The development of the heart and body wall muscles of the Drosophila embryo provides a powerful model for investigating general principles of gene regulation and organ formation, with wide applicability of the findings to other biological systems, including the comparable processes that occur in mammals. We are taking an interdisciplinary approach to this problem by combining genetics, genomics, biochemical, molecular, cellular and computational methods that enable us to obtain a systems-level understanding of organogenesis.? ? Both the heart and somatic muscles are mesodermal derivatives that arise from progenitor cells that undergo progressive determination under the influence of a well-orchestrated set of intrinsic and extrinsic regulatory influences. The former include several tissue-restricted transcription factors that reflect the early developmental histories of the cells in which they are expressed, while the latter include inputs from multiple intercelluar signals. These signals are mediated by Wnt, BMP, receptor tyrosine kinase (RTK)/Ras and Notch pathways acting in various cell type-specific combinations. We have found that Ras and Notch activities are dynamically modulated by reciprocal cross-talk and feedback mechanisms. Furthermore, by studying one particular enhancer that is associated with a muscle- and heart progenitor identity gene, we established that the Wnt, BMP and Ras pathways converge at the transcriptional level. These findings support a model in which generic signals yield mesoderm-specific outputs by functioning together with the aforementioned intrinsic transcription factors to regulate the expression of common target gene enhancers. To test this hypothesis on a larger scale, we have turned to a more comprehensive analysis of the transcriptional regulatory networks that govern muscle and heart development.? ? In these studies, we first define the groups of genes that are co-expressed in particular subsets of mesodermal cells, and then we determine the molecular mechanisms by which these genes are co-regulated. Microarray-based gene expression profiling of RNA isolated from subpopulations of wild-type embryonic cells purified by flow cytometry led to the initial characterization of several hundred genes having enriched expression in the mesoderm. These findings were then validated by in situ hybridization. Application of this approach to a set of informative genetic perturbations of heart and muscle development, followed by statistical meta-analysis of the pooled microarray results and further in situ hybridization experiments, enabled us to identify large numbers of genes expressed in subsets of mesodermal cells that contribute to both organ types.? ? Using the structure of the abovementioned progenitor identity gene enhancer as a predictive model, we and our collaborators have developed several strategies for computationally searching the entire Drosophila genome for candidate sequences with related regulatory functions. We have validated several predicted enhancers, and have used other computational methods to identify shared functional motifs that bind additional transcription factors. More recently, we have refined our enhancer prediction strategy by incorporating comparative genomics, gene expression profiling, and novel transcription factors for which binding specificities are determined using protein binding microarrays. We also have applied computational methods to systematically examine the relative contributions of each transcription factor to and its specificity for different combinatorial models. Using this integrated approach, we have identified and validated a variant somatic muscle regulatory model. Current work is focused on uncovering additional sequence features that confer cell subtype specificity to gene expression patterns. For example, we discovered that homeodomain selector proteins regulate the activities of two muscle founder cell enhancers, leading to additional experimental and computational experiments that support the involvement of this class of selectors in muscle gene expression on a genome-wide scale.? ? Having identified large numbers of genes with muscle and cardiac expression, we are systematically studying the developmental functions of these genes using a double-stranded RNA interference (RNAi) approach in intact embryos. This screening strategy has led us to discover previously uncharacterized genes that are involved in myoblast fusion, myotube projection, myofiber attachment, muscle contraction, determination of cardiac cell number, and regulation of heart tube closure. Genes of particular phenotypic interest are then selected for more detailed analysis. Two genes whose functions we have investigated in the past year serve to illustrate our efforts in this area.? ? One gene that is expressed in both founder and fusion-competent myoblasts yielded a defect in myoblast fusion in RNAi and genetic loss-of-function experiments. This gene, named singles bar (sing) after its muscle phenotype, encodes a MARVEL domain protein. Additional work, undertaken in collaboration with colleagues at the University of Cambridge, revealed that (1) myoblast specification, attraction, migration, recognition and adhesion occur normally in the absence of sing, (2) sing is dispensable for the first round of myoblast fusion, but is essential for the second and subsequent rounds, and (3) myoblasts are blocked at the pre-fusion complex stage in sing mutant embryos. This latter phenotypecharacterized by a symmetrical accumulation of electron-dense vesicles near the apposed plasma membranes of adhering myoblastsis unique among known muscle fusion genes. Given the involvement of other MARVEL domain proteins in cell biological processes in which, like myoblast fusion, close membrane apposition occurs, our results suggest that sing could be directly involved in the steps leading to vesicle fusion with the myoblast plasma membrane.? ? A second gene, which we named perdido (perd, lost in Spanish), is expressed in a subset of muscle founder cells and is required for the formation of myotendinous junctions. Loss of Perd function causes abnormal myotube projections and subsequent muscle detachment from the ectodermal tendon cell. Perd encodes a transmembrane protein containing extracellular laminin domains and an intracellular PDZ domain-binding sequence, which interacts biochemically with one of the PDZ domains of Glutamate Receptor Interacting Protein (Grip), another factor that is involved in the formation of muscle projections. Through this mechanism, Perd is required for the normal subcellular localization of Grip to the plasma membrane at muscle attachment sites. Using a newly developed whole-embryo RNAi assay, we also demonstrated that muscle-specific Perd interacts genetically, not only with Grip, but also with the aPS1 integrin subunit that is expressed specifically on tendon cells. Additional experiments uncovered a previously unrecognized role for this integrin in the formation of myotube projections. Collectively, these results led us to formulate a model in which Perd regulates the proper extension of myofilopodial processes and subsequent differentiation of the myotendinous junction by priming formation of a protein complex through its intracellular interaction with Grip in myotubes, and its extracellular interaction with the laminin-binding aPS1-bPS integrin on tendon cell targets.? ? In summary, our investigations are providing an increased understanding of the genetic regulatory networks that orchestrate specific aspects of embryogenesis. Moreover, these studies serve as an instructive experimental paradigm for related investigations in other developmental systems.
| Estrada, Beatriz; Gisselbrecht, Stephen S; Michelson, Alan M (2007) The transmembrane protein Perdido interacts with Grip and integrins to mediate myotube projection and attachment in the Drosophila embryo. Development 134:4469-78 |
| Estrada, Beatriz; Maeland, Anne D; Gisselbrecht, Stephen S et al. (2007) The MARVEL domain protein, Singles Bar, is required for progression past the pre-fusion complex stage of myoblast fusion. Dev Biol 307:328-39 |
| Philippakis, Anthony A; Busser, Brian W; Gisselbrecht, Stephen S et al. (2006) Expression-guided in silico evaluation of candidate cis regulatory codes for Drosophila muscle founder cells. PLoS Comput Biol 2:e53 |
