The ability of p53 to induce apoptosis plays an important role in tumor suppression. Here, we describe a previously unknown posttranslational modification of the DNA-binding domain of p53. This modification, acetylation of lysine 120 (K120), occurs rapidly after DNA damage and is catalyzed by the MYST family acetyltransferases hMOF and TIP60. Mutation of K120 to arginine, as occurs in human cancer, debilitates K120 acetylation and diminishes p53-mediated apoptosis without affecting cell-cycle arrest. The K120R mutation selectively blocks the transcription of proapoptotic target genes such as BAX and PUMA while the nonapoptotic targets p21 and hMDM2 remain unaffected. Consistent with this, depletion of hMOF and/or TIP60 inhibits the ability of p53 to activate BAX and PUMA transcription. Furthermore, the acetyllysine 120 (acetyl-K120) form of p53 specifically accumulates at proapoptotic target genes. These data suggest that K120 acetylation may help distinguish the cell-cycle arrest and apoptotic functions of p53.

Abnormal epigenetic regulation has been implicated in oncogenesis. We report here the identification of somatic mutations by exome sequencing in acute monocytic leukemia, the M5 subtype of acute myeloid leukemia (AML-M5). We discovered mutations in DNMT3A (encoding DNA methyltransferase 3A) in 23 of 112 (20.5%) cases. The DNMT3A mutants showed reduced enzymatic activity or aberrant affinity to histone H3 in vitro. Notably, there were alterations of DNA methylation patterns and/or gene expression profiles (such as HOXB genes) in samples with DNMT3A mutations as compared with those without such changes. Leukemias with DNMT3A mutations constituted a group of poor prognosis with elderly disease onset and of promonocytic as well as monocytic predominance among AML-M5 individuals. Screening other leukemia subtypes showed Arg882 alterations in 13.6% of acute myelomonocytic leukemia (AML-M4) cases. Our work suggests a contribution of aberrant DNA methyltransferase activity to the pathogenesis of acute monocytic leukemia and provides a useful new biomarker for relevant cases.

BackgroundChemoresistance has long been recognized as a major obstacle in cancer therapy. Clarifying the underlying mechanism of chemoresistance would result in novel strategies to improve patient’s response to chemotherapeutics.MethodslncRNA expression levels in gastric cancer (GC) cells was detected by quantitative real-time PCR (qPCR). MALAT1 shRNAs and overexpression vector were transfected into GC cells to down-regulate or up-regulate MALAT1 expression. In vitro and in vivo assays were performed to investigate the functional role of MALAT1 in autophagy associated chemoresistance.ResultsWe showed that chemoresistant GC cells had higher levels of MALAT1 and increased autophagy compared with parental cells. Silencing of MALAT1 inhibited chemo-induced autophagy, whereas MALAT1 promoted autophagy in gastric cancer cells. Knockdown of MALAT1 sensitized GC cells to chemotherapeutics. MALAT1 acts as a competing endogenous RNA for miR-23b-3p and attenuates the inhibitory effect of miR-23b-3p on ATG12, leading to chemo-induced autophagy and chemoresistance in GC cells.ConclusionsTaken together, our study revealed a novel mechanism of lncRNA-regulated autophagy-related chemoresistance in GC, casting new lights on the understanding of chemoresistance.Electronic supplementary materialThe online version of this article (10.1186/s12943-017-0743-3) contains supplementary material, which is available to authorized users.

Benefitting from large-scale training datasets and the complex training network, Convolutional Neural Networks (CNNs) are widely applied in various fields with high accuracy. However, the training process of CNNs is very time-consuming, where large amounts of training samples and iterative operations are required to obtain high-quality weight parameters. In this paper, we focus on the time-consuming training process of large-scale CNNs and propose a Bi-layered Parallel Training (BPT-CNN) architecture in distributed computing environments. BPT-CNN consists of two main components: (a) an outer-layer parallel training for multiple CNN subnetworks on separate data subsets, and (b) an inner-layer parallel training for each subnetwork. In the outer-layer parallelism, we address critical issues of distributed and parallel computing, including data communication, synchronization, and workload balance. A heterogeneousaware Incremental Data Partitioning and Allocation (IDPA) strategy is proposed, where large-scale training datasets are partitioned and allocated to the computing nodes in batches according to their computing power. To minimize the synchronization waiting during the global weight update process, an Asynchronous Global Weight Update (AGWU) strategy is proposed. In the inner-layer parallelism, we further accelerate the training process for each CNN subnetwork on each computer, where computation steps of convolutional layer and the local weight training are parallelized based on task-parallelism. We introduce task decomposition and scheduling strategies with the objectives of thread-level load balancing and minimum waiting time for critical paths. Extensive experimental results indicate that the proposed BPT-CNN effectively improves the training performance of CNNs while maintaining the accuracy.