Gold and silver plasmonic nanoprobes trace the positions of histone codes

We visualized the distribution of heterochromatin in a single nucleus using plasmonic nanoparticle-conjugated H3K9me3 and H3K27me3 antibodies. Due to distance-dependent plasmonic coupling effects between nanoprobes, their scattering spectra shift to longer wavelengths as the distance between heterochromatin histone markers reduced during oncogene-induced senescence (OIS). These observations were supported by simulating scattering profiles based on considerations of particle numbers, interparticle distances, and the spatial arrangements of plasmonic nanoprobes. Using this plasmon-based colourimetric imaging, we estimated changes in distances between H3K9me3 and H3K27me3 during the formation of senescence-associated heterochromatin foci in OIS cells. We anticipate that the devised analytical technique combined with high-spatial imaging and spectral simulation will eventually lead to a new means of diagnosing and monitoring disease progression and cellular senescence.     

Angiogenic factor AGGF1 blocks neointimal formation after vascular injury via interaction with integrin α7 on vascular smooth muscle cells

Angiogenic factor AGGF1 (AngioGenic factor with G-patch and FHA (Forkhead-Associated) domain 1) blocks neointimal formation (formation of a new or thickened layer of arterial intima) after vascular injury by regulating phenotypic switching of vascular smooth muscle cells (VSMCs). However, the AGGF1 receptor on VSMCs and the underlying molecular mechanisms of its action are unknown. In this study, we used functional analysis of serial AGGF1 deletions to reveal the critical AGGF1 domain involved in VSMC phenotypic switching. This domain was required for VSMC phenotypic switching, proliferation, cell cycle regulation, and migration, as well as the regulation of cell cycle inhibitors cyclin D, p27, and p21. This domain also contains an RDDAPAS motif via which AGGF1 interacts with integrin α7 (ITGA7), but not α8. In addition, we show that AGGF1 enhanced the expression of contractile markers MYH11, α-SMA, and SM22 and inhibited MEK1/2, ERK1/2, and ELK phosphorylation in VSMCs, and that these effects were inhibited by knockdown of ITGA7, but not by knockdown of ITGA8. In vivo, deletion of the VSMC phenotypic switching domain in mice with vascular injury inhibited the functions of AGGF1 in upregulating α-SMA and SM22, inhibiting MEK1/2, ERK1/2, and ELK phosphorylation, in VSMC proliferation, and in blocking neointimal formation. Finally, we show the inhibitory effect of AGGF1 on neointimal formation was blocked by lentivirus-delivered shRNA targeting ITGA7. Our data demonstrate that AGGF1 interacts with its receptor integrin α7 on VSMCs, and this interaction is required for AGGF1 signaling in VSMCs and for attenuation of neointimal formation after vascular injury.     

EGFR/MET promotes hepatocellular carcinoma metastasis by stabilizing tumor cells and resisting to RTKs inhibitors in circulating tumor microemboli

The receptor tyrosine kinases (RTKs) family is well-recognized as vital targets for the treatment of hepatocarcinoma cancer (HCC) clinically, whereas the survival benefit of target therapy sorafenib is not satisfactory for liver cancer patients due to metastasis. EGFR and MET are two molecules of the RTK family that were related to the survival time of liver cancer patients and resistance to targeted therapy in clinical reports. However, the mechanism and clinical therapeutic value of EGFR/MET in HCC metastasis are still not completely clarified. The study confirmed that EGFR/MET was highly expressed in HCC cells and tissues and the phosphorylation was stable after metastasis. The expression of EGFR/MET was up-regulated in circulating tumor microemboli (CTM) to accelerate IL-8 production and resistance to the lethal effect of leukocytes. Meanwhile, highly expressed EGFR/MET effectively regulated the Ras/MAPK pathway and stabilized suspended HCC cells by facilitating proliferation and inhibiting apoptosis. Moreover, EGFR/MET promoted phosphorylation of hetero-RTKs, which was dependent on high-energy phosphoric acid compounds rather than their direct interactions. In conclusion, highly expressed EGFR/MET could be used in CTM identification and suitable for preventing metastasis of HCC in clinical practice.Subject terms: Liver cancer, Metastasis     

基于转录组的肝豆状核变性调控网络的构建和分析

本研究旨在利用生物信息学方法构建经铜诱导的ATP7B基因敲除HepG2细胞系的转录调控网络。探讨关键转录因子在肝豆状核变性发生、发展中的潜在作用机制。收集公共基因表达数据库(gene expression omnibus, GEO)中包含野生型、ATP7B基因敲除型、铜诱导的野生型和铜诱导的ATP7B基因敲除型HepG2细胞系数据。筛选由铜诱导产生的差异表达基因(differentially expressed genes,DEGs)后进行基因本体论(gene ontology,GO)、京都基因和基因组百科全书(Kyoto encyclopedia of genes and genomes, KEGG)富集分析。基于蛋白相互作用网络,识别疾病关键基因和功能模块,并对关键功能模块中的基因进行富集分析。最后,构建转录调控网络,筛选核心转录因子。共筛选出1 034个差异表达基因,其中上调525个,下调509个。上、下调关键功能模块分别包括了3785个和3931个基因。关键功能模块中的基因主要定位于细胞-基质连接、染色体、剪接复合体、核糖体等区域,共同参与了mRNA加工、组蛋白修饰、RNA剪切...     

Patient contrastive learning: A performant,expressive, and practical approach to electrocardiogram modeling

Supervised machine learning applications in health care are often limited due to a scarcity of labeled training data. To mitigate the effect of small sample size, we introduce a pre-training approach, Patient Contrastive Learning of Representations (PCLR), which creates latent representations of electrocardiograms (ECGs) from a large number of unlabeled examples using contrastive learning. The resulting representations are expressive, performant, and practical across a wide spectrum of clinical tasks. We develop PCLR using a large health care system with over 3.2 million 12-lead ECGs and demonstrate that training linear models on PCLR representations achieves a 51% performance increase, on average, over six training set sizes and four tasks (sex classification, age regression, and the detection of left ventricular hypertrophy and atrial fibrillation), relative to training neural network models from scratch. We also compared PCLR to three other ECG pre-training approaches (supervised pre-training, unsupervised pre-training with an autoencoder, and pre-training using a contrastive multi ECG-segment approach), and show significant performance benefits in three out of four tasks. We found an average performance benefit of 47% over the other models and an average of a 9% performance benefit compared to best model for each task. We release PCLR to enable others to extract ECG representations at https://github.com/broadinstitute/ml4h/tree/master/model_zoo/PCLR.