as the implantation of an embryo outside the normal uterine cavity. It most commonly occurs in the fallopian tube, hence termed a tubal ectopic pregnancy (tEP). It is a gynaecological emergency and remains the leading cause of direct maternal mortality related to the first trimester of pregnancy worldwide. This article explores the emergence of additional risk factors for tEP, showing new evidence for identifying patient risk factors and highlighting potential areas of research. Additionally, we discuss the up-to-date patient-centred approach for the diagnosis, management and counselling of patients with tEP and ongoing clinical trials for the improvement of medical management.

Debiased ambient vibrations optical coherence elastography to profile cell, organoid and tissue mechanical properties.

The role of the mechanical environment in defining tissue function, development and growth has been shown to be fundamental. Assessment of the changes in stiffness of tissue matrices at multiple scales has relied mostly on invasive and often specialist equipment such as AFM or mechanical testing devices poorly suited to the cell culture workflow.In this paper, we have developed a unbiased passive optical coherence elastography method, exploiting ambient vibrations in the sample that enables real-time noninvasive quantitative profiling of cells and tissues. We demonstrate a robust method that decouples optical scattering and mechanical properties by actively compensating for scattering associated noise bias and reducing variance. The efficiency for the method to retrieve ground truth is validated in silico and in vitro, and exemplified for key applications such as time course mechanical profiling of bone and cartilage spheroids, tissue engineering cancer models, tissue repair models and single cell. Our method is readily implementable with any commercial optical coherence tomography system without any hardware modifications, and thus offers a breakthrough in on-line tissue mechanical assessment of spatial mechanical properties for organoids, soft tissues and tissue engineering.

A Study on Truncated Newton Methods for Linear Classification.

Truncated Newton (TN) methods have been a useful technique for large-scale optimization. Instead of obtaining the full Newton direction, a truncated method approximately solves the Newton equation with an inner conjugate gradient (CG) procedure (TNCG for the whole method). These methods have been employed to efficiently solve linear classification problems. However, even in this deeply studied field, various theoretical and numerical aspects were not completely explored. The first contribution of this work is to comprehensively study the global and local convergence when TNCG is applied to linear classification. Because of the lack of twice differentiability under some losses, many past works cannot be applied here. We prove various missing pieces of theory from scratch and clarify many proper references. The second contribution is to study the termination of the CG method. For the first time when TNCG is applied to linear classification, we show that the inner stopping condition strongly affects the convergence speed. We propose using a quadratic stopping criterion to achieve both robustness and efficiency. The third contribution is that of combining the study on inner stopping criteria with that of preconditioning. We discuss how convergence theory is affected by preconditioning and finally propose an effective preconditioned TNCG.

The Effect of an Er, Cr: YSGG Laser Combined with Implantoplasty Treatment on Implant Surface Roughness and Morphologic Analysis: A Pilot In Vitro Study.

Although laser irradiation and implantoplasty (IP) are both treatment options for peri-implantitis, no studies have yet combined these two treatment solutions. The aim of this study was to identify the effect of an Er, Cr: YSGG laser on the IP surface. In experiment 1, TiUnite anodized surface implants were treated with an Er, Cr: YSGG laser at 0.5 to 2 W on the panel energy setting and 20 Hz under water irrigation. In experiment 2, all implant surfaces were treated with the IP procedure first, then irradiated with the Er, Cr: YSGG laser. All samples were analyzed by stereomicroscopy, scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDS), and surface topography. Stereomicroscopy and SEM revealed no obvious surface change at any energy setting once the surface was polished with the IP procedure, whereas damage was caused to the TiUnite original implant surface when the Er, Cr: YSGG laser panel energy was set at 1 W or higher. EDS showed no significant difference in element composition once the surface was polished with the IP procedure, while a compositional change was detected when the Er, Cr: YSGG laser panel energy was set to 0.5 W or higher to irradiate the original TiUnite surface. Surface roughness may be related to laser irradiation energy, but no significant changes occurred following IP. These results indicated that the Er, Cr: YSGG laser may have little effect on the post-IP surface compared with the virgin TiUnite surface.