Gradients of signaling proteins are essential for inducing tissue morphogenesis. However, mechanisms of gradient formation remain controversial. Here we characterized the distribution of fluorescently-tagged signaling proteins, FGF and FGFR, expressed at physiological levels from the genomic knock-in alleles in Drosophila. FGF produced in the larval wing imaginal-disc moves to the air-sac-primordium (ASP) through FGFR-containing cytonemes that extend from the ASP to contact the wing-disc source. The number of FGF-receiving cytonemes extended by ASP cells decreases gradually with increasing distance from the source, generating a recipient-specific FGF gradient. Acting as a morphogen in the ASP, FGF activates concentration-dependent gene expression, inducing pointed-P1 at higher and cut at lower levels. The transcription-factors Pointed-P1 and Cut antagonize each other and differentially regulate formation of FGFR-containing cytonemes, creating regions with higher-to-lower numbers of FGF-receiving cytonemes. These results reveal a robust mechanism where morphogens self-generate precise tissue-specific gradient contours through feedback regulation of cytoneme-mediated dispersion.

Fibroblast growth factors (FGF) are essential signaling proteins that regulate diverse cellular functions in developmental and metabolic processes. In Drosophila, the FGF homolog, branchless (bnl) is expressed in a dynamic and spatiotemporally restricted pattern to induce branching morphogenesis of the trachea, which expresses the Bnl-receptor, breathless (btl). Here we have developed a new strategy to determine bnl-expressing cells and study their interactions with the btl-expressing cells in the range of tissue patterning during Drosophila development. To enable targeted gene expression specifically in the bnl expressing cells, a new LexA based bnl enhancer trap line was generated using CRISPR/Cas9 based genome editing. Analyses of the spatiotemporal expression of the reporter in various embryonic stages, larval or adult tissues and in metabolic hypoxia, confirmed its target specificity and versatility. With this tool, new bnl expressing cells, their unique organization and functional interactions with the btl-expressing cells were uncovered in a larval tracheoblast niche in the leg imaginal discs, in larval photoreceptors of the developing retina, and in the embryonic central nervous system. The targeted expression system also facilitated live imaging of simultaneously labeled Bnl sources and tracheal cells, which revealed a unique morphogenetic movement of the embryonic bnl- source. Migration of bnl- expressing cells may create a dynamic spatiotemporal pattern of the signal source necessary for the directional growth of the tracheal branch. The genetic tool and the comprehensive profile of expression, organization, and activity of various types of bnl-expressing cells described in this study provided us with an important foundation for future research investigating the mechanisms underlying Bnl signaling in tissue morphogenesis.

Bnl controls tracheal development in Drosophila, but it is unclear how this fibroblast growth factor is prepared for tissue-specific dispersal. Sohr et al. find that Furin1 cleaves Bnl in the Golgi, which polarizes its sorting to the basal surface of the source cells and determines its range of cytoneme-mediated intercellular dispersion, signaling, and branching morphogenesis.

Binary transcription systems are powerful genetic tools widely used for visualizing and manipulating cell fate and gene expression in specific groups of cells or tissues in model organisms. These systems contain two components as separate transgenic lines. A driver line expresses a transcriptional activator under the control of tissue-specific promoters/enhancers, and a reporter/effector line harbors a target gene placed downstream to the binding site of the transcription activator. Animals harboring both components induce tissue-specific transactivation of a target gene expression. Precise spatiotemporal expression of the gene in targeted tissues is critical for unbiased interpretation of cell/gene activity. Therefore, developing a method for generating exclusive cell/tissue-specific driver lines is essential. Here we present a method to generate highly tissue-specific targeted expression system by employing a "Clustered Regularly Interspaced Short Palindromic Repeat/CRISPRassociated" (CRISPR/Cas)-based genome editing technique. In this method, the endonuclease Cas9 is targeted by two chimeric guide RNAs (gRNA) to specific sites in the first coding exon of a gene in the Drosophila genome to create double-strand breaks (DSB). Subsequently, using an exogenous donor plasmid containing the transactivator sequence, the cell-autonomous repair machinery enables homology-directed repair (HDR) of the DSB, resulting in precise deletion and replacement of the exon with the transactivator sequence. The knocked-in transactivator is expressed exclusively in cells where the cis-regulatory elements of the replaced gene are functional. The detailed step-by-step protocol presented here for generating a binary transcriptional driver expressed in Drosophila fgf/branchless-producing epithelial/neuronal cells can be adopted for any gene-or tissue-specific expression.

During development, a handful of signals sculpt diverse tissue architectures. How the same signal produces different tissue/context-specific information and outcomes is poorly understood. We explored the basis that programs tissue-specific FGF dispersion and interpretation by cytoneme-mediated contact-dependent communication. Although a Drosophila FGF was thought to be freely secreted, we discovered that it is glypiated and GPI-anchored on the source cell surface, which inhibits non-specific secretion but facilitates tissue-specific cytoneme contact formation and contact-dependent release. For long-distance signaling, source and recipient cells extend FGF-containing and FGFR-containing cytonemes that contact and recognize each other by CAM-like receptor-ligand binding. FGF-FGFR binding reciprocally induces forward and reverse signaling in recipient and source cells, responses of which polarize their cytonemes toward each other to mutually self-sustain contacts. FGFR-bound FGF is then unanchored to hand-deliver FGF to receiving cytonemes, and cytoneme contacts dissociate. Thus, while cytonemes spatiotemporally control FGF dispersion/interpretation, FGF self-regulates its tissue-specific signaling by controlling cytonemes.