NET-CAGE characterizes the dynamics and topology of human transcribed cis-regulatory elements

Hirabayashi, Shigeki; Bhagat, Shruti; Matsuki, Yasuhiko; Takegami, Yujiro; Uehata, Takuya; Kanemaru, Ai; Itoh, Masayoshi; Shirakawa, Kotaro; Takaori‐Kondo, Akifumi; Takeuchi, Osamu; Carninci, Piero; Katayama, Shintaro; Hayashizaki, Yoshihide; Kere, Juha; Kawaji, Hideya; Murakawa, Yuji

doi:10.1038/s41588-019-0485-9

Cited by 89 publications

(102 citation statements)

References 86 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…5c). Altogether, we identified more than 10,000 transcribed enhancers in our Dox (+) DUX4 TetOn hESCs and ~ 90% of these enhancers have not been identified previously 16–18 (Fig. 3b, Supplementary Information 3).…”

Section: Resultsmentioning

confidence: 90%

See 1 more Smart Citation

DUX4 regulates oocyte to embryo transition in human

Vuoristo

Hydén-Granskog

Yoshihara

et al. 2019

Preprint

Self Cite

View full text Add to dashboard Cite

45During the human oocyte-to-embryo transition, the fertilized oocyte undergoes 46 final maturation and the embryo genome is gradually activated during the first 47 three cell divisions. How this transition is coordinated in humans is largely 48 unknown. We show that the double homeodomain transcription factor DUX4 49 contributes to this transition. DUX4 knockdown in human zygotes caused 50 insufficient transcriptome reprogramming as observed three days after 51 fertilization. Induced DUX4 expression in human embryonic stem cells activated 52 transcription of thousands of newly identified bi-directional transcripts, including 53 putative enhancers for embryonic genome activation genes such as LEUTX. DUX4 54 protein interacted with transcriptional modifiers that are known to couple 55 enhancers and promoters. Taken together, our results reveal that DUX4 is a 56 pioneer regulating oocyte-to-embryo transition in human through activation of 57 intergenic genome, especially enhancers, and hence setting the stage for early 58 human embryo development. 59 60 61 62 63 64 65 4 Mammalian pre-implantation development commences with oocyte-to-embryo 66 transition, which involves fundamental changes in the epigenetic landscapes, 67 modulation of cell cycle control, and translation or clearance of selected maternal 68 mRNAs, culminating to embryonic genome activation 1,2 . The pioneer regulators 69 orchestrating the oocyte-to-embryo transition and first embryo genome activation steps 70 in human remain poorly understood. The conserved double homeodomain transcription 71factor DUX4 represents a plausible candidate regulating the oocyte-to-embryo 72 transition in humans, given its capacity to activate germline genes and genomic repeat 73 elements 3-5 . Here we show that DUX4 is able to launch the first reprogramming steps 74 from oocyte to embryo in human by activating thousands of novel enhancers and 75 therein, modulating the transcriptome and chromatin. Human DUX4 knockdown 76 embryos are viable until the third day of development, but their transcriptome is 77 severely altered. Our proteomics approaches suggest that DUX4 binds to Mediator 78 complex and chromatin modifiers through its C-terminal domains, providing a likely 79 explanation to how DUX4 may extensively modulate the genome. This study implies a 80 wider role for DUX4 as a cellular gate keeper acting both as a general genomic modifier 81 of cell fate as well as a specific inducer of first wave embryo genome activation genes. 82 83Results 84 Quantification of DUX4 in human embryos 85The DUX4 induced gene network is highly conserved 6 and recent reports showed that 86 DUX4 is expressed in early human embryos 3,4 . However, details of this dynamic 87 process, including initiation of DUX4 expression, remained ambiguous. Therefore, we 88Given the short-term and precise manifestation of DUX4 mRNA and protein in human 107 zygotes and early cleavage stage embryos, we next asked how DUX4 regulates the first 108 steps of human embryo development. We microinjected either DUX4...

show abstract

Section: Resultsmentioning

confidence: 90%

“…3a). For this, we investigated native elongating transcripts using cap analysis of gene expression (NET-CAGE) 16 , which sensitively identifies unstable transcripts such as enhancer RNAs (Extended Data Fig. 5a, b).…”

Section: Resultsmentioning

confidence: 99%

DUX4 regulates oocyte to embryo transition in human

Vuoristo

Hydén-Granskog

Yoshihara

et al. 2019

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…We developed MENTR ML models by combining deep convolutional neural networks (from 2-kb sequence bin to 2002 epigenetic features, using publicly available pre-trained models (Zhou and Troyanskaya, 2015;Zhou et al, 2018)) and binary classifiers using non-linear, gradient boosting trees (from the many epigenetic features in +/-100-kb sequence to accurate transcription probability; see METHOD DETAILS) (Chen and Guestrin, 2016). The binary classifier outputs a probability of expression for each tissue or cell-type, chosen because we focus here on predicting lowly-expressed RNAs whose quantitative measurement might be not reliable (Hirabayashi et al, 2019)). 5 We trained the MENTR ML models using the autosomal mRNA and ncRNA promoterand enhancer-level transcripts profiled by CAGE except for chromosome 8 and tested the accuracy using those of chromosome 8 ( Figure S1B).…”

Section: Strategy To Predict Mutations' Effects On Ncrnasmentioning

confidence: 99%

“…Considering the quite low expression levels of many ncRNAs, we were concerned that MENTR could learn anomalous read patterns, noise, or artifacts related to transcript mappability that vary depending on sequence context, rather than reproducible transcriptional status; if this were the case, it would be unexpected if 5 MENTR could accurately predict the expression patterns of RNAs that are in fact transcribed, but are not detected due to low depth of sequencing or degradation. To evaluate this possibility, we took advantage of enhancer RNA transcription data in five ENCODE cell lines profiled using NET-CAGE, a sensitive method for measuring nascent RNA transcription (Hirabayashi et al, 2019). We then evaluated false positive 10 (FP) predictions, in which a transcript was not detected by the standard CAGE method, but MENTR predicted a probability of expression >0.5 (based on ML models trained using standard CAGE data).…”

Section: Accurate Prediction Of Cell-type-specific Promoter-and Enhanmentioning

confidence: 99%

“…We developed two MENTR ML models from 2 replicates of CAGE transcriptome for each cell line and used FP predictions in either ML models in the following analysis. We 15 evaluated the expression levels of FP predictions in expression levels of the nascent elongating RNAs from NET-CAGE samples treated by 2 M urea lysis buffer (Hirabayashi et al, 2019). We noted that the NET-CAGE could capture transcribed CAGE peaks before degradation and therefore detect lowly expressed transcripts including enhancers at the same sequence depth.…”

Section: Evaluation Of Fp Predictions Using Nascent Rna Profiling 10mentioning

confidence: 99%

See 1 more Smart Citation

Predicting cell-type-specific non-coding RNA transcription from genome sequence

Koido

Hon

Koyama

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

Transcription is regulated through complex mechanisms involving non-coding RNAs (ncRNAs).However, because transcription of ncRNAs, especially enhancer RNAs, is often low and cell type-specific, its dependency on genotype remains largely unexplored. Here, we developed mutation effect prediction on ncRNA transcription (MENTR), a quantitative machine learning 5 framework reliably connecting genetic associations with expression of ncRNAs, resolved to the level of cell type. MENTR-predicted mutation effects on ncRNA transcription were concordant with estimates from previous genetic studies in a cell type-dependent manner. We inferred reliable causal variants from 41,223 GWAS variants, and proposed 7,775 enhancers and 3,548 long-ncRNAs as complex trait-associated ncRNAs in 348 major human primary cells and tissues, 10 including plausible enhancer-mediated functional alterations in single-variant resolution in Crohn's disease. In summary, we present new resources for discovering causal variants, the biological mechanisms driving complex traits, and the sequence-dependency of ncRNA regulation in relevant cell types. (145/150 words)

show abstract