Final Report Summary - SPIM MOUSE (Characterization of the cellular mechanics of lineage segregation during mouse blastocyst morphogenesis by SPIM-based 4D-imaging and micromanipulation)
In mammalian pre-implantation development, the cells forming the blastocyst are segregated into two cell lineages: the inner cell mass (ICM) in the inner-cell layer and the trophectoderm (TE) in the outer cell layer. The ICM and TE give rise to embryonic and extra-embryonic tissues respectively. The aim of this project is to identify the physical and/or molecular cues that segregate cells into either inner or outer cell layer.
Cell division and cell sorting along in-out axis are two putative mechanisms segregating ICM cells from TE cells. The patterns of cell division and cell sorting are different from embryo to embryo in mice, making it difficult to extract cues that orient these two processes along in-out axis. To overcome this difficulty, we first established the methods of high-resolution fluorescent live-cell imaging and 4D digital membrane segmentation. To minimize photo-bleaching and damage during live-imaging, we applied selective plane illumination microscopy (SPIM), which has been extensively developed at the host institution, European Molecular Biology Laboratory (EMBL), (Krzic et al., 2012; Strnad et al., 2016). In collaboration with Petr Strnad, Lars Hufnagel, and Jan Ellenberg, we decided to use an inverted SPIM, since it allowed us to visualize many embryos simultaneously in high spatio-temporal resolution. We visualized the nuclei and the cell membrane by using the microscope and transgenic fluorescent reporter available in the Hiiragi group. We optimized our imaging protocol by modulating laser intensity and minimizing z-spacing and the time interval, while keeping embryos alive. The spatial and temporal resolutions for the imaging from 4-cell stage to the blastocyst stage has reached 1 micrometer z-spacing and 10 minutes interval respectively, which was sufficient to track all the cell divisions and cell sorting segregating ICM cells from TE cells. We also optimized spinning-disc microscope to visualize the mouse embryo for experiments that do not require in-toto imaging.
To segment the cell membrane in big image data from either the inverted SPIM or the spinning-disk microscope, I developed automated image-processing pipelines. I initially established a program to segment the cell membrane in 2D images, with which we can quantify and track the local curvature of the cell membrane, and the local signal intensity of the fluorescently labeled membrane protein. A 3D membrane segmentation pipeline was also established by combining the active-contour method described in Mikula et al. (2011) and the sparse-field method. This combination accelerated the Mikula’s algorithm by the factor of more than 10. We ran the pipeline in the PC cluster, which allowed the simultaneous segmentation of the all the cells in our time-lapse in a few hours. Overall, I established both imaging and image processing pipeline to track the dynamics of cell division and cell sorting, which was an essential part of the proposed research.
We applied the 2D image-processing pipeline to analyze the mechanism of the compaction at the 8-cell stage. The compaction refers to rounding-up of the 8-cell aggregate, and is thus the first morphological event of the mouse embryo. In the mouse pre-implantation embryo, the compaction starts from the 8-cell stage and is important for the embryonic morphology in the blastocyst stage. My colleague, Jean-Leon Maitre found that the surface tension at the cell-medium interface, which depends on the contractility generated by myosin, increases during compaction at the 8-cell stage whereas it decreases at the cell-cell contact, and showed it is sufficient to explain the compaction. To characterize the contractility, Jean-Leon Maitre imaged isolated cells from 8-cell stage embryos in 2D with high temporal resolution. His images showed the dynamic motion of the cell surface in a actin and myosin dependent manner. Our image analysis showed that cortical domains with low/high curvature have high/low amount of the cortical actin respectively, and that such domains propagate either clockwise or counterclockwise all over the surface with a periodicity of about 80 seconds. This dynamic contractility that is novel to the mouse development was reported in Nature Cell Biology (Maitre, Niwayama et al., doi:10.1038/ncb3185). Furthermore, our analysis established that one of daughter cells in 8-to-16 division inherits the dynamic contractility, and that a cell with higher contractility tends to be covered by the sister cell with lower contractility. Combining this knowledge with other experimental results and simulation, we proposed that asymmetric inheritance of the contractility is an essential event spatially segregating ICM from TE cells. This study is in press in Nature (Maitre et al., 2016).
Through this project, I established the imaging and image-processing pipelines useful for biologists analyzing the early mouse development. Their application contributed to establish that the cortical contractility is essential not only for cellular morphogenesis but also for lineage segregation, achieving the major objective of the project. The image analysis to quantify the motion of cells due to contractility might be able to determine which embryonic cells in vitro tend to differentiate into pluripotent cells. This might have some applications in basic and/or clinical researches, and therefore have a high potential to contribute to improve the health and wellness of human being.
Cell division and cell sorting along in-out axis are two putative mechanisms segregating ICM cells from TE cells. The patterns of cell division and cell sorting are different from embryo to embryo in mice, making it difficult to extract cues that orient these two processes along in-out axis. To overcome this difficulty, we first established the methods of high-resolution fluorescent live-cell imaging and 4D digital membrane segmentation. To minimize photo-bleaching and damage during live-imaging, we applied selective plane illumination microscopy (SPIM), which has been extensively developed at the host institution, European Molecular Biology Laboratory (EMBL), (Krzic et al., 2012; Strnad et al., 2016). In collaboration with Petr Strnad, Lars Hufnagel, and Jan Ellenberg, we decided to use an inverted SPIM, since it allowed us to visualize many embryos simultaneously in high spatio-temporal resolution. We visualized the nuclei and the cell membrane by using the microscope and transgenic fluorescent reporter available in the Hiiragi group. We optimized our imaging protocol by modulating laser intensity and minimizing z-spacing and the time interval, while keeping embryos alive. The spatial and temporal resolutions for the imaging from 4-cell stage to the blastocyst stage has reached 1 micrometer z-spacing and 10 minutes interval respectively, which was sufficient to track all the cell divisions and cell sorting segregating ICM cells from TE cells. We also optimized spinning-disc microscope to visualize the mouse embryo for experiments that do not require in-toto imaging.
To segment the cell membrane in big image data from either the inverted SPIM or the spinning-disk microscope, I developed automated image-processing pipelines. I initially established a program to segment the cell membrane in 2D images, with which we can quantify and track the local curvature of the cell membrane, and the local signal intensity of the fluorescently labeled membrane protein. A 3D membrane segmentation pipeline was also established by combining the active-contour method described in Mikula et al. (2011) and the sparse-field method. This combination accelerated the Mikula’s algorithm by the factor of more than 10. We ran the pipeline in the PC cluster, which allowed the simultaneous segmentation of the all the cells in our time-lapse in a few hours. Overall, I established both imaging and image processing pipeline to track the dynamics of cell division and cell sorting, which was an essential part of the proposed research.
We applied the 2D image-processing pipeline to analyze the mechanism of the compaction at the 8-cell stage. The compaction refers to rounding-up of the 8-cell aggregate, and is thus the first morphological event of the mouse embryo. In the mouse pre-implantation embryo, the compaction starts from the 8-cell stage and is important for the embryonic morphology in the blastocyst stage. My colleague, Jean-Leon Maitre found that the surface tension at the cell-medium interface, which depends on the contractility generated by myosin, increases during compaction at the 8-cell stage whereas it decreases at the cell-cell contact, and showed it is sufficient to explain the compaction. To characterize the contractility, Jean-Leon Maitre imaged isolated cells from 8-cell stage embryos in 2D with high temporal resolution. His images showed the dynamic motion of the cell surface in a actin and myosin dependent manner. Our image analysis showed that cortical domains with low/high curvature have high/low amount of the cortical actin respectively, and that such domains propagate either clockwise or counterclockwise all over the surface with a periodicity of about 80 seconds. This dynamic contractility that is novel to the mouse development was reported in Nature Cell Biology (Maitre, Niwayama et al., doi:10.1038/ncb3185). Furthermore, our analysis established that one of daughter cells in 8-to-16 division inherits the dynamic contractility, and that a cell with higher contractility tends to be covered by the sister cell with lower contractility. Combining this knowledge with other experimental results and simulation, we proposed that asymmetric inheritance of the contractility is an essential event spatially segregating ICM from TE cells. This study is in press in Nature (Maitre et al., 2016).
Through this project, I established the imaging and image-processing pipelines useful for biologists analyzing the early mouse development. Their application contributed to establish that the cortical contractility is essential not only for cellular morphogenesis but also for lineage segregation, achieving the major objective of the project. The image analysis to quantify the motion of cells due to contractility might be able to determine which embryonic cells in vitro tend to differentiate into pluripotent cells. This might have some applications in basic and/or clinical researches, and therefore have a high potential to contribute to improve the health and wellness of human being.