Organ development is a process that operates at both local and global cellular levels. Locally, individual cells undergo proliferation and are responsible for all biomass production. Furthermore, the function of a cell over the course of development and in the adult animal is largely dictated by the state of its transcriptome and proteome. However, globally, all cells must collectively work together to build an organ of proper size, shape, and function. To this end, distinct types of cells must be present in the correct number and location within the organ. Furthermore, organ development must also be robust, capable of withstanding potential setbacks such as tissue damage and the emergence of aberrant cells. Linking these local and global scales of organ development likely intertwines signaling pathways with emergent properties of the developing organ such as physical forces, cellular heterogeneity, and local heterotypic signaling interactions.
The common fruit fly Drosophila melanogaster is a powerful workhorse in the field of developmental biology, and boasts a massive genetic toolkit for in vivo experimentation. Specifically for organ development, the Drosophila wing-imaginal disc present an ideal system for studying the genetic regulation and heterotypic interactions that govern organ growth and patterning. The wing disc is embryonically derived from approximately 30 cells, and by late larval development, forms a simple sac-like structure consisting of two epithelial layers called the disc proper and peripodial epithelium. Significant literature exists surrounding the disc proper, which develops into the adult dorsal thorax and wing blade. Underneath the proximal-most region of the disc proper are the adult flight muscle precursors (AMPs), the myoblasts that give rise to nearly all adult flight muscles within the dorsal thorax. Despite the variety of morphologically- and physiologically-distinct thoracic muscle fibers created by the AMPs, there exists little knowledge on how the AMPs achieve such diversity. Experiments performed in the 1980s and 1990s in which wing-disc AMPs were transplanted into ectopic locations revealed that these cells readily fused with myoblasts in their new environment, suggesting that the AMPs are a largely naïve population shaped by external cues. Since then, only two epithelial-to-myoblasts interactions have been characterized, involving Wingless and Notch signaling.
The work described in Chapter 2 of my thesis identifies additional epithelial-to-myoblast interactions using single-cell transcriptomics. Single-cell RNA sequencing (scRNAseq) enables sequencing of thousands of individual cells, providing an unprecedented opportunity to understand cellular heterogeneity and infer heterotypic interactions. Using this technique, my colleagues and I catalogued the cell-type composition of the disc epithelium and AMPs at two developmental time points, mid and late larval 3rd instar (L3) development, and built a multi-layered model of the wing disc. While we observed a consolidation of epithelial identities by mid-L3 larval development, AMP identities were largely established later in development. We examined our data for the expression of receptors and ligands that could be mediating heterotypic interactions between the disc proper and AMPs, and characterized two such pathways. The first, fibroblast growth factor (FGF) signaling, was required to localize the AMPs to their developmental niche along the wing disc. Cell-type specific expression of FGF ligands, Thisbe and Pyramus, restricted the AMPs to the proximal regions of the disc proper, and overexpressing either ligand was sufficient to induce AMP growth and proliferation in ectopic regions. Second, we found that Hedgehog ligand secreted from the posterior disc proper induced a cell-type identity within the posterior-most AMPs, defined by AMP-specific Hedgehog targets Neurotactin and midline. While perturbing Hedgehog transduction within the AMPs had no dramatic effect on AMP numbers during larval development, we observed severe anomalies in adult muscle formation, suggesting that this Hedgehog-conferred identity is needed for proper muscle development.
Organ development can easily go awry, and Drosophila has provided many insights into the process of tissue regeneration. While early imaginal disc regeneration studies dating back into the mid-1900s required transplantation experiments, the Drosophila community has developed a number of in vivo ablation systems for the wing disc. Wing disc regeneration is accomplished via the formation of a blastema, proliferative region characterized by heightened cellular plasticity that develops in response to damage. The regenerative process is complex, requiring cells surrounding the wound site to respond and undergo some level of cell-fate change to recuperate the lost tissue. The heterotypic interactions responsible for activating downstream genetic programs in both near and distant cells, and those that reestablish cellular repatterning, are not fully understood. Furthermore, many of the genes known to be crucial for regeneration are used in earlier developmental processes. Whether there exists a regulatory network designed solely for tissue regeneration, with a negligible role in normal development, would be fascinating, but such a pathway has not been described.
In Chapter 3 of my thesis, my colleagues and I applied scRNAseq to study wing disc regeneration. We observed unappreciated heterogeneity within the blastema at 24 hours into regeneration and characterized two distinct populations, Blastema1 and Blastema2. Both populations express genetic markers of regeneration, such as Wingless and Insulin-like Peptide 8 (Ilp8), but Blastema1 is an inner subset of Blastema2 and uniquely expressed a number of secreted molecules. We examined our scRNAseq for transcription factors that might regulate the regenerative process and identified the gene Ets21C as being specifically expressed within both blastema populations. Mutant analysis revealed that while Ets21C is largely dispensable for normal development (having no observable aberrant phenotype), it is absolutely crucial for proper regeneration, suggesting that it regulates a regeneration-specific gene regulatory network. Indeed, loss of Ets21C abolishes the expression of Blastema1 markers, decreases Ilp8 expression within the blastema, and induces premature pupariation of regenerating larvae. Additionally, we identified blastema-like cells that upregulated components of the Ets21C-dependent gene network within scRNAseq data collected from wing disc tumors, highlighting a mechanism in which tumors may co-opt regenerative processes.
