Journal: eLife

Article Title: Deep learning detects cardiotoxicity in a high-content screen with induced pluripotent stem cell-derived cardiomyocytes

doi: 10.7554/eLife.68714

Figure Lengend Snippet: ( A ) Schematic of knockin strategy of the blasticidin (Bsd) selection cassette into the endogenous MYH6 locus and subsequent Cre-excision of the Puro-selectable marker. ( B ) Single-nucleotide polymorphism karyotyping of the WTC-Bsd line showed no aberrant karyotypic abnormalities after genome engineering and clonal expansion. ( C ) Induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) cultured in Tenaya’s cardiomyocyte (TCM) media showed significantly upregulated maturity markers (MYH7 and TNNI3) and downregulated neonatal myosin isoform (MYH6) compared with iPSC-CMs cultured in RPMI/LT3 media. Error bars = SD. ns: not significant. ( D, E ) iPSC-CMs recovered and cultured in TCM media showed a reduced spontaneous beat rate and increased velocity of contractions. Data collected from five independent differentiations (n = 40 technical replicates per differentiation). Error bars = SD. ( F ) A representative batch of iPSC-CMs recovered in TCM media, leading to greater intensity of cardiac sarcomere markers (MYBP3 and TNNT2) without affecting cell survival in culture. ( G ) Representative immunostaining of iPSC-CMs recovered in PRMI/LT3 media vs. TCM media. TCM media improved recovery of iPSC-CMs after thaw. Scale bars = 100 μm. Figure 1—figure supplement 1—source data 1. Characterization and culturing conditions of WTC-Bsd iPSC line and differentiated cells.

Article Snippet: Sequence-based reagent , MYH7 , Thermo Fisher Scientific , Hs01110632_m1 , TaqMan qPCR probe, cardiac marker.

Techniques: Knock-In, Selection, Marker, Derivative Assay, Cell Culture, Immunostaining

Journal: eLife

Article Title: Deep learning detects cardiotoxicity in a high-content screen with induced pluripotent stem cell-derived cardiomyocytes

doi: 10.7554/eLife.68714

Figure Lengend Snippet:

Article Snippet: Sequence-based reagent , MYH7 , Thermo Fisher Scientific , Hs01110632_m1 , TaqMan qPCR probe, cardiac marker.

Techniques: Sequencing, Marker, Drug discovery, Solvent

Journal:

Article Title: Acute Humanin Therapy Attenuates Myocardial Ischemia and Reperfusion Injury in Mice

doi: 10.1161/ATVBAHA.110.205997

Figure Lengend Snippet: HNG improves cardiomyocyte survival in vitro and decreases apoptosis in response to Daunorubicin. Effect of HNG on A) Cardiomyocyte survival as assessed by cell viability assay, * p < 0.05 Dauno compared to Dauno+HNG, and B) early markers of apoptosis as assessed by CaspASE FITC -VAD-FMK in situ apoptosis assay.

Article Snippet: Effect of HNG on cardiomyocyte survival and apoptosis in vitro Rat heart myocardium cell line H9c2 (2-1) (cat# CRL-1446) was purchased from ATCC.

Techniques: In Vitro, Viability Assay, In Situ, Apoptosis Assay

Journal: Bioinformatics

Article Title: De novo pattern discovery enables robust assessment of functional consequences of non-coding variants

doi: 10.1093/bioinformatics/bty826

Figure Lengend Snippet: Different genomics patterns of DVAR-clusters C1∼C5 in 2 000 000 non-coding variants randomly sampled from the 1000 Genomes Project. (a) The percentage of each cluster of the non-coding variants. (b) The distributions of DVAR-scores of non-coding variants in different clusters. (c) Enrichment of DVAR-clusters with gene-based regions: Splice, 3′ UTR, 5′ UTR, Enhancer, Promoter Intron and Intergenic region, extracted from UCSC database. The darkness of the blue color indicates the relative enrichment with darker blue representing higher significance [−10log10 (Fisher exact P-value)] and vice versa. (d) Enrichment with chromatin states of GM12878 cell line identified by ChromHMM. (e) The boxplot of the numbers of active epigenomes in DVAR-clusters C1-C5 across 127 Roadmap epigenomes. (f) The boxplot of the numbers of active tissue-groups in DVAR-clusters C1∼C5 across 15 Roadmap tissue-groups (Color version of this figure is available at Bioinformatics online.)

Article Snippet: To test that, we grouped the cell types into 17 tissues according to the grouping by the Roadmap Epigenomics (‘IMR90’, ‘ES cell’, ‘iPSC’, ‘ES-deriv’, ‘Blood &T cell’, ‘HSC &B cell’, ‘Mesench’, ‘Myosat’, ‘Epithelial’, ‘Neurosph’, ‘Brain’, ‘Adipose’, ‘Muscle’, ‘Heart’, ‘Smooth muscle’ and ‘Digestive’).

Techniques:

Journal: Antioxidants

Article Title: Desmin Reorganization by Stimuli Inducing Oxidative Stress and Electrophiles: Role of Its Single Cysteine Residue

doi: 10.3390/antiox12091703

Figure Lengend Snippet: Effect of H 2 O 2 on desmin wt or C333S network organization and oligomerization in H9c2 cardiomyoblasts. Cells were transfected with desmin wt or C333S and treated with 1 mM H 2 O 2 for 2 h, as indicated. ( A ) Non-transfected H9c2 cells show undetectable levels of desmin either by immunofluorescence ( upper panel) or by western blot ( lower panel). NT, non-transfected cells; Desmin WT, cells transfected with desmin wt. ( B ) Cells grown on coverslips were fixed and stained for desmin and f-actin, and nuclei were counterstained with DAPI. Overall projections showing the distribution of desmin ( left images) and the overlays with f-actin and DAPI ( right images) are shown. Bars, 25 µm. ( C ) The graph displays the ratios between the area of the cells with detectable desmin filaments and that occupied by f-actin. The results shown are average values from at least 15 determinations ± SEM. * p < 0.05 by Student’s t -test. ( D ) Cells on cell culture wells were lysed, proteins were analyzed by SDS-PAGE under non-reducing or reducing conditions, as indicated, and desmin was detected by western blot. The results are representative of four experiments. Dotted lines indicate where lanes from the same gel have been cropped.

Article Snippet: H9c2 cardiomyoblasts (rat heart myoblasts derived from embryonic BD1X rat heart myocardium) from the European Collection of Authenticated Cell Cultures (ECACC General Cell Collection, originally deposited by the ATCC) were cultured in DMEM plus antibiotics (100 U/mL penicillin, 100 μg/mL streptomycin), supplemented with FBS (10% v / v , Gibco, New York, NY, USA), routinely in a humidified atmosphere with 5% CO 2 .

Techniques: Transfection, Immunofluorescence, Western Blot, Staining, Cell Culture, SDS Page

Journal: Antioxidants

Article Title: Desmin Reorganization by Stimuli Inducing Oxidative Stress and Electrophiles: Role of Its Single Cysteine Residue

doi: 10.3390/antiox12091703

Figure Lengend Snippet: Effect of 15d-PGJ 2 on desmin network organization and protein levels in H9c2 cardiomyoblasts. Cells transfected as above were treated with 10 µM 15d-PGJ 2 for 2 h. ( A ) Distribution of desmin and f-actin was assessed as in . ( B ) Ratios between the cell areas covered by detectable structures of desmin and actin. Results shown are average values from at least 18 determinations ± SEM. *** p < 0.001, ** p < 0.01 by Student’s t -test. ( C ) Proportion of cells showing desmin accumulations. Results are average values ± SEM from four different assays totaling between 50 and 100 cells per experimental condition. ( D ) Lysates from cells treated as in ( A ) were analyzed by gel electrophoresis under non-reducing conditions, and desmin was detected by western blot. Results are representative of four experiments. The dotted line indicates where lanes from the same gel have been cropped.

Article Snippet: H9c2 cardiomyoblasts (rat heart myoblasts derived from embryonic BD1X rat heart myocardium) from the European Collection of Authenticated Cell Cultures (ECACC General Cell Collection, originally deposited by the ATCC) were cultured in DMEM plus antibiotics (100 U/mL penicillin, 100 μg/mL streptomycin), supplemented with FBS (10% v / v , Gibco, New York, NY, USA), routinely in a humidified atmosphere with 5% CO 2 .

Techniques: Transfection, Nucleic Acid Electrophoresis, Western Blot

Journal: Antioxidants

Article Title: Desmin Reorganization by Stimuli Inducing Oxidative Stress and Electrophiles: Role of Its Single Cysteine Residue

doi: 10.3390/antiox12091703

Figure Lengend Snippet: Effect of chemical hypoxia on the organization of desmin wt or C333S in H9c2 cells. Cells were transfected with desmin wt or C333S, as indicated, and treated with the specified concentrations of CoCl 2 for 24 h. ( A ) Cells were fixed and processed for detection of desmin (immunofluorescence), f-actin (phalloidin-TRITC staining), and nuclei (DAPI). ( B ) The proportion of cells showing any type of desmin condensation or aggregate was obtained by visual inspection from three different experiments. Results shown are average values ± SEM. * p < 0.05. Bars, 20 µm. ( C ) Lysates from cells treated as in ( A ) were analyzed by SDS-PAGE under non-reducing conditions and immunoblotted with an anti-desmin antibody. Results are representative of three assays with similar observations.

Article Snippet: H9c2 cardiomyoblasts (rat heart myoblasts derived from embryonic BD1X rat heart myocardium) from the European Collection of Authenticated Cell Cultures (ECACC General Cell Collection, originally deposited by the ATCC) were cultured in DMEM plus antibiotics (100 U/mL penicillin, 100 μg/mL streptomycin), supplemented with FBS (10% v / v , Gibco, New York, NY, USA), routinely in a humidified atmosphere with 5% CO 2 .

Techniques: Transfection, Immunofluorescence, Staining, SDS Page

Journal: Antioxidants

Article Title: Desmin Reorganization by Stimuli Inducing Oxidative Stress and Electrophiles: Role of Its Single Cysteine Residue

doi: 10.3390/antiox12091703

Figure Lengend Snippet: Importance of C333 in desmin assembly in several cellular models. ( A ) H9c2 cardiomyoblasts were transfected with mEmerald-desmin wt or C333S, as indicated. The morphology of the desmin network was assessed 48 h later by direct fluorescence visualization; f-actin was stained with Phalloidin-TRITC, and nuclei were counterstained with DAPI. The proportion of cells displaying desmin aggregates is shown in ( B ). Results are average values ± SEM from three different experiments. ** p < 0.01 by Student’s t -test. ( C ) SW13/cl.2 cells were transfected with the indicated plasmids, and the distribution of the desmin fluorescent construct was assessed by confocal microscopy. Images are representative of four assays with similar results. ( D ) SW13/cl.2 cells expressing desmin wt or C333S were treated with H 2 O 2 , as indicated. The morphology of the desmin network was assessed by immunofluorescence. The proportion of cells showing extended desmin filaments is shown in ( E ). Results are average values ± SEM from three independent experiments. * p < 0.05 by Student’s t -test.

Article Snippet: H9c2 cardiomyoblasts (rat heart myoblasts derived from embryonic BD1X rat heart myocardium) from the European Collection of Authenticated Cell Cultures (ECACC General Cell Collection, originally deposited by the ATCC) were cultured in DMEM plus antibiotics (100 U/mL penicillin, 100 μg/mL streptomycin), supplemented with FBS (10% v / v , Gibco, New York, NY, USA), routinely in a humidified atmosphere with 5% CO 2 .

Techniques: Transfection, Fluorescence, Staining, Construct, Confocal Microscopy, Expressing, Immunofluorescence

Journal: Genomics Data

Article Title: Identification of transcripts regulated by CUG-BP, Elav-like family member 1 (CELF1) in primary embryonic cardiomyocytes by RNA-seq

doi: 10.1016/j.gdata.2015.08.014

Figure Lengend Snippet:

Article Snippet: Organism/cell line/tissue , Gallus gallus /primary embryonic cardiomyocytes/heart muscle.

Techniques: Knockdown, Isolation, Transfection, Incubation, Cell Culture

Journal: Clinical and Experimental Immunology

Article Title: Revolutionizing immune research with organoid-based co-culture and chip systems

doi: 10.1093/cei/uxae004

Figure Lengend Snippet: Examples of microfluidic-chip systems that can be used for co-culture with immune cells or other cells in more physiologically relevant environments to study cell-cell interactions

Article Snippet: The human-on-chip (service company, that does not sell devices) , Hesperos , Multi-organ chips Liver Muscle Heart Pre-neurons Kidney Monocytic cells , Y e.g. Biopsy-derived hepatocytes, iPSC-derived cardiomyocytes and arm or leg skeletal muscle-derived myoblasts , https://hesperosinc.com/ [ ] .

Techniques:

Journal: PLoS Computational Biology

Article Title: Tox_(R)CNN: Deep learning-based nuclei profiling tool for drug toxicity screening

doi: 10.1371/journal.pcbi.1006238

Figure Lengend Snippet: HL1 cells treated or not (-) with the indicated concentrations of compounds (μM) or DMSO were processed as described in the Materials and Methods. (A) Representative fluorescence microscopy images of DAPI-stained cells treated or not (Ctrl) with the highest concentrations of the indicated compounds in a reference experiment (Experiment #1) used for CNN training. (B-D) Boxplots of per-well toxicity assessments from established measurements: nucleus count (Num Nuc) (B), Caspase 3/7 nucleus:cytoplasm ratio (Casp nuc/cyto) (C), and Mitotracker cytoplasmic intensity (Mito) (D). (E) CNN architecture for predicting health status from single-cell image crops, as described in Materials and Methods. (F) Cropping strategies; representative image crops are shown of nucleus (Nuc), nucleus+cytoplasm (Cell), nucleus+margin (Nuc_Ring), and nucleus+cytoplasm+background+neighboring cells (All). (G) Boxplot of per-well toxicity assessment from CNN Nuc predictions (percentage of cells classified as healthy). (H) Plot displaying mean toxicity readouts of replicate wells, obtained from the percentage of healthy cells predicted by the different CNN models (CNN Nuc, Nuc_Ring, Cell, All, 4crops) and the standard nuclei counting (Num Nuc) for the different treatments indicated. For each well, toxicity readouts were obtained by computing Z-scores (normalizing to DMSO-treated wells) with adjustment of the sign to display toxic effects as positive values. Points and corresponding error bars represent the mean and standard error of the mean, respectively, of results obtained by evaluating the 5 different CNN models trained for each cropping strategy. (I) Evaluation performance of the different CNN models for predicting toxic effects of staurosporine assessed using Caspase 3/7 fluorescent reporter as reference, as described in the Materials and Methods. Boxplots display AUC values obtained with the 5 models trained for each cropping strategy. (J) Correlations between cell density and CNN predictions obtained with the different models for untreated cells. Boxplots represent Pearson correlation coefficient, R, obtained with all 5 models trained with each cropping strategy (top), and exemplary dotplots for each strategy (selecting the model with the median AUC among the five) including regression line, R value and significance (bottom). p-value: *<0.01, **<0.001, ***<0.0001.

Article Snippet: Mouse cardiac muscle HL1 cells purchased from Merck Millipore were grown on fibronectin(25μg/ml)/gelatin(1mg/ml) coated dishes with 10% fetal bobine serum (FBS) in Claycomb medium (Sigma-Aldrich).

Techniques: Fluorescence, Microscopy, Staining

Journal: PLoS Computational Biology

Article Title: Tox_(R)CNN: Deep learning-based nuclei profiling tool for drug toxicity screening

doi: 10.1371/journal.pcbi.1006238

Figure Lengend Snippet: HL1 cells treated or not (-) with DMSO or the indicated concentrations of drugs (μM) from Experiment #1 were processed as described in the Materials and Methods. Plots display individual well toxicity readouts (top) and the 5-Fluorouracil dose-response curve fitted from well-averages, including the EC50 (bottom). Representative model results out of the 5 independent ones trained for each cropping strategy is shown. (A-C) CNN-based toxicity readouts: CNN Nuc (A), CNN Cell (B), and CNN All (C). (D-F) Standard toxicity readouts: Nuclei counting by standard image segmentation (Num Nuc) (D), mean Caspase 3/7 nucleus:cytoplasm ratio (Casp Nuc/Cyto) (E), and mean Mitotracker cytoplasmic intensity (Mito) (F). For each well, toxicity readouts were obtained by computing Z-scores (normalizing to DMSO-treated wells) with adjustment of the sign to display toxic effects as positive values. Z-scores > 3 represent toxic hits.

Article Snippet: Mouse cardiac muscle HL1 cells purchased from Merck Millipore were grown on fibronectin(25μg/ml)/gelatin(1mg/ml) coated dishes with 10% fetal bobine serum (FBS) in Claycomb medium (Sigma-Aldrich).

Techniques:

Journal: PLoS Computational Biology

Article Title: Tox_(R)CNN: Deep learning-based nuclei profiling tool for drug toxicity screening

doi: 10.1371/journal.pcbi.1006238

Figure Lengend Snippet: (A) RCNN architecture for automated detection of cells and prediction of their health status from micrographs of DAPI fluorescence, as described in the Materials and Methods. (B) Example of nuclei bounding boxes resulting from the region proposal network included in the RCNN framework. (C) HL1 cells (Experiment #8) were seeded at the indicated densities (cells/well). (D,E) HL1 (Experiment #1) and EAHY926 (Experiment #9) cells were seeded at 5000 cells/well. (C-E) 24h after seeding, cells were treated or not (-) with the indicated concentrations of DMSO (%) or the indicated drugs (μM) and processed as described in the Materials and Methods. Representative images are shown of untreated cells at the indicated cell-seeding densities (cells/well). Plots display mean toxicity readouts of four replicate wells, obtained from the percentage of healthy cells predicted by the CNN Nuc (Tox_CNN) and RCNN (Tox_RCNN_balanced and Tox_RCNN) mixed models, and from nuclei counting by standard image segmentation (Num Nuc), or by using RCNN-based automated detection (Num Nuc RCNN) from Tox_RCNN training. For each well, toxicity readouts were obtained by computing Z-scores (normalizing to DMSO-treated wells) with adjustment of the sign to display toxic effects as positive values.

Article Snippet: Mouse cardiac muscle HL1 cells purchased from Merck Millipore were grown on fibronectin(25μg/ml)/gelatin(1mg/ml) coated dishes with 10% fetal bobine serum (FBS) in Claycomb medium (Sigma-Aldrich).

Techniques: Fluorescence

Journal: PLoS Computational Biology

Article Title: Tox_(R)CNN: Deep learning-based nuclei profiling tool for drug toxicity screening

doi: 10.1371/journal.pcbi.1006238

Figure Lengend Snippet: HL1 cells treated with one of 24 compounds or DMSO at the concentrations indicated (μM) (Experiments #11–14) were processed as described in the Materials and Methods. (A) Plots displaying mean toxicity readouts of four replicate wells, obtained from percentage of healthy cells predicted by the CNN Nuc (Tox_CNN) or RCNN (Tox_RCNN) mixed models, and from nuclei counting by standard image segmentation (Num Nuc), or by using RCNN-based automated detection (Num Nuc RCNN). For each well, toxicity readouts were obtained by computing Z-scores (normalizing to DMSO-treated wells) with adjustment of the sign to display toxic effects as positive values. (B) Hierarchical clustering of features obtained with the Tox_CNN model from HL1 cells treated with 25μM of the indicated drugs or 0.78μM Taxol; untreated (-); DMSO, control. Colors highlight mechanism of toxicity associated with compounds.

Article Snippet: Mouse cardiac muscle HL1 cells purchased from Merck Millipore were grown on fibronectin(25μg/ml)/gelatin(1mg/ml) coated dishes with 10% fetal bobine serum (FBS) in Claycomb medium (Sigma-Aldrich).

Techniques: Control