Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Trans-channel fluorescence learning improves high-content screening for Alzheimer’s disease therapeutics

Nature Machine Intelligence volume 4, pages 583–595 (2022)Cite this article

Subjects

A preprint version of the article is available at bioRxiv.

Abstract

In microscopy-based drug screens, fluorescent markers carry critical information on how compounds affect different biological processes. However, practical considerations, such as the labour and preparation formats needed to produce different image channels, hinder the use of certain fluorescent markers. Consequently, completed screens may lack biologically informative but experimentally impractical markers. Here we present a deep learning method for overcoming these limitations. We accurately generated predicted fluorescent signals from other related markers and validated this new machine learning (ML) method on two biologically distinct datasets. We used the ML method to improve the selection of biologically active compounds for Alzheimer’s disease from a completed high-content high-throughput screen (HCS) that only contained the original markers. The ML method identified novel compounds that effectively blocked tau aggregation, which had been missed by traditional screening approaches unguided by ML. The method improved triaging efficiency of compound rankings over conventional rankings by raw image channels. We reproduced this ML pipeline on a biologically independent cancer-based dataset, demonstrating its generalizability. The approach is disease-agnostic and applicable across diverse fluorescence microscopy datasets.

This is a preview of subscription content, access via your institution

Access options

Rent or buy this article

Get just this article for as long as you need it

$39.95

Learn more

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: A cellular Tau(P301S)-YFP aggregation assay that models prion propagation in vitro with an additional AT8-pTau channel for ML training.
Fig. 2: The ML model generated a phosphorylated AT8-pTau channel, solely from DAPI and YFP-tau channels.
Fig. 3: The ML model outperformed all others.
Fig. 4: Transforming the archival HCS data with ML-predicted AT8-pTau images revealed previously unknown active compounds.
Fig. 5: ML triaged active compounds better than other methods.
Fig. 6: The trans-fluorescence learning method generated accurate cyclin-B1 signal from Hoechst signal, on an independent and biologically unrelated dataset.

Data availability

All image data are freely available at https://osf.io/xntd658 and https://doi.org/10.17605/OSF.IO/XNTD6.

Code availability

The full source code and fully trained models are available at https://github.com/keiserlab/trans-channel-paper59 and https://doi.org/10.5281/zenodo.6336183.

References

  1. Li, Z., Cvijic, M. E. & Zhang, L. in Comprehensive Medicinal Chemistry III (eds Chackalamannil, S., Rotella, D. & Ward, S. E.) 362–387 (Elsevier, 2017); https://doi.org/10.1016/B978-0-12-409547-2.12328-5

  2. Kim, S.-W., Roh, J. & Park, C.-S. Immunohistochemistry for pathologists: protocols, pitfalls and tips. J. Pathol. Transl. Med. 50, 411–418 (2016).

    Article  Google Scholar 

  3. Cardoso, M. C. in Encyclopedic Reference of Genomics and Proteomics in Molecular Medicine 583–586 (Springer, 2006); https://doi.org/10.1007/3-540-29623-9_5560

  4. Lao, K. et al. Drug development for Alzheimer’s disease. J. Drug Target. 27, 164–173 (2019).

    Article  Google Scholar 

  5. Cummings, J., Lee, G., Ritter, A., Sabbagh, M. & Zhong, K. Alzheimer’s disease drug development pipeline: 2019. Alzheimers Dement. 5, 272–293 (2019).

    Article  Google Scholar 

  6. Iqbal, K., Liu, F. & Gong, C.-X. Tau and neurodegenerative disease: the story so far. Nat. Rev. Neurol. 12, 15–27 (2016).

    Article  Google Scholar 

  7. Eckermann, K. et al. The β-propensity of tau determines aggregation and synaptic loss in inducible mouse models of tauopathy. J. Biol. Chem. 282, 31755–31765 (2007).

    Article  Google Scholar 

  8. Fatouros, C. et al. Inhibition of tau aggregation in a novel Caenorhabditis elegans model of tauopathy mitigates proteotoxicity. Hum. Mol. Genet. 21, 3587–3603 (2012).

    Article  Google Scholar 

  9. Goedert, M., Clavaguera, F. & Tolnay, M. The propagation of prion-like protein inclusions in neurodegenerative diseases. Trends Neurosci. 33, 317–325 (2010).

    Article  Google Scholar 

  10. Jucker, M. & Walker, L. C. Self-propagation of pathogenic protein aggregates in neurodegenerative diseases. Nature 501, 45–51 (2013).

    Article  Google Scholar 

  11. Clavaguera, F. et al. Transmission and spreading of tauopathy in transgenic mouse brain. Nat. Cell Biol. 11, 909–913 (2009).

    Article  Google Scholar 

  12. Aoyagi, A. et al. Aβ and tau prion-like activities decline with longevity in the Alzheimer’s disease human brain. Sci. Transl. Med. 11, eaat8462 (2019).

    Article  Google Scholar 

  13. Sanders, D. W. et al. Distinct tau prion strains propagate in cells and mice and define different tauopathies. Neuron 82, 1271–1288 (2014).

    Article  Google Scholar 

  14. Jackson, S. J. et al. Short fibrils constitute the major species of seed-competent tau in the brains of mice transgenic for human P301S tau. J. Neurosci. 36, 762–772 (2016).

    Article  Google Scholar 

  15. Furman, J. L. et al. Widespread tau seeding activity at early Braak stages. Acta Neuropathol. 133, 91–100 (2017).

    Article  Google Scholar 

  16. Despres, C. et al. Identification of the tau phosphorylation pattern that drives its aggregation. Proc. Natl. Acad. Sci. USA 114, 9080–9085 (2017).

    Article  Google Scholar 

  17. Lai, R., Harrington, C. & Wischik, C. Absence of a role for phosphorylation in the tau pathology of Alzheimer’s disease. Biomolecules 6, 19 (2016); erratum 6, 35 (2016).

  18. Goedert, M. et al. Assembly of microtubule-associated protein tau into Alzheimer-like filaments induced by sulphated glycosaminoglycans. Nature 383, 550–553 (1996).

    Article  Google Scholar 

  19. Grundke-Iqbal, I. et al. Abnormal phosphorylation of the microtubule-associated protein τ (tau) in Alzheimer cytoskeletal pathology. Proc. Natl. Acad. Sci. USA 83, 4913–4917 (1986).

    Article  Google Scholar 

  20. Sengupta, A. et al. Phosphorylation of tau at both Thr 231 and Ser 262 is required for maximal inhibition of its binding to microtubules. Arch. Biochem. Biophys. 357, 299–309 (1998).

    Article  Google Scholar 

  21. Alonso, A. C., Zaidi, T., Grundke-Iqbal, I. & Iqbal, K. Role of abnormally phosphorylated tau in the breakdown of microtubules in Alzheimer disease. Proc. Natl. Acad. Sci. USA 91, 5562–5566 (1994).

    Article  Google Scholar 

  22. Lindwall, G. & Cole, R. D. Phosphorylation affects the ability of tau protein to promote microtubule assembly. J. Biol. Chem. 259, 5301–5305 (1984).

    Article  Google Scholar 

  23. Johnson, G. V. W. & Stoothoff, W. H. Tau phosphorylation in neuronal cell function and dysfunction. J. Cell Sci. 117, 5721–5729 (2004).

    Article  Google Scholar 

  24. Gong, C.-X. & Iqbal, K. Hyperphosphorylation of microtubule-associated protein tau: a promising therapeutic target for Alzheimer disease. Curr. Med. Chem. 15, 2321–2328 (2008).

    Article  Google Scholar 

  25. Grandjean, J.-M. M. et al. Discovery of 4-piperazine isoquinoline derivatives as potent and brain-permeable tau prion inhibitors with CDK8 activity. ACS Med. Chem. Lett. 11, 127–132 (2020).

    Article  Google Scholar 

  26. Preuss, U., Döring, F., Illenberger, S. & Mandelkow, E. M. Cell cycle-dependent phosphorylation and microtubule binding of tau protein stably transfected into Chinese hamster ovary cells. Mol. Biol. Cell 6, 1397–1410 (1995).

    Article  Google Scholar 

  27. Malia, T. J. et al. Epitope mapping and structural basis for the recognition of phosphorylated tau by the anti-tau antibody AT8. Proteins 84, 427–434 (2016).

    Google Scholar 

  28. Goedert, M., Jakes, R. & Vanmechelen, E. Monoclonal antibody AT8 recognises tau protein phosphorylated at both serine 202 and threonine 205. Neurosci. Lett. 189, 167–169 (1995).

    Article  Google Scholar 

  29. Duka, V. et al. Identification of the sites of tau hyperphosphorylation and activation of tau kinases in synucleinopathies and Alzheimer’s diseases. PLoS ONE 8, e75025 (2013).

    Article  Google Scholar 

  30. Jensen, E. C. Overview of live-cell imaging: requirements and methods used. Anat. Rec. 296, 1–8 (2013).

    Article  Google Scholar 

  31. Sung, M.-H. & McNally, J. G. Live cell imaging and systems biology. Wiley Interdiscip. Rev. Syst. Biol. Med. 3, 167–182 (2011).

    Article  Google Scholar 

  32. Wang, C. et al. Scalable production of iPSC-derived human neurons to identify tau-lowering compounds by high-content screening. Stem Cell Rep. 9, 1221–1233 (2017).

    Article  Google Scholar 

  33. Azorsa, D. O. et al. High-content siRNA screening of the kinome identifies kinases involved in Alzheimer’s disease-related tau hyperphosphorylation. BMC Genomics 11, 25 (2010).

    Article  Google Scholar 

  34. Narayan, P. J. et al. Assessing fibrinogen extravasation into Alzheimer’s disease brain using high-content screening of brain tissue microarrays. J. Neurosci. Methods 247, 41–49 (2015).

    Article  Google Scholar 

  35. Vatansever, S. et al. Artificial intelligence and machine learning-aided drug discovery in central nervous system diseases: state-of-the-arts and future directions. Med. Res. Rev. 41, 1427–1473 (2021).

    Article  Google Scholar 

  36. Kim, S.-H. et al. Prediction of Alzheimer’s disease-specific phospholipase c gamma-1 SNV by deep learning-based approach for high-throughput screening. Proc. Natl. Acad. Sci. USA 118, e2011250118 (2021).

    Article  Google Scholar 

  37. Bermudez-Lugo, A, J., Rosales-Hernandez, M. C., Deeb, O., Trujillo-Ferrara, J. & Correa-Basurto, J. In silico methods to assist drug developers in acetylcholinesterase inhibitor design. Curr. Med. Chem. 18, 1122–1136 (2011).

    Article  Google Scholar 

  38. Basile, L. in Computational Modeling of Drugs Against Alzheimer’s Disease (ed. Roy, K.) 107–137 (Humana Press, 2018); https://doi.org/10.1007/978-1-4939-7404-7_4

  39. Carpenter, K. A. & Huang, X. Machine learning-based virtual screening and its applications to Alzheimer’s drug discovery: a review. Curr. Pharm. Des. 24, 3347–3358 (2018).

    Article  Google Scholar 

  40. Christiansen, E. M. et al. In silico labeling: predicting fluorescent labels in unlabeled images. Cell 173, 792–803 (2018).

    Article  Google Scholar 

  41. Ounkomol, C., Seshamani, S., Maleckar, M. M., Collman, F. & Johnson, G. R. Label-free prediction of three-dimensional fluorescence images from transmitted-light microscopy. Nat. Methods 15, 917–920 (2018).

    Article  Google Scholar 

  42. Moen, E. et al. Deep learning for cellular image analysis. Nat. Methods 16, 1233–1246 (2019).

    Article  Google Scholar 

  43. Ronneberger, O., Fischer, P. & Brox, T. U-Net: convolutional networks for biomedical image segmentation. In: Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015. MICCAI 2015. Lecture Notes in Computer Science (ed. Navab, N., et al.) 9351. (Springer, Cham, 2015). https://doi.org/10.1007/978-3-319-24574-4_28

  44. Liu, G. et al. Image inpainting for irregular holes using partial convolutions. In Proc. European Conference on Computer Vision (ECCV) 85–100 (Springer, 2018).

  45. Caicedo, J. C. et al. Data-analysis strategies for image-based cell profiling. Nat. Methods 14, 849–863 (2017).

    Article  Google Scholar 

  46. Hughes, J. P., Rees, S., Kalindjian, S. B. & Philpott, K. L. Principles of early drug discovery. Br. J. Pharmacol. 162, 1239–1249 (2011).

    Article  Google Scholar 

  47. Huang, N., Shoichet, B. K. & Irwin, J. J. Benchmarking sets for molecular docking. J. Med. Chem. 49, 6789–6801 (2006).

    Article  Google Scholar 

  48. Carpenter, A. E. et al. CellProfiler: image analysis software for identifying and quantifying cell phenotypes. Genome Biol. 7, R100 (2006).

    Article  Google Scholar 

  49. Müllers, E., Cascales, H. S., Burdova, K., Macurek, L. & Lindqvist, A. Residual Cdk1/2 activity after DNA damage promotes senescence. Aging Cell 16, 575–584 (2017).

    Article  Google Scholar 

  50. Chuang, K. V. & Keiser, M. J. Comment on ‘Predicting reaction performance in C-N cross-coupling using machine learning’. Science 362, eaat8603 (2018).

    Article  Google Scholar 

  51. Soekhoe, D., van der Putten, P. & Plaat, A. in Advances in Intelligent Data Analysis XV (ed. Boström, H., et al.) 50–60 (Springer, 2016); https://doi.org/10.1007/978-3-319-46349-0_5

  52. Williams, E. et al. Image Data Resource: a bioimage data integration and publication platform. Nat. Methods 14, 775–781 (2017).

    Article  Google Scholar 

  53. Allen, B. et al. Abundant tau filaments and nonapoptotic neurodegeneration in transgenic mice expressing human P301S tau protein. J. Neurosci. 22, 9340–9351 (2002).

    Article  Google Scholar 

  54. Lee, I. S., Long, J. R., Prusiner, S. B. & Safar, J. G. Selective precipitation of prions by polyoxometalate complexes. J. Am. Chem. Soc. 127, 13802–13803 (2005).

    Article  Google Scholar 

  55. Wager, T. T., Hou, X., Verhoest, P. R. & Villalobos, A. Central nervous system multiparameter optimization desirability: application in drug discovery. ACS Chem. Neurosci. 7, 767–775 (2016).

    Article  Google Scholar 

  56. Wager, T. T., Hou, X., Verhoest, P. R. & Villalobos, A. Moving beyond rules: the development of a central nervous system multiparameter optimization (CNS MPO) approach to enable alignment of druglike properties. ACS Chem. Neurosci. 1, 435–449 (2010).

    Article  Google Scholar 

  57. Bray, M.-A. et al. Cell Painting, a high-content image-based assay for morphological profiling using multiplexed fluorescent dyes. Nat. Protoc. 11, 1757–1774 (2016).

    Article  Google Scholar 

  58. Wong, D. & Keiser, M. Trans-channel Fluorescence Learning (OSFHOME, 2020); https://doi.org/10.17605/OSF.IO/XNTD6

  59. Keiser, M. keiserlab/trans-channel-paper: v1.0.0 (Zenodo, 2022); https://doi.org/10.5281/zenodo.6336183

Download references

Acknowledgements

This work was supported by grant no. 2018-191905 from the Chan Zuckerberg Initiative DAF, an advised fund of the Silicon Valley Community Foundation (M.J.K.), the National Institutes of Health (AG002132; S.B.P.), as well as by support from the Brockman Foundation (S.B.P.) and the Sherman Fairchild Foundation (S.B.P.).

Author information

Author notes

  1. These authors contributed equally: Jay Conrad, Noah R. Johnson.

Authors and Affiliations

  1. Institute for Neurodegenerative Diseases, University of California, San Francisco, CA, USA

    Daniel R. Wong, Jay Conrad, Noah R. Johnson, Jacob Ayers, Joanne C. Lee, Jisoo Lee, Stanley B. Prusiner, Nick A. Paras & Michael J. Keiser

  2. Bakar Computational Health Sciences Institute, University of California, San Francisco, CA, USA

    Daniel R. Wong, Atul J. Butte & Michael J. Keiser

  3. Department of Pharmaceutical Chemistry, University of California, San Francisco, CA, USA

    Daniel R. Wong & Michael J. Keiser

  4. Department of Bioengineering and Therapeutic Sciences, University of California, San Francisco, CA, USA

    Daniel R. Wong, Annelies Laeremans, Sourav Bandyopadhyay & Michael J. Keiser

  5. Kavli Institute for Fundamental Neuroscience, University of California, San Francisco, CA, USA

    Daniel R. Wong & Michael J. Keiser

  6. Department of Pediatrics, University of California, San Francisco, CA, USA

    Daniel R. Wong & Atul J. Butte

  7. Center for Data-Driven Insights and Innovation, University of California, Office of the President, Oakland, CA, USA

    Daniel R. Wong & Atul J. Butte

  8. University of Colorado Alzheimer’s and Cognition Center, Linda Crnic Institute for Down Syndrome, University of Colorado, Anschutz Medical Campus, Aurora, CO, USA

    Noah R. Johnson

  9. Department of Neurology, University of Colorado, Anschutz Medical Campus, Aurora, CO, USA

    Noah R. Johnson

  10. Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA, USA

    Annelies Laeremans & Sourav Bandyopadhyay

Authors
  1. Daniel R. Wong
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Jay Conrad
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Noah R. Johnson
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Jacob Ayers
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Annelies Laeremans
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Joanne C. Lee
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Jisoo Lee
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Stanley B. Prusiner
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Sourav Bandyopadhyay
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Atul J. Butte
    View author publications

    You can also search for this author in PubMed Google Scholar

  11. Nick A. Paras
    View author publications

    You can also search for this author in PubMed Google Scholar

  12. Michael J. Keiser
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Conceptualization was provided by D.R.W., J.C., N.J., N.A.P. and M.J.K., methodology by D.R.W., J.C., N.J., N.A.P. and M.J.K., software by D.R.W., validation by D.R.W., A.L. and S.B.P., formal analysis by D.R.W., investigations by D.R.W., J.C., N.J., J.C.L., J.A., J.L. and A.L., resources by S.B.P., S.B., A.J.B., N.A.P. and M.J.K. and data curation by D.R.W. The original draft was written by D.R.W. Reviewing and editing was carried out by D.R.W. and M.J.K. Visualization was provided by D.R.W., supervision by S.B.P., S.B., A.J.B., N.A.P. and M.J.K., project administration by D.R.W. and M.J.K. and funding acquisition by S.B.P. and M.J.K.

Corresponding author

Correspondence to Michael J. Keiser.

Ethics declarations

Competing interests

The authors declare no competing interests. S.B.P. is a member of the Scientific Advisory Board of ViewPoint Therapeutics and a member of the Board of Directors of Trizell, Ltd, neither of which have contributed financial or any other support to the studies discussed here.

Peer review

Peer review information

Nature Machine Intelligence thanks Florian Heigwer and Hao Zhu for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Supplementary Figs. 1–13.

Reporting Summary

Supplementary Table 1

Supplementary Tables

Rights and permissions

Reprints and Permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Wong, D.R., Conrad, J., Johnson, N.R. et al. Trans-channel fluorescence learning improves high-content screening for Alzheimer’s disease therapeutics. Nat Mach Intell 4, 583–595 (2022). https://doi.org/10.1038/s42256-022-00490-8

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/s42256-022-00490-8

This article is cited by

Search

Advanced search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing