The major finding of this paper is the presence in GIT1 KO mice of profound abnormalities in lung morphogenesis including decreased pulmonary vascular density, parenchymal hemorrhage and increased alveolar spaces. Organ development and function in the GIT1 KO mouse was otherwise normal, suggesting that there is a unique requirement for GIT1 in the lung. Interestingly, Premont's lab showed that GIT1 expression in the lung was confined to EC and bronchi, while GIT2 was ubiquitous3
. While we observed no compensation by GIT2 in the lung, the ubiquitous expression of this protein may be sufficient to compensate for GIT1 deficiency in most tissues. The pulmonary findings in the GIT1 KO mice phenocopy the VEGF120 mice13
, suggesting that GIT1 is a critical mediator of VEGF signaling in the lung. Support for this concept is provided by impaired signaling downstream of VEGF including reduced activation of PLCγ and ERK1/2 in the lungs of GIT1 KO. We show further that this pathway is critical for EC tube formation and proliferation. These data support a novel mechanism for VEGF signaling in the lung mediated by a VEGF-GIT1-PLCγ-MEK1-ERK1/2 pathway. Importantly, this phenotype resembles the human diseases bronchopulmonary dysplasia (BPD) and alveolar capillary dysplasia (ACD), in which impaired vascular development retards pulmonary development chronically.
The abnormalities in pulmonary vascular development in GIT1 KO mice were most striking in the decreased numbers of distal arteriolar branches and capillaries with diameters of 0-50 μm as shown by X-ray, micro-CT, and fluorescein angiography. Correspondingly, total EC number was obviously reduced as demonstrated by decreased expression of the EC markers, vWF and VEGFR2, in lungs of GIT1 KO mice. However, embryonic lung morphogenesis (E14.5) of GIT1 KO was normal suggesting that the primary problem in the GIT1 KO is a failure of pulmonary angiogenesis rather than epithelial morphogenesis. In addition we found that tight junctions in arteries and veins were normal in GIT1 KO based on electron microscopy. Bleeding time and platelet number were also normal in GIT1 KO mice. Thus the impressive pulmonary hemorrhage we observed was likely not due to abnormalities in vascular permeability or hemostasis. Therefore the essential requirement for GIT1 in pulmonary vascular development likely reflects unique developmental and perinatal events in the pulmonary vasculature (and possibly alveoli as discussed below). Future studies using EC-specific knockout of GIT1 will be necessary to define the roles of GIT1 more precisely.
Because of reduced pulmonary vascularity with chronic hypoxemia, we hypothesized that adult GIT1 KO mice may develop pulmonary hypertension. Surprisingly, adult GIT1 KO mice showed PAP and RV pressures similar to WT mice. Furthermore, components of the pulmonary vasoconstrictor pathway - endothelin-1 (ET-1)17,18
and the endothelin-1 receptor A were unchanged in GIT1 KO mice (Supplemental Fig. 6
). These results suggest that compensatory mechanisms exist in the GIT1 KO mice that survive to adulthood. It will be interesting in future studies to characterize these compensatory pathways.
The finding that the GIT1 KO has many morphologic similarities to the VEGF120 mouse suggests that GIT1 is a novel mediator for VEGF signaling in the lung13
. There is close coordination of airway and vessel growth that is essential for normal lung development, especially during the saccular and alveolar stages of development. Angiogenesis has been shown to be necessary for alveolarization during lung development, and inhibiting pulmonary vascular growth impedes alveolar growth12 19
. Thus we propose that inhibiting pulmonary vascular growth during a critical period of postnatal lung growth impairs alveolarization, suggesting that endothelial-epithelial cross talk, especially via VEGF signaling, is critical for normal lung growth following birth. This concept is supported by decreased SP-C positive cell numbers and decreased expression of VEGF mRNA in GIT1 KO mice (Supplemental Fig. 7
,). We believe that a requirement for GIT1 in pneumocyte development is unlikely since these cells express primarily GIT23
. In addition to VEGF, fibroblast growth factor-10 (FGF-10) has been shown to be essential for lung morphogenesis. We believe it is unlikely that FGF-10 is involved in the GIT1 KO pulmonary phenotype since mRNA expression of FGF-10 was normal in GIT1 KO mice (Supplemental Fig. 6
Both VEGF164 and VEGF18820
have been shown to be essential to development of the lung vasculature. The disruption of even a single copy of the VEGF gene results in embryonic lethality due to the failure of endothelial differentiation and blood vessel formation20
. To overcome this limitation, Galambos et al13
engineered knock-in transgenic mice expressing only VEGF120, which led to a selective impairment in the development of the distal pulmonary arteriolar branches, strongly resembling the vascular defects of GIT1-deficient animals13
. VEGF contributes to blood vessel formation in the developing lung 21,22
, including the arteriolar branches and capillary formation. Postnatal lung development (alveolar stage) is primarily dependent upon angiogenesis, which correlates with increased expression of GIT1 () and VEGF13,22-24
in the lung. Based on these findings we suggest that GIT1 function is required for VEGF signaling. VEGF signaling involves activation of c-Src, PLC-γ, and ERK1/2, which are important in angiogenesis6,25
. Our previous data showed that GIT1 is a substrate of c-Src and functions as a scaffold protein for PLCγ and MEK16,26,27
. Thus we believe that GIT1 is required for PLCγ-MEK1-ERK1/2 activation by VEGF and for lung microvasculature development. In cultured EC, knockdown of GIT1 by siRNA obviously decreased the activation of PLC-γ, MEK1 and ERK1/2 () which play important roles in EC proliferation and migration 6,25
. This role of GIT1 was confirmed by the findings of significantly decreased EC proliferation and tube formation after GIT1 knockdown (, ).
Among the components of the VEGF-GIT1-PLCγ-MEK1-ERK1/2 signaling pathway we believe that activation of PLCγ is most critical to the pulmonary vasculature. Neither the MEK1 or ERK1/2 KO mice have a significant vascular phenotype, while PLCγ KO mice die at embryonic day 8-9 due to deficiencies in EC function14
. PLCγ phosphorylation, a measure of activity, exhibited a time course in GIT1 WT lungs that matched GIT1 and VEGF expression, suggesting a role in postnatal lung maturation (). Importantly, phosphorylation of VEGFR2 at tyrosine 1175 which is required for activation of PLCγ was not decreased by deletion of GIT1 ()16
. Thus the difference between GIT1 and PLCγ KO in terms of time of fetal death likely reflects a requirement for PLCγ protein expression (rather than activity) and/or some compensation by GIT2 during development.
The diminished numbers of EC and lack of capillary in-growth into the alveolar septae found in the GIT1 KO lungs represent characteristic features of BPD and ACD. BPD is a chronic lung disease that develops in premature infants treated with ventilation and supplemental oxygen28
. As a consequence of impaired alveolar capillary in-growth, normal air-blood barriers fail to form29,30
. BPD has been associated with decreased expression of VEGF and VEGFR228
, suggesting mechanisms similar to GIT1 KO. ACD occurs in infants with persistent pulmonary hypertension refractory to vasodilator treatment. ACD also is characterized by deficiency of alveolar capillary in-growth and failure to form normal air-blood barriers between EC and type I epithelial cells. While the gene defect is unknown, a vascular etiology appears likely since children with ACD often display misalignment of pulmonary veins, with both venous and arterial vessels sharing a common adventitial sheath31
. It is possible that abnormal GIT1 expression and function might contribute to impaired alveolar vascularization in BPD and ACD patients. In conclusion, the present report provides novel evidence of a previously unrecognized contribution of GIT1 to pulmonary vascular development and maturation. Thus, pharmacological or genetic manipulation of the GIT1 signaling pathway (especially PLCγ) in the perinatal period may represent a therapeutic strategy to improve respiratory function in premature infants and in clinical conditions associated with abnormalities in lung morphogenesis.