The Effect of ATP INDUCED CALCIUM DYNAMICS ON EPITHELIAL TO MESENCHYMAL TRANSITIONS

Cells respond to a multitude of external triggers by a limited number of signaling pathways activated by receptors on plasma membrane, such as receptor tyrosine kinases (RTKs) or G protein-coupled receptors (GPCRs). These pathways do not simply convey the downstream signal, but instead the signal is very often processed by encoding and integrated with the current state of the cell.
A traditional transcriptional analysis tends to provide an averaged output measured in a population, what often masks the behavior of individual cells. However, with recent single cell techniques developments, it is possible to investigate transcription in individual living cells. This contributed tremendously to the understanding of development and progression of many diseases including cancer. The more we understand about this high complexity of signaling mechanisms and multitude of cellular safety countermeasures, the more we see cancer as a microevolution state of “rebellious cells” (cells entering the fate opposite to the one intended) following a patch through a discreet system.
This thesis specifically focused on the temporal aspect of signaling in the context of the epithelia-to-mensenchymal transition (EMT) by combining single cell experiments and bioinformatics analysis. We investigated cellular signaling changes in response to different dynamical profiles of the stimuli. In particular, we used the HMLER cell line, which is a metastatic breast cancer model for the epithelial to mesenchymal transition. By applying stochastic or oscillatory pulses of extracellular ATP-induced Ca2+ signals with different interspike intervals, we were investigating different transcription states from those evoked by constant ATP-induced Ca2+ dose responses. In order to precisely apply those stimulation profiles, we have developed and established a perfusion system. This device allows to treat population of cells simultaneously with the exact same dynamical profiles. Cells treated by these well controlled signals were subsequently processed by the single cell RNA-seq technique Drop-seq for transcriptional analysis. The resulting high dimensional digital gene expression matrices were analyzed by a developed high-throughput computational analysis pipeline. This analysis includes the identification of differentially expressed genes and cellular clusters (states) by dimensionality reduction methods (PCA, t-SNE) and pathway analysis. We evaluated changes and trends of genes from difference dynamical profiles by investigating their involvement in stress, stemness and regulation of motility.
First, we confirmed that oscillatory stimulation with extracellular ATP (eATP) tends to lower the burden of cellular stress and apoptosis related pathways while maintaining its other effector functions compared to constant eATP stimulation. Interestingly, stochastic spiking of extracellular ATP in our setup led to a massive (~80%) increase in overall differential gene expression compared to deterministic oscillatory stimulation with the same period. Consequentially, stochastic signaling seems to activate a much wider range of biological pathways, which indicates the much higher complexity in information processing capability of producing rebellious cells during cancer progression and metastasis. On the other hand, our findings suggests that oscillatory eATP stimulation could contribute to EMT by lowering ID3 expression compared to stochastic stimulation where we observed a stronger upregulation of IRS2. Finally, we integrated the DEGs into biological processes involved in each conditions and put these new insights into the context of the eATP-induced Ca2+ induced epithelial to mesenchymal transition.
Overall, this thesis has applied recent single cell technologies to characterize underlying principles of cellular heterogeneity induced by cell signaling and specifically investigated the complex mechanisms of cell fate in the context of EMT