When a female fruit fly makes an egg, she packs it full of everything a developing embryo needs for the earliest stages of its life: a yolk to feed it, and proteins and RNAs to drive its vital cellular processes. Fueled by these maternally deposited molecules, development begins with a series of rapid cell divisions during which there is little, if any, activation of the embryo’s own genome.
However, within a few hours, after around 14 cell divisions, the fertilized egg breaks free of its mother’s influence. The embryo’s genome kicks in, and the process of development begins in earnest.
“There is nothing that we know of quite like Zelda in terms of the scale of its activity and its particular role.”
Michael B. Eisen
Although this transition – known as the maternal-to-zygotic transition (MZT) – is a critical developmental milestone, relatively little is known about the molecular events that govern the handoff from mother's influence to an animal's own genome.
Now, in a study published October 20, 2011, in the journal PLoS Genetics, a team led by Howard Hughes Medical Institute investigator Michael B. Eisen of the University of California, Berkeley has sketched out the broad influence of a protein that works as a genomic "on” switch at the MZT.
The first hint that such a protein exists came almost a decade ago from John ten Bosch, then a graduate student studying a small set of Drosophila genes activated prior to the MZT in Tom Cline’s lab at UC Berkeley. He noticed that all of these genes shared a specific nucleotide sequence – CAGGTAG – in their promoters, and that when he removed the sequence, the genes did not turn on.
Intrigued by this result, a group led by Chris Rushlow at NYU (who have an accompanying paper in the October 20th issue of PLoS Genetics) identified a protein that binds to CAGGTAG, and showed that removal of this protein from eggs affects the activation of hundreds of genes and massively disrupts the earliest stages of development. They named the protein Zelda.
Around the same time, Eisen’s group stumbled upon CAGGTAG in their own research. They study a class of sequences known as enhancers that control the spatial patterns of gene expression that choreograph the differentiation of distinct cells and tissues in the developing embryo. Having used genomic techniques to identify thousands of enhancers active at the MZT, they noticed that virtually all of these regions also contained CAGGTAG.
Inspired by Cline and Rushlow’s work, Eisen, Berkeley colleague Michael Botchan, and members of their labs sought to demonstrate that Zelda was also acting at enhancers, and, if it were, to begin figuring out how it works. Postdoctoral fellow Melissa Harrison and senior scientist Xiao-Yong Li set out to experimentally determine where along the genome Zelda actually binds.
They first had to ensure they were looking at the right developmental stages. So Harrison and Li spent weeks staring down a microscope, hand sorting tiny Drosophila embryos to get pure samples from before, during and after the MZT. Using an antibody raised against the protein, they isolated fragments of DNA bound by Zelda at each stage, and precisely mapped them using high-throughput sequencing, joining forces with computational biology fellow Tommy Kaplan to analyze the results.
What they found was striking. After only eight cycles of cell division – an hour before the MZT -- they observed Zelda bound to thousands of sites across the genome. As expected, these regions included the promoters of the early-activated genes already shown by Cline and Rushlow to be affected by Zelda. But they also found Zelda bound to the promoters of thousands of genes that do not turn on until the MZT. What’s more, Zelda was sitting on the thousands of enhancers whose high concentration of CAGGTAG sites led them to investigate its binding in the first place.
“We were surprised to see Zelda bound early to so many regions that are not active until much later,” said Eisen. His team’s analysis further demonstrated that early Zelda binding predicted which promoters and enhancers would be functional at the MZT.
“Zelda appears to be acting as a kind of gatekeeper,” Eisen said. “While most of the rest of the genome remains dormant at the MZT, places where Zelda binds are poised for activation.”
Eisen’s analysis of other insect genomes reveals that they all have their own Zelda protein, highlighting its importance to insect development. But these sequences offer few clues to how Zelda might be working.
“It’s a new class of protein. There is nothing that we know of quite like Zelda in terms of the scale of its activity and its particular role,” Eisen added. “But all animal embryos also undergo an MZT, and we are excited by the possibility that humans have their own, as of yet undiscovered, Zelda.”