Tracheal occlusion increases the rate of epithelial branching of embryonic mouse lung via the FGF10-FGFR2b-Sprouty2 pathway

Abstract

Tracheal occlusion during lung development accelerates growth in response to increased intraluminal pressure. In order to investigate the role of internal pressure on murine early lung development, we cauterized the tip of the trachea, to occlude it, and thus to increase internal pressure. This method allowed us to evaluate the effect of tracheal occlusion on the first few branch generations and on gene expression. We observed that the elevation of internal pressure induced more than a doubling in branching, associated with increased proliferation, while branch elongation speed increased 3-fold. Analysis by RT-PCR showed that Fgf10, Vegf, Sprouty2 and Shh mRNA expressions were affected by the change of intraluminal pressure after 48h of culture, suggesting mechanotransduction via internal pressure of these key developmental genes. Tracheal occlusion did not increase the number of branches of Fgfr2b-/- mice lungs nor of wild type lungs cultured with Fgfr2b antisense RNA. Tracheal occlusion of Fgf10(LacZ/-) hypomorphic lungs led to the formation of fewer branches than in wild type. We conclude that internal pressure regulates the FGF10-FGFR2b-Sprouty2 pathway and thus the speed of the branching process. Therefore pressure levels, fixed both by epithelial secretion and boundary conditions, can control or modulate the branching process via FGF10-FGFR2b-Sprouty2.

Figure 1

Panel A shows the morphological effect of increased internal pressure on cauterized lungs as compared to control lungs. Black arrow shows the cauterized trachea. A significant increase (3 fold) in the number of branches occurs in cauterized lungs. Panel B shows the average number of total new branches per day in cauterized and control lungs. A 2 or 3 fold increase of branching events occurs due to the increase of intraluminal pressure by cauterization. Panel C shows details of distal branches morphology of the control and cauterized lungs after 48h corresponding to the red box in panel A. More branches are visible in the cauterized lungs and the inter-branch length appears shorter than in control lungs.

Figure 2

Panel A displays the evolution in time, each image being separated by an interval of 1 hour and 30 minutes, of the distal tip of a cauterized lung branch. To study the branching process, we defined a length L corresponding to the distance from the centre of the apex of a branch to the capsule (the border of the organ). This distance L is represented on image 1 of the panel A. Distinct phases can be observed during lung branch growth. From panel 1 to 4, the branch grows towards the capsule and the distance L decreases, corresponding to the elongation phase. From panel 5 to 8, the branch stops and its sides progressively dilate, corresponding to the stop phase. Finally, from panel 9 to 12, the branch divides in two and the centre of the branch apex goes backwards relative to the pleura and the distance L increases, corresponding to the branching phase. Panel B displays the measurements of the distance L of a control lung branch. The elongation phase (in green), stop phase (in blue) and branching phase (in red) are clearly distinct. The measurements were fitted linearly and the average growth speed of each phase corresponds to the slope. Panel C displays the average growth speed in each phase of cauterized and control lung branches. In the elongation phase, the average growth speed is 2.7 ± 1.1 microns/hour for control lungs and 9.5 ± 2 microns/hour for clamped lungs. In the stop phase, the average growth speed is 0.4 ± 0.4 microns/hour for control lungs and −0.1 ± 0.8 microns/hour for cauterized lungs. During the branching phase, the average growth speed is 5.1 ± 0.8 microns/hour for control lungs and 5.0 ± 2.2 microns/hour for cauterized lungs. The increase of internal pressure leads to a three-fold increase in the elongation phase growth speed.

Figure 3

Phosphorylated Histone-H3 and phosphorylated ERK immunostainings of 5 μm sections of control and cauterized lungs cultured during 48h. A high increase in the number of proliferative cells can be observed in cauterized lungs as compared to controls. Mitotic cells are both mesenchymal and epithelial. The increase of intraluminal pressure in the cauterized lungs induces a raise of the percentage of phosphorylated Histone-H3 positive cells. 7.59 % ± 1.7 % epithelial cells and 8.63 % ± 1.07 % mesenchymal cells were stained in the control lungs and 13.1 % ± 1.56 % epithelial cells and 12.5 % ± 1.4% mesenchymal cells were stained in the control lungs. Phosphorylated ERK staining shows a similar increase in epithelial proliferation. 3.47 % ± 1.04 % of epithelial cells were stained in the cauterized lungs and 16.21 % ± 3.11 % of epithelial cells were stained in the control lungs.

Figure 4

Tracheal occlusion affects gene expression of cauterized lungs as compared to controls. After 48h of culture semi quantitative RT-PCR (panel A) showed an increase of Fgf10 (53%), Shh (45%) and Vegf (37%) mRNA expression and a decrease of Sprouty2 (27%) mRNA expression in E12.5 cauterized embryonic mouse lungs cultured during 48h as compared to controls. β-actin mRNA expression level was used for quantification. Panel B summarizes these variations of gene expression. Control lung mRNA expression is fixed at 100% and variations in cauterized lungs genetic expression are represented correspondingly.

Figure 5

LacZ staining of control (panel A) and cauterized lungs (panel B) of Flk1^nlacZ/+ E11.5 transgenic mice after 24h of culture. Blood vessel network appears more ramified and developed in cauterized lung as compared to control. The increase of intraluminal pressure by cauterization leads to an acceleration of blood vessel network formation and development.

Figure 6

Fgf10 and Sprouty2 whole mount in situ hybridization of control (left panel) and cauterized (right panel) lungs after 48 hours of culture (lungs were put in culture at E12.5). As observed with RT-PCR, the Fgf10 expression is higher in cauterized lungs than in controls while the expression of Sprouty2 is decreased in cauterized lungs. In the control, Fgf10 is principally expressed in the distal mesenchyme in front of the growing branches. The cauterized lung displays a high increase of Fgf10 expression and a change in its localization, Fgf10 being more widespread in the internal mesenchyme and surrounds the newly forming branches.

Figure 7

E12.5 Fgfr2IIIb^−/− transgenic mice lungs display a very strong phenotype with no branch formation. Only the primordial buds are formed and are unable to branch. Cauterization leads to a dilation of these primordial buds with no branch formation. Intraluminal pressure elevation is not able to rescue Fgfr2-IIIb^−/− phenotype. The observed dilation of cauterized Fgfr2IIIb^−/− mice lungs is consistent with an increase of lung internal pressure. FGF10 signaling by epithelial cells via FGFR2IIIb is needed for lung branching morphogenesis and pressure is unable by itself to induce branching events.

Figure 8

Panels A to C show the growth of an E12.5 cauterized lung every 24 hours. Panel D displays an E12.5 cauterized lung. After 24 hours of culture, antisense Fgfr2-IIIb RNA was added to its culture medium (see Panel D). From 24 to 48 hours of culture this lung formed only 2 new branches (see Panel E) while cauterized lungs form an average 33 new branches during this period. The addition of antisense Fgfr2-IIIb RNA leads to the inhibition of Fgfr2-IIIb expression and of the lung branching despite the increase of internal pressure by trachea cauterization.

Figure 9

Cauterization experiments of a wild type lung (left panels) and of a Fgf10^LacZ/− lung at E12.5 put in culture for 48 hours. The Fgf10^LacZ/− lungs express 70% less Fgf10 than wild type lungs. Cauterization of these lungs decreased the number of formed branches. During the 48 hours of culture a 3.5 ± 0.2 fold increase in the number of branches is observed in the wild type cauterized lungs, while Fgf10^LacZ/− lungs exhibit a 2.4 ± 0.1 fold increase in the number of branches. The low level of Fgf10 expression decreases the number of newly formed branches.