© 2003 American Thoracic Society DOI: 10.1165/rcmb.F259
Lung Vascular DevelopmentBreathing New Life Into An Old ProblemDepartment of Pediatrics, Developmental Lung Biology, and Department of Medicine, Cardiovascular Pulmonary Research, University of Colorado Health Sciences Center, Denver, Colorado Address correspondence to: Kurt R. Stenmark, M.D., University of Colorado Health Sciences Center, 4200 East 9th Avenue, Campus Box B131, Denver, CO 80262. E-mail: Kurt.Stenmark{at}UCHSC.edu
Abbreviations: insulin-like growth factor, IGF • IGF-binding proteins, IGF-BP • vascular endothelial growth factor, VEGF The lung originates as a pair of evaginations from the foregut endoderm. These endodermal buds branch and differentiate within the surrounding mesoderm ultimately giving rise to the airways, blood vessels, and alveoli of the mature lung. Lung branching morphogenesis and epithelial development have been the subject of intense investigation over the past 50 yr (1–5), yet studies of the mechanisms regulating pulmonary vascular development are relatively recent and limited in scope. This is surprising, given the central role the pulmonary circulation plays in mediating the critical gas exchange function of the lung. Current information regarding pulmonary vascular development is largely descriptive in nature and often based on model systems and techniques that may preclude accurate assessment of the diverse processes governing pulmonary vascular development, especially with regard to the origin, differentiation, and maturation of the various cell types within the pulmonary vascular wall. Nevertheless, available studies appear to have definitively established that the lung vasculature develops via at least two processes that likely occur concurrently: vasculogenesis, in which new blood vessels form in situ from angioblasts, and angiogenesis, which involves sprouting of new vessels from existing ones. The idea that some blood vessels arise de novo in the mesoderm surrounding the protruding endoderm was initially raised by Chang 70 yr ago (6). Confirmation of this idea has been eloquently shown by recent studies using molecular markers of endothelial progenitor cells (angioblasts) as well as endothelial differentiation markers (7–11). These studies have established that vessel development begins at the earliest stages of lung development, continues throughout lung development, and requires epithelial–mesenchymal crosstalk. They challenge old notions that development of the pulmonary circulation passively follows that of the airways. In addition, studies combining light and transmission electron microscopy with scanning electron microscopy of Mercox vascular casts suggest that the large pulmonary arteries originate via the process of angiogenesis and imply that a third process, fusion, is necessary to ultimately "connect" the angiogenic and vasculogeneic vessels (11, 12). Thus, we assume that assembly of a normal, functional pulmonary vascular network is dependent on maintaining some yet to be defined balance between the angiogenic and vasculogenic processes, and that an imbalance may have pathologic consequences during fetal, neonatal, or adult life. To expand our current knowledge of factors dictating pulmonary vascular development we need to define when, where, and how these two vascular beds connect to accommodate flow from the heart which begins beating early in embryonic life. The study by Han and coworkers in this issue of AJRCMB provides further evidence that both vasculogenesis and angiogenesis contribute even to early human lung vascular development (13). Their observations demonstrate that distinct subpopulations of endothelial cells (based on SMC association or lack thereof) are observed early in the developing lung and that these populations may arise from distinct genetic lineages (vasculogenic versus angiogenic). Further, these data provide evidence that different embryonic origins of vascular cells within the fetal lung may account for some of the endothelial (and ?smooth muscle) heterogeneity that characterizes the adult lung (14, 15). These observations are important because identifying the epigenetic origins of the different cells that comprise the pulmonary circulation will doubtless contribute significantly to understanding the unique physiology and pathophysiology of the pulmonary vascular bed. Recent studies demonstrate marked differences in the functional capabilities of endothelial cells from the large versus small pulmonary arteries and that these differences are maintained in culture (14, 16, 17). These observations oppose the commonly held perception that the phenotypic heterogeneity of endothelial cells along the pulmonary vascular tree is primarily induced and maintained by their disparate biochemical, mechanical, and anatomic environs. Based on the different processes involved in establishing the lung circulation, it is apparent that endothelium (and potentially SMC) of the large and small vessels arise from embryologically distinct origins. Small vessel endothelial cells appear to derive from the lung mesenchyme through vasculogenesis, whereas the macrovascular endothelium derives from the pulmonary truncus through angiogenesis. Thus, in addition to environmental stimuli that modulate differentiation, endothelial phenotype may also be partly generated through the cell's genetic program. If genetically programmed determinants do in fact contribute to endothelial cell differentiation and phenotypic distinction, then cells isolated from large versus small vessel segments should retain the phenotypic characteristics governed by their genetic program, even when placed in culture. Indeed, it has been reported that microvascular endothelial cells express more vascular endothelial cadherin (VE-cadherin) and less endothelial cell nitric oxide synthase (e-NOS) than do main pulmonary artery endothelial cells (14, 16). These phenotypes can further be discriminated on the basis of distinct lectin-binding profiles, and recent studies indicate that the different lectin-binding patterns are established in the early stages of embryonic lung development (14, 16). Microvascular endothelial cells also possess enhanced barrier function compared with their macrovascular counterparts and do not similarly change shape in response to inflammatory calcium agonists (16). Thus, a better understanding of the processes involved in lung vascular development (angiogenesis and vasculogenesis) may ultimately lead to a better understanding of the genetically preprogrammed determinates that control the function of fully differentiated adult lung endothelial cells both in the macro- and microcirculations. However, the idea that there are only two endothelial cell phenotypes in the lung (macro- and microvessel-derived) is probably overly simplistic. For instance, there is evidence to suggest that control of blood flow through supernumary pulmonary arteries is different than in adjacent pulmonary arteries of the same size (18). Current thinking would predict that both vessels have the same "genetic" (angiogenic) origins and thereby the same programmed physiology (12), yet another level of heterogeneity must account for their different responses. Other recent studies demonstrate that endothelial cells isolated even from discrete regions within the main pulmonary artery are heterogeneous with regard to functional capabilities, including proliferation and transdifferentiation (19, 20). Further, differences between endothelial cells in the venous and arterial systems within the lung have been noted (21). Based on studies in other systems, these differences may reflect stable genetic differences between endothelial (and SMC) of the arterial and venous circulations, including differential expression of ephrin B2 (22). Apparently, this arterial versus venous cell fate decision is made by early angioblasts and directed by a Notch gridlock signaling pathway (23). Other factors directly related to different genetic origins of vascular precursor cells may also contribute to the vast heterogeneity that characterizes the pulmonary circulation. Noden and others have pointed out that angioblasts can stream to developing organ beds from great distances (24). Angioblasts from different sites could be incorporated into developing vessels in a clonal-like fashion, and could lead to heterogeneity of endothelial cells in a small vascular region, as has been described for coronary vessels (25). Small patches of monoclonal endothelium could result and contribute under pathologic conditions to unique expansion of a specific endothelial phenotype (26). Lastly, once the connection between the proximal (angiogenic) and distal (vasculogenic) circulation is made, circulating precursor cells become another potential source of vascular cells and may add to the vast phenotypic heterogeneity of cells in the lung. The possibility that marrow-derived precursor cells can act as progenitors of lung alveolar epithelium has recently been addressed (27). Vascular development is marked by the carefully timed and cell specific expression of numerous growth factors and receptors (28, 29). Among the many factors previously thought to be important in cell proliferation and differentiation during embryogenesis are the insulin-like growth factors (IGF-1/2) (30–32). The findings of Han and colleagues with respect to IGF-1/2, IGF/R immunolocalization are consistent with the concept that the IGFs act either as autocrine or as paracrine mediators of human fetal lung vascular development. These studies corroborate and extend previous observations in the developing lung that have implicated a role for IGF family members in lung development (33). Studies of retinal development in IGF-/- mice also demonstrate that IGF-1 is critical for normal vascular development and that, in the absence of IGF-1, vascular endothelial growth factor (VEGF), a key factor mediating vasculogenesis in the lung and other organs, cannot stimulate normal vascular development (34). Further, IGF-1 is required for maximum VEGF stimulation of the MAPK pathway, which is important for endothelial cell proliferation (34). Thus, the expression of IGF-1 (and probably IGF-2) and the Type-1 IGF receptor throughout gestation in the lung supports an important role for the IGF system in lung growth and development. In addition, previous studies have also emphasized the importance of the IGF-binding proteins (IGFBP) in regulating the effects of IGF and demonstrated that there is a complex pattern of IGFBP expression during lung development (33). Thus, a complete understanding of the role of IGF in vascular development within the lung will require knowledge of the cells producing and the factors regulating IGFBP expression. Because IGFs also have significant effects on mesenchymal cell differentiation and SMC migration and proliferation, it will also be important to evaluate the role of IGFs in the acquisition of the smooth muscle phenotype by the developing vasculature. In addition to defining the spatio-temporal expression of IGF and its receptors in the developing human lung, Han and coworkers used an established explant culture method to study the role of IGF signaling in fetal lung vascularization. The data from these studies suggest an important role for the IGFs in survival of endothelial and probably other mesenchymal cells in the developing lung. These observations are consistent with the previously described role of IGFs in cell survival, a response mediated in large part through the Akt pathway (35). IGF may act in concert with VEGF and other local growth factors to maintain phosphorylated Akt levels in the endothelial cell. The fetal lung explant culture system has long been used to investigate the role that growth factors, matrix molecules, and tissue interactions play in lung epithelial branching morphogenesis and is now being used to characterize factors dictating fetal lung vascular development. As Han and colleagues demonstrate, this approach has been and will continue to be a useful tool when investigating the progression of fetal lung epithelial and mesenchymal development. However, its usefulness for studying vascular development may be limited because many have observed that vessel growth does not proceed at nearly the same extent as in vivo. We should be ever cognizant of the wise observations of Wahl and Noden, who cautioned that "surgical intervention in avian embryos may have substantial effects upon tissues within, adjacent to and distant to those that are being manipulated"(24). This issue of long-range effects is germane to the discussion of potential problems/limitations of lung explant culture in particular as a tool to investigate mediators of pulmonary vascular development. As suggested above, the marked heterogeneity of cells comprising the pulmonary vascular wall and the unique physiology and morphology pulmonary vascular bed suggests that the lung vasculature may be assembled from a variety of cells of unique embryonic and/or fetal origins, some of which may be quite distant from the lung. Remote sources of cells include neural crest and other neural cells; endothelial, smooth muscle and fibroblasts of cardiac origin; and circulating precursors. Thus, we should be cautious when drawing conclusions about pulmonary vascular development in a setting where the lung has been separated from the fetus and thereby many of the potential long-range effectors. In vivo studies will be needed to identify the embryonic origins of the various cells comprising the pulmonary vascular wall, and new transgenic approaches for marking cells at distinct stages of lung vascular development could be used to conduct classic cell-fate mapping studies. These approaches will allow us to better determine where, when, and how these assorted cell phenotypes arise. With this information in hand, in vitro studies can be designed to investigate the effects of various agents on inductive, proliferative, and assembly events occurring within and between the distinct cell populations. Oxygen tension also appears to be an important physiologic mediator of embryonic and fetal development, and hypoxia is known to be an important regulator of both vasculogenesis and angiogenesis. As such, it is not surprising that studies of mammalian development in vitro demonstrate that proper embryonic development is dependent on low oxygen tensions (3–5%) and even short exposures to normoxic environments (20%) can be detrimental to embryonic development (36, 37). Many of the genes that regulate vascular development, such as VEGF, Flk-1, Flt-1, Tie-2, PDGFb, bFGF, iNOS, and endothelin are known to be affected directly or indirectly by hypoxia (36, 38, 39). In addition, and relevant to the findings presented in the paper by Han and coworkers, is the fact that hypoxia also appears to regulate most members of the IGF axis of signaling proteins (40). These observations make it surprising that so few studies have been performed focusing on the role of hypoxia in the regulation of lung development.
To investigate the role of fetal oxygen tension in lung morphogenesis, we recently used a system in which fetal rat lungs were cultured at either 3 or 21% oxygen concentration. We found that branching of the terminal epithelium and cell proliferation in both epithelial and mesenchymal compartments increased significantly in explants cultured at 3% oxygen (Figure 1). Culture at 3% oxygen also increased the mRNA expression of VEGF. Interestingly, surfactant protein C mRNA was increased in 3% oxygen explants and was localized at the distal tips of the new branches, perhaps consistent with the newly described role for HIF-2
Recent studies have begun to elucidate the mechanisms through which hypoxia might modulate vascular cell proliferation and angiogenesis. With regard to hypoxia-induced angiogenesis, it is clear that PI3-kinase activity is essential and that downstream targets independent of Akt are important. Activation of mTOR appears essential for hypoxia-mediated amplification of cell proliferation in angiogenesis and activation may occur independent of Akt activation under hypoxic conditions, demonstrating a unique signaling aspect of hypoxia (46). In addition, recent studies also suggest that purinergic signaling is critical in hypoxia-induced fibroblast proliferation and differentiation as well as smooth muscle cell migration. This purinergic signaling loop may be autocrine in nature, because hypoxia induces release of ATP from vascular wall cells (47). In addition, sympathetic and/or sensory nerves may be an additional source of purine nucleotides, and may thus exert significant trophic effects on both the nascent and mature vasculature (48). The role of nerves in directing vascular development in embryonic skin has been recently eloquently demonstrated (49). An important role for nerves in airway SMC development and function has also been recently described (50). The role of nerves and purine nucleotides, especially under hypoxic conditions, in lung vascular development needs to be considered. Thus, it appears unequivocal that hypoxia activates unique cellular signaling pathways that are critical for growth and differentiation of cells within the pulmonary vasculature. Studies in the future need to be directed at using the effects of hypoxia to study lung vascular development. In summary, many important new studies have begun to elucidate the mechanisms through which the lung vasculature develops. These studies provide insight into the factors which control development and also provide insight into the genetic diversity of pulmonary vascular wall cells. These findings begin to provide explanations for the tremendous functional heterogeneity of the pulmonary vascular cells under both normal and pathophysiologic conditions. In the future, we will need to focus more attention on understanding from where and when endothelial and smooth muscle cells arise in the course of pulmonary arterial, bronchial, and pulmonary venous development. We will need to identify the environmental signals and signaling molecules that contribute to the terminal differentiation of specific vascular cells at the local level and which confer unique properties to these cells. We will need to use model systems that allow us to accurately mark and follow cell fates within the complex environment that obviously contributes to the ultimate phenotype of the pulmonary vascular cell of interest, as well as model systems where cell migration, cell–cell interaction, and proper environmental cues remain intact. We will need to take into account the fact that angioblasts may arise from many distant sites and at certain stages of lung development could even come from the bone marrow–derived pool of circulating stem cells. Because it is clear that oxygen tension plays such a critical role in directing development of many organs, we need to take into account the oxygen tension at which experiments are performed. Further, we need to address the role that the nervous system may play in directing vascular development within the lung. In doing all of the above, we will come to a better understanding of the unique origins of the macro- and microcirculations of the lung, and may also provide new insight into the unique expansion and function of the selective cell types that play critical roles in many pulmonary diseases.
The authors wish to thank Dr. John Shannon, University of Cincinnati, for the insights into pulmonary vascular development which he has provided to the authors over the years. In addition we thank Dr. John T. Reeves for the invaluable direction that he provides with regard to our understanding of lung vascular development. This study was supported in part by NIH Grants: HL-57144 (SCOR) and HL-14985 (PPG). Received in original form December 17, 2002
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
and VEGF in alveolar type II cell function (
Top
References
