At Janelia Farm, Philipp Keller has been applying some uncommonly revealing imaging techniques to help answer one of the grand old questions of biology: How in the world does a single cell develop into the complicated arrangement of zillions of cells that make up a fruit fly, let alone a human being? Keller’s approach to answering this question is to watch the transformation happen in fruit flies and other model organisms.
The power of his favorite techniques—called simultaneous multiview light-sheet microscopy—depends on a thin sheet of laser light that beams, stepwise, into different planes of the living, developing, speck-sized fly embryo, which Keller nestles into a bed of agarose gel, where it carries on developing.
The sheet of light induces the cells in the illuminated plane to fluoresce while a set of two or four exquisitely sensitive CMOS cameras gather snapshots of every cell in the plane from several different angles. (CMOS is electronic engineers’ shorthand for complementary metal-oxide semiconductor, a workhorse technology for making integrated circuits, and refers to the type of chip that serves as the camera’s light sensor). By stepping the embryo from one edge to the other through the light beam, the system assembles a set of planar views. A computer then registers and fuses those images into a dynamic three-dimensional depiction of the embryo at any given time during its 21 or more hours of development, according to Keller’s paper published June 3, 2012, in Nature Methods.
Over the course of that time, the system takes an image set every 30 seconds or so, amounting ultimately to more than 1 million images and a total data set sized at 10 terabytes. That’s comparable to the estimated 11 terabytes worth of books in the Library of Congress (LOC). So far, Keller notes, his team has recorded hundreds of LOC’s worth of embryogenesis data in fruit flies, zebrafish, and mice. Their images and movies routinely elicit exclamations of “wow” or less-articulate utterances of amazement.
“You can look at everything at the same time; you can follow every cell in the embryo,” Keller says with excitement. “You don’t have to focus on this cell cluster or that one.” And by applying additional image analysis and computational tools to the data sets, he can highlight patterns of cell migration, waves of cell division, and other dynamic features of the marvelous one-cell-to-organism transformation.
-- Ivan Amato