The lungs and vascular system are tubular networks that transport oxygen and blood throughout the body. The walls of the tubes are sheets of cells (an epithelium) or in some cases individual cells rolled up into a tube. Each organ contains thousands or millions of tubes, and the proper size and pattern of tubes and connections between them are crucial for efficient flow through the networks. Our goal is to identify the genes and molecular pathways that control embryonic development of such complex structures. We want to answer three basic questions: (1) What specifies the complex pattern of branching—where each branch sprouts, the direction it grows, and when it sprouts again to form the next generation of branches—and how is this patterning information encoded in the genome? (2) How does an epithelium migrate and assemble into tubes of the appropriate size and shape? (3) How does oxygen influence the process? Our work focuses on the Drosophila melanogaster respiratory system and the mouse lung because the excellent genetics and molecular biology in these model organisms provide powerful tools to address these questions in molecular detail. Ultimately we hope to understand how the developmental program goes awry in lung cancer and other lung diseases, and how we can restart the program in diseased lungs to generate healthy new tissue.
Genetic Dissection of Airway Development in Drosophila
During development of most branched organs, a small unbranched epithelial tube or sac initially forms and then new branches successively sprout from it to form a tree-like structure of interconnected tubes. Despite the similar overall appearance of the branches, our analysis of Drosophila airway development showed that each successive generation of branches uses a different set of genes and a different cellular mechanism of tube formation.
The Drosophila airways (the tracheal system) arise from 20 unbranched sacs of ~80 cells. The six main (primary) branches begin to form when one or two cells at six positions in each sac migrate out in specific directions. A small number of cells follow the lead cells, organizing into tubes as they migrate. Several hours later, secondary branches sprout from the tips of the primary branches. These are formed by individual tracheal cells that roll up into unicellular tubes. Subsequently, secondary branches ramify into dozens of terminal branches, which are long cytoplasmic extensions that form small-bore tubes that directly contact the tissues. In this way, each of the 20 sacs generates a small network of tubes. During the process, specific branches in each sac find and fuse to branches in the neighboring sacs to form an interconnected network of more than 5,000 tubes that transports oxygen throughout the body.
Our initial screens for mutants with defective airways identified more than 50 genes that control various aspects of airway development. Different sets of genes are required for each of the three levels of branching. The three sets are organized into a regulatory hierarchy, with genes required for one level of branching also required to trigger expression of genes that control the next level of branching. Thus, although each level of branching is genetically distinct, their development is coupled by a gene-regulatory hierarchy that ensures that branching occurs in a defined sequence with bigger branches forming before the next generation of smaller branches. A fourth set of genes controls the fusions of tracheal tubes. Yet another set of genes is necessary for proper lumen formation and to maintain tube size and shape.
We are conducting screens that so far have identified several hundred new tracheal mutants. We are also using DNA microarrays to monitor expression of all genes in the developing respiratory system. Our goal is to discover all the critical genes and to elucidate the full genetic program for respiratory development in Drosophila, and use this information to recreate a tracheal system in vitro.
An FGF Pathway Patterns Airway Branching
To understand how the identified genes control airway development, we are characterizing their protein products. This has begun to reveal at a molecular level how the pattern of tracheal branches is specified. The branchless gene encodes a homolog of a major family of vertebrate signaling molecules called fibroblast growth factors (FGFs). The gene is expressed in a remarkably complex and dynamic pattern—near the position where each new branch is sprouting. The secreted branchless FGF functions as a chemoattractant that guides the migrating tracheal cells to their correct destinations. It does so by activating a transmembrane receptor tyrosine kinase called Breathless, which is expressed on the developing tracheal cells.
Some of the other characterized gene products regulate production of the FGF signal. Others function in the receiving cells to transduce the signal from the receptor to the cellular machinery responsible for cell migration and tube formation. Another gene we discovered, called sprouty, is a negative regulator of the pathway. It restricts the range of the FGF signal; in its absence too many tracheal cells receive the signal and too many branches sprout. The biochemical studies of the FGF pathway and its regulation now under way should reveal how binding of the ligand to its receptor leads to formation of new tubes at the appropriate positions in the developing airway.
Oxygen Regulation of Airway Branching
Although the major branches that form in the embryo are stereotyped, the fine terminal branches that sprout later in development are variable and regulated by oxygen need of the target tissues. We discovered that branchless also plays the critical role in this physiological control of branching. Larval cells respond to oxygen starvation by turning on expression of branchless. The secreted FGF signals nearby tracheal cells to grow toward the signaling cells and supply them with more oxygen. The process is dynamic: other cells become hypoxic and the process repeats, generating a complex pattern of branches that matches the oxygen needs of the tissue.
Other genes required specifically for terminal branching encode components of the oxygen-sensing pathway used by oxygen-starved cells to recognize and initiate the response to the crisis. One encodes a transcription factor that becomes active only under low-oxygen conditions. We are examining how low-oxygen tension activates the transcription factor and how this leads to increased expression of branchless. We hope to identify the oxygen sensor that monitors oxygen tension and activates the transcription factor when oxygen is low, and all the target genes the transcription factor regulates. (A grant from the National Institutes of Health provides support for this project.)
Making and Shaping Tubes
The molecular mechanisms that create tubes from cells are not well understood for any organ. We have begun to elucidate the mechanisms that make and shape terminal branch tubes. These tiny tubes (less than 1 micron in diameter) form in a surprising way: cytoplasmic extensions from tracheal cells develop a long membrane-bound lumen that runs the length of each extension, forming a channel through which oxygen flows. Mutations in tube morphogenesis genes block lumen formation or alter the shape or number of lumens that form. Molecular characterization of the first tube morphogenesis genes suggests that lumens are created by formation and fusion of cytoplasmic vesicles that are aligned or stabilized in the cell through associations with the actin cytoskeleton.
Genetic Dissection of Lung Development
We are extending our studies of respiratory system development to the mammalian lung. Lung development is considerably more complex than airway development in Drosophila, as there are more than 15 million airway branches in the human lung, plus a similar number of blood vessels. We have begun by mapping the pattern of airway and blood vessel sprouting in the developing mouse lung. Some of the same mechanisms and types of molecules that are used to pattern the Drosophila airways are also used in the mammalian lung, including FGFs, FGF receptors, and Sprouty proteins. We are investigating the functions of these genes as well as homologs of other Drosophila tracheal genes, and we have also begun genome-wide searches for lung development genes in mouse. We are also developing methods to manipulate the activity of individual genes in localized regions within the lung, to help define their functions. Elucidation of the genes and genetic pathways that control mammalian lung development should lead to a molecular understanding of the process. It may also provide insights into human lung diseases and suggest ways to regenerate lung tissue. (A grant from the National Institutes of Health provides support for part of this project.)