CoChat

Anti-senescent drug screening by deep learning-based morphology senescence scoring

Open access

The DOI matches its Crossref record; 32 of the 32 references it cites were found in a registry or an index.

Abstract

Advances in deep learning technology have enabled complex task solutions. The accuracy of image classification tasks has improved owing to the establishment of convolutional neural networks (CNN). Cellular senescence is a hallmark of ageing and is important for the pathogenesis of ageing-related diseases. Furthermore, it is a potential therapeutic target. Specific molecular markers are used to identify senescent cells. Moreover senescent cells show unique morphology, which can be identified. We develop a successful morphology-based CNN system to identify senescent cells and a quantitative scoring system to evaluate the state of endothelial cells by senescence probability output from pre-trained CNN optimised for the classification of cellular senescence, Deep Learning-Based Senescence Scoring System by Morphology (Deep-SeSMo). Deep-SeSMo correctly evaluates the effects of well-known anti-senescent reagents. We screen for drugs that control cellular senescence using a kinase inhibitor library by Deep-SeSMo-based drug screening and identify four anti-senescent drugs. RNA sequence analysis reveals that these compounds commonly suppress senescent phenotypes through inhibition of the inflammatory response pathway. Thus, morphology-based CNN system can be a powerful tool for anti-senescent drug screening.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/s41467-020-20213-0

Cites

32 of 32 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 CNN training to classify control and senescent cells. a Representative images of input images. Input images of control and H2O2-induced senescent cells were cropped from phase-contrast microscopy images at single-cell resolution by the OpenCV-based script. Scale bar, 7.1 µm. Data are representative of over three independent experiments. b Learning curve through the CNN training. Accuracy and loss in the training data and validation accuracy and validation loss in the validation data show the process of training. c Indexes: F1 score, accuracy, precision, and recall in the final setting of training. d AUC of the ROC curve in the final setting of training. e Protocol for the evaluation of CNN generalisability. Each CNN was trained by either the images of H2O2-induced senescent cells, the images of CPT-induced senescent cells, or the mixed images of H2O2- and CPT-induced senescent cells. Newly acquired data were used as test data. In the test data, cellular senescence was induced by H2O2, CPT, or replication. Test data were evaluated by three pre-trained CNNs. f A heatmap shows the accuracy of CNN prediction in each test dataset. Three independent experiments and evaluations were conducted for each senescence induction method. g Macro-averaged accuracy for each evaluation (n = 3 independent experiments). h AUC of the receiver operating characteristic (ROC) curve in the test data evaluated by CNNs, which were pre-trained by the data from H2O2-, CPT-, or replication-induced senescent cells. Data are representative of three independent experiments. i Grad-CAM shows an important region for the prediction of healthy or senescent cells. Data are representative of three independent experiments. CNN convolutional neural network, CPT camptothecin, Rep replication.
Fig. 1 CNN training to classify control and senescent cells. a Representative images of input images. Input images of control and H2O2-induced senescent cells were cropped from phase-contrast microscopy images at single-cell resolution by the OpenCV-based script. Scale bar, 7.1 µm. Data are representative of over three independent experiments. b Learning curve through the CNN training. Accuracy and loss in the training data and validation accuracy and validation loss in the validation data show the process of training. c Indexes: F1 score, accuracy, precision, and recall in the final setting of training. d AUC of the ROC curve in the final setting of training. e Protocol for the evaluation of CNN generalisability. Each CNN was trained by either the images of H2O2-induced senescent cells, the images of CPT-induced senescent cells, or the mixed images of H2O2- and CPT-induced senescent cells. Newly acquired data were used as test data. In the test data, cellular senescence was induced by H2O2, CPT, or replication. Test data were evaluated by three pre-trained CNNs. f A heatmap shows the accuracy of CNN prediction in each test dataset. Three independent experiments and evaluations were conducted for each senescence induction method. g Macro-averaged accuracy for each evaluation (n = 3 independent experiments). h AUC of the receiver operating characteristic (ROC) curve in the test data evaluated by CNNs, which were pre-trained by the data from H2O2-, CPT-, or replication-induced senescent cells. Data are representative of three independent experiments. i Grad-CAM shows an important region for the prediction of healthy or senescent cells. Data are representative of three independent experiments. CNN convolutional neural network, CPT camptothecin, Rep replication.CC BY · PubMed Central
Fig. 2 Development of Deep-SeSMo. a Representative input images of HUVECs treated with various concentration of H2O2. Scale bar, 7.1 µm. Data are representative of three independent experiments. b–d qRT-PCR analysis to determine the mRNA expression of P21 under various b concentrations of H2O2, c concentrations of CPT, and d passage numbers (n = 3 biological replicates). e Density plots of all senescence probabilities, which were outputs of CNN prediction, in HUVECs treated with various H2O2 concentrations. Data are representative of three independent experiments. f A rate graph shows the ratio of prediction for each senescence probability with various H2O2 concentrations. Data are representative of three independent experiments. g–i Senescence score calculated by CNNs trained by H2O2- and CPT-induced senescent HUVECs, in HUVECs treated with various H2O2 concentrations, CPT concentrations, and passage numbers. The thin line and bold line indicate each score and the average score, respectively. j A heatmap shows Pearson correlations for each set of test data. Three independent experiments of H2O2- and CPT-induced senescence and one experiment for replication-induced senescence, evaluated by three trained CNNs, were conducted. k Macro-averaged Pearson correlation, evaluated by three trained CNNs (n = 3 independent experiments for H2O2- and CPT-induced test data, n = 1 for rep-induced test data). CNN convolutional neural network, CPT camptothecin, Rep Replication. Data are shown as mean ± s.e.m.
Fig. 2 Development of Deep-SeSMo. a Representative input images of HUVECs treated with various concentration of H2O2. Scale bar, 7.1 µm. Data are representative of three independent experiments. b–d qRT-PCR analysis to determine the mRNA expression of P21 under various b concentrations of H2O2, c concentrations of CPT, and d passage numbers (n = 3 biological replicates). e Density plots of all senescence probabilities, which were outputs of CNN prediction, in HUVECs treated with various H2O2 concentrations. Data are representative of three independent experiments. f A rate graph shows the ratio of prediction for each senescence probability with various H2O2 concentrations. Data are representative of three independent experiments. g–i Senescence score calculated by CNNs trained by H2O2- and CPT-induced senescent HUVECs, in HUVECs treated with various H2O2 concentrations, CPT concentrations, and passage numbers. The thin line and bold line indicate each score and the average score, respectively. j A heatmap shows Pearson correlations for each set of test data. Three independent experiments of H2O2- and CPT-induced senescence and one experiment for replication-induced senescence, evaluated by three trained CNNs, were conducted. k Macro-averaged Pearson correlation, evaluated by three trained CNNs (n = 3 independent experiments for H2O2- and CPT-induced test data, n = 1 for rep-induced test data). CNN convolutional neural network, CPT camptothecin, Rep Replication. Data are shown as mean ± s.e.m.CC BY · PubMed Central
Fig. 3 Drug screening using Deep-SeSMo. a Representative images of SA-β-gal activity in senescent HUVECs treated with metformin or NMN. Scale bar, 100 μm. DAPI indicates the cell nuclei. Data are representative of two independent experiments. b Percentage of SA-β-gal-positive cells per total cells in senescent HUVECs treated with metformin or NMN (n = 12 images over two independent experiments). c Western blotting of P21, P53, Ser15 phosphorylation of P53, and P16INK4a in senescent HUVECs treated with metformin or NMN. GAPDH was used as an internal control. Data are representative of two independent experiments. d, e Senescence score calculated by Deep-SeSMo of senescent HUVECs treated with d metformin and e NMN. Data are representative of two independent experiments. f Eighty kinase inhibitors were added to HUVECs, and cellular senescence was induced by H2O2. The senescence score was calculated by Deep-SeSMo and normalised to a control score. g Senescence score ranking for 80 kinase inhibitors and control. Rankings were calculated for three stressors with two replications and evaluated by three pre-trained CNNs. A surface plot shows senescence score ranking sorted by the median value of senescence score ranking. h Top ten compounds detected by Deep-SeSMo. i A heatmap demonstrates a senescence score ranking for each condition evaluated by CNNs trained by H2O2- and CPT-induced senescent HUVECs. The right map shows an anti-senescence cluster that strongly suppresses senescence. CPT camptothecin, Rep replication, NMN nicotinamide mononucleotide. Data are shown as mean ± s.e.m. p values by two-sided Student’s t-test.
Fig. 3 Drug screening using Deep-SeSMo. a Representative images of SA-β-gal activity in senescent HUVECs treated with metformin or NMN. Scale bar, 100 μm. DAPI indicates the cell nuclei. Data are representative of two independent experiments. b Percentage of SA-β-gal-positive cells per total cells in senescent HUVECs treated with metformin or NMN (n = 12 images over two independent experiments). c Western blotting of P21, P53, Ser15 phosphorylation of P53, and P16INK4a in senescent HUVECs treated with metformin or NMN. GAPDH was used as an internal control. Data are representative of two independent experiments. d, e Senescence score calculated by Deep-SeSMo of senescent HUVECs treated with d metformin and e NMN. Data are representative of two independent experiments. f Eighty kinase inhibitors were added to HUVECs, and cellular senescence was induced by H2O2. The senescence score was calculated by Deep-SeSMo and normalised to a control score. g Senescence score ranking for 80 kinase inhibitors and control. Rankings were calculated for three stressors with two replications and evaluated by three pre-trained CNNs. A surface plot shows senescence score ranking sorted by the median value of senescence score ranking. h Top ten compounds detected by Deep-SeSMo. i A heatmap demonstrates a senescence score ranking for each condition evaluated by CNNs trained by H2O2- and CPT-induced senescent HUVECs. The right map shows an anti-senescence cluster that strongly suppresses senescence. CPT camptothecin, Rep replication, NMN nicotinamide mononucleotide. Data are shown as mean ± s.e.m. p values by two-sided Student’s t-test.CC BY · PubMed Central
Fig. 4 Anti-senescent effects of the top four compounds. a Representative images of SA-β-gal activity in senescent HUVECs treated with the top four compounds: terreic acid, PD-98059, daidzein, and Y-27632. Scale bar: 100 μm. Data are representative of two independent experiments. b Percentage of SA-β-gal-positive cells (n = 12 images over two independent experiments). c Western blotting of P21, P53, Ser15 phosphorylation of P53, and P16INK4a. The top four compounds were added to HUVECs, and cellular senescence was induced by H2O2. GAPDH was used as an internal control. Data are representative of two independent experiments. d A heatmap shows the top 10 genes selected by differential gene expression analysis among all four compounds and control. The genes designated in red are associated with NFκB function. e GSEA of genes associated with the inflammatory response. f GSEA of genes associated with TNFα signalling via NFκB. g GO analysis, categorised as biological process, of genes upregulated in HUVECs treated with terreic acid compared with control. h, i GSEA of genes associated with h positive regulation of ATPase activity and i oxidative phosphorylation. NES normalised enrichment score. Data are shown as mean ± s.e.m. p values by two-sided Student’s t-test.
Fig. 4 Anti-senescent effects of the top four compounds. a Representative images of SA-β-gal activity in senescent HUVECs treated with the top four compounds: terreic acid, PD-98059, daidzein, and Y-27632. Scale bar: 100 μm. Data are representative of two independent experiments. b Percentage of SA-β-gal-positive cells (n = 12 images over two independent experiments). c Western blotting of P21, P53, Ser15 phosphorylation of P53, and P16INK4a. The top four compounds were added to HUVECs, and cellular senescence was induced by H2O2. GAPDH was used as an internal control. Data are representative of two independent experiments. d A heatmap shows the top 10 genes selected by differential gene expression analysis among all four compounds and control. The genes designated in red are associated with NFκB function. e GSEA of genes associated with the inflammatory response. f GSEA of genes associated with TNFα signalling via NFκB. g GO analysis, categorised as biological process, of genes upregulated in HUVECs treated with terreic acid compared with control. h, i GSEA of genes associated with h positive regulation of ATPase activity and i oxidative phosphorylation. NES normalised enrichment score. Data are shown as mean ± s.e.m. p values by two-sided Student’s t-test.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 32 references

Cited by

See all 143

Metadata

Authors
Kusumoto, D., Seki, T., Sawada, H., Kunitomi, A., Katsuki, T., Kimura, M., Ito, S., Komuro, J., Hashimoto, H., Fukuda, K., & Yuasa, S.
Affiliations
  • Keio University (JP)
  • University of Tokyo Hospital (JP)
  • Kyoto University (JP)
Corresponding author
Yuasa, Shinsuke
Venue
Nature Communications
Published
2021-01-11
Volume
12
Issue
1
Pages
257-257
ISSN
2041-1723 (ISSN-L)
OpenAlex
W3119109946
Cited by
143
References
32
Citations by year
2026: 26 · 2025: 31 · 2024: 36 · 2023: 22 · 2022: 19 · 2021: 9
FWCI
13.51
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Communications (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, other-oa)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
  • MEXT | Japan Society for the Promotion of Science (19K08549)
  • Keio University
  • Japanese Circulation Society
  • Japan Agency for Medical Research and Development (JP18bm0404026)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Cell Biology › Telomeres, Telomerase, and Senescence
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Biophysics › Cell Image Analysis Techniques
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Aging › Genetics, Aging, and Longevity in Model Organisms
Keywords
cellular senescence, drug screening, endothelial cells, kinase inhibitors, inflammation reduction, aging
MeSH terms
Deep Learning, beta-Galactosidase, Drug Evaluation, Preclinical, Humans, Hydrogen Peroxide, Neural Networks, Computer, Cellular Senescence, Cell Shape, Human Umbilical Vein Endothelial Cells

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Anti-senescent drug screening by deep learning-based morphology senescence scoring. https://app.cochat.ai/papers/10.1038/s41467-020-20213-0
The paper
Kusumoto, D., Seki, T., Sawada, H., Kunitomi, A., Katsuki, T., Kimura, M., Ito, S., Komuro, J., Hashimoto, H., Fukuda, K., & Yuasa, S. (2021). Anti-senescent drug screening by deep learning-based morphology senescence scoring. Nature Communications. https://doi.org/10.1038/s41467-020-20213-0

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.