Skip Internal Navigation
Our goal is to understand how linear information encoded in genomic DNA functions to control cell fates during development. The Drosophila genome is about twenty times smaller than the human genome. Despite its smaller size, most developmental genes and at least half of the disease- and cancer-causing genes in man are conserved in Drosophila, making Drosophila a particularly important model system for the study of human development and disease. One of the important groups of conserved developmental genes are the homeotic genes. In Drosophila, the homeotic genes specify cell identities at both the embryonic and adult stages. The genes encode homeodomain-containing transcription factors that control cell fates by regulating the transcription of downstream target genes. The homeotic genes are expressed in precise spatial patterns that are crucial for the proper determination of cell fate. Both loss of expression and ectopic expression in the wrong tissues lead to changes in cell fate. The changes provide powerful assays for identifying the trans-acting factors that regulate the homeotic genes and the cis-acting sequences through which they act. The trans-acting factors are also conserved between Drosophila and human and have important functions, not only in development but also in stem-cell maintenance and cancer.
cis-acting sequences for transcriptional regulation of the Sex combs reduced homeotic gene
Figure 1. Chromosomal aberrations in the distal half of the Antennapedia complex
The transcription units are shown above the genomic DNA, while chromosomal aberrations are shown below (solid triangles indicate insertions of transposable elements and upward arrows indicate breakpoints of translocations and inversions). Chromosomal aberrations (shown in red) interfere with silencing in the adult second and third legs. The regions that include the proximal and distal MES are indicated by horizontal arrows.
Back to top
Assays in transgenes in Drosophila previously identified cis-acting transcriptional regulatory elements from homeotic genes. The assays identified tissue-specific enhancer elements as well as cis-regulatory elements that are required for the maintenance of activation or repression throughout development. While these transgenic assays have been important in defining the structure of the cis-regulatory elements and identifying trans-acting factors that bind to them, their functions within the context of the endogenous genes remain poorly understood. We used a large number of existing chromosomal aberrations in the Sex combs reduced homeotic gene to investigate the functions of the cis-acting elements within the endogenous gene. The chromosomal aberrations identified an imaginal leg enhancer about 35 kb upstream of the Sex combs reduced promoter. The enhancer is not only able to activate transcription of the Sex combs reduced promoter that is 35 kb distant but can also activate transcription of the Sex combs reduced promoter on the homologous chromosome. Although the imaginal leg enhancer can activate transcription in all three pairs of legs, it is normally silenced in the second and third pairs of legs. The silencing requires the Polycomb-group proteins. We are currently trying to identify the cis-regulatory DNA sequences in the Sex combs reduced gene that are required for transcriptional activation in the first leg and for Polycomb-group silencing in the second and third legs. Characterization of the chromosomal rearrangements shown in Figure 1 also revealed that two genetic elements (proximal and distal MES) about 70 kb apart in the Sex combs reduced gene must be in cis to maintain proper repression. When not physically linked, these elements interact with elements on the homologous chromosome and cause derepression of its wild-type Sex combs reduced gene. We tested DNA fragments from the Sex combs reduced gene in transgenic assays to identify endogenous cis-regulatory elements that could interact. From the regions that include the regulatory elements required for the maintenance of silencing, we identified two clusters of DNA fragments that promote pairing-sensitive silencing (an assay for interaction of cis-regulatory DNA fragments). To assay their role in transcriptional regulation, we are using targeted gene replacement to delete the clusters of elements from the endogenous gene. We are also using these pairing-sensitive silencing elements to screen for the trans-acting factors that are required for the clusters' function.
Figure 2. Homeotic phenotypes of new pharate-adult lethal mutants
Panel B shows a wild-type haltere on the left and transformations of anterior and posterior haltere to anterior and posterior wing in the middle and right, respectively. Panel C shows the first legs from a wild-type male on the left and three different mutants with reduced sex combs on the right. Panel D shows a mutant male with sex combs on both the first and second tarsal segments. Panel E shows a mutant male with sex combs on all three pairs of legs. Panel F shows, on the left, the abdominal segments from a wild-type male and, on the right, a mutant male with transformation of the fifth abdominal segment to a more anterior identity.
Trans-acting activators and repressors of homeotic genes
Figure 3. Genetic screen for new mutations that disrupt pairing-sensitive silencing
Flies homozgous for transposons carrying the mini-white reporter gene and a pairing-sensitive silencing element have white eyes. Clones of cells homozygous for newly-induced mutants are generated using the yeast site-specific recombinase (FLP recombinase) and its target site (FRT). The clones of mutant cells are able to express the mini-white reporter gene and are pigmented (shown in the eye on the lower right of the figure).
Figure 4. New mutations that disrupt pairing-sensitive silencing
Nine examples of new mutations that disrupt pairing-sensitive silencing are shown. The homozygous mutant cells are able to express the mini-white reporter gene in the transposon and are pigmented.
Back to top
The initial domains of homeotic gene repression are set by the segmentation proteins, which also divide the embryo into segments. Genetic studies have identified the trithorax group of genes that are required for expression or function (such as maintenance of transcriptional activation) of the homeotic genes. Maintenance of transcriptional repression requires the proteins encoded by the Polycomb-group genes. To identify new trithorax-group activators and Polycomb-group repressors, we are screening for new mutations that mimic the following phenotypes: loss of function or ectopic expression of the homeotic genes. We generated over 4,000 lethal mutants and, among those that die late in development, have identified two dozen mutants with homeotic phenotypes. Some of the homeotic phenotypes are shown in Figure 2. The mutants identify genes required for expression or function of the homeotic genes.
We are also using pairing-sensitive silencing elements from the Sex combs reduced gene to screen for recessive mutations that interfere with homeotic gene silencing. In flies homozygous for transgenes with pairing-sensitive silencing elements, we are making clones of cells in the eye that are homozygous for newly-induced mutations, using the yeast FLP/FRT site-specific recombination system (Figure 3). Silencing mutations are detected by the appearance of pigmented spots in the white-eyed flies. Several examples of the new mutations recovered on the second chromosome are shown in Figure 4. About 40% of the mutations recovered are in known Polycomb-group genes. The remaining mutations identify at least nine new genes required for silencing.
Structure and function of the Drosophila genome
Figure 5. Molecular map of the genomic region that includes Chd3
The approximately 640 kb of genomic DNA that includes the Chd3 gene is broken into three parts (A, B, and C) and is represented by the horizontal black arrows at the top of each part. The annotated transcription units are represented by colored thick horizontal arrows. The Chd3 transcription unit in Panel A is depicted in black, the essential transcription units in red, and the remaining transcription units in blue. The two regions that include the five essential genes (76BDg, 76BDh, 76BDd, 76BDj, and 76BDt) for which the transcription units have not been identified are indicated by red horizontal brackets above the candidate transcription units (Panel B). A 31.4 kb tandem duplication (distal copy and proximal copy) flanked by Doc transposable elements in the sequenced iso-1 strain (but not in other wild-type strains) is shown on the genomic DNA at the left of Panel C, with the Doc elements represented by inverted brown triangles.
Back to top
As part of a long-term project to understand the function and organization of the Drosophila genome, we have tried to identify all essential genes in two regions of the genome that span almost 1 megabases of DNA (shown in Figures 5 and 6). We identified 10 gene clusters that appear to have arisen by tandem duplication. These clusters include 34 of the 137 predicted genes. We identified 48 genes essential for zygotic viability, including two of the pairs of tandemly-duplicated genes. Surprisingly, when we deleted a region that spanned 55 kb and included 7 predicted genes we found no effects on either viability or fertility. This nonessential gene desert includes 48 DNA sequences between 12 and 33 base pairs each that are identical in 12 different Drosophila species. This strong conservation indicates that these sequences must have some function that is beneficial to flies living outside the laboratory.
We were also able to assess the progress of the Drosophila Gene Disruption Project. While the Gene Disruption Project has tagged about two-thirds of the annotated genes with transposon insertions, these insertions only disrupt the function of 45% of the genes. We were also able to estimate the frequency of base-pair changes in the DNA after EMS treatment, which is 5–8 times the frequency of mutations that actually affect gene function sufficiently to cause a mutant phenotype.
Figure 6. Molecular map of the genomic region deleted in Df(3L)th102
The approximately 320 kb of genomic DNA (from 3L: 15918k to 16240k, Release 5.23) is broken into three parts (A, B, and C), and is represented by the horizontal black arrows at the top of each part. The annotated transcription units are represented by colored thick horizontal arrows. The essential transcription units are red and orange. The clusters of transcription units encoding related proteins are brown (the cluster in A), grey (the cluster in B), purple (the cluster in B), dark green (the cluster in C), light green (the cluster in C), and orange (the essential genes CG32155 and CG32154 in C). All other transcription units are blue. The two regions that include the five essential genes (72CDc, 72De, 72Df, 72Dg, and 72Di) for which the transcription units have not been identified are indicated by red horizontal brackets below the candidate transcription units. The DNA missing in flies trans-heterozygous for the overlapping deletions Df(3L)Exel6128 and Df(3L)BSC559 is indicated by the horizontal black bar at the bottom of C.
Back to top