HomeResearchRegulatory Sequence Function and Evolution

Our Scientists

Regulatory Sequence Function and Evolution

Research Summary

Michael Eisen is interested in understanding how the spatial and temporal patterns of gene expression that choreograph development are established in the early Drosophila embryo and in characterizing the ways that microorganisms manipulate the behavior of animals, including humans.

Early Activation of the Embryonic Genome inDrosophila
My lab has a long-standing interest in understanding the enhancer sequences that control spatial and temporal patterns of gene expression in early Drosophila development. These sequences play a major role in pattern formation—establishing segments along the anterior-posterior axis and tissue layers along the dorsal-ventral axis—but their precise mechanism of activity has remained elusive.

For more than a decade, we have been using a combination of experimental and computational techniques to catalog enhancers, document their activities, characterize their interactions with the DNA binding transcription factors that link them to the molecular machinery of transcription, and study their evolution. Several years ago, we discovered that the early-acting enhancers we identified were massively enriched with a single DNA motif, CAGGTAG, unrelated to any known transcription factor but previously implicated in early transcriptional activation in the fly sex determination pathway. Subsequent experiments and computational modeling in the lab suggested that this motif plays a major role in determining which parts of theDrosophila genome have enhancer activity.

Figure 1: Individual embryos used for a sex-specific RNA time course...

After a colleague identified a protein, Zelda, that binds CAGGTAG, we used a combination of careful embryo sorting, chromatin immunoprecipitation, and high-throughput sequencing to characterize Zelda binding at different early embryonic stages. Strikingly, within 90 minutes of fertilization, before widespread transcriptional activation, Zelda is bound to essentially all the enhancers and promoters that will become active later in development.

These and other data from the lab suggest that Zelda is a new class of protein—a master regulator of genome activation—and that it acts by altering how the chromatin state is established early in development, switching regions to which it binds from a default inactive destiny to an active one. Precisely how it does this, however, remains a mystery and is a major focus of ongoing studies in the lab.

Dosage Compensation in Early Development
When Drosophila melanogaster embryos initiate zygotic transcription, the dose-sensitive expression of specialized genes on the X chromosome triggers a sex-determination cascade that, among other things, compensates for differences in sex chromosome dose by hypertranscribing the single X chromosome in males. However, a delay of approximately 90 minutes occurs between the onset of zygotic transcription and the establishment of canonical dosage compensation near the end of mitotic cycle 14. During this time, zygotic transcription drives segmentation, cellularization, and other important developmental events.

Because many of the genes involved in these processes are on the X chromosome, we wondered whether they are transcribed at higher levels in females and whether this might lead to sex-specific early embryonic patterning. To investigate this possibility, we developed methods to precisely stage, sex, and characterize the transcriptomes of individual embryos. We measured genome-wide mRNA abundance in male and female embryos at eight timepoints, spanning mitotic cycle 10 through late cycle 14 (see Figure) using polymorphisms between parental lines to distinguish maternal and zygotic transcription. We found limited sex-specific zygotic transcription, with a weak tendency for genes on the X chromosome to be expressed at higher levels in females. However, transcripts derived from the single X chromosome in males were more abundant that those derived from either X chromosome in females, demonstrating that there is widespread dosage compensation before the activation of the canonicalDrosophila dosage compensation system, which results in hypertranscription of the single male X.

For a variety of reasons, we believe that SXL, the master regulator of sex in D. melanogaster, is likely to directly downregulate transcripts in females. We are currently investigating this hypothesis by, among other things, looking at gene expression in SXL-deficient females and trying to isolate the RNAs bound in vivo by SXL.

Microbes and Animal Behavior
Every animal lives in an environment filled with microorganisms, and it is becoming increasingly clear that the bacteria, fungi, and other microbes that live in, on, and around them play an important role in vital processes such as feeding, digestion, and immunity. We have long been interested in the possibility that microbes also actively influence animal behavior. Just as animals depend on microbes, many microbes depend on animals for homes, food, protection. and places to mate. In addition, the things an animal does—what it eats, where it travels, who it interacts with—can have a dramatic effect on a microbe's fitness. Thus, it seems almost impossible that microbes have not evolved ways to influence the behavioral choices animals make. Indeed, many systems have now been described in which microbes clearly alter the behavior of animals that they live in or on. However, the detailed molecular basis for this influence has not been worked out in any of these cases, something my lab hopes to address by working on several microbe-animal interactions.

Sacchromyces and DrosophilaAlthough studied separately in the lab, the fly D. melanogaster and the yeast Sacchromyces cerevisiae rely on each other for their survival in nature. D. melanogaster larvae feed primarily on S. cerevisiae growing on rotting fruit, and have adapted to live in the high ethanol environment they produce. S. cerevisiae, in turn, are dependent on adult D. melanogaster for transport to fresh substrates. We are studying the role that the myriad volatile compounds produced by S. cerevisiae during fermentation (the compounds that impart distinct tastes and aromas to fermented beverages) play in mediating interactions between these species. Our governing hypotheses are that yeast have evolved to produce compounds that manipulate fly behavior in ways that benefit them, and that flies have evolved to "read" the aroma of yeast to glean information about the nutritional suitability of potential oviposition sites.

Drosophila gut microbes. There is increasing evidence that the gut microflora of humans influence our behavior, and several recent studies have hinted at similar effects in Drosophila. But, in neither species has the molecular basis for these effects been elucidated. We are using the relatively simple gut microflora of D. melanogaster to look for small molecules produced by gut microbes and to identify potential targets of these molecules in the D. melanogaster nervous system, using a variety of different techniques, including isotopic labeling, dissection coupled with GC and LC-MS, and manipulations of the microbial composition of the gut.

Toxoplasma and mouse behavior. The unicellular parasiteToxoplasma gondii has a multihost lifecycle. It infects most warm-blooded animals (mammals and birds) but can reproduce sexually exclusively in the digestive systems of cats. Several years ago, Robert Sapolsky's lab demonstrated that mice infected with T. gondii lose their innate fear of cats, presumably because this effect increases the chances that the mouse will be eaten by a cat, allowing the parasite to mate. We have replicated this phenomenon many times and are working to determine how the parasite accomplishes this striking change in host behavior.

The lab also receives support from the National Human Genome Research Institute and the National Science Foundation.

As of December 12, 2012

Scientist Profile

University of California, Berkeley
Computational Biology, Evolutionary Biology