Segmental pattern formation during Drosophila development

We are interested in the molecular mechanisms that establish the body plans of multi-cellular animals.  In Drosophila, previous genetic screens have identified many of the genes that provide patterning information to the developing embryo.  Most of these genes encode transcription factors that are expressed in broad bands or striped patterns at specific positions along the anterior posterior axis.  Our specific focus is on gradients of transcription factors provided by maternal and gap genes.  Studies in our lab and others suggest that these gradients function by directly binding to specific sites in the regulatory elements (enhancers) of downstream genes.  This causes the activation or repression of transcription via concentration-dependent and combinatorial mechanisms.  Our goals are to better understand how the Drosophila body plan is formed, and to identify the general rules governing the on/off transcriptional decisions made by developmentally important genes.

The role of repression gradients in segmental patterning.  Previous genetic experiments suggests that the gap genes hunchback (hb), giant (gt), Kruppel (Kr), and knirps (kni) act as repressive gradients that set on/off boundaries of stripes of even-skipped (eve) expression.   To test this hypothesis further, we have developed a misexpression strategy that places an ectopic domain in the ventral region of the embryo.  This domain creates a ventral to dorsal gradient that intersects all eve stripes.  We find that misexpression of a single gap gene is sufficient for repression of specific stripes, and that stripes controlled by different regulatory elements (enhancers) that are differentially sensitive to the ectopic gradients.  Using a bio-informatics approach, we show that differential enhancer sensitivity is controlled in large part by the composition of DNA-binding sites in the individual enhancers.  We have also used computational approaches to predictably change enhancer sensitivity to repression, and these predictions have been confirmed by in vivo experiments with reporter genes.  Our data suggest that multiple expression boundaries can be established by differential enhancer sensitivity, and that embryonic complexity can be achieved by relatively simple gradient mechanisms.

Genome-wide analysis of the Bicoid (Bcd) regulatory network.  The maternal morphogen Bcd is expressed in an AP gradient, which controls various aspects of anterior body patterning.  Bcd is thought to differentially position target genes by binding directly to enhancers that contain clusters of Bcd-binding sites of different “strengths”.  Here we use a combination of Bcd-site cluster analysis and evolutionary conservation to predict novel enhancers near genes known to be expressed in anterior regions of the developing embryo.  We tested fourteen predicted enhancers by in vivo reporter gene assays; eleven show Bcd-dependent activation, which brings the total number of known Bcd target elements to twenty-one.  Some enhancers drive expression patterns that are restricted to the anterior-most part of the embryo, while others extend into middle and posterior regions.  However, we do not detect a strong correlation between AP position of target gene expression and the strength of Bcd-site clusters alone.  Rather, we find that combinatorial inputs with other activators including Hunchback (Hb) and Caudal (Cad) are critical for extending expression into middle and posterior regions.  Also, many patterns are restricted to anterior regions by a centrally localized gradient of the gap protein Kruppel (Kr).  We propose that combinatorial interactions are critical for positioning the myriad of cell fates that comprise the AP body plan.

Recombination-mediated transgenesis.  Analyzing and comparing multiple reporter gene constructs has always been plagued by genomic position effects.  To address this issue, we have developed an injection-based method for transferring multiple transgenes into the exact same genomic locus.  The concept is to create a landing site that contains a marker gene (y) flanked by lox sites.  Embryos containing the landing site are then injected with a shuttle vector containing the reporter construct and a second marker gene (w) (also flanked by lox sites), along with a helper plasmid that provides Cre activity.  The method is very reliable and relatively efficient (5-10% of fertile crosses give offspring with the replaced transgene).