Resolving transcriptional dynamics of the epithelial-mesenchymal transition using single-cell RNA sequencing
- 1. Resolving transcriptional dynamics of the epithelial-mesenchymal transition using single-cell RNA sequencing David Cook (@DavidPCook) PhD Candidate, Vanderhyden Lab February 2nd, 2018 OHRI 10x Genomics Seminar
- 3. Accuracy—average measurements from complex distributions are meaningless
- 4. Biological systems are complex – Tissue Heterogeneity ~100,000 single cells from 20 mouse tissues Glossary tSNE (t-distributed stochastic neighbor embedding): A dimensionality reduction technique (like PCA) that enables visualization of complex relationships between cells in two dimensions The Tabula Muris Consortium, BioRxiv, 2017
- 5. The Tabula Muris Consortium, BioRxiv, 2017 Biological systems are complex – Tissue Heterogeneity
- 6. Biological systems are complex – Cellular Heterogeneity Cell Cycle Apoptosis/Senescence Phenotypic spectra (eg Th17 T-cells) Gaublomme et al., Cell, 2015 Xu et al., Blood, 2006 The Human Cell Atlas, eLife, 2017
- 7. If money weren’t a factor, there’s no reason not to do it
- 8. Resolving transcriptional dynamics of the epithelial-mesenchymal transition using single-cell RNA sequencing David Cook (@DavidPCook) PhD Candidate, Vanderhyden Lab February 2nd, 2018 OHRI 10x Genomics Seminar
- 10. (Maybe) The EMT contributes to the progression of ovarian cancer
- 11. Thiery et al., Cell, 2009 Lamouille et al., Nat Rev Mol Cell Biol, 2014 What about this??? The transcriptional determinants of the EMT are poorly defined
- 12. What are the transcriptional determinants associated with the EMT?
- 13. ** p < 0.01 OVCA420 ovarian cancer cells undergo a typical EMT in response to TGFB1
- 14. Enhanced invasion Enhanced migration Reduced proliferation OVCA420 ovarian cancer cells undergo a typical EMT in response to TGFB1
- 15. 1 day 3 days 7 days 0 days 10,000 OVCA420 cells undergoing a TGFB1-induced EMT ~1.5 Billion Reads 1. Untreated Controls 2. TGFB1 1d 3. TGFB1 3d 4. TGFB1 7d
- 16. Transcriptional trajectory of the EMT
- 17. 1,777 differentially expressed genes Gene expression dynamics throughout the EMT
- 18. Gene expression dynamics throughout the EMT
- 19. Gene expression dynamics throughout the EMT Metabolic changes? Immunoregulatory changes?
- 20. Trajectory analysis can miss interesting nuances in the data DC1 DC2 Glossary DC (Diffusion Component): Component of a diffusion map—a dimensionality reduction approach based on transition probabilities between cells 1 day 3 days 7 days 0 days
- 21. Predicting a cell’s future state La Manno et al., BioRxiv, 2017 Glutaminergic neuronal development (Forebrain from 10wk human embryo)
- 22. Mapping cell dynamics along a manifold DC1 DC2
- 23. Mapping cell dynamics along a manifold DC1 DC2
- 24. Mapping cell dynamics along a manifold DC1 DC2
- 25. Unspliced vs. spliced ratio Total expression Mapping cell dynamics along a manifold
- 26. But how do we find expression patterns? DC1 DC2
- 27. But how do we find expression patterns? Svennson et al., BioRxiv, 2017
- 28. What coordinates these changes?
- 29. Using ATAC-Seq to identify putative drivers of these dynamics Buenrostro et al., Nature Methods, 2013 1kb FOXS1 Control TGFB1 1d TGFB1 3d TGFB1 7d Scan all peaks for presence of transcription factor motifs Quantify differences in accessibility across samples Find all motifs with significant changes in accessibility across samples
- 30. Putative drivers of changes associated with the EMT U ntreated TG FB1 1d TG FB1 3d TG FB1 7d 115 Transcription Factor Motifs NFKB1 RELA HNF1A CTCF FOXA2 E2F1 ELF3 EHF ETS2 CEBPA JUNB FOSB SNAI2 TP63 SMAD3 AccessibilityZ-score
- 31. Putative drivers of changes associated with the EMT U ntreated TG FB1 1d TG FB1 3d TG FB1 7d 115 Transcription Factors Can we link these motifs to the regulatory regions of these genes? 1,777 Differentially Expressed Genes
- 32. Next Steps Validate transcriptional patterns Assess importance of identified transcription factors Explore the generalizability of this EMT across other contexts • RNA-FISH for markers of distinct stages of the EMT • CRISPR-mediated deletion of several transcription factors followed by • Scaling up and performing scRNA-Seq on a variety of epithelial lines (normal & cancer) undergoing an EMT
- 33. Why?
- 34. Dr. Barb Vanderhyden Cristina Andronic Lauren Carter Dr. David Wang (McGill) Dr. Ioannis Ragoussis (McGill) The Vanderhyden Lab Dr. Ken Garson Dr. Ati Atefeh Dr. Galaxia Rodriguez Dr. Shan Zhao Liz Macdonald Rose Vuong Pascale Charette Curtis McCloskey Laura Forrest Kristianne Galpin Lisa Oliviero Omar Salah Salah Brendan Kelly Acknowledgements The Developers CellRanger: 10x Genomics scater: Davis McCarthy et al. Seurat: Rahul Satija et al. scanpy: Alex Wolfe et al. velocyto: Gioele La Manno et al. spatialDE: Valentine Svennson et al. chromVAR: Alicia Schep et al. Anyone missed and the rest of the comp bio community! The StemCore Crew Caroline Vergette Cindy Gadoury Kathy Sheikheleslamy Pearl Campbell