Where are Enhancers? An Analysis of the Zebrafish Enhancer Trap Screen

Colin Orr, Dr. Stephen Johnson, Genetics Department, Washington University, St Louis, MO

The zebrafish is a relatively new genetic model that scientist have just begun to explore. The transparent nature of the zebrafish makes it possible to manipulate their genomes with sequences that code for fluorescent proteins.

My experiment exploits the transparent nature of zebrafish. We introduced a sequence that codes for the green fluorescent protein (GFP) via an enhancer trap. The fish are then bred and sorted based on a specific GFP expression pattern. Those fish that exhibited the expression pattern were classified as expressing fish. The fish that did not have the GFP expression pattern of interest were called non-expressing fish. The GFP fish were bred and the sorting process continued until we reached a generation where we could safely assume that we had isolated a single enhancer trap responsible for the GFP expression that we had observed. The fish that had the specific expression pattern of GFP had the section of their genome that contained the enhancer analyzed; the GFP negative fish were also analyzed in this manner. Based on the analysis of both lines of fish we decided to further analyze the fish to determine what conditions must be met in order for a given fish to express GFP.

We chose four factors to study; these included: the orientation of the GFP molecule relative to the surrounding genes, associated of the GFP with a gene, distance of the insertion site from the promoter, and the frequency of expression of the nearest gene. To analyze these criteria I used the databases of Pre Ensembl and NCBI. For each insertion site I would use the Pre Ensemble Blastview and receive a Contigview which would show the relative position and orientation of the insertion site, surrounding genes from various organism and gene models. I used NCBI as a resource to determine the accuracy of Pre Ensembl’s gene model. I would blast the sequence of the possible zebrafish gene model against the sequence of a known gene model, with the intent of receiving a graph that was linear. Although this did not guarantee that the gene model was correct, it did provide evidence that the gene model was plausible. Another feature of NCBI vital to the experiment included the EST blast program, which I used to determine the frequency of expression of the gene. Other programs of NCBI that were useful included the human Refseq Blast and the BlastX program, both of which provided more evidence that the gene model was feasible.

In conclusion, it appears that the mechanisms that govern transcription and expression are more complex than the four factors that we chose to study. The study of these factors did not give any clear indication of the mechanisms responsible for the observed phenomenon.