The interaction of proteins encoded by Drosophila homeotic and segmentation genes with specific DNA sequences

Genomics of the HOX gene cluster

The Hox family of homeobox genes encode transcription factors that control different aspects of metazoan development. They appear clustered in the genomes of those animals in which their relative positions have been mapped. Although clustering is assumed to be a general property of Hox genes in all bilaterians, just a few species have been studied in sufficient detail to support this claim. Linear duplication of genes inside the cluster, as well as full-cluster duplications account for the actual complexity of HOX clusters in the different animal groups that have been studied (mainly vertebrates). Understanding how the Hox genes are regulated during development will depend, ultimately, on the generation of more powerful tools for cloning intact HOX clusters and for elucidating their cis-regulatory components. To clarify the roles of the Hox genes themselves, we will need to characterize in detail their downstream targets, and some progress in this direction is coming mainly from the recent use of arrayed libraries. Moreover, a comprehensive study of Hox target genes in tissues and organisms promises, in the long term, to give us a clear idea of the role that Hox genes play during development and how they have evolved over time.

Identification of novel functional regions important for the activity of HOXB7 in mammalian cells

Members of the HOX family of homeobox transcription factors play a role in pattern formation in diverse developmental systems. The clearly documented role of HOX genes in the proliferation and differentiation of primary hematopoietic cells and cell lines provides a convenient system to pursue a biochemical analysis of HOX gene function in mammalian cells. To explore the role of HOXB7 in myeloid hematopoiesis, a number of mutations and deletions in the gene were constructed that targeted sequences with known functions or in regions that had not been examined previously. The wild-type and mutant B7 constructs were introduced into the murine myelomonocytic cell line, 32D, and assayed for their effects on G-CSF-induced myeloid differentiation. Wild-type HOXB7 inhibited the differentiation of 32D cells, whereas mutations in the Pbx-binding pentapeptide motif or the DNA-binding homeodomain, as well as internal deletions of the N-terminal unique region, blocked this effect. Interestingly, mutations eliminating two target sites for casein kinase II, the glutamate-rich C terminus, or the first 14 amino acids of HOXB7, led to enhanced 32D differentiation. A model proposing a role for these regions of HOXB7 is presented.

Establishment and maintenance of parasegmental compartments

Embryos of higher metazoans are divided into repeating units early in development. In Drosophila, the earliest segmental units to form are the parasegments. Parasegments are initially defined by alternating stripes of expression of the fushi-tarazu and even-skipped genes. How fushi-tarazu and even-skipped define the parasegment boundaries, and how parasegments are lost when fushi-tarazu or even-skipped fail to function correctly, have never been fully or properly explained. Here we show that parasegment widths are defined early by the relative levels of fushi-tarazu and even-skipped at stripe junctions. Changing these levels results in alternating wide and narrow parasegments. When shifted by 30% or more, the enlarged parasegments remain enlarged and the reduced parasegments are lost. Loss of the reduced parasegments occurs in three steps; delamination of cells from the epithelial layer, apoptosis of the delaminated cells and finally apoptosis of inappropriate cells remaining at the surface. The establishment and maintenance of vertebrate metameres may be governed by similar processes and properties.

Mechanisms regulating target gene selection by the homeodomain-containing protein Fushi tarazu

Homeodomain proteins are DNA-binding transcription factors that control major developmental patterning events. Although DNA binding is mediated by the homeodomain, interactions with other transcription factors play an unusually important role in the selection and regulation of target genes. A major question in the field is whether these cofactor interactions select target genes by modulating DNA binding site specificity (selective binding model), transcriptional activity (activity regulation model) or both. A related issue is whether the number of target genes bound and regulated is a small or large percentage of genes in the genome. In this study, we have addressed these issues using a chimeric protein that contains the strong activation domain of the viral VP16 protein fused to the Drosophila homeodomain-containing protein Fushi tarazu (Ftz). We find that genes previously thought not to be direct targets of Ftz remain unaffected by FtzVP16. Addition of the VP16 activation domain to Ftz does, however, allow it to regulate previously identified target genes at times and in regions that Ftz alone cannot. It also changes Ftz into an activator of two genes that it normally represses. Taken together, the results suggest that Ftz binds and regulates a relatively limited number of target genes, and that cofactors affect target gene specificity primarily by controlling binding site selection. Activity regulation then fine-tunes the temporal and spatial domains of promoter responses, the magnitude of these responses, and whether they are positive or negative.

A target of the HoxB5 gene from the mouse nervous system

An epitope-specific antibody against the protein product of the murine HoxB5 gene was used to select an enriched library of Hox target sequences. Genomic DNA was purified by immunoaffinity chromatography, using glutaraldehyde-cross-linked chromatin from CNS of mouse embryos at gestational day 15. Screening was done by colony hybridization with TAAT-containing oligonucleotides, filter DNA-protein binding, and gel mobility shift assay. Nucleotide sequencing identified a 910 bp DNA fragment, containing a consensus Antennapedia-like binding site, and identical in 640 bps at 3' end of the clone to the promoter of the SPI3 gene, which encodes a serine protease inhibitor protein [Sun, J., Rose, J.B. and Bird, P., J. Biol. Chem., 270 (1995) 16089-16096]. In situ hybridization experiments were performed to see if a correlation could be found between the expression patterns the SPI3 and the HoxB5 genes. Using a 120 bp cDNA fragment as probe, SPI3 expression was detected mainly in the CNS of 15 day mouse embryos, a pattern which is similar to that of the HoxB5 gene at this stage [Hogan, B.L., Holland, P.W. and Lumsden, A., Cell Diff. Dev., 25 Suppl. (1988) 39-44; Sakach, M. and Safaei, R., Int. J. Dev. Neurosci., 14 (1996) 567-573]. In conclusion, data presented here suggest that the SPI3 gene is a candidate target of the HoxB5 gene in vertebrate embryos.

Chromosomal binding sites of Ultrabithorax homeotic proteins

Drosophila homeotic genes and their vertebrate cognates, the Hox genes, encode homeodomain proteins that are thought to control segment-specific morphogenesis by regulating subordinate target genes. Although expression of many genes is thought to be influenced by homeotic/Hox function, little is known about the genes they directly regulate in the developing embryo. One of the Drosophila homeotic genes is Ultrabithorax (Ubx) that specifies the identity of specific thoracic and abdominal metameres. Towards identifying genes directly regulated by Ubx we have mapped the binding sites of Ubx proteins (UBX) in polytene chromosomes. We found that the UBX isoforms Ia and IVa accumulate in about 100 discrete chromosomal sites. Most, if not all, the sites are the same for the two UBX isoforms. These sites are all euchromatic, include both bands and interbands and are reproducible from chromosome to chromosome. Some of these sites correspond to the locations of known genes that are good candidates, or are known to be, under direct Ubx control.

Two homeo domain proteins bind with similar specificity to a wide range of DNA sites in Drosophila embryos

We have used in vivo UV cross-linking to directly measure DNA binding by the homeo domain proteins even-skipped (eve) and fushi tarazu (ftz) in Drosophila embryos. Strikingly, these two proteins bind at uniformly high levels throughout the length of their genetically identified target genes and at lower, but significant, levels to genes that they are not expected to regulate. The data also suggest that these two proteins have very similar DNA-binding specificities in vivo. In contrast, a non-homeo domain transcription factor, zeste, is only detected on short DNA elements within a target promoter and not on other genes. These results are consistent with the in vitro properties of these various proteins, their respective concentrations in the nucleus, and with earlier predictions of how transcription factors bind DNA in vivo. We propose that these data favor the model that eve, ftz, and closely related homeo domain proteins act by directly regulating mostly the same target genes.

A universal target sequence is bound in vitro by diverse homeodomains

To determine the number of DNA binding proteins capable of binding a consensus Engrailed binding site, this consensus sequence was used to screen a library of Drosophila cDNA clones in a bacteriophage expression vector. We retrieved clones encoding 20 distinct DNA binding domains, 17 of which are homeodomains. Binding to a variety of oligonucleotides confirms the related sequence specificity of the retrieved binding domains. Nonetheless, the homeodomains have remarkably diverse amino acid sequences. We conclude that during the evolutionary divergence of homeodomains, the specificity of DNA binding has been much more highly conserved than the amino acid sequence.

Production of mouse Hox-2.1 protein in Escherichia coli: characterisation of in vitro binding to DNA

The developmentally regulated mouse Hox-2.1 gene encodes a homeodomain-containing (Hox) protein which is likely to function as a transcription factor. We expressed the DNA coding for full-length Hox-2.1 protein in a T7 promoter-containing vector in bacteria, which produced low levels of protein showing weak DNA-binding activity. Synthesis of a truncated polypeptide lacking all the sequence upstream from the homeodomain enabled us to produce greater amounts of protein and demonstrate its sequence-dependent DNA binding. The tetranucleotide ATTA is necessary for binding, but a single copy is not by itself sufficient. Flanking sequences are important; in particular a cytosine immediately 5' to the ATTA enhances binding.

Isolation and in vitro characterization of temperature-sensitive mutants of the bacteriophage f1 gene V protein

In vivo selections were used to isolate 43 temperature-sensitive gene V mutants of the bacteriophage f1 from a collection of mutants constructed by saturation mutagenesis of the gene. The sites of temperature-sensitive substitutions are found in both the beta-sheets and the turns of the protein, and some sites are exposed to the solvent while others are not. Thirteen of the variant proteins were purified and characterized to evaluate their free energy changes upon unfolding and their affinities for single-stranded DNA, and eight were tested for their tendencies to aggregate at 42 degrees C. Each of the three temperature-sensitive mutants at buried sites and six of ten at surface sites had free energy changes of unfolding substantially lower (less stabilizing) than the wild-type at 25 degrees C. A seventh mutant at a surface site had a substantially altered unfolding transition and its free energy of unfolding was not estimated. The affinities of the mutant proteins for single-stranded DNA varied considerably, but two mutants at a surface site, Lys69, had much weaker binding to single-stranded DNA than any of the other mutants, while two mutants at another surface site, Glu30, had the highest DNA-binding affinities. The wild-type gene V protein is stable at 42 degrees C, but six of the eight mutants tested aggregated within a few minutes and the remaining two aggregated within 30 minutes at this temperature. Overall, each of the temperature-sensitive proteins tested had a tendency to aggregate at 42 degrees C, and most also had either a low free energy of unfolding (at 25 degrees C), or weak DNA binding. We suggest that any of these properties can lead to a temperature-sensitive gene V phenotype.

Targets of homeotic gene control in Drosophila

The homeotic gene Ultrabithorax (Ubx) encodes homeodomain-containing transcription factors that determine segmental identity in Drosophila. Here, an immunopurification procedure is described that enriches for embryonic chromatin fragments containing binding sites for Ubx protein. In two cases these binding sites are located near embryonic transcription units regulated by the Ubx locus in vivo. Thus, these transcripts may correspond to Ubx target genes involved in elaborating segment-specific developmental pathways.

The specificities of Sex combs reduced and Antennapedia are defined by a distinct portion of each protein that includes the homeodomain

The sequence requirements for distinguishing the functional specificities of two homeodomain proteins, Antennapedia and Sex combs reduced, involved in the specification of segmental identities in Drosophila, have been determined. A series of deletions and hybrid proteins was generated and assayed for their function in vivo after heat shock-induced ectopic expression during development. A distinct portion of each protein, including the residues within and adjacent to both ends of the homeodomain, has been found to almost entirely determine its functional specificity as measured by diagnostic cuticular transformations of embryonic and adult head structures. The remaining sequences contribute to the potency with which the proteins act in different cells and are to a limited extent functionally transferable from one protein to the other.

Analysis of the ftz upstream element: germ layer-specific enhancers are independently autoregulated

The Drosophila fushi tarazu (ftz) upstream element is an enhancer-like element that is required for the correct expression of ftz in developing embryos and that directs transcription from a minimal promoter in a ftz-like seven-striped pattern. Using a deletion analysis, we have identified several independent cis-regulatory elements in the upstream element. A distal enhancer directs fusion gene expression in seven stripes primarily in the mesoderm. A more complex proximal enhancer contains a mesodermally active element and a second element with which it interacts to generate seven stripes in the ectoderm. Striped expression directed by each enhancer is ftz-dependent, and each contains binding sites for purified ftz homeo domain. We suggest that ftz protein acts in combination with germ layer-restricted transcription factors directly and positively to regulate the transcription of its own gene.