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.

Application of machine learning to large in vitro databases to identify drug–cancer cell interactions: azithromycin and KLK6 mutation status

Abstract

Recent advances in machine learning promise to yield novel insights by interrogation of large datasets ranging from gene expression and mutation data to CRISPR knockouts and drug screens. We combined existing and new algorithms with available experimental data to identify potentially clinically relevant relationships to provide a proof of principle for the promise of machine learning in oncological drug discovery. Specifically, we screened cell line data from the Cancer Dependency Map for the effects of azithromycin, which has been shown to kill cancer cells in vitro. Our findings demonstrate a strong relationship between Kallikrein Related Peptidase 6 (KLK6) mutation status and the ability of azithromycin to kill cancer cells in vitro. While the application of azithromycin showed no meaningful average effect in KLK6 wild-type cell lines, statistically significant enhancements of cell death are seen in multiple independent KLK6-mutated cancer cell lines. These findings suggest a potentially valuable clinical strategy in patients with KLK6-mutated malignancies.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.

from$8.99

All prices are NET prices.

Fig. 1: R2 (https://en.wikipedia.org/wiki/Coefficient_of_determination) for predicting CRISPR knockout effects on holdout datasets for the 730 genes selected to be both efficacious and selective through the application of the arbitrarily selected thresholds efficacy = −0.56, selectivity = 1 as defined by Shimada et al. [28].
Fig. 2: The effect of azithromycin on cancer cell lines.

References

  1. 1.

    Travis J. Making the cut. Science. 2015;350:1456–7.

    CAS  Article  Google Scholar 

  2. 2.

    Jinek M, Chylinski K, Fonfara I, Hauer M, Doudna JA, Charpentier EA. Programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity. Science. 2012;337:816–21.

    CAS  Article  Google Scholar 

  3. 3.

    Lander ES, Linton LM, Birren B, Nusbaum C, Zody MC, Baldwin J, et al. (International Human Genome Sequencing Consortium (IHGSC)). Finishing the euchromatic sequence of the human genome. Nature. 2004;431:931–45.

    Article  Google Scholar 

  4. 4.

    Qiao X, Wang X, Shang Y, Li Y, Chen S. Azithromycin enhances anticancer activity of TRAIL by inhibiting autophagy and up- regulating the protein levels of DR4/5 in colon cancer cells in vitro and in vivo. Cancer Commun. 2018;38:43.

    Article  Google Scholar 

  5. 5.

    Fiorillo M, Toth F, Sotgia F, Lisanti MP. Doxycycline, azithromycin and vitamin C (DAV): a potent combination therapy for targeting mitochondria and eradicating cancer stem cells (CSCs). Aging. 2019;11:2202.

    CAS  Article  Google Scholar 

  6. 6.

    Lamb R, Ozsvari B, Lisanti CL, Tanowitz HB, Howell A, Martinez-Outschoorn UE, et al. Antibiotics that target mitochondria effectively eradicate cancer stem cells, across multiple tumor types: Treating cancer like an infectious disease. Oncotarget. 2015;6:4569.

    Article  Google Scholar 

  7. 7.

    Li F, Huang J, Ji D, Meng Q, Wang C, Chen S, et al. Azithromycin effectively inhibits tumor angiogenesis by suppressing vascular endothelial growth factor receptor 2 mediated signaling pathways in lung cancer. Oncol Lett. 2017;14:89–96.

    CAS  Article  Google Scholar 

  8. 8.

    Tsherniak A, Vazquez F, Montgomery PG, Weir BA, Kryukov G, Cowley GS, et al. Defining a cancer dependency map. 2017. Cell. 2017;170:564.

    CAS  Article  Google Scholar 

  9. 9.

    Ghandi M, Huang FW, Jané-Valbuena J, Kryukov GV, Lo CC, McDonald ER, et al. Next-generation characterization of the Cancer Cell Line Encyclopedia. Nature. 2019;569:503.

    CAS  Article  Google Scholar 

  10. 10.

    Pedregosa F, Varoquaux G, Gramfort A, Michel V, Thirion B, Grisel O, et al. Scikit-learn: Machine Learning in Python. J Mach Learn Res. 2011;12:2825.

    Google Scholar 

  11. 11.

    Lundberg SM, Lee S. A unified approach to interpreting model predictions. In: Advances in neural information processing systems. Long Beach, CA, 30;2017.

  12. 12.

    Lundberg SM, Erion G, Chen H, DeGrave A, Prutkin JM, Nair B, et al. From local explanations to global understanding with explainable AI for trees. Nat Mach Intell. 2020;2:56.

    Article  Google Scholar 

  13. 13.

    Raileanu LE, Stoffel K. Theoretical Comparison between the Gini Index and Information Gain Criteria. Ann Math Artif Intell. 2004;41:77.

    Article  Google Scholar 

  14. 14.

    DepMap, Broad. DepMap 21Q1 Public. figshare. 2020. https://doi.org/10.6084/m9.figshare.13681534.v1.

  15. 15.

    Bazaga A, Leggate D, Weisser H. Genome-wide investigation of gene-cancer associations for the prediction of novel therapeutic targets in oncology. Sci Rep. 2020;10:10787.

    CAS  Article  Google Scholar 

  16. 16.

    Dezső Z, Ceccarelli M. Machine learning prediction of oncology drug targets based on protein and network properties. BMC Bioinform. 2020;21:104.

    Article  Google Scholar 

  17. 17.

    Madhukar NS, Khade PK, Huang L, Gayvert K, Galletti G, Stogniew M, et al. A Bayesian machine learning approach for drug target identification using diverse data types. Nat Commun. 2019;10:5221.

    Article  Google Scholar 

  18. 18.

    Lord CJ, Ashworth A. PARP inhibitors: synthetic lethality in the clinic. Science. 2017;355:1152.

    CAS  Article  Google Scholar 

  19. 19.

    Tamir A, Jag U, Sarojini S, Schindewolf C, Tanaka T, Gharbaran R, et al. Kallikrein family proteases KLK6 and KLK7 are potential early detection and diagnostic biomarkers for serous and papillary serous ovarian cancer subtypes. J Ovarian Res. 2014;7:109.

    Article  Google Scholar 

  20. 20.

    Haritos C, Michaelidou K, Mavridis K, Missitzis J, Ardavanis A, Griniatsos J, et al. Kallikrein-related peptidase 6 (KLK6) expression differentiates tumor subtypes and predicts clinical outcome in breast cancer patients. Clin Exp Med. 2018;18:203–13.

    CAS  Article  Google Scholar 

  21. 21.

    Ahmed N, Dorn J, Napieralski R, Drecoll E, Kotzsch M, Goettig P, et al. Clinical relevance of kallikrein-related peptidase 6 (KLK6) and 8 (KLK8) mRNA expression in advanced serous ovarian cancer. Biol Chem. 2016;397:1265.

    CAS  Article  Google Scholar 

  22. 22.

    Wang SM, Mao J, Li B, Wu W, Tang LL. Expression of KLK6 protein and mRNA in primary breast cancer and its clinical significance. Chin J Cell Mol Immunol. 2008;24:1087.

    CAS  Google Scholar 

  23. 23.

    Chalmers ZR, Connelly CF, Fabrizio D, Gay L, Ali SM, Ennis R, et al. Analysis of 100,000 human cancer genomes reveals the landscape of tumor mutational burden. Genome Med. 2017;9:34.

    Article  Google Scholar 

  24. 24.

    Jiang X, Baucom C, Elliott RL. Mitochondrial toxicity of azithromycin results in aerobic glycolysis and DNA damage of human mammary epithelia and fibroblasts. Antibiotics. 2019;8:E110.

    Article  Google Scholar 

  25. 25.

    Corsello SM, Nagari RT, Spangler RD, Rossen J, Kocak M, Bryan JG, et al. Discovering the anticancer potential of non-oncology drugs by systematic viability profiling. Nat Cancer. 2020;1:235–48.

    Article  Google Scholar 

  26. 26.

    Schrader CH, Kolb M, Zaoui K, Flechtenmacher C, Grabe N, Weber KJ, et al. Kallikrein-related peptidase 6 regulates epithelial-to-mesenchymal transition and serves as prognostic biomarker for head and neck squamous cell carcinoma patients. Mol Cancer. 2015;14:107.

    Article  Google Scholar 

  27. 27.

    Wang H, Unternaehrer JJ. Epithelial-mesenchymal transition and cancer stem cells: at the crossroads of differentiation and dedifferentiation. Developmental Dyn. 2019;248:10–20.

    Article  Google Scholar 

  28. 28.

    Shimada K, Muhlich JL, Mitchison TJ. A tool for browsing the Cancer Dependency Map reveals functional connections between genes and helps predict the efficacy and selectivity of candidate cancer drugs. 2019. https://www.biorxiv.org/content/10.1101/2019.12.13.874776v1.

Download references

Acknowledgements

The authors would like to thank Eoin McDonnell, Kris Wood, and Jonathan Mizrahi for relevant discussions.

Author information

Affiliations

Authors

Corresponding author

Correspondence to Jeff Sherman.

Ethics declarations

Conflict of interest

JS, GV, and YB have roles at Red Cell Partners and Zephyr AI that involve the application of AI to cancer drug discovery. The authors all report no known financial interest in azithromycin.

Additional information

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

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Sherman, J., Verstandig, G., Rowe, J.W. et al. Application of machine learning to large in vitro databases to identify drug–cancer cell interactions: azithromycin and KLK6 mutation status. Oncogene 40, 3766–3770 (2021). https://doi.org/10.1038/s41388-021-01807-4

Download citation

Search

Quick links