11/12/2002 08:00:00

Cellular choreography, not molecular prepattern, creates repeated segments of vertebrate embryo

In a study that combines state-of-the-art biological imaging with gene expression analysis, scientists at the California Institute of Technology have uncovered a fundamental insight into the way embryonic cells and tissue move about to form key structures along the vertebrate axis. The study, which could lead to a better understanding of human development, takes advantage of the accessibility of chick embryos to embryonic manipulation.

The study by Caltech biologists Scott Fraser and Paul Kulesa, appearing in the November 1 issue of the journal Science, centers on segments known as somites, which form along either side of the future spinal cord of an embryo. Somites give rise to mature structures such as ribs, individual vertebrae, and even skin. The key role of somite segmentation in the patterning of the nervous system and the vertebral column has been long known. But the question of precisely how an individual somite buds off from a block of tissue in a pattern that is repeated all along the animal's torso, from head to tail, is poorly understood.

"Developmental biologists have had a difficult time getting a handle on how cell movements and gene expression patterns are coordinated to form complex structures, in this case the segmented units called somites," says Kulesa, a postdoctoral scholar in Fraser's lab and lead author of the paper. "The problems have been due to limitations in obtaining cellular resolution of tissue deep within living vertebrate embryos and difficulty in coordinating the cell movements and tissue shaping in living tissue with gene expression patterns typically obtained at one time point from fixed, non-living tissue."

The new insight of the paper is that the factors that determine the embryo's ultimate form as well as the eventual position of its cells involve a complicated set of motions of the cells themselves. Previous models of embryonic patterning had suggested that there was a molecular prepattern that subdivided the tissues, somewhat like a "paint-by-numbers" piece of art. The study thus shows the action of a more complex coordination between physical forces within the tissue and gene expression patterns that determine where an embryonic cell will go and what type of structure it will help form.

Kulesa and Fraser's study is made possible with a new culture technique combined with confocal time-lapse microscopy, an advanced form of imaging that allows the tissue of a living, developing embryo to be studied in intricate detail at the cellular level. Time-lapse imaging involves, first, labeling the tissue so that it will fluoresce when exposed to laser light, then passing a laser through the tissue, then reconstructing the fluorescent patterns of individual cells to form a three-dimensional microscopic image. The laser scans over the tissue of the developing embryo every minute or so, which allows the researchers to gather the hundreds of images taken during a several-hour run into a time-lapse video.

Using fertilized eggs, the researchers placed an embryo into a specially designed chamber to allow for high-resolution time-lapse imaging, and afterwards performed gene expression analyses on the same embryo. Thus, they were both videotaping cell movements for 6-to-12 hours as well as analyzing the expression of several genes, including EphA4 and c-Meso1, both thought to play a role in determining future somite boundary sites.

The results showed that the straight-line patterns of gene expression, which were thought to correlate with a simple, periodic slicing of the tissue into blocks, did not predict the complex cell movements. Time-lapse imaging showed that a ball-and-socket separation of tissue takes places in a series of six repeatable steps.

"It turns out that a somite pulls apart from the block of tissue, and cells move in anterior and posterior directions near the forming somite boundary," Kulesa says. "This is contrary to many models of somite segmentation which assume that gene expression boundaries that correlate with presumptive somite boundaries allocate cells into a particular block with very little cell movement.

"This study tells us that we have to be careful about assuming that gene expression patterns strictly determine a cell's fate and position."

Kulesa says the next step is to do the work in mouse embryos, which pose considerably more difficult challenges for developmental imaging, but have a tremendous advantage over chick-embryo imaging in attempting to isolate the role of key genes through gene manipulation.



Written by Robert Tindol