[Frontiers in Bioscience 7, d1609-1623, July 1, 2002]

THE ROLE OF VASCULAR GROWTH FACTORS IN HYPEROXIA-INDUCED INJURY TO THE DEVELOPING LUNG

Carl T. D'Angio and William M. Maniscalco

Strong Children's Research Center, Department of Pediatrics, University of Rochester School of Medicine and Dentistry, Rochester NY

TABLE OF CONTENTS

1. Abstract
2. Introduction
3. Vascular changes in neonatal lung injury
4. Vascular development and its control
4.1. Processes of vascular development
4.2. Vasculogenic, angiogenic and angiostatic factors
4.2.1. Vasculogenic growth factors
4.2.2. Vascular endothelial growth factor (VEGF)
4.2.3. The angiopoietin system
4.2.4. Chemokines
4.2.5. Angiostatic factors
4.3. Vascular growth factors in normal lung development
4.3.1. Lung vascular development
4.3.2. VEGF in lung development
4.3.3. Angiopoietins in lung development
5. Vascular growth factors in lung injury
5.1. Endothelial alterations in animal lung injury
5.2. VEGF and the angiopoietins in lung injury <
5.2.1. VEGF in animal lung injury
5.2.2. The angiopoietin system in animal lung injury
5.2.3. VEGF in human BPD
5.3. Chemokines in lung injury
5.3.1. Angiogenic chemokines in animal lung injury
5.3.2. Interleukin-8 in human BPD
6. Therapy for endothelial loss during lung injury
7. Summary and perspective
8. Acknowledgments
9. References

1. ABSTRACT

Normal pulmonary vascular development is the result of a complex interplay of growth factors, including vascular endothelial growth factor (VEGF) and the angiopoietins. Injury to the developing lung, whether due to hyperoxia or mechanical ventilation, results in disordered vascular development, ranging from an apparent arrest of microvascular development in milder injury to extensive microvascular derangement in more severe injury. Alterations in vascular growth factors may participate in these injuries. During injury to the developing animal lung, VEGF abundance is markedly decreased. In models of post-injury recovery, up-regulation of VEGF accompanies the re-establishment of normal vasculature. Alterations in lung VEGF levels in human premature infants are less clear cut. However, among humans premature newborns who later go on to develop bronchopulmonary dysplasia (BPD), VEGF production is decreased in comparison to those newborns who recover. Other angiogenic factors, such as the CXC ELR+ chemokines, are also altered in injury to the developing lung, but their specific roles in vascular injury are less clear. Strategies that enhance microvascular integrity, whether through attenuating alterations in vascular growth factors or by other means, also improve the outcome of lung injury. Such therapies may eventually offer hope in human BPD.

2. INTRODUCTION

Bronchopulmonary dysplasia (BPD), the chronic lung disease of newborn infants, is characterized by pulmonary inflammation, fibrosis and architectural changes, including disordered vascular development (1,2). The pathogenesis of the disease is complex and multifactorial. Increasing immaturity, more aggressive mechanical ventilation and higher inspired concentrations of oxygen all predispose to the condition. Exposure of newborn animals to hyperoxia recapitulates many of the findings found in human BPD (3). This review will focus on the vascular abnormalities seen in animal models of BPD, and describe the changes in vascular growth factors accompanying these alterations.

3. VASCULAR CHANGES IN NEONATAL LUNG INJURY

Fetal and early neonatal life are times of rapid angiogenesis. This is particularly true of the lung. During the first 6 weeks after birth, lung endothelial cell mass in the rat increases six-fold (4). Pulmonary injury during this critical period of pulmonary vascular development can have far-reaching consequences (5,6). Newborn rats exposed to 100% oxygen for six days show markedly decreased capillary density when compared to air-breathing controls (7). If these animals are allowed to recover in air for 1-2 weeks, they develop normal to increased numbers of capillaries. Extending similar 80%-100% oxygen exposures in rats for 8 days to 2 weeks makes the decrease in pulmonary vascularity irreversible, even if the animals are allowed to recover thereafter in air (8,9). Neonatal mice exposed to 100% oxygen for 6 weeks also have decreased capillary surface area (10).

Premature lambs delivered at 120-130 days of a 147 day gestation (80-85% of gestation) and exposed to mechanical ventilation and a fractional concentration of inspired oxygen (FiO2) of 0.40-0.50 for 3 weeks develop a chronic lung disease similar to human BPD (11). These animals show decreased small arteries and veins when compared to term newborn lambs or 3 week old lambs. In addition, the capillary surface density of the animals is decreased in comparison to controls, while the epithelial surface density is unchanged.

Premature baboons delivered at 140 days (75% of a 185 day gestation, roughly equivalent to 30 weeks gestation in a human) and exposed to high FiO2 and mechanical ventilation develop a pulmonary lesion very similar to the "classic," fibrotic human BPD (1). These animals display decreased endothelial cell densities in areas of damaged lung when compared to less aggressively ventilated infant baboons (12). Extremely premature baboons, delivered at 125 days (approximately 67% of gestation, roughly equivalent to a human gestation of 26-27 weeks), develop chronic pulmonary changes even if treated with exogenous surfactant and managed with the minimum oxygen and positive pressure ventilation possible (13). Although these animals display less fibrosis than aggressively ventilated more mature animals, they still have markedly deranged architecture, a finding typical of the "new" BPD seen in extremely premature human infants (14). Extremely premature baboons treated in this manner and sacrificed at 1-2 months of age display severe capillary hypoplasia, with decreased volume density of capillaries when compared to term animals or 2-month-old animals born at term (13). Adult baboons exposed to mechanical ventilation and 100% oxygen for 11 days also manifest decreased endothelial cell numbers (15).

Human infants who die as a result of BPD also have disordered pulmonary vasculature. Much of the work in humans has focused on abnormalities in larger vessels, with mixed results (16). One analysis of a small number of patients dying of BPD showed increased arterial concentrations in two patients (17). Another study evaluated premature infants dying either during their initial neonatal illnesses or 4-15 months after having apparently recovered from their neonatal illnesses (2). The lungs of both latter sets of infants showed normal arterial-to-alveolar ratios, but reduced alveolar numbers and reduced overall numbers of arteries.

Functionally, human newborns with severe BPD often display evidence of pulmonary hypertension, although it is unclear how this relates to the pathologic microvascular findings in these infants (18,19). In one study in which both pre-mortem cardiac catheterization data and post-mortem morphometric data were available on 3 children, the infants had high pulmonary vascular resistance and raised intrapulmonary shunts during life (18). Medial thickness in the small pulmonary arteries was increased at post-mortem, and, in one child, some vessels were completely obliterated by fibrosis. Although alveolar numbers were decreased, no comment was made on the microvasculature in these children. Another group of investigators followed 10 children with BPD who resided at an altitude of about 1500 m (5,000 feet) through the age of about 6 years. In the four children who underwent repeat cardiac catheterization, pulmonary artery pressure and pulmonary vasculature resistance, although improved, continued to be elevated throughout follow-up (20).

4. VASCULAR DEVELOPMENT AND ITS CONTROL

4.1. Processes of vascular development

New blood vessel creation and growth occurs by two processes. The first of these, termed vasculogenesis, is primarily limited to early embryonic life (21). In vasculogenesis, endothelial cell precursors known as angioblasts differentiate and form tubes in the absence of pre-existing vessels (Figure 1). Tissues of endodermal origin, including the lung, heart tube and major vessels, are among those vascularized by vasculogenesis (22). Angiogenesis describes the formation of small vessels and capillaries by budding or sprouting from existing larger vessels (23). New vessels can be formed by the growth of an interstitial column of cells that eventually divides a vessel lumen into two parallel lumens (a process known as intussusception) or by a new vessel sprouting at an angle from an existing vessel (known as sprouting angiogenesis).

Sprouting angiogenesis entails an initial lysis of the basement membrane of the original vessel and of the surrounding extracellular matrix, followed by the migration of endothelial cells into the resulting space, proliferation and tube formation by the endothelial cells, and the formation of loops by anastomoses of growing sprouts.

Nascent vessels formed by sprouting angiogenesis are then "stabilized" by the re-formation of a basement membrane and the enclosing of the vessel by pericytes (for a capillary) or smooth muscle cells (for a vessel). Immature capillaries that are not stabilized experience rapid apoptosis of the endothelium and regress (24). Angiogenesis is the primary mode of vascularization in ectoderm- and mesoderm-derived tissues (25). Unlike vasculogenesis, angiogenesis continues throughout fetal development, and plays an important role in such normal processes of postnatal life as wound healing and the female reproductive cycle.

During development and throughout life, vessels are also subject to remodeling. Many vascular beds begin as a plexus of relatively undifferentiated capillary tubes. As development progresses, the vessels in these networks acquire the differing lumen sizes and wall thicknesses that characterize the mature, arboreal pattern of vessels seen in the adult (25). Remodeling also involves a dynamic balance between the growth of new vessels and the regression of others. In some circumstances, such as the female reproductive cycle, this results in marked alterations in vessel number over time.

4.2. Vasculogenic, angiogenic and angiostatic factors

As is common in biological systems, the net potential for new vessel growth and development is the product of a balance between factors promoting vasculogenesis and/or angiogenesis and those inhibiting or reversing this activity (Figure 2). These include a wide variety of molecules responsible for matrix resorption and reconstitution, cellular chemoattraction, cellular proliferation, cellular differentiation, and cell adhesion, as well as for the blockade or downregulation of each of these processes. These factors have been extensively reviewed elsewhere (21,25,26). This review will focus primarily on several of the vascular growth factors regulating endothelial cell function in angiogenesis.

4.2.1. Vasculogenic growth factors

The control of vasculogenesis is, as yet, poorly understood. Fibroblast growth factor-2 (FGF-2, bFGF) is member of a large family of heparin-binding, 18-30 kDa growth factors (27). In numerous in vitro and in vivo models, FGF-2 is capable of stimulating angiogenic activity. In vitro, it also causes primitive mesodermal cells to differentiate into hemangioblast-like cells, implying involvement in the earliest steps of vasculogenesis (28). However, FGF-2 deficient mice develop without any evident early defects in vasculogenesis or angiogenesis (29). The transforming growth factor beta (TGF-beta) family contains a number of homodimeric peptides involved in a wide variety of cellular growth and differentiation processes (30). TGF-beta1-deficient mice die in early embryogenesis with extensive derangement or absence of vasculogenesis (31). This may represent an indirect effect of TGF-beta through its regulation of matrix accumulation or integrin expression (25).

The vascular endothelial growth factor (VEGF) system, described in detail below, also has an apparent role in vasculogenesis. Mice deficient in VEGF die early in gestation with disordered vasculogenesis, even if only a single VEGF allele is missing (32,33). Mice lacking the VEGF receptor 2 (VEGFR-2) have defects in early angioblast formation (34), while those lacking VEGFR-1 form angioblasts but do not develop usable blood vessels (35). Both receptor knockout animals die in early embryogenesis.

4.2.2. Vascular endothelial growth factor (VEGF)

Vascular endothelial growth factor (VEGF or VEGF-A) is the prototype and best-described member of a family of endothelial-specific, heparin-binding, angiogenic growth factors. The family also contains four closely related proteins labeled VEGF-B, -C, and -D, and placental growth factor (PlGF). VEGF itself undergoes alternative mRNA splicing of exons 6 and 7 of the gene to produce 5 isoforms varying from 121 to 206 amino acids in length (and varying slightly in size among species). The relative abundance of the alternatively spliced variants of VEGF mRNA presumably affects the species of VEGF protein present in a tissue. VEGF121 lacks a heparin binding domain and is soluble, while the other isoforms are increasingly heparin bound with increasing size. VEGF165 is the predominant isoform in most tissues. VEGF stimulates endothelial cells to degrade extracellular matrix, proliferate, migrate and form tubes, and may act as an endothelial cell survival factor. In vivo, VEGF also increases vascular permeability, leading to its alternative name, vascular permeability factor (VPF) (36).

VEGF mRNA transcription is upregulated by a number of other growth factors, including platelet derived growth factor (PDGF), epidermal growth factor (EGF), and TGF-beta, and the cytokines tumor necrosis factor alpha (TNF-alpha), and interleukin-1 beta (IL-1 beta) (37-39). In addition, VEGF abundance is increased by hypoxia and decreased by normoxia. The effect of hypoxia is modulated in part through the hypoxia-sensing transcription factor hypoxia inducible factor-1 (HIF-1), which binds to a site in the VEGF promoter (40). Hypoxia also stabilizes VEGF mRNA (41).

The direct effects of hyperoxia on VEGF mRNA expression are somewhat unclear. Although VEGF mRNA expression following hyperoxia exposure is generally decreased in the lung (as detailed in section 5.2.1. below) and retina (42,43), hyperbaric hyperoxia exposure has been associated with increased VEGF protein levels in a rat wound model (44).

VEGF interacts primarily with two receptors, VEGFR-1 (also known as flt-1) and VEGFR-2 (also known as flk-1 or KDR). Binding of VEGF to VEGFR-1 stimulates endothelial cell migration, and may mediate vascular organization (35). VEGFR-2 activation stimulates endothelial cell differentiation, proliferation and migration, and has been implicated in angioblast differentiation (34). A related receptor, VEGFR-3 (flt-4), binds VEGF-C and VEGF-D. The cell surface proteins neuropilin-1 and neuropilin-2 also bind the VEGF165 isoform, but these proteins have short intracellular domains, and it is unclear whether they act independently to transduce any signal (45). In the embryo, VEGFR-1, VEGFR-2 and VEGFR-3 are expressed in developing vessels and capillaries throughout vascular maturation (46,47).

4.2.3. The angiopoietin system

The angiopoietins, angiopoietin-1 (Ang-1) and Ang-2, are ligands for the endothelial cell-specific receptor TIE (tyrosine kinase with immunoglobulin- and EGF-like domains)-2. TIE-2 and a closely related receptor, TIE-1, were described several years before the angiopoietins were found to be the ligands for TIE-2. At present, the ligand of TIE-1 remains unknown. Ang-1 interacts with TIE-2 to transduce a signal, while Ang-2 occupies the receptor without signal transduction, acting as an antagonist of Ang-1 (48). Ang-1 mediates vascular remodeling and endothelial cell interactions with supporting cells, making it an important factor in vascular stabilization following angiogenic sprouting (48). The balance between Ang-1 and Ang-2 determines the degree of vascular stability (49). In quiescent vasculature, Ang-1 is bound to TIE-2, and vessels are stabilized. In hypoxia or following VEGF stimulation, Ang-2 is upregulated and displaces Ang-1 from TIE-2 (50). As a result, endothelial cells lose their association with pericytes and the matrix. In the presence an angiogenic factor such as VEGF, endothelial mitogenesis and migration ensues, resulting in new vessel formation. In the absence of such a factor, endothelial apoptosis and vascular regression occur (49).

Mice lacking either Ang-1 or TIE-2 die in early embryogenesis with vascular abnormalities (48,51,52). The embryos have extensive vascular disorganization. The vascular plexuses remain immature without progression to a normal network of large and small vessels. The hearts remain immature, and in some TIE-2 knockout animals the aorta is completely disrupted (52). These findings all support the importance of the angiopoietin system in completing the work of angiogenesis.

4.2.4. Chemokines

The chemokines are a superfamily of heparin binding cytokines originally described as having chemotactic and activating effects on leukocytes (53). Members of the superfamily share a highly conserved pattern of 4 cysteine residues crucial to the secondary structure of the molecules. Members of the CXC subfamily of chemokines, so called due to an interposed amino acid between the first two conserved cysteine residues, also have apparent effects on angiogenesis (53). Among CXC chemokines, some members contain a three-amino-acid pattern of Glu-Leu-Arg ("ELR") immediately N-terminal to the CXC motif. These members of the CXC family (CXC ELR+) display angiogenic capacity. Other CXC members lack the three amino acid pattern (CXC ELR-) and possess angiostatic properties.

The CXC ELR+ chemokines include interleukin-8 (IL-8) and the growth related proteins (GRO alpha, beta and gamma). CXC ELR+ proteins induce endothelial cell chemotaxis and mitogenesis in vitro (54). They also cause angiogenesis in a variety of in vivo systems in a manner that appears divorced from their ability to cause inflammation (53,55). CXC ELR+ chemokines are expressed by a variety of tumor cells, and depletion or blocking of the CXC ELR+ chemokines limits the angiogenic activity of the tumors (56). Although the receptor that mediates the angiogenic activity is unknown, the CXC chemokine receptor 2 (CXCR2) is a possible candidate. Mice deficient in the CXCR2, which binds all CXC ELR+ chemokines, have been generated. These mice do not display angiogenesis in response to the exogenous administration of CXC ELR+ cytokines, implying that the CXCR2 is responsible for transducing the angiogenic signal of these chemokines (57). However, other than decreased neovascularization during wound healing, CXCR2 deficient mice do not display alterations in normal angiogenesis (58). Whether this is due to the extensive duplication in the chemokine-chemokine receptor pathways or whether this implies that CXC ELR+ chemokines have limited roles in developmental angiogenesis is unknown.

The CXC ELR- chemokines include several chemokines induced by interferon gamma, such as monokine induced by interferon gamma (MIG) and interferon-gamma-inducible protein (IP-10). These interferon-inducible CXC ELR- chemokines block the angiogenic effects of CXC ELR+ chemokines and other angiogenic factors, including VEGF, in a variety of in vitro and in vivo models (53). In tumor models, CXC ELR- chemokines limit tumor growth by reducing tumor-associated angiogenesis (59,60). The angiostatic effects attributed to interferon gamma (IFN gamma) may be mediated through its induction of CXC ELR- chemokines. Evidence accumulated with the CXC ELR- chemokine platelet factor 4 (PF4) suggests that some of its angiostatic effects may be due to its binding to VEGF165 and FGF-2, which prevents their association with their receptors (61,62).

The CXC ELR- chemokine stromal cell-derived factor (SDF-1) has been reported to have both angiostatic (63) and angiogenic (64) activities in vitro and in vivo. However, deficiency of the receptor for SDF-1, CXCR4, in mice leads to defective vascular development, hematopoiesis and cardiogenesis, and to death in utero (65). This implies that the contribution to angiogenesis by CXC chemokines may have complexities beyond our current models of the process.

4.2.5. Angiostatic factors

Several cytokines and growth factors have been credited with being angiostatic. As discussed above, the CXC ELR- chemokines prevent new vessel growth. Ang-2, by virtue of its capacity for disrupting endothelial cell interactions, can result in vascular destruction if its actions are unaccompanied by those of an angiogenic factor. Endothelial monocyte activating peptide II (EMAP-II) is a recently described cytokine with angiostatic properties both in vitro and in vivo (66,67). It apparently operates by specifically causing endothelial cell apoptosis (66). In the developing mouse lung, EMAP II markedly inhibits neovascularization, and also inhibits alveolar type II cell development (67). It is unclear whether the alterations of lung epithelial morphogenesis are primary or secondary effects of EMAP II.

4.3. Vascular growth factors in normal lung development

4.3.1. Lung vascular development

The normal early development of the lung vasculature involves three processes (68). The central pulmonary vessels branch from the existing systemic vasculature. Smaller vessels originate by vasculogenesis within the lung mesenchyme. Finally, connections between the two systems are formed by lysis of some existing vessel walls and fusion of the lumens. In the human embryo, the first pulmonary vascular structures to form are vascular lakes of hematopoietic cells surrounded by mesenchyme, which are present by about 4-5 weeks of gestation (69). Central pulmonary vessels are present and beginning to branch into the lung mesenchyme by 7-8 weeks of gestation. The pulmonary venous network forms earlier than the pulmonary arterial tree, with central pulmonary veins having connections with the intrapulmonary vasculature by 7-8 weeks of gestation. The pace of venous development exceeds the pace of branching of the developing airways. By contrast, the formation of the pulmonary arterial circulation lags behind the branching of the developing airways. By the close of the pseudoglandular phase of lung development (16 weeks), both pulmonary arterial and pulmonary venous branches accompany all airways. However, the process of fusion of the central arteries to the existing venous and capillary networks continues throughout the canalicular phase (16-23 weeks). At the close of the canalicular stage, capillaries remain centrally located in the relatively thick septa of the distal airways (70). As the septa thin during the saccular phase of lung development at 24 weeks and beyond, the capillaries become closely apposed to the developing alveolar epithelium.

4.3.2. VEGF in lung development

VEGF is expressed abundantly in the developing lung of humans and animals, and appears to be important in the genesis and maintenance of the developing vasculature (71-74). The lung is one of the chief organs expressing VEGF mRNA in the mid-trimester human fetus (73). VEGF is expressed in the developing mouse lung, as well as other tissues, by the early third trimester of murine fetal development at embryonic day 14 (E14) and E17 (of a 20 day gestation) (75). VEGF-D, however, is expressed almost exclusively in the lung, and does not appear until the late third trimester (E17) (76). VEGF-B and VEGF-C do not appear to be expressed in the mouse lung during fetal development (75). In normal lung development in the mouse, VEGF and VEGFR-2 mRNAs undergo coordinate, 3-fold increases in expression during the canalicular and saccular stages of lung development between E13 and E18, and a further 2-to-3-fold increase by two weeks following delivery (77). Both messages increase twofold again by adulthood. VEGF mRNA expression appears localized primarily to the alveolar epithelium from E18 onward, with specific expression in alveolar type II epithelial cells by 8 days postnatal age. VEGF protein immunostaining is more diffuse, with protein detected in epithelial, mesenchymal and vascular smooth muscle cells (77).

In the developing baboon, mRNA expression for VEGF and its receptors VEGFR-1 and VEGFR-2 increase throughout the third trimester of pregnancy, as the lung progresses from saccular to alveolar stages (78). VEGF mRNA expression reaches adult levels by shortly before term, and VEGFR-1 mRNA expression in late gestation exceeds adult levels. VEGF mRNA is expressed in a scattered fashion throughout the septa in the distal airways in the early third trimester baboon fetus, and becomes more localized to the alveolar epithelium as gestation progresses. The less soluble VEGF189 splice variant becomes progressively more expressed in the lung during the third trimester, while the more soluble VEGF121 variant experiences a relative decrease. A similar pattern of increasing VEGF189 to VEGF121 ratio during the third trimester is present in the lung of the developing rabbit fetus (79).

Explanted embryonic mouse lungs display a pattern of lung vascular development similar to the in utero murine fetus, including the progressive organization and extension of originally mesenchymal vascular plexes to surround the developing airways (80). In the explanted mouse lung, VEGF is diffusely distributed in the airway epithelium and subepithelial matrix during the initial vascular-lake phase of vessel development at E11.5. As the vessels organize and surround the airways during E13.5-E15.5, VEGF expression becomes limited to the branching tips of distal lung airways. Similarly, the VEGF receptor, VEGFR-2 is present in the lung mesenchyme immediately underlying the developing epithelium in the rat as soon as the lung evaginates from the foregut, and remains present in clusters of cells immediately under the epithelium throughout lung vascular development (81). If lung mesenchyme is removed from proximity to the epithelium, VEGFR-2 expression rapidly fades. However, if mesenchyme and epithelium are explanted together, the normal in vivo pattern of VEGFR-2 expression persists (81).

Newborn rabbits express VEGF mRNA in the alveolar epithelium from birth through 5 weeks of age (72). Pulmonary alveolar epithelial cells are major VEGF-expressing cells in both newborn and adult rabbit lung (72,82). As a result, VEGF is also present in alveolar fluid from these animals (79). VEGF is also apparently necessary even when active angiogenesis is not taking place, since moderate levels of VEGF mRNA are expressed in pulmonary alveolar epithelial cells of normal adult rats and rabbits, and VEGF mRNA is expressed diffusely in adult mouse lungs (75,82,83).

Lungs from human fetuses of 16-20 weeks' gestation display VEGF mRNA and protein expression localized primarily to distal airway epithelial cells, with some expression in connective tissue (84). Over the brief age span covered in the study, the authors did not note an effect of gestational age on VEGF expression. In a single study of VEGF protein in lung fluid collected on the first day after birth from newborn humans born at 24-33 weeks' gestation, we have documented a trend of decreasing VEGF with increasing gestational age (85). Since the VEGF in lung lavage fluid is likely to consist predominantly of the more soluble VEGF121 isoform, this finding may be more reflective of changes in isoform ratios (the proportion of VEGF121 falls in most animals during gestation) than of overall lung VEGF levels (which tend to rise in animals during gestation).

VEGF has also been examined immunohistochemically in human newborns dying in the first 48 hours after birth from pulmonary hypoplasia due to congenital diaphragmatic hernia or from non-pulmonary causes (86). In lungs from control infants, VEGF immunostaining is present in bronchial epithelial cells and in arterial medial smooth muscle cells. In the hypoplastic lungs, the immunostaining in the bronchial epithelium and smooth muscle is more intense. In addition, VEGF is detected in the media of smaller diameter arteries and in the arterial endothelium in hypoplastic lungs.

4.3.3. Angiopoietins in lung development

Ang-1 and Ang-2 are both expressed in the developing mouse lung throughout its prenatal development from E9.5 until term (87). However, spatial co-localization of these growth factors with the developing lung vasculature has not yet been confirmed. In the developing baboon fetus, mRNA expression for both Ang-1 and its receptor TIE-2 each increase about two-fold in a coordinate fashion between 125 days gestation (equivalent to about 27 weeks gestation in the human fetus) and term (185 days) (78). Unlike VEGF and the VEGF receptors, there does not appear to be a specific cellular localization for Ang-1 and TIE-2 mRNAs. Messenger RNA expression for the orphan receptor, TIE-1, appears to peak in late gestation (175 days), and to decrease after birth in the baboon.

5. VASCULAR GROWTH FACTORS IN LUNG INJURY

5.1. Endothelial alterations in animal lung injury

Hyperoxia is toxic to isolated endothelial cells, although the specific mechanisms of oxygen-induced toxicity in these cells remain unclear. Hyperoxia and the resulting free radicals can induce direct biochemical injury through such mechanisms as lipid peroxidation, protein sulfhydryl oxidation, enzyme inactivation, mitochondrial injury and DNA damage (88). In isolated endothelial cells, DNA synthesis is decreased by as little as 24 hours in 100% oxygen (88). Hyperoxia also mediates changes in multiple growth factors and cytokines expressed by endothelial cells, including such factors as heme oxygenase I and TGF beta (88,89). In the hyperoxia-exposed animal, endothelial cells may induce secondary damage to themselves by upregulation of such leukocyte adhesion factors as intracellular adhesion molecule-1 (90).

In the adult mouse lung, overall DNA synthesis is markedly decreased by 72 hours of 100% oxygen exposure (91). Lung microvascular endothelial cells show swelling and disruption following 90 hours of hyperoxia. The total mass of endothelial cells in the lung falls progressively during a 90 hour hyperoxia exposure. In the adult rat, there is electron microscopic evidence of endothelial cell apoptosis as early as 24 hours into a >95% oxygen exposure (92). Successful recovery from hyperoxia-induced lung injury depends, in part, on the successful proliferation of endothelial cells to repopulate the denuded capillary basement membrane (93).

In the premature baboon delivered at 125 days gestation and managed with the minimum of mechanical ventilation and increased fraction of inspired oxygen necessary to maintain normal partial pressures of O2 and CO2 in arterial blood, significant lung injury still supervenes (13). Levels of protein and mRNA expression of the endothelial marker platelet endothelial cell adhesion molecule (PECAM)-1 in the lung normally rise progressively during the third trimester of baboon gestation (78). In animals delivered at 125 days, PECAM-1 protein levels and mRNA expression cease rising following delivery, and following 14 days of ventilation (139 days equivalent gestational age) fall significantly below the levels found in utero in animals at 140 days' gestation. Histologically, there appears to be a delay in maturation of the capillary network of the lung. Capillary numbers are decreased, and the capillaries remain centrally located in the developing septa, a pattern more consistent with the early saccular pattern of development of the 125 day fetus than the subepithelial capillary pattern becoming apparent in the late saccular lung of the 140 day fetus. This delay in maturation is apparent after as little as 6 days of ex utero ventilation.

5.2. VEGF and the angiopoietins in lung injury

5.2.1. VEGF in animal lung injury

Acute lung injury depresses pulmonary VEGF mRNA expression in adult animals. Adult rats exposed to polytetrafluoroethylene (PTFE) fumes experience a fall in VEGF mRNA expression to 20% of control levels within 4 hours (94). In adult rats exposed to >95% oxygen, VEGF mRNA expression falls to about 70% of control levels by 24 hours of exposure, and to 50% of control levels by 48 hours (92). VEGF mRNA expression in adult rats appears localized to alveolar type II epithelial cells. VEGF mRNA expression in these cells is markedly decreased following 24 hours of hyperoxia, and is virtually absent by 48 hours of hyperoxia. VEGF protein is more diffusely present in the bronchial and alveolar epithelium, but undergoes a similar pattern of decrease by 48 hours of hyperoxia. The mRNAs for the receptors VEGFR-1 and VEGFR-2 also decrease to approximately half control levels during a 48-hour hyperoxia exposure (92). Adult rabbits exposed to 100% oxygen for 64 hours experience a marked depression of pulmonary VEGF mRNA expression (82). Again, VEGF expression appears to be limited largely to the type II alveolar epithelial cell and to disappear with hyperoxia exposure. If 64-hour exposed rabbits are allowed to recover from hyperoxic injury, their type II epithelial cells express supranormal levels of VEGF mRNA during the first 3-5 days after their return to room air. Secreted VEGF protein, measured in lung lavage from oxygen-injured adult rabbits, is mildly elevated during a 64-hour hyperoxia exposure, but falls at the close of the exposure, before undergoing a 40-fold rise during recovery (79). Other investigators have also reported an increase in bronchoalveolar lavage VEGF from adult mice exposed to hyperoxia for 72 hours, but do not have data for recovering mice (95).

In the adult rat lung, the relative abundances of mRNA for the splice variants VEGF188 and VEGF164, which code for heavily and intermediately heparin-bound isoforms of the protein, do not change over a 48 hour hyperoxia exposure (92). In the adult rabbit lung, however, the relative abundance of VEGF189 mRNA (homologous to the 188 amino acid protein in the rat) decreases by half during a 64-hour hyperoxia exposure, while the relative expression of the mRNA for the soluble VEGF121 species increases twofold. The expression of VEGF189 does not return to control levels until 5-7 days following hyperoxic lung injury.

Newborn animals of many species are more resistant to hyperoxia than adults, but eventually succumb to oxygen-induced lung injury (3). Hyperoxia-exposed newborn rabbits exhibit changes in VEGF expression similar to the adult, albeit with a longer time frame. In the newborn rabbit, pulmonary VEGF mRNA expression is depressed during an acute, 9-day, 100% oxygen exposure and returns toward normal during a recovery period in 60% oxygen (72). In addition, VEGF protein in lung lavage from newborn rabbits, although mildly increased early in hyperoxia exposure, falls as the inflammation peaks (Figure 3). VEGF protein in lavage then increases more than 8-

In the fetal baboon, VEGF mRNA expression progresses from a diffuse pattern throughout the alveolar septa at 125 days gestation to a more localized expression in cells at the alveolar surface by 140 days gestation (78). If the animal is delivered at 125 days gestation, and treated with oxygen and mechanical ventilation as needed for 14 days, overall expression of VEGF mRNA is markedly decreased (Figure 4), and the localized epithelial pattern of VEGF expression does not develop. A similar decrease in overall abundance and inhibited pattern of maturation is seen if immunoreactive VEGF protein is measured. Expression of mRNA for VEGFR-1 also decreases 30-40% in treated animals compared to 140-day-gestation controls, while the VEGFR-2 mRNA expression remains unchanged. The decreases in VEGF and its receptor parallel the stunted capillary development in these animals (78).

5.2.2. The angiopoietin system in animal lung injury

In contrast to the findings with VEGF, premature baboon lung injury does not alter the normal gestational pattern of increasing mRNA expression for the vascular stabilization factor Ang-1 and its receptor, TIE-2, in the lung (78). However, the mRNA expression for the orphan TIE-1 receptor is decreased about 40% in the same injury. Ang-1 and TIE-2 alterations have not yet been examined in other newborn lung injuries.

5.2.3. VEGF in human BPD

Several studies have evaluated levels of VEGF in lung fluid obtained from intubated human premature newborns born at 23-33 weeks' gestation (85,96,97). Although the methods varied somewhat between studies, all investigators found that VEGF protein measured in lung fluid rose at least 3-to-4-fold over the first 3-10 days after birth (Figure 5). These findings appear to be at odds with information gathered in animals, particularly that gathered in the roughly comparable premature baboon (section 5.2.1.). In the baboon, lung VEGF mRNA expression falls about 30% in the six days following delivery at 125 days' gestation (approximately equivalent to 26-27 weeks human gestation), and barely returns to 125-day levels by 14 days following delivery (Figure 4). Lung VEGF protein appears to follow suit. One potential explanation of the discrepancy may lie in the differences between the compartments sampled in the baboon and the human. The human studies all depended upon measurement of VEGF protein in lung lavage fluid. The VEGF measured in this fluid may over-represent soluble VEGF121. As detailed above in section 5.2.1., this splice variant of VEGF is increased relative to other VEGF isoforms during lung injury in adult and newborn rabbits. In addition, measurements of lavage VEGF in both adult and newborn animals show initial increases during hyperoxia, culminating only later in decreases as inflammation reaches critical levels (79,95). The measurements available in human newborns may reflect results from a specific compartment during a relatively mild, early injury.

More consonant with the animal findings are the later patterns of VEGF expression in human newborns with more severe lung injury, i.e. those who go on to develop BPD. Lassus and colleagues (96), in a study of 44 intubated premature infants with a mean gestational age of 27 weeks, found that although tracheal aspirate fluid VEGF levels rose over the first week of life, infants destined to develop BPD had depressed VEGF levels between days 4 and 7, when compared to those infants who would recover without BPD. In a study of 40 premature infants born at 24-33 weeks' gestation, we were unable to delineate an association between VEGF levels at any time over the first week after birth with the subsequent development of BPD (85). There was a trend, however, toward gradual elevation of VEGF over time in children who recovered without BPD, while children who did develop BPD or died appeared to have a decrease in VEGF with age. This suggests that the subset of human infants who develop BPD may have lung VEGF alterations similar to those in the premature baboon model, where VEGF is markedly depressed over the first week following delivery.

The confusing picture of lavage VEGF measurements in humans may also be due in part to additional variables not present in animal models. In premature humans, elevated levels of lung fluid VEGF are associated with prolonged or premature rupture of the membranes, and with chorioamnionitis, factors which predispose to the later development of BPD (85,96). Corticosteroid administration may also be a factor. Our cohort of infants was drawn from a trial comparing the administration of dexamethasone at 12 and 24 hours of age to placebo for the prevention of BPD (98). There were no differences in VEGF levels in lung fluid between dexamethasone-treated and placebo groups at 1 or 3 days after birth, but dexamethasone subjects had increased VEGF levels at 7 days after birth. Although experiments in isolated cells (99,100) have indicated that VEGF expression is decreased following glucocorticoid administration, animal data suggest otherwise. In neonatal mice, VEGF mRNA is increased in a dose-dependent fashion by a 3-day course of dexamethasone, with a 2-fold increase at the highest dose administered (5 mg/kg/day) (77). VEGF protein, however, shows no corresponding increase over that time frame. It may be that the increase in VEGF protein seen in our human subjects followed an earlier increase in mRNA expression akin to that seen in the animal.

5.3. Chemokines in lung injury

5.3.1. Angiogenic chemokines in animal lung injury

Although IL-8 and other CXC ELR+ chemokines have been implicated in pulmonary angiogenesis (101), it is difficult in observational studies of pulmonary injury to dissociate the angiogenic effects of these molecules from their pro-inflammatory functions. A number of investigators have evaluated the role of CXC ELR+ chemokines in newborn hyperoxia-induced lung injury. In a newborn rabbit model of >95% oxygen exposure for 8-9 days, followed by recovery in 60% oxygen until 5 weeks of age, histologic evidence of acute lung injury and neutrophil infiltration peak at the close of the >95% oxygen exposure, and recede thereafter (102,103). Signs of chronic lung injury, including simplification of alveolar architecture, persist through 3-5 weeks of age. IL-8 mRNA expression coincides very closely with the temporal pattern of neutrophil infiltration, and is largely limited to neutrophils. IL-8 mRNA expression returns to levels similar to those in air-exposed, age-matched controls during the period of recovery and repair. In a similar model of 8-day hyperoxia exposure in the newborn rat, Auten and colleagues have shown 10-fold increases in the CXC ELR+ chemokines cytokine-induced neutrophil chemoattractant protein-1 (CINC-1) and macrophage inflammatory protein-2 (MIP-2) by the close of exposure (104). These increases are accompanied by inflammatory cell infiltrates, septal thickening and decreases in alveolar volume and surface density (104,105). Treatment with anti-CINC-1 or anti-MIP-2 antibodies or with a CXCR2 blocker prevents many of these hyperoxia-related changes (104-106). Perhaps most importantly for the purposes of this discussion, anti-CINC-1 antibody therapy decreases CINC-1 levels to those of air-exposed controls and preserves normal alveolar architecture (105). The close correlation of CXC ELR+ chemokines to neutrophil infiltration and the preservation of normal alveolar architecture in their absence argue that their major roles in hyperoxic lung injury are proinflammatory rather than angiogenic. If their roles were indeed angiogenic, one might expect them to be present during recovery, and for their absence to lead to disordered alveolar repair. However, neither the complementary network of chemokines and chemokine receptors nor the possible roles of anti-angiogenic CXC ELR- chemokines have been adequately addressed in neonatal hyperoxic lung injury. Further study may reveal as-yet-unanticipated effects of chemokines upon the pulmonary vasculature during hyperoxia.

5.3.2. Interleukin-8 in human BPD

Premature infants who later develop bronchopulmonary dysplasia display elevated airway fluid levels of several proinflammatory cytokines. Elevations in the CXC ELR+ chemokine IL-8 in infants who subsequently develop BPD have been reported in a number of studies (107-110). Peak levels of IL-8 appear to be present at 1-2 weeks of age (107,108,111), a time at which clinical signs heralding BPD are beginning to emerge. A small number of studies have suggested that changes in IL-8 within the first few days of life may also predict the development of BPD (110,112,113). The pattern of IL-8 elevation in human newborns appears to be the inverse of that described for VEGF. This implies that, as in the animal, IL-8 may be less important in vascular growth or repair than in inflammation. The absence of either therapeutic trials or extensive histology in humans makes any definite inference difficult, however.

6. THERAPY FOR ENDOTHELIAL LOSS DURING LUNG INJURY

Several recent publications have presented fascinating glimpses of the potential for endothelial preservation during hyperoxic lung injury. Keratinocyte growth factor (KGF, also known as FGF-7) administered intratracheally prior to hyperoxia exposure protects adult mice against lung injury, with both prevention of histologic injury and improvement in survival (91,114). KGF has well-described mitogenic effects on pulmonary epithelial cells, and much research has focused on these effects. However, a recent study of prophylactic KGF administration in the rat also showed significant preservation of capillary endothelial structure and of total endothelial volume during a 90-hour hyperoxia exposure (91). It is unclear whether this effect is a direct effect of KGF on the endothelium, an indirect effect through the epithelium, or a less-specific effect of KGF-mediated attenuation of pro-apoptotic factors such as bax and bcl-x (91).

Interleukin-13 (IL-13) is a cytokine with a variety of pro- and anti-inflammatory properties that also has mitogenic effects on endothelial cells (95,115). Adult transgenic mice that overexpress IL-13 under the control of the Clara cell 10-kDa protein (CC10) promoter are markedly protected during hyperoxia, with a tripling of survival time (95). A more modest survival advantage is incurred by intranasal administration of exogenous IL-13. The IL-13 transgenic mice have increased amounts of VEGF in bronchoalveolar lavage, with immunohistochemical evidence of VEGF protein in airway and alveolar epithelium, smooth muscle, alveolar macrophages and blood vessels. This increase is further augmented in hyperoxia, with VEGF levels following 72 hours of hyperoxia some fourfold higher than in non-transgenic, hyperoxia-exposed mice. While the predominant VEGF isoform in either air- or hyperoxia-exposed wild type mice is VEGF164, hyperoxia-exposed transgenic mice express increased amounts of VEGF120 and VEGF188. Neutralization of VEGF using anti-VEGF antibody shifts the survival in hyperoxia of IL-13 transgenic mice back toward that of wild type mice, implying that a significant portion of the IL-13 survival effect is mediated through VEGF. The specific effect of increased VEGF on the capillary endothelium in this model remains unknown.

Taken together, the effects of these therapeutic trials imply that preservation of the capillary endothelium, perhaps through a VEGF-mediated mechanism, may be a promising avenue of investigation for the prevention of oxygen-induced lung injury in humans. Further research will be necessary to determine whether such an approach would be useful in the multi-factorial injury that leads to BPD in immature humans.

7. SUMMARY AND PERSPECTIVE

Multiple vascular growth factors, including VEGF and the angiopoietins, interact to induce normal vasculogenesis and angiogenesis in the developing lung. Normal patterns of vascular development are disrupted by early delivery and the attendant lung injury that leads to bronchopulmonary dysplasia (figure 6). Hyperoxia exposure and/or mechanical ventilation of newborn animals, which mimics BPD, leads to an apparent arrest of microvascular development in milder injury or extensive microvascular derangement in more severe injury. Some of these alterations may be due to disruption of the elaboration of vascular growth factors. Of these factors, VEGF has been examined the most extensively. In most adult, newborn or premature animal models of hyperoxic or ventilator-associated injury, VEGF mRNA expression and protein production are markedly attenuated by the injury. In models of post-injury recovery, there appears to be an up-regulation of VEGF that accompanies the re-establishment of normal vasculature. The direction of changes in VEGF secretion in human BPD is less certain. VEGF, as measured in bronchoalveolar lavage, appears to rise in the first few days following premature delivery in newborn humans, a finding difficult to reconcile with the decreases over the same time period in most animal models. Some of the differences, however, may be due to the different compartment (lung fluid) sampled in human studies. Several animal studies have documented similar increases in lavage VEGF early in hyperoxia exposure, although levels tend to fall as injury worsens. In human premature newborns who later go on to develop BPD, there is evidence that VEGF production may be decreased in comparison to those newborns who recover. Overall, the data imply that in human BPD, as in animal lung injuries, decreases in vascular growth factors such as VEGF may be associated with poorer vascular development and poorer outcome. Preliminary evidence from animal studies suggests that strategies that enhance microvascular integrity also improve the outcome of lung injury. This gives hope that such therapies may eventually prove useful for the prevention or treatment of human BPD.

8. ACKNOWLEDGMENTS

Supported in part by NHLBI K08 HL-03493.

9. REFERENCES

1. Northway W. H., Jr.: Bronchopulmonary dysplasia: then and now. Arch Dis Child 65, 1076-1081 (1990)

2. Hislop A. A. & S. G. Haworth: Pulmonary vascular damage and the development of cor pulmonale following hyaline membrane disease. Pediatr Pulmonol 9, 152-61 (1990)

3. Frank L., J. R. Bucher & R. J. Roberts: Oxygen toxicity in neonatal and adult animals of various species. J Appl Physiol: Respir Envir Exercise Physiol 45, 699-704 (1978)

4. Burri P. H., J. Dbaly & E. R. Weibel: The postnatal growth of the rat lung. I. Morphometry. Anat Rec 178, 711-730 (1974)

5. Crapo J. D.: Morphologic changes in pulmonary oxygen toxicity. Ann Rev Physiol 48, 721-731 (1986)

6. Fracica P. J., M. J. Knapp, C. A. Piantadosi, K. Takeda, W. J. Fulkerson, R. E. Coleman, W. G. Wolfe & J. D. Crapo: Responses of baboons to prolonged hyperoxia: physiology and qualitative pathology. J Appl Physiol 71, 2352-2362 (1991)

7. Roberts R. J., K. M. Weesner & J. R. Bucher: Oxygen-induced alterations in lung vascular development in the newborn rat. Pediatr Res 17, 368-375 (1983)

8. Wilson W. L., M. Mullen, P. M. Olley & M. Rabinovitch: Hyperoxia-induced pulmonary vascular and lung abnormalities in young rats and potential for recovery. Pediatr Res 19, 1059-1067 (1985)

9. Shaffer S. G., O. N. D, S. K. Bradt & D. W. Thibeault: Chronic vascular pulmonary dysplasia associated with neonatal hyperoxia exposure in the rat. Pediatr Res 21, 14-20 (1987)

10. Obara H., C. T. Pappas, W. H. Northway, Jr. & K. G. Bensch: Comparison of the effect of two and six week exposure to 80% and 100% oxygen on the lung of the newborn mouse: a quantitative SEM and TEM correlative study. Int J Radiat Oncol Biol Phys 11, 285-298 (1985)

11. Bland R. D., K. H. Albertine, D. P. Carlton, L. Kullama, P. Davis, S. C. Cho, B. I. Kim, M. Dahl & N. Tabatabaei: Chronic lung injury in preterm lambs: abnormalities of the pulmonary circulation and lung fluid balance. Pediatr Res 48, 64-74 (2000)

12. Coalson J. J., V. T. Winter, D. R. Gerstmann, S. Idell, R. J. King & R. A. Delemos: Pathophysiologic, morphometric, and biochemical studies of the premature baboon with bronchopulmonary dysplasia. Am Rev Respir Dis 145, 872-881 (1992)

13. Coalson J. J., V. T. Winter, T. Siler-Khodr & B. A. Yoder: Neonatal chronic lung disease in extremely immature baboons. Am J Respir Crit Care Med 160, 1333-1346 (1999)

14. Jobe A. H. & E. Bancalari: Bronchopulmonary dysplasia. Am J Respir Crit Care Med 163, 1723-1729 (2001)

15. Coalson J. J., R. J. King, V. T. Winter, T. J. Prihoda, A. R. Anzueto, J. I. Peters & W. G. Johanson, Jr.: O2- and pneumonia-induced lung injury. I. Pathological and morphometric studies. J Appl Physiol 67, 346-356 (1989)

16. Pariente D., J. de Blic, C. Ernest, D. Lallemand & P. Lacombe: Evaluation of pulmonary vascularity in children as seen on digital angiography. Annales de Radiologie 27, 166-172 (1984)

17. Tomashefski J. F., Jr., H. C. Oppermann, G. F. Vawter & L. M. Reid: Bronchopulmonary dysplasia: a morphometric study with emphasis on the pulmonary vasculature. Pediatr Pathol 2, 469-487 (1984)

18. Bush A., C. M. Busst, W. B. Knight, A. A. Hislop, S. G. Haworth & E. A. Shinebourne: Changes in pulmonary circulation in severe bronchopulmonary dysplasia. Arch Dis Child 65, 739-745 (1990)

19. Goodman G., R. M. Perkin, N. G. Anas, D. R. Sperling, D. A. Hicks & M. Rowen: Pulmonary hypertension in infants with bronchopulmonary dysplasia. J Pediatr 112, 67-72 (1988)

20. Berman W., Jr., R. Katz, S. M. Yabek, T. Dillon, R. R. Fripp & L. A. Papile: Long-term follow-up of bronchopulmonary dysplasia. J Pediatr 109, 45-50 (1986)

21. Vailhe B., D. Vittet & J. J. Feige: In vitro models of vasculogenesis and angiogenesis. Lab Investig 81, 439-452 (2001)

22. Pardanaud L., F. Yassine & F. Dieterlen-Lievre: Relationship between vasculogenesis, angiogenesis and haemopoiesis during avian ontogeny. Development 105, 473-485 (1989)

23. Risau W.: Mechanisms of angiogenesis. Nature 386, 671-674 (1997)

24. Benjamin L. E., I. Hemo & E. Keshet: A plasticity window for blood vessel remodelling is defined by pericyte coverage of the preformed endothelial network and is regulated by PDGF-B and VEGF. Development 125, 1591-1598 (1998)

25. Beck L., Jr. & P. A. D'Amore: Vascular development: cellular and molecular regulation. FASEB J 11, 365-373 (1997)

26. Liekens S., E. De Clercq & J. Neyts: Angiogenesis: regulators and clinical applications. Biochem Pharmacol 61, 253-270 (2001)

27. Florkiewicz R. Z. & A. Sommer: Human basic fibroblast growth factor gene encodes four polypeptides: three initiate translation from non-AUG codons. Proc Nat Acad Sci USA 86, 3978-3981 (1989)

28. Flamme I. & W. Risau: Induction of vasculogenesis and hematopoiesis in vitro. Development 116, 435-439 (1992)

29. Ortega S., M. Ittmann, S. H. Tsang, M. Ehrlich & C. Basilico: Neuronal defects and delayed wound healing in mice lacking fibroblast growth factor 2. Proc Nat Acad Sci USA 95, 5672-5677 (1998)

30. Kingsley D. M.: The TGF-beta superfamily: new members, new receptors, and new genetic tests of function in different organisms. Genes Devel 8, 133-146 (1994)

31. Dickson M. C., J. S. Martin, F. M. Cousins, A. B. Kulkarni, S. Karlsson & R. J. Akhurst: Defective haematopoiesis and vasculogenesis in transforming growth factor-beta 1 knock out mice. Development 121, 1845-1854 (1995)

32. Carmeliet P., V. Ferreira, G. Breier, S. Pollefeyt, L. Kieckens, M. Gertsenstein, M. Fahrig, A. Vandenhoeck, K. Harpal, C. Eberhardt, C. Declercq, J. Pawling, L. Moons, D. Collen, W. Risau, & A. Nagy: Abnormal blood vessel development and lethality in embryos lacking a single VEGF allele. Nature 380, 435-439 (1996)

33. Ferrara N., K. Carver-Moore, H. Chen, M. Dowd, L. Lu, O. S. KS, L. Powell-Braxton, K. J. Hillan & M. W. Moore: Heterozygous embryonic lethality induced by targeted inactivation of the VEGF gene. Nature 380, 439-442 (1996)

34. Shalaby F., J. Rossant, T. P. Yamaguchi, M. Gertsenstein, X. F. Wu, M. L. Breitman & A. C. Schuh: Failure of blood-island formation and vasculogenesis in Flk-1-deficient mice. Nature 376, 62-66 (1995)

35. Fong G. H., J. Rossant, M. Gertsenstein & M. L. Breitman: Role of the Flt-1 receptor tyrosine kinase in regulating the assembly of vascular endothelium. Nature 376, 66-70 (1995)

36. Heiss J. D., E. Papavassiliou, M. J. Merrill, L. Nieman, J. J. Knightly, S. Walbridge, N. A. Edwards & E. H. Oldfield: Mechanism of dexamethasone suppression of brain tumor-associated vascular permeability in rats. Involvement of the glucocorticoid receptor and vascular permeability factor. J Clin Investig 98, 1400-1408 (1996)

37. Veikkola T. & K. Alitalo: VEGFs, receptors and angiogenesis. Sem Cancer Biol 9, 211-220 (1999)

38. Akagi Y., W. Liu, K. Xie, B. Zebrowski, R. M. Shaheen & L. M. Ellis: Regulation of vascular endothelial growth factor expression in human colon cancer by interleukin-1beta. Brit J Cancer 80, 1506-1511 (1999)

39. Enholm B., K. Paavonen, A. Ristimaki, V. Kumar, Y. Gunji, J. Klefstrom, L. Kivinen, M. Laiho, B. Olofsson, V. Joukov, U. Eriksson & K. Alitalo: Comparison of VEGF, VEGF-B, VEGF-C and Ang-1 mRNA regulation by serum, growth factors, oncoproteins and hypoxia. Oncogene 14, 2475-2483 (1997)

40. Kimura H., A. Weisz, Y. Kurashima, K. Hashimoto, T. Ogura, F. D'Acquisto, R. Addeo, M. Makuuchi & H. Esumi: Hypoxia response element of the human vascular endothelial growth factor gene mediates transcriptional regulation by nitric oxide: control of hypoxia-inducible factor-1 activity by nitric oxide. Blood 95, 189-197 (2000)

41. Ikeda E., M. G. Achen, G. Breier & W. Risau: Hypoxia-induced transcriptional activation and increased mRNA stability of vascular endothelial growth factor in C6 glioma cells. J Biol Chem 270, 19761-19766 (1995)

42. Yamada H., E. Yamada, S. F. Hackett, H. Ozaki, N. Okamoto & P. A. Campochiaro: Hyperoxia causes decreased expression of vascular endothelial growth factor and endothelial cell apoptosis in adult retina. J Cell Physiol 179, 149-156 (1999)

43. Pournaras C. J., J. W. Miller, E. S. Gragoudas, D. Husain, J. L. Munoz, M. J. Tolentino, M. Kuroki & A. P. Adamis: Systemic hyperoxia decreases vascular endothelial growth factor gene expression in ischemic primate retina. Arch Ophthal 115, 1553-1558 (1997)

44. Sheikh A. Y., J. J. Gibson, M. D. Rollins, H. W. Hopf, Z. Hussain & T. K. Hunt: Effect of hyperoxia on vascular endothelial growth factor levels in a wound model. Arch Surg 135, 1293-1297 (2000)

45. Neufeld G., T. Cohen, S. Gengrinovitch & Z. Poltorak: Vascular endothelial growth factor (VEGF) and its receptors. FASEB J 13, 9-22 (1999)

46. Breier G., M. Clauss & W. Risau: Coordinate expression of vascular endothelial growth factor receptor-1 (flt-1) and its ligand suggests a paracrine regulation of murine vascular development. Devel Dynam 204, 228-239 (1995)

47. Kaipainen A., J. Korhonen, K. Pajusola, O. Aprelikova, M. G. Persico, B. I. Terman & K. Alitalo: The related FLT4, FLT1, and KDR receptor tyrosine kinases show distinct expression patterns in human fetal endothelial cells. J Exp Med 178, 2077-2088 (1993)

48. Suri C., P. F. Jones, S. Patan, S. Bartunkova, P. C. Maisonpierre, S. Davis, T. N. Sato & G. D. Yancopoulos: Requisite role of angiopoietin-1, a ligand for the TIE2 receptor, during embryonic angiogenesis. Cell 87, 1171-1180 (1996)

49. Holash J., S. J. Wiegand & G. D. Yancopoulos: New model of tumor angiogenesis: dynamic balance between vessel regression and growth mediated by angiopoietins and VEGF. Oncogene 18, 5356-5362 (1999)

50. Oh H., H. Takagi, K. Suzuma, A. Otani, M. Matsumura & Y. Honda: Hypoxia and vascular endothelial growth factor selectively up-regulate angiopoietin-2 in bovine microvascular endothelial cells. J Biol Chem 274, 15732-15739 (1999)

51. Dumont D. J., G. Gradwohl, G. H. Fong, M. C. Puri, M. Gertsenstein, A. Auerbach & M. L. Breitman: Dominant-negative and targeted null mutations in the endothelial receptor tyrosine kinase, tek, reveal a critical role in vasculogenesis of the embryo. Genes Devel 8, 1897-1909 (1994)

52. Sato T. N., Y. Tozawa, U. Deutsch, K. Wolburg-Buchholz, Y. Fujiwara, M. Gendron-Maguire, T. Gridley, H. Wolburg, W. Risau & Y. Qin: Distinct roles of the receptor tyrosine kinases Tie-1 and Tie-2 in blood vessel formation. Nature 376, 70-74 (1995)

53. Belperio J. A., M. P. Keane, D. A. Arenberg, C. L. Addison, J. E. Ehlert, M. D. Burdick & R. M. Strieter: CXC chemokines in angiogenesis. J Leukocyte Biol 68, 1-8 (2000)

54. Strieter R. M., P. J. Polverini, S. L. Kunkel, D. A. Arenberg, M. D. Burdick, J. Kasper, J. Dzuiba, J. Van Damme, A. Walz & D. Marriott: The functional role of the ELR motif in CXC chemokine-mediated angiogenesis. J Biol Chem 270, 27348-27357 (1995)

55. Norrby K.: Interleukin-8 and de novo mammalian angiogenesis. Cell Prolif 29, 315-323 (1996)

56. Arenberg D. A., S. L. Kunkel, P. J. Polverini, M. Glass, M. D. Burdick & R. M. Strieter: Inhibition of interleukin-8 reduces tumorigenesis of human non-small cell lung cancer in SCID mice. J Clin Investig 97, 2792-2802 (1996)

57. Addison C. L., T. O. Daniel, M. D. Burdick, H. Liu, J. E. Ehlert, Y. Y. Xue, L. Buechi, A. Walz, A. Richmond & R. M. Strieter: The CXC chemokine receptor 2, CXCR2, is the putative receptor for ELR+ CXC chemokine-induced angiogenic activity. J Immunol 165, 5269-5277 (2000)

58. Devalaraja R. M., L. B. Nanney, J. Du, Q. Qian, Y. Yu, M. N. Devalaraja & A. Richmond: Delayed wound healing in CXCR2 knockout mice. J Investig Derm 115, 234-244 (2000)

59. Sgadari C., J. M. Farber, A. L. Angiolillo, R. Liao, J. Teruya-Feldstein, P. R Burd, L. Yao, G. Gupta, C. Kanegane & G. Tosato: MIG, the monokine induced by interferon-gamma, promotes tumor necrosis in vivo. Blood 89, 2635-2643 (1997)

60. Teruya-Feldstein J., E. S. Jaffe, P. R Burd, H. Kanegame, D. W. Kingma, W. H. Wilson, D. L. Longo & G. Tosato: The role of MIG, the monokine induced by interferon-gamma, and IP-10, the interferon-gamma-inducible protein-10, in tissue necrosis and vascular damage associated with Epstein-Barr virus-positive lymphoproliferative disease. Blood 90, 4099-4105 (1997)

61. Gengrinovitch S., S. M. Greenberg, T. Cohen, H. Gitay-Goren, P. Rockwell, T. E. Maione, B. Z. Levi & G. Neufeld: Platelet factor-4 inhibits the mitogenic activity of VEGF121 and VEGF165 using several concurrent mechanisms. J Biol Chem 270, 15059-15065 (1995)

62. Perollet C., Z. C. Han, C. Savona, J. P. Caen & A. Bikfalvi: Platelet factor 4 modulates fibroblast growth factor 2 (FGF-2) activity and inhibits FGF-2 dimerization. Blood 91, 3289-3299 (1998)

63. Arenberg D. A., P. J. Polverini, S. L. Kunkel, A. Shanafelt & R. M. Strieter: In vitro and in vivo systems to assess role of C-X-C chemokines in regulation of angiogenesis. Methods Enzymol 288, 190-220 (1997)

64. Salcedo R., K. Wasserman, H. A. Young, M. C. Grimm, O. M. Howard, M. R. Anver, H. K. Kleinman, W. J. Murphy & J. J. Oppenheim: Vascular endothelial growth factor and basic fibroblast growth factor induce expression of CXCR4 on human endothelial cells: In vivo neovascularization induced by stromal-derived factor-1alpha. Am J Pathol 154, 1125-1135 (1999)

65. Tachibana K., S. Hirota, H. Iizasa, H. Yoshida, K. Kawabata, Y. Kataoka, Y. Kitamura, K. Matsushima, N. Yoshida, S. Nishikawa, T. Kishimoto & T. Nagasawa: The chemokine receptor CXCR4 is essential for vascularization of the gastrointestinal tract. Nature 393, 591-594 (1998)

66. Schwarz M. A., J. Kandel, J. Brett, J. Li, J. Hayward, R. E. Schwarz, O. Chappey, J. L. Wautier, J. Chabot, P. Lo Gerfo & D. Stern: Endothelial-monocyte activating polypeptide II, a novel antitumor cytokine that suppresses primary and metastatic tumor growth and induces apoptosis in growing endothelial cells. J Exp Med 190, 341-354 (1999)

67. Schwarz M. A., F. Zhang, S. Gebb, V. Starnes & D. Warburton: Endothelial monocyte activating polypeptide II inhibits lung neovascularization and airway epithelial morphogenesis. Mech Devel 95, 123-132 (2000)

68. deMello D. E., D. Sawyer, N. Galvin & L. M. Reid: Early fetal development of lung vasculature. Am J Respir Cell Mol Biol 16, 568-581 (1997)

69. deMello D. E. & L. M. Reid: Embryonic and early fetal development of human lung vasculature and its functional implications. Pediatr Devel Pathol 3, 439-449 (2000)

70. Burri P. H.: Fetal and postnatal development of the lung. Ann Rev Physiol 46, 617-628 (1984)

71. Cheung C. Y.: Vascular endothelial growth factor - possible role in fetal development and placental function. J Soc Gynecol Invest 4, 169-177 (1997)

72. Maniscalco W. M., R. H. Watkins, C. T. D'Angio & R. M. Ryan: Hyperoxic injury decreases alveolar epithelial cell expression of vascular endothelial growth factor (VEGF) in neonatal rabbit lung. Am J Respir Cell Mol Biol 16, 557-567 (1997)

73. Shifren J. L., N. Doldi, N. Ferrara, S. Mesiano & R. B. Jaffe: In the human fetus, vascular endothelial growth factor is expressed in epithelial cells and myocytes, but not vascular endothelium: implications for mode of action. J Clin Endocr Metab 79, 316-322 (1994)

74. Cheung C. Y. & R. A. Brace: Ovine vascular endothelial growth factor: nucleotide sequence and expression in fetal tissues. Growth Factors 16, 11-22 (1998)

75. Lagercrantz J., F. Farnebo, C. Larsson, T. Tvrdik, G. Weber & F. Piehl: A comparative study of the expression patterns for vegf, vegf-b/vrf and vegf-c in the developing and adult mouse. Biochimica et Biophysica Acta 1398, 157-163 (1998)

76. Farnebo F., F. Piehl & J. Lagercrantz: Restricted expression pattern of vegf-d in the adult and fetal mouse: high expression in the embryonic lung. Biochem Biophys Res Communic 257, 891-894 (1999)

77. Bhatt A. J., S. B. Amin, P. R. Chess, R. H. Watkins & W. M. Maniscalco: Expression of vascular endothelial growth factor and Flk-1 in developing and glucocorticoid-treated mouse lung. Pediatr Res 47, 606-613 (2000)

78. Maniscalco W. M., R. H. Watkins, G. S. Pryhuber, A. Bhatt, C. Shea & H. Huyck: Angiogenic factors and the alveolar vasculature: development and alterations by injury in very premature baboons. Am J Physiol 282, L811-23 (2002)

79. Watkins R. H., C. T. D'Angio, R. M. Ryan, A. Patel & W. M. Maniscalco: Differential expression of VEGF mRNA splice variants in newborn and adult hyperoxic lung injury. Am J Physiol 276, L858-L867 (1999)

80. Healy A. M., L. Morgenthau, X. Zhu, H. W. Farber & W. V. Cardoso: VEGF is deposited in the subepithelial matrix at the leading edge of branching airways and stimulates neovascularization in the murine embryonic lung. Devel Dynam 219, 341-352 (2000)

81. Gebb S. A. & J. M. Shannon: Tissue interactions mediate early events in pulmonary vasculogenesis. Devel Dynam 217, 159-169 (2000)

82. Maniscalco W. M., R. H. Watkins, J. N. Finkelstein & M. H. Campbell: Vascular endothelial growth factor mRNA increases in alveolar epithelial cells during recovery from oxygen injury. Am J Respir Cell Mol Biol 13, 377-386 (1995)

83. Monacci W. T., M. J. Merrill & E. H. Oldfield: Expression of vascular permeability factor/vascular endothelial growth factor in normal rat tissues. Am J Physiol 264, C995-1002 (1993)

84. Acarregui M. J., S. T. Penisten, K. L. Goss, K. Ramirez & J. M. Snyder: Vascular endothelial growth factor gene expression in human fetal lung in vitro. Am J Respir Cell Mol Biol 20, 14-23 (1999)

85. D'Angio C. T., W. M. Maniscalco, R. M. Ryan, N. E. Avissar, K. Basavegowda & R. A. Sinkin: Vascular endothelial growth factor in pulmonary lavage fluid from premature infants: effects of age and postnatal dexamethasone. Biol Neonate 76, 266-273 (1999)

86. Shehata S. M., W. J. Mooi, T. Okazaki, I. El-Banna, H. S. Sharma & D. Tibboel: Enhanced expression of vascular endothelial growth factor in lungs of newborn infants with congenital diaphragmatic hernia and pulmonary hypertension. Thorax 54, 427-431 (1999)

87. Colen K. L., C. A. Crisera, M. I. Rose, P. R. Connelly, M. T. Longaker & G. K. Gittes: Vascular development in the mouse embryonic pancreas and lung. J Pediatr Surg 34, 781-785 (1999)

88. Park M. S. & H. M. Wallace: Hyperoxia influences mRNA expression of cytokines in cultured human umbilical vein endothelial cells. Yonsei Med J 39, 1-12 (1998)

89. Fogg S., A. Agarwal, H. S. Nick & G. A. Visner: Iron regulates hyperoxia-dependent human heme oxygenase 1 gene expression in pulmonary endothelial cells. Am J Respir Cell Mol Biol 20, 797-804 (1999)

90. Suzuki Y., K. Nishio, K. Takeshita, O. Takeuchi, K. Watanabe, N. Sato, K. Naoki, H. Kudo, T. Aoki & K. Yamaguchi: Effect of steroid on hyperoxia-induced ICAM-1 expression in pulmonary endothelial cells. Am J Physiol: Lung Cell Mol Physiol 278, L245-L252 (2000)

91. Barazzone C., Y. R. Donati, A. F. Rochat, C. Vesin, C. D. Kan, J. C. Pache & P. F. Piguet: Keratinocyte growth factor protects alveolar epithelium and endothelium from oxygen-induced injury in mice. Am J Pathol 154, 1479-1487 (1999)

92. Klekamp J. G., K. Jarzecka & E. A. Perkett: Exposure to hyperoxia decreases the expression of vascular endothelial growth factor and its receptors in adult rat lungs. Am J Pathol 154, 823-831 (1999)

93. Crapo J. D., B. E. Barry, H. A. Foscue & J. Shelburne: Structural and biochemical changes in rat lungs occurring during exposures to lethal and adaptive doses of oxygen. Am Rev Respir Dis 122, 123-143 (1980)

94. Johnston C. J., J. N. Finkelstein, R. Gelein, R. Baggs & G. Oberdorster: Characterization of the early pulmonary inflammatory response associated with PTFE fume exposure. Toxicol Appl Pharmacol 140, 154-163 (1996)

95. Corne J., G. Chupp, C. G. Lee, R. J. Homer, Z. Zhu, Q. Chen, B. Ma, Y. Du, F. Roux, J. McArdle, A. B. Waxman & J. A. Elias: IL-13 stimulates vascular endothelial cell growth factor and protects against hyperoxic acute lung injury. J Clin Investig 106, 783-791 (2000)

96. Lassus P., A. Ristimaki, O. Ylikorkala, L. Viinikka & S. Andersson: Vascular endothelial growth factor in human preterm lung. Am J Respir Crit Care Med 159, 1429-1433 (1999)

97. Perkett E. A., R. B. Cotton & U. Lappalainen: Vascular endothelial growth factor (VEGF) increases in tracheobronchial aspirates from intubated low birthweight infants during the first 10 days of life [abstract]. Pediatr Res 43, 293A (1998)

98. Sinkin R. A., H. S. Dweck, M. J. Horgan, K. J. Gallaher, C. Cox, W. M. Maniscalco, P. R. Chess, C. T. D'Angio, R. Guillet, J. W. Kendig, R. M. Ryan & D. L. Phelps: Early dexamethasone-attempting to prevent chronic lung disease. Pediatr 105, 542-548 (2000)

99. Klekamp J. G., K. Jarzecka, R. L. Hoover, M. L. Summar, N. Redmond & E. A. Perkett: Vascular endothelial growth factor is expressed in ovine pulmonary vascular smooth muscle cells in vitro and regulated by hypoxia and dexamethasone. Pediatr Res 42, 744-749 (1997)

100. Finkenzeller G., A. Technau & D. Marme: Hypoxia-induced transcription of the vascular endothelial growth factor gene is independent of functional AP-1 transcription factor. Biochem Biophys Res Communic 208, 432-439 (1995)

101. Keane M. P., D. A. Arenberg, J. P. Lynch, 3rd, R. I. Whyte, M. D. Iannettoni, M. D. Burdick, C. A. Wilke, S. B. Morris, M. C. Glass, B. DiGiovine, S. L. Kunkel & R. M. Strieter: The CXC chemokines, IL-8 and IP-10, regulate angiogenic activity in idiopathic pulmonary fibrosis. J Immunol 159, 1437-1443 (1997)

102. D'Angio C. T., J. N. Finkelstein, M. B. Lomonaco, A. Paxhia, S. A. Wright, R. B. Baggs, R. H. Notter & R. M. Ryan: Changes in surfactant protein gene expression in a neonatal rabbit model of hyperoxia-induced fibrosis. Am J Physiol: Lung Cell Mol Physiol 16, L720-L730 (1997)

103. D'Angio C. T., M. B. LoMonaco, S. A. Chaudhry, A. Paxhia & R. M. Ryan: Discordant pulmonary proinflammatory cytokine expression during acute hyperoxia in the newborn rabbit. Exp Lung Res 25, 443-465 (1999)

104. Deng H., S. N. Mason & R. L. Auten, Jr.: Lung inflammation in hyperoxia can be prevented by antichemokine treatment in newborn rats. Am J Respir Crit Care Med 162, 2316-2323 (2000)

105. Auten R. L., Jr., S. N. Mason, D. T. Tanaka, K. Welty-Wolf & M. H. Whorton: Anti-neutrophil chemokine preserves alveolar development in hyperoxia-exposed newborn rats. Am J Physiol: Lung Cell Mol Physiol 281, L336-L344 (2001)

106. Auten R. L., R. M. Richardson, J. R. White, S. N. Mason, M. A. Vozzelli & M. H. Whorton: Nonpeptide CXCR2 antagonist prevents neutrophil accumulation in hyperoxia-exposed newborn rats. J Pharm Exp Ther 299, 90-95 (2001)

107. Groneck P., B. Gotze-Speer, M. Oppermann, H. Eiffert & C. P. Speer: Association of pulmonary inflammation and increased microvascular permeability during the development of bronchopulmonary dysplasia: a sequential analysis of inflammatory mediators in respiratory fluids of high-risk preterm neonates. Pediatr 93, 712-718 (1994)

108. Kotecha S., B. Chan, N. Azam, M. Silverman & R. J. Shaw: Increase in interleukin-8 and soluble intercellular adhesion molecule-1 in bronchoalveolar lavage fluid from premature infants who develop chronic lung disease. Arch Dis Child Fetal Neonat Ed 72, F90-F96 (1995)

109. McColm J. R. & N. McIntosh: Interleukin-8 in bronchoalveolar lavage samples as predictor of chronic lung disease in premature infants [letter]. Lancet 343, 729 (1994)

110. Takasaki J. & Y. Ogawa: Interleukin 8 and granulocyte elastase in the tracheobronchial aspirate of infants without respiratory distress syndrome or intrauterine infection and development of chronic lung disease. Acta Paediatrica Japonica 39, 437-441 (1997)

111. Little S., T. Dean, S. Bevin, M. Hall, M. Ashton, M. Church, J. Warner & J. Shute: Role of elevated plasma soluble ICAM-1 and bronchial lavage fluid IL-8 levels as markers of chronic lung disease in premature infants. Thorax 50, 1073-1079 (1995)

112. Buron E., J. A. Garrote, E. Arranz, P. Oyaguez, J. L. Fernandez Calvo & A. Blanco Quiros: Markers of pulmonary inflammation in tracheobronchial fluid of premature infants with respiratory distress syndrome. Allergologia et Immunopathologia 27, 11-17 (1999)

113. D'Angio C. T., K. Basawegowda, N. E. Avissar, J. N. Finkelstein & R. A. Sinkin: Comparison of tracheal aspirate and bronchoalveolar lavage specimens from premature infants. Biol Neonate in press (2002)

114. Panos R. J., P. M. Bak, W. S. Simonet, J. S. Rubin & L. J. Smith: Intratracheal instillation of keratinocyte growth factor decreases hyperoxia-induced mortality in rats. J Clin Investig 96, 2026-2033 (1995)

115. Halloran M. M., C. J. Haskell, J. M. Woods, S. Hosaka & A. E. Koch: Interleukin-13 is an endothelial chemotaxin. Pathobiology 65, 287-292 (1997)

Key Words: Endothelial Growth Factors, Angiogenesis Factor, Vascular Endothelial Growth Factor, Lung Growth and Development, Animal Disease Models, Bronchopulmonary Dysplasia, Hyperoxia, Review

Send correspondence to: Carl T. D' Angio, MD, Box 651, Neonatology, Golisano Children's Hospital at Strong, University of Rochester, 601 Elmwood Avenue, Rochester NY 14642, Tel: 585-273-4911, Fax: 585-461-3614, Email: carl_dangio@urmc.rochester.edu