News & Views | Published:

Developmental biology

Order in the lung

Nature volume 453, pages 733735 (05 June 2008) | Download Citation

Given the lung's thousands of branching airways, its development might be expected to be a highly complex process. Yet a surprisingly simple picture now emerges of when, where and in what order these branches form.

Elaborate branching is everywhere in nature. From riverbeds to oilfields, from trees to blood vessels, branching connects the large to the small. The lung is also a prime example of a reproducible branching system, allowing gas to be transported from the air to tissues deep within an animal. Without it — or without the simpler branched ducts found in less complex organisms — oxygen transport by diffusion probably would have limited the evolution of terrestrial animals to less than one millimetre in size. But how does such a sophisticated network develop? Metzger et al.1 (page 745 of this issue) provide a remarkable, yet simple picture that explains the orderly development of the more than a million branches in the mammalian lung.

In mammals, air enters through the nasal and oral cavities and passes through the larynx and trachea before reaching the lung. The trachea branches into two primary bronchi, which, within the lung, further branch into secondary and tertiary bronchi and finally into bronchioles. To investigate the sequence of events leading to this complex, yet highly reproducible network of branches, Metzger et al. studied the early bronchial tree in three dimensions by examining chemically fixed lung tissue from mouse embryos using microscopy.

The authors parse bronchial branching beyond the primary branch into three geometrical modes, which they call domain branching, planar bifurcation and orthogonal bifurcation. In domain branching, daughter branches form in rows along the parent branch, like bristles on a bottle brush. This branching mode forms the main secondary branches. Next, planar bifurcation is used for the formation of tertiary and later-generation branches; this mode is characterized by the splitting of a branch tip into two. Finally, orthogonal bifurcation involves two rounds of branching. Both rounds involve planar bifurcation, but, between the two, a 90° rotation of the bifurcation plane leads to the arrangement of the resulting branches into a rosette (Fig. 2f on page 746). These simple branching modes are used iteratively to give rise to the labyrinthine network that constitutes the bronchial tree.

The repetitive nature of the branching modules, together with their hierarchical control and the fact that they are coupled, suggests that the genetic 'hard-wiring' for bronchial branching could actually be quite simple. Thus, determining how the genome encodes the early development of the lung might be more tractable than previously thought. It also gives hope that, some day, regeneration or engineering of damaged lung tissue might be possible.

How are the three branching modes regulated? Metzger et al. infer that, once left–right laterality of the lung is established, airway branching is driven by a 'master' branch generator, with three 'slaves' in the form of subroutines (series of discrete patterning events). Of these, one subroutine seems to instruct a periodicity clock, which times the appearance of subsequent branches; another determines the rotational orientation of the branches around the axis of the parent airway; and the third mediates bifurcation (Fig. 1).

Figure 1: A master and three slaves.
Figure 1

Studying early lung morphogenesis in the mouse embryo, Metzger et al.1 show that this organ's airways form in a sequential manner in three series of events, or subroutines, which are all driven by one master branch generator. a, At embryonic day (E) 10.5, the primary bronchial branch (1) forms, followed by (b) the development of the left upper-lobe branch (2) by E11. c, The first two segmental branches of the left upper-lobe branch (2.2 and 2.3) and the subsequent formation of branches 3–6 occur at E12. The master branch generator is active throughout these events, and the inferred sites of action of the periodicity clock and bifurcator subroutines are shown. At E12, the rotator subroutine has not yet begun to function.

The authors identify a protein called Sprouty2 as a candidate component of the periodicity-clock subroutine in mice. Sprouty2 is named after the excessive tracheal branching seen in fly mutants that lack this gene. In flies, branching of the tracheal airway is initiated and controlled by the branchless gene, the closest mammalian equivalent of which is the gene that encodes the signalling protein FGF10. Also, the receptor for the protein product of branchless is Breathless, whose mammalian counterparts are FGF receptors. Sprouty2 is an evolutionarily conserved, inducible downstream inhibitor of FGF-receptor signalling from flies to mice.

The Fgf10 gene is expressed in the mesenchymal tissue, which overlies the epithelial-cell layer lining the emerging branch tip. The FGF receptor FGFR2 is expressed throughout the epithelium, and Sprouty2 is expressed locally at the branch tips2. Also, mutations in the Fgf10 or Fgfr2b genes that prevent their expression completely abrogate lung branching, and either decreased FGF10 expression or enhanced expression of Sprouty2 produces a small, poorly branched lung. Thus, in both flies and mice and probably in humans — Sprouty2 mediates fine regulation of FGF signalling at the correct time, place and dose to induce and control orderly airway branching.

Metzger and colleagues' observations further suggest that the balance between FGF expression, FGF-receptor activation and Sprouty2-mediated inhibition of FGF signalling is possibly a central component not only of the master branch generator but also of the periodicity-clock subroutine. The periodicity clock can be speeded up by increasing the internal pressure in cultured embryonic mouse lung tissue3. This gain-of-function effect, which involves a significant increase in the rate of branch extension, a reduction in inter-branch length and a shift from bifurcation to trifurcation of branch tips, is mediated by a pathway requiring FGF10, FGFR2b and Sprouty23,4,5,6. Thus, the often-overlooked connection between physics and biology in developmental processes is clearly important.

Other crucial players involved in lung branching include a long list of gene transcription factors, such as Nkx2.1; major signalling pathways, including those mediated by retinoids, bone morphogenetic protein, Hedgehog and Wnt; and essential components of the extracellular matrix, especially fibronectin and laminin6,7. All these protein factors are expressed at the right time and place, and function in and around the tips of the airway branches. Equally important for structural reproducibility is suppression of branching at regions away from the tips. The signal protein Sonic hedgehog accomplishes this by negatively regulating Fgf10 expression proximal to the tips, thus suppressing out-of-place branching.

Genes related to those encoding most, if not all, of these morphogenetic factors are thought to have been present even in Urbilateria, the common ancestor of the planarian flatworms8. So it is possible that conditions of relative oxygen shortage exerted strong evolutionary pressure on pre-existing groups of such genes, to select for tube formation and hence gas transport. In support of this idea, FGF signalling in flies is regulated by oxygen levels so as to match terminal tracheal branching to the local oxygen needs of a tissue9. In mice, and presumably in humans, the signalling pathway mediated by molecules such as hypoxia-inducible factor and vascular endothelial growth factor also plays a crucial developmental part with the FGF–FGFR–Sprouty2 pathway. Together, these pathways match the capillary vasculature to the epithelial layer in branches of the early lung, a process that is crucial for determining the eventual gas-diffusing capacity of the organ10.

Whether a master branch generator controlling a select few slave subroutines represents a general developmental strategy that has been reused over evolutionary time, in different branched organs, remains an intriguing possibility. Also, solving the specific problem of gas diffusion as a limit on size, and discovering how simplified, genetically controlled branching routines interact with physical and biological factors to direct complex yet reproducible patterns of development, will be matters of great interest. To quote Charles Darwin as interpreted by biologist Sean Carroll, they will aid our understanding of how “endless forms most beautiful” have evolved from a relatively simple tool-box of genetic modules.


  1. 1.

    , , & Nature 453, 745–750 (2008).

  2. 2.

    et al. Mech. Dev. 102, 81–94 (2001).

  3. 3.

    et al. Mech. Dev. 125, 314–324 (2008).

  4. 4.

    et al. Dev. Cell 4, 11–18 (2003).

  5. 5.

    et al. Curr. Biol. 9, 219–222 (1999).

  6. 6.

    et al. Pediatr. Res. 57, 26R–37R (2005).

  7. 7.

    & Development 133, 1611–1624 (2006).

  8. 8.

    Cell 132, 185–195 (2008).

  9. 9.

    et al. Dev. Cell 14, 547–558 (2008).

  10. 10.

    et al. Dev. Biol. 290, 177–188 (2006).

Download references

Author information


  1. David Warburton is at the Saban Research Institute, Childrens Hospital Los Angeles, University of Southern California, 4650 Sunset Boulevard, Los Angeles, California 90027, USA.

    • David Warburton


  1. Search for David Warburton in:

About this article

Publication history



Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Newsletter Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing