So far, the clearest picture of living embryos comes from zebrafish and fruit flies. Ten years ago, Philipp Keller, a physicist and biologist at the Jenice Research Institute in the Howard Hughes Medical Institute, and his colleagues developed the first “digital embryo” of zebrafish, in which the zebrafish was a transparent striped fish that is usually provided to scientists for research. They scanned the zebrafish embryos with a light sheet microscope. Keller designed a computer program that collects and analyzes all the imaging data, and the result is a high-resolution observation of the first 24 hours of zebrafish embryo development.
In 2014, Keller and his colleagues reported digital fruit fly embryos in the Nature Methods (Nature Methods, 2014, doi:10.1038/nmeth.3036). Keller said that imaging these animals is relatively straightforward, especially zebrafish. They are transparent and insensitive to light, making them “the target of easy observation by the microscope.”
Mice are different from zebrafish and fruit flies. Keeping mouse embryos alive in the laboratory—even for a short time—requires a range of conditions. First, mouse embryos must remain sterile; they need to be immersed in a nutrient solution; gas and temperature levels must be precisely controlled. More importantly, the cells in the embryo are very sensitive to light, the embryonic tissue is dense and opaque, and the embryo cannot remain stationary under the microscope. Instead, it is only fixed at one point, so it “driers like a small balloon.”
Finally, during the time that the Keller team wanted to observe, that is, six and a half days to eight and a half days after fertilization, the mouse embryos grew more than an order of magnitude—a diameter of almost 3 mm, roughly equivalent to a sesame seed length. For light-illuminated microscopes, mouse embryos are a moving target that constantly changes size and position.
In a new study, in order to solve these problems, the Keller team adopted a different strategy: they designed a smart microscope that could do all the work. The relevant research results were published online in the Cell, and the title of the paper is “In Toto Imaging and Reconstruction of Post-Implantation Mouse Development at the Single-Cell Level.”
In the center of this intelligent microscope, a clear acrylic cubic structure houses the embryonic imaging chamber. Two light sheets illuminate the mouse embryo and two cameras record the image. These components allow these researchers to spy on the world of early organ development that was once invisible, revealing dynamic events with high-resolution details that have never been seen before.
The head of the microscope is equipped with a set of algorithms that track the position and size of the embryo. These algorithms map how the light sheet moves through the sample and then find out how to get the best image—keeping the mouse embryo focused in the field of view and in the middle of the field of view.
As mouse embryo is constantly changing, the microscope must constantly adapt to make decisions on hundreds of images at hundreds of different time points in millisecond intervals. Keller said, “I won’t say that our microscope is smarter than humans, but it can do things that human operators can’t do.”
Using this intelligent microscope, the Keller team is now able to peer into the living mouse embryo for the first time, observing the beginning of the formation of the intestines, and the heart cells begin to try to beat for the first time. In a critical 48-hour window—the period in which the primary organs begin to form—they can track each embryonic cell and determine where they are going, which genes they open, and which cells they encounter on the road.
This new study is actually a cellular resolution building plan for the entire mouse. They are making such smart microscopes and calculation tools, and all imaging data is free and open.
Katie McDole, Léo Guignard, Fernando Amat et al. In Toto Imaging and Reconstruction of Post-Implantation Mouse Development at the Single-Cell Level. Cell, Published Online: 11 October 2018, doi:10.1016/j.cell.2018.09.031.