Oxidative protein folding in the mammalian endoplasmic reticulum.

Native disulphide bonds are essential for the structure and function of many membrane and secretory proteins. Disulphide bonds are formed, reduced and isomerized in the endoplasmic reticulum of mammalian cells by a family of oxidoreductases, which includes protein disulphide isomerase (PDI), ERp57, ERp72, P5 and PDIR. This review will discuss how these enzymes are maintained in either an oxidized redox state that allows them to form disulphide bonds in substrate proteins or a reduced form that allows them to perform isomerization and reduction reactions, how these opposing pathways may co-exist within the same compartment and why so many oxidoreductases exist when PDI alone can perform all three of these functions.

Disrupted pulmonary vasculature and decreased vascular endothelial growth factor, Flt-1, and TIE-2 in human infants dying with bronchopulmonary dysplasia.

An abnormal pulmonary vasculature may be an important component of bronchopulmonary dysplasia (BPD). We examined human infant lung for the endothelial cell marker PECAM-1 and for angiogenic factors and their receptors. Lung specimens were collected prospectively at approximately 6 h after death. The right middle lobe was inflation fixed and part of the right lower lobe was flash frozen. We compared lungs from infants dying with BPD (n = 5) with lungs from infants dying from nonpulmonary causes (n = 5). The BPD group was significantly more premature and had more days of ventilator and supplemental oxygen support, but died at a postconceptional age similar to control infants. PECAM-1 protein and mRNA were decreased in the BPD group. PECAM-1 immunohistochemistry showed the BPD group had decreased staining intensity and abnormal distribution of alveolar capillaries. The dysmorphic capillaries were frequently in the interior of thickened alveolar septa. The BPD group had decreased vascular endothelial growth factor (VEGF) mRNA and decreased VEGF immunostaining, compared with infants without BPD. Messages for the angiogenic receptors Flt-1 and TIE-2 were decreased in the BPD group. We conclude that infants dying with BPD have abnormal alveolar microvessels and that disordered expression of angiogenic growth factors and their receptors may contribute to these abnormalities.

The cyclin-dependent kinase inhibitor p21 protects the lung from oxidative stress.

The lung is a major target tissue for oxidative stress, including hyperoxia used to relieve tissue hypoxia. Unfortunately, severe hyperoxia damages DNA, inhibits proliferation, and kills cells, resulting in morbidity and mortality. Although hyperoxia induces the tumor suppressor p53 and its downstream target, the cyclin-dependent kinase inhibitor p21(Cip1/WAF1/Sdi1) (p21), their role in pulmonary injury remains unknown. Using p53- and p21-deficient mice we demonstrate that hyperoxia induces p21 in the absence of p53, suggesting that previous conclusions that p53 does not modify hyperoxic lung injury cannot be extrapolated to p21. In fact, mean survival of p21-deficient mice decreased by 40% and was associated with terminal deoxyribonucleotidyl transferase-mediated deoxyuridine triphosphate-biotin nick-end labeling staining of alveolar debris, indicative of DNA fragmentation and cell death. Ultrastructural analyses revealed that alveolar endothelial and type I epithelial cells died rapidly by necrosis. Although hyperoxia decreased DNA replication in p21-wild-type lungs, it had no effect on replication in p21-deficient lungs. Our findings suggest that p21 protects the lung from oxidative stress, in part, by inhibiting DNA replication and thereby allowing additional time to repair damaged DNA. Our findings have implications for patients suffering from the toxic effects of supplemental oxygen therapies.

Maturation of carbonic anhydrase IV expression in rabbit kidney.

Carbonic anhydrase (CA) IV facilitates renal acidification by catalyzing the dehydration of luminal H(2)CO(3). CA IV is expressed in proximal tubules, medullary collecting ducts, and A-intercalated cells of the mature rabbit kidney (Schwartz GJ, Kittelberger AM, Barnhart DA, and Vijayakumar S. Am J Physiol 278: F894-F904, 2000). In view of the maturation of HCO transport in the proximal tubule and collecting duct, the ontogeny of CA IV expression was examined. During the first 2 wk, CA IV mRNA was expressed in maturing cortex and medulla at ~20% of adult levels. The maturational increase was gradual in cortex over 3-5 wk of age but surged in the medulla, so that mRNA levels appeared higher than those in the adult medulla. In situ hybridization showed very little CA IV mRNA at 5 days, with increases in deep cortex and medullary collecting ducts by 21 days. Expression of CA IV protein in the cortex and medulla was minimal at 3 days of age but then apparent in the juxtamedullary region, A-intercalated cells and medullary collecting ducts by 18 days; there was little labeling of the proximal straight tubules of the medullary rays. Thus CA IV expression may be regulated to accommodate the maturational increase in HCO absorption in the proximal tubule. In the medullary collecting duct, there is a more robust maturation of CA IV mRNA and protein, commensurate with the high rate of HCO absorption in the neonatal segment.

Expression of vascular endothelial growth factor and Flk-1 in developing and glucocorticoid-treated mouse lung.

Although the endothelial cell is the most abundant cell type in the differentiated lung, little is known about regulation of lung developmental vasculogenesis. Vascular endothelial growth factor (VEGF) is an endothelial cell mitogen and angiogenic factor that has putative roles in vascular development. Mitogenic actions of VEGF are mediated by the tyrosine kinase receptor KDR/murine homologue fetal liver kinase Flk-1. HLF (hypoxia-inducible factor-like factor) is a transcription factor that increases VEGF gene transcription. Dexamethasone augments lung maturation in fetal and postnatal animals. However, in vitro studies suggest that dexamethasone blocks induction of VEGF. The objectives for the current study were to measure VEGF mRNA and Flk-1 mRNA in developing mouse lung and to measure the effects of dexamethasone treatment in vivo on VEGF and Flk-1 in newborn mouse lung. Our results show that VEGF and Flk-1 messages increase in parallel during normal lung development (d 13 embryonic to adult) and that the distal epithelium expresses VEGF mRNA at all ages examined. Dexamethasone (0.1-5.0 mg x kg(-1) x d(-1)) treatment of 6-d-old mice resulted in significantly increased VEGF, HLF, and Flk-1 mRNA. Dexamethasone did not affect cell-specific expression of VEGF, VEGF protein, or proportions of VEGF mRNA splice variants. These data suggest that the developing alveolar epithelium has an important role in regulating alveolar capillary development. In addition, unlike effects on cultured cells, dexamethasone, even in relatively high doses, did not adversely affect VEGF expression in vivo. The relatively high levels of VEGF and Flk-1 mRNA in adult lung imply a role for pulmonary VEGF in endothelial cell maintenance or capillary permeability.

p53-independent induction of GADD45 and GADD153 in mouse lungs exposed to hyperoxia.

Previous studies have shown that lungs of adult mice exposed to >95% oxygen have increased terminal deoxyribonucleotidyltransferase dUTP nick end-label staining and accumulate p53, the expression of which increases in cells exposed to DNA-damaging agents. The present study was designed to determine whether hyperoxia also increased expression of the growth arrest and DNA damage (GADD) gene 45 and GADD153, which are induced by genotoxic stress through p53-dependent and -independent pathways. GADD proteins have been shown to inhibit proliferation and stimulate DNA repair and/or apoptosis. GADD45 and GADD153 mRNAs were not detected in lungs exposed to room air but were detected after 48 and 72 h of exposure to hyperoxia. In situ hybridization and immunohistochemistry revealed that hyperoxia increased GADD45 and GADD153 expression in the bronchiolar epithelium and GADD45 expression predominantly in alveolar cells that were morphologically consistent with type II cells. Hyperoxia also increased GADD expression in p53-deficient mice. Terminal deoxyribonucleotidyltransferase dUTP nick end-label staining of lung cells from p53 wild-type and p53-null mice exposed to hyperoxia for 48 h revealed that hyperoxia-induced DNA fragmentation was not modified by p53 deficiency. These studies are consistent with the hypothesis that hyperoxia-induced DNA fragmentation is associated with the expression of GADD genes that may participate in DNA repair and/or apoptosis.

Bcl-2 family gene expression during severe hyperoxia induced lung injury.

Exposure of the lung to severe hyperoxia induces terminal transferase dUTP end-labeling (TUNEL) indicative of DNA damage or apoptosis and increases expression of the tumor suppressor p53 and of members of the Bcl-2 gene family. Because cell survival and apoptosis are regulated, in part, by the relative abundance of proteins of the Bcl-2 family, we hypothesized that lung cells dying during exposure would show increased expression of pro-apoptotic members, such as Bax, whereas surviving cells would have increased expression of anti-apoptotic members, such as Bcl-X(L). The hypothesis is tested in the current study by determining which Bcl-2 genes are regulated by hyperoxia, with specific focus on correlating expression of Bax and Bcl-X(L) with morphologic evidence of apoptosis or necrosis. Adult mice exposed to greater than 95% oxygen concentrations for 48 to 88 hours had increased whole-lung mRNA levels of Bax and Bcl-X(L), no change in Bak, Bad, or Bcl-2, and decreased levels of Bcl-w and Bfl-1. In situ hybridization revealed that hyperoxia induced Bax and Bcl-X(L) mRNA in uniform and overlapping patterns of expression throughout terminal bronchioles and parenchyma, coinciding with TUNEL staining. Electron microscopy and DNA electrophoresis, however, suggested relatively little classical apoptosis. Unexpectedly, Western analysis demonstrated increased Bcl-X(L), but not Bax, protein in response to hyperoxia. Bax and Bfl-1 were not altered by hyperoxia in p53 null mice; however, oxygen toxicity was not lessened by p53 deficiency. These findings suggest that oxygen-induced lung injury does not depend on the relative expression of these Bcl-2 members.

Carotenoid composition of marigold (Tagetes erecta) flower extract used as nutritional supplement.

Commercially prepared marigold flower (Tagetes erecta) extract was saponified and analyzed for carotenoid composition. HPLC analyses were performed on two normal-phase columns (beta-Cyclobond and silica) and on a C(30) reversed-phase column. The extract contained 93% utilizable pigments (detected at 450 nm), consisting of all-trans and cis isomers of zeaxanthin (5%), all-trans and cis isomers of lutein, and lutein esters (88%). All were identified by chromatographic retention, UV-visible spectra, and positive ion electrospray mass spectrometry in comparison to authentic standards. Contrary to previous findings, insignificant levels (<0.3%) of lutein oxidation products were detected in the saponified extract. This compositional determination is important for the application of marigold extract in nutritional supplements and increases its value as a poultry feed colorant because it contains more biologically useful lutein compounds than previously believed.

Differential expression of VEGF mRNA splice variants in newborn and adult hyperoxic lung injury.

Lung development and repair of hyperoxic injury require closely regulated growth and regeneration of alveolar capillaries. Vascular endothelial growth factor (VEGF), a mitogen for endothelial cells, is expressed by alveolar epithelial cells. Alternative splicing of VEGF mRNA results in isoforms of varying mitogenicity and solubility. We examined changes in the proportions of the VEGF splice variant mRNAs in rabbit lung development and in control, oxygen-injured, and recovering newborn and adult rabbit lungs. The proportion of the 189-amino acid VEGF mRNA, which codes for an isoform that binds to the extracellular matrix, increased fivefold during development (from 8% of total VEGF message at 22 days gestation to 40% in 10-day newborn lungs; P < 0.001). During neonatal oxygen injury, its expression declined from 38 to 8% of VEGF message (P < 0.002) and returned to the control value in recovery. A similar pattern was observed in adults. VEGF protein in lung lavage fluid increased slightly during hyperoxia, declined to barely detectable levels at the 50% lethal dose time point, and increased 10-fold (newborn) or up to 40-fold (adult) in recovering animals. We conclude that alternative splicing may have important roles in the regulation of VEGF activity in developing and injured lungs.

Rat primary hepatocytes and H4 hepatoma cells display differential sensitivity to cyclic AMP at the level of expression of tyrosine aminotransferase.

We have shown that the sensitivity of isolated hepatocytes and H4 hepatoma cells to cyclic AMP is different. In terms of activation of tyrosine aminotransferase at mRNA and activity level in response to cyclic AMP, isolated hepatocytes are 10-fold more sensitive. Hepatocytes and H4 hepatoma cells show similar sensitivities to cyclic AMP at the level of protein kinase A activation and phosphorylation of cyclic AMP response element binding protein (CREB) and the differential sensitivity must reside at other sites. The consequences of these findings to studies of control phenomena at the transcriptional level is discussed.

Accumulation of p21(Cip1/WAF1) during hyperoxic lung injury in mice.

Hyperoxic lung injury results in decreased cell proliferation, DNA damage, and cell death. Because the cyclin-dependent kinase inhibitor p21(Cip1/WAF1) (p21) inhibits cell proliferation in G1/S, enhances DNA repair, and regulates apoptosis in some cells, we hypothesized that the expression of p21 would increase in lungs of C57Bl/6J male mice exposed to and recovered from > 95% oxygen. A low level of p21 messenger RNA (mRNA) expression was detected by Northern blot analysis of room air-exposed lungs. Exposure to hyperoxia resulted in a modest increase in p21 mRNA expression by 24 h, followed by a marked induction by 48 to 72 h. In situ hybridization revealed that p21 mRNA abundance increased in bronchiolar epithelium and in resident alveolar cells, but not in smooth-muscle cells or large airway epithelium. Hyperoxia increased the expression of p21 protein by 24 h and continued to increase at 48 and 72 h. Immunohistochemical staining showed that p21 protein accumulated in the bronchiolar epithelium and in alveolar regions that had increased p21 mRNA expression. In contrast, the expression of the cyclin-dependent kinase inhibitor p27(Kip1) was not altered by hyperoxia. To determine whether p21 expression was altered during the repair process, mice were exposed to hyperoxia for 64 h and allowed to recover for up to 4 d in room air. The abundance of p21 mRNA and protein decreased by 1 to 2 d of recovery and returned to room air-exposed levels by 3 to 4 d of recovery. These findings support the concept that bronchiolar epithelial and alveolar cells damaged by hyperoxia express molecules such as p21, which may participate in regulating cell proliferation, DNA repair, and cell death.

Interleukin-1 expression during hyperoxic lung injury in the mouse.

An important component of the pathophysiologic response to hyperoxia (O2) is pulmonary inflammation, although the roles of specific inflammatory mediators during pulmonary O2 toxicity are not completely known. Interleukin-1 (IL-1) is an early inflammatory mediator and is sufficient to elicit many of the responses associated with acute injury. The IL-1 family comprises two bioactive proteins, IL-1alpha and IL-1beta, and their natural antagonist IL-1ra. Here we report studies of IL-1 regulation during hyperoxic lung injury in the adult mouse. When assayed by Northern blot, increases in IL-1beta mRNA were seen after 2 days of hyperoxia. In contrast, IL-1alpha mRNA was barely detectable before 4 days of hyperoxia. To further understand the cellular origin of IL-1beta expression in lungs, in situ hybridization and immunohistochemical analyses were performed. IL-1beta mRNA or protein was not detected in the lungs of unexposed animals. At 3 days, we observed the accumulation of IL-1beta transcripts in pulmonary interstitial macrophages and in a subset of neutrophils, and immunodetectable IL-1beta protein was co-localized in adjacent sections. At 4 days of exposure, IL-1beta transcripts were widespread in lung tissue, but many areas rich in IL-1beta mRNA were devoid of immunodetectable IL-1beta. However, it is not known whether increased synthesis of IL-1beta or the uncoupling of IL-1beta protein and mRNA accumulation has a role in pathophysiology of pulmonary O2 toxicity.

Cell-specific expression of fibronectin and EIIIA and EIIIB splice variants after oxygen injury.

Cellular fibronectin (cFN) expression is characteristic of injured tissues. Unlike plasma FN, cFN mRNA often contains the EIIIA or EIIIB domains. We examined the lung cell-specific expression of total cFN mRNA and the EIIIA and EIIIB splice variants in rabbits after acute oxygen injury. By in situ hybridization, control lung had low cFN mRNA. After exposure to > 95% oxygen, mRNAs for total cFN and EIIIA were noted primarily in alveolar macrophages and large-vessel endothelial cells. By 3-5 days recovery, cFN and EIIIA mRNA abundance was increased in alveolar septal cells (i.e., alveolar epithelial, interstitial, or endothelial cells) and in some large-vessel endothelial cells but was low in bronchial epithelial cells. During recovery, EIIIB mRNA was low in alveolar septal cells but was noted mainly in chondrocytes. Immunostaining for EIIIA increased during recovery, paralleling the in situ hybridizations. Because FN may modulate alveolar type II cell phenotype, we investigated type II cell cFN mRNA expression in vivo. During recovery, neither isolated type II cells nor cells with surfactant protein C mRNA in vivo contained FN mRNA. In summary, these data suggest that cFN with the EIIIA domain has a role in alveolar cell recovery from oxygen injury and that type II cells do not express cFN during recovery.

Bixin and norbixin in human plasma: determination and study of the absorption of a single dose of Annatto food color.

A procedure was developed for the detection and determination of bixin and norbixin in human plasma by reversed-phase HPLC with a sensitivity limit of 5 micrograms l-1. A group of seven volunteers ingested a single dose of 1 ml of a commercial Annatto Food Color (16 mg of cis-bixin in soybean oil). The presence of bixin (cis and trans) and norbixin (cis and trans) was demonstrated in the plasma at average levels of 11.6, 10.1, 2.8 and 0 micrograms l-1 of bixin and 48, 58, 53 and 29 micrograms l-1 of norbixin after 2, 4, 6 and 8 h, respectively. Considerable individual variations were observed. Complete plasma clearance generally occurred for bixin by 8 h and for norbixin by 24 h after ingestion of cis-bixin.

Hyperoxic injury decreases alveolar epithelial cell expression of vascular endothelial growth factor (VEGF) in neonatal rabbit lung.

Normal neonatal lung growth requires a substantial increase in microvascular endothelial cells. Oxygen injury to neonatal lung destroys endothelial cells and alters the normal process of alveolarization, including development of the microvasculature. The mechanisms that regulate lung alveolar capillary growth and development are not known. Vascular endothelial growth factor (VEGF) is a specific mitogen for endothelial cells that is often expressed by epithelial cells in close proximity to capillary beds. VEGF expression is induced by hypoxia and may be inhibited by hyperoxia. We examined the cell-specific expression of VEGF during normal postnatal lung development and the effects of hyperoxic lung injury on VEGF mRNA and protein in vivo. Normal newborn rabbits between 1 day and 5 wk of age had VEGF transcripts located mainly in alveolar epithelial cells, with little or no VEGF mRNA noted in smooth muscle or endothelial cells. A subpopulation of freshly isolated, normal type II cells, but not mesenchymal cells, expressed VEGF mRNA. Newborn rabbits exposed to 100% oxygen for 4 days had no change in VEGF mRNA abundance, transcript location, or immunostaining. Animals exposed to 100% oxygen for an average of 9 days had an 80% decrease in lung VEGF mRNA abundance, decreased alveolar epithelial cell VEGF expression, and decreased VEGF immunostaining. Recovery of VEGF expression to control levels occurred during a 5-day recovery period. We conclude that alveolar epithelial cells in postnatal lung express VEGF, suggesting epithelial regulation of alveolar capillary formation. Furthermore, hyperoxic injury decreases neonatal lung VEGF mRNA and protein, which may be a contributory mechanism of impaired postnatal microvascular development in oxygen injury.

Changes in decorin expression with hyperoxic injury to developing rat lung.

Proteoglycans are extracellular matrix components that appear to play important roles in lung development and in the response to injury. Decorin, a small extracellular matrix-associated proteoglycan, is known to be involved in collagen fibrillogenesis and is a likely participant in the pathogenesis of lung injury. We hypothesized that chronic exposure of the developing lung to hyperoxia would result in temporal and spatial changes in decorin expression. To determine the expression of decorin in normal and oxygen-injured lung, newborn rats were exposed to hyperoxia for 6 wk. Decorin mRNA abundance was determined using Northern hybridization analyses, and decorin expression was localized by in situ hybridization and immunohistochemistry. Decorin mRNA expression in type II pneumocytes was studied using reverse transcription-polymerase chain reaction. Oxygen exposure is associated with a 77% reduction in decorin mRNA in whole lung and a decrease in decorin immunoreactivity in connective tissues surrounding large airways and blood vessels, but an increase in decorin mRNA and protein expression at the tips of alveolar septa. Studies using isolated cells indicate that macrophages and polymorphonuclear neutrophils contain decorin core protein but not decorin mRNA. Type II pneumocytes do not contain either decorin mRNA or core protein. These findings demonstrate that hyperoxic lung injury is associated with localized changes in decorin expression, changes that are not reflected in whole lung RNA studies. It is likely that regional changes in lung decorin expression are influenced by factors produced and acting locally, and that such changes may contribute to the morphologic alterations characteristic of oxygen-induced lung injury.

Expression of fibronectin mRNA splice variants by rabbit lung in vivo and by alveolar type II cells in vitro.

Fibronectin (FN) is a multidomain glycoprotein with putative functions in tissue development and repair. In repair of alveolar injury, FN may promote the transition of type II epithelial cells to type I epithelial cells. Alternative splicing of FN mRNA, including the EIIIA and EIIIB exons, results in protein isoforms that have cell, tissue, and developmental specificity. The present work found that FN mRNA with the EIIIA exon was in fetal, adult, and oxidant-injured lung. The EIIIB splice variant, however, was restricted to fetal lung and adult lung recovering from oxidant injury. Because alveolar type II cells in vitro express FN, we examined the splice variants in two conditions that induce FN [transforming growth factor-beta 1 (TGF-beta 1) treatment and time in culture]. TGF-beta 1 increased both EIIIA and EIIIB mRNA abundance by 10-fold. Increased EIIIA isoform immunostaining was also noted. Type II cells that spontaneously express FN at 72 h in vitro had increased EIIIA and EIIIB mRNA and increased immunostaining for EIIIA. Nuclear runoff showed induction of FN gene transcription at 72 h in vitro. Together, these data show differential FN splice variant expression in lung, with EIIIB mRNA restricted to fetal and recovering oxidant-injured lung. Furthermore, the transition of type II cells to a type I-like cell is accompanied by increased FN gene transcription and induction of both EIIIA and EIIIB mRNA.

Extracellular glutathione peroxidase in human lung epithelial lining fluid and in lung cells.

The epithelial cells of the lower respiratory tract are exposed to high levels of inhaled oxygen and other oxidants. We hypothesized that lung cells would secrete the antioxidant enzyme, extracellular glutathione peroxidase (eGPx), into epithelial lining fluid (ELF). To investigate this hypothesis, we used specific immunoprecipitations of GPx enzymes from ELF, specific immunoprecipitations of 75Se metabolically labeled proteins from lung cells in culture, and in situ hybridization, Northern blot, and reverse transcription-polymerase chain reaction (RT-PCR) analyses. Fifty-seven percent of ELF GPx activity was due to eGPx and 40% was due to cellular GPx (cGPx). Primary bronchial epithelial cells (BEC), primary alveolar macrophages (AM), and two human bronchial epithelial cell lines, BEP2D and A549, synthesized both eGPx and cGPx and secreted eGPx into the medium. Freshly isolated human AM and BEC expressed eGPx mRNA, while freshly isolated rabbit type 2 pneumocytes did not. In lung tissue, eGPx mRNA was found mainly in interstitial cells of tissue surrounding airways. It is concluded that more than half of GPx activity in BAL is due to eGPx, and that BEC, AM, and interstitial cells are potential sources of pulmonary eGPx.

Retinal vascular endothelial growth factor (VEGF) mRNA expression is altered in relation to neovascularization in oxygen induced retinopathy.

The temporal and spatial expression of vascular endothelial cell growth factor (VEGF) mRNA was studied in normal developing cat retina, and in oxygen induced retinopathy. Unexposed control and oxygen-exposed animals (80 h of 80% oxygen from day 3, n = 16) were studied at 1, 2, 4, and 6 weeks after birth. India ink injected retinal flat mounts were used to study vessel progression, and in situ hybridizations using retinal cross sections were used to assess VEGF mRNA accumulation. In controls, as the retina matured, VEGF mRNA hybridization was evident in the ganglion cell layer in a scattered line of distinct cells prior to the ingrowth of vessels, involved the most cells in regions just peripheral to invading vessels and persisted in a fewer positive cells, widely spaced in the vascularized retinas of control, six week animals. In the inner nuclear layer, hybridization initially appeared diffusely and later became localized to a narrow portion of that layer and persisted there. In animals with oxygen induced retinopathy, a substantial increase in hybridization was observed in both the ganglion cell and inner nuclear layers of the avascular retina anterior to the advancing neovascularization. VEGF hybridization decreased abruptly to background levels in both layers at the point were neovascularization met avascular retina. By six weeks, when the neovascularization reached the ora, there was a return of VEGF mRNA in the inner nuclear layer which was similar to normal control expression. A low level of unchanging expression was also observed in the retinal pigment epithelium in both groups at all ages. These results indicate that VEGF mRNA abundance is regulated during retinal vascularization and is increased in relation to oxygen induced neovascularization, suggesting that VEGF may play an important role in both normal retinal vessel development and in the pathophysiology of retinopathy of prematurity.

Vascular endothelial growth factor mRNA increases in alveolar epithelial cells during recovery from oxygen injury.

Destruction of pulmonary endothelial cells is characteristic of hyperoxic lung injury. During recovery from hyperoxia, pulmonary endothelial cells proliferate to regenerate the vascular endothelium. Vascular endothelial growth factor (VEGF) is a peptide growth factor that is mitogenic specifically for endothelial cells. We hypothesized that VEGF messenger RNA (mRNA) increases during recovery from acute hyperoxic lung injury. Adult rabbits were exposed to 100% oxygen for 64 h and allowed to recover in air for 0, 1, 3, and 5 days. In situ hybridization showed increased VEGF expression in alveolar epithelial cells beginning at 1 day recovery. By 3 days recovery the message was in alveolar epithelial cells throughout the lung. Compared with alveolar epithelial cells, little or no expression was noted in large vessel endothelial cells, airway cells, or smooth muscle cells. Combined in situ hybridization for VEGF and immunostaining for macrophages and other mesenchymal cells found no VEGF message in those cell types. Isolated alveolar macrophages had no detectable VEGF message. Cells expressing VEGF mRNA were enriched in alveolar type II cell preparations from recovering lung. Double in situ hybridization for VEGF and surfactant protein-C (SP-C) showed co-expression in a population of type II cells, but with an inverse relationship: cells with abundant VEGF mRNA did not have abundant SP-C mRNA. Type II cells in vitro expressed VEGF message, but only when the SP-C message abundance was relatively low. We conclude that alveolar type II cells express increased VEGF mRNA during recovery from acute hyperoxia. These findings are consistent with a role for VEGF in regulating microvascular endothelial repair after oxidant injury.