Image processing to automate stem cell tracking
GFP actin expressed in human umbilical vascular endothelial cells. Picture courtesy of Liyuan Wang, PhD candidate
The three major questions in stem cell research are (a) how do stem cells differentiate into specialised tissue cell types (b) how do cells know where to go in the body and (c) how do cells know when to stop dividing. An exciting new method to learn about stem cells and answer these important questions involves recording and analysing videos of their lifecycle. A major problem with this approach is the time it takes to manually track thousands of cells in movies; a few hundred cell lifecycles may take weeks. The cellular dynamics lab at the Graduate School of Biomedical Engineering has developed live cell imaging to study stem cell lifecycles in a precisely controlled culture environment. Our lab is using live cell imaging to (a) study the process of stem cell development and (b) quantify the motion of cells in the presence of chemical attractants.
Computer vision software
The arduous task of cell tracking has been streamlined by applying object tracking and face recognition algorithms. Cell trajectory maps are entered into a hierarchical database representing cell kinship relationships, with sophisticated statistical analysis of cell fate outcomes. Future developments include machine learning methods (e.g., hidden Markov chains) to optimise tracking parameters for greater automation and accuracy.
Real time imaging of subcellular signalling and transcription using fluorescent protein reporters
Fluorescent protein reporters are used to establish the causative link between the molecular machinery inside the cell and cell fate outcomes as observed by live cell imaging. We have developed software to track the FUCCI cell cycle reporter system (Draper et al, McMasters Stem Cell and Cancer Research Institute, Hamilton, Canada), allowing one to study the effect of genetic or external cues on cell cycle control.
When fluorescent protein expression is driven by transcription factors, it is possible to understand how transcription factors regulate cell fate decisions. Fluorescent reporter transgenic mice models are used to track the expression of transcriptional proteins in cardiac-derived stem cells (Harvey et al. Victor Chang Cardiac Research Centre, Sydney, Australia).
Live cell tracking also allows one to study the activity of intra-cellular signalling molecules that direct cell locomotion. The rho GTPases, key mechanotransducers that control cytoskeletal remodelling, are studied using Raichu probes (Matsuda et al., Osaka University, Japan). These live cell sensors detect rho GTPase activation by a change in their fluorescent colour (Forster energy transfer). This knowledge will lead to a better understanding of the molecular machinery controlling guidance of cells through tissues along chemical gradients.