SUMMARY Alternative splicing (AS) of pre-mRNA is utilized by higher eukaryotes to achieve increased transcriptome and proteomic complexity. The serine/arginine (SR) splicing factors regulate tissue- or cell-type-specific AS in a concentration- and phosphorylation-dependent manner. However, the mechanisms that modulate the cellular levels of active SR proteins remain to be elucidated. In the present study, we provide evidence for a role for the long nuclear-retained regulatory RNA (nrRNA), MALAT1 in AS regulation. MALAT1 interacts with SR proteins and influences the distribution of these and other splicing factors in nuclear speckle domains. Depletion of MALAT1 or overexpression of an SR protein changes the AS of a similar set of endogenous pre-mRNAs. Furthermore, MALAT1 regulates cellular levels of phosphorylated forms of SR proteins. Taken together, our results suggest that MALAT1 regulates AS by modulating the levels of active SR proteins. Our results further highlight the role for an nrRNA in the regulation of gene expression.

The long noncoding MALAT1 RNA is upregulated in cancer tissues and its elevated expression is associated with hyper-proliferation, but the underlying mechanism is poorly understood. We demonstrate that MALAT1 levels are regulated during normal cell cycle progression. Genome-wide transcriptome analyses in normal human diploid fibroblasts reveal that MALAT1 modulates the expression of cell cycle genes and is required for G1/S and mitotic progression. Depletion of MALAT1 leads to activation of p53 and its target genes. The cell cycle defects observed in MALAT1-depleted cells are sensitive to p53 levels, indicating that p53 is a major downstream mediator of MALAT1 activity. Furthermore, MALAT1-depleted cells display reduced expression of B-MYB (Mybl2), an oncogenic transcription factor involved in G2/M progression, due to altered binding of splicing factors on B-MYB pre-mRNA and aberrant alternative splicing. In human cells, MALAT1 promotes cellular proliferation by modulating the expression and/or pre-mRNA processing of cell cycle–regulated transcription factors. These findings provide mechanistic insights on the role of MALAT1 in regulating cellular proliferation.

SUMMARY Origin recognition complex (ORC) plays critical roles in the initiation of DNA replication and cell-cycle progression. In metazoans, ORC associates with origin DNA during G1 and with heterochromatin in postreplicated cells. However, what regulates the binding of ORC to chromatin is not understood. We have identified a highly conserved, leucine-rich repeats and WD40 repeat domain-containing protein 1 (LRWD1) or ORC-associated (ORCA) in human cells that interacts with ORC and modulates chromatin association of ORC. ORCA colocalizes with ORC and shows similar cell-cycle dynamics. We demonstrate that ORCA efficiently recruits ORC to chromatin. Depletion of ORCA in human primary cells and embryonic stem cells results in loss of ORC association to chromatin, concomitant reduction of MCM binding, and a subsequent accumulation in G1 phase. Our results suggest ORCA-mediated association of ORC to chromatin is critical to initiate preRC assembly in G1 and chromatin organization in post-G1 cells.

The origin recognition complex (ORC) is a DNA replication initiator protein also known to be involved in diverse cellular functions including gene silencing, sister chromatid cohesion, telomere biology, heterochromatin localization, centromere and centrosome activity, and cytokinesis. We show that, in human cells, multiple ORC subunits associate with hetereochromatin protein 1 (HP1) α-and HP1β-containing heterochromatic foci. Fluorescent bleaching studies indicate that multiple subcomplexes of ORC exist at heterochromatin, with Orc1 stably associating with heterochromatin in G1 phase, whereas other ORC subunits have transient interactions throughout the cell-division cycle. Both Orc1 and Orc3 directly bind to HP1α, and two domains of Orc3, a coiled-coil domain and a mod-interacting region domain, can independently bind to HP1α; however, both are essential for in vivo localization of Orc3 to heterochromatic foci. Direct binding of both Orc1 and Orc3 to HP1 suggests that, after the degradation of Orc1 at the G1/S boundary, Orc3 facilitates assembly of ORC/HP1 proteins to chromatin. Although depletion of Orc2 and Orc3 subunits by siRNA caused loss of HP1α association to heterochromatin, loss of Orc1 and Orc5 caused aberrant HP1α distribution only to pericentric heterochromatin-surrounding nucleoli. Depletion of HP1α from human cells also shows loss of Orc2 binding to heterochromatin, suggesting that ORC and HP1 proteins are mutually required for each other to bind to heterochromatin. Similar to HP1α-depleted cells, Orc2 and Orc3 siRNA-treated cells also show loss of compaction at satellite repeats, suggesting that ORC together with HP1 proteins may be involved in organizing higherorder chromatin structure and centromere function.centromere | origin recognition complex | pericentric heterochromatin | chromatin structure

It is well known that DNA sequence contains a certain amount of transcription factors (TF) binding sites, and only part of them are identified through biological experiments. However, these experiments are expensive and time-consuming. To overcome these problems, some computational methods, based on k-mer features or convolutional neural networks, have been proposed to identify TF binding sites from DNA sequences. Although these methods have good performance, the context information that relates to TF binding sites is still lacking. Research indicates that standard recurrent neural networks (RNN) and its variants have better performance in time-series data compared with other models. In this study, we propose a model, named KEGRU, to identify TF binding sites by combining Bidirectional Gated Recurrent Unit (GRU) network with k-mer embedding. Firstly, DNA sequences are divided into k-mer sequences with a specified length and stride window. And then, we treat each k-mer as a word and pre-trained word representation model though word2vec algorithm. Thirdly, we construct a deep bidirectional GRU model for feature learning and classification. Experimental results have shown that our method has better performance compared with some state-of-the-art methods. Additional experiments about embedding strategy show that k-mer embedding will be helpful to enhance model performance. The robustness of KEGRU is proved by experiments with different k-mer length, stride window and embedding vector dimension.