Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Respiratory syncytial virus (RSV), a member of the Paramyxoviridae virus family, is a major respiratory pathogen. It is most common in infants and children, where it causes a range of illnesses from asymptomatic infection through upper respiratory tract infection, bronchiolitis, and pneumonia. Approximately 30% of children have an RSV-induced, medically attended illness, usually bronchiolitis or pneumonia, in the first year of life and virtually all infants are infected by RSV by the time they are 2 yr of age (reviewed in Reference 1). RSV is also the major cause of bronchiolitis and pneumonia in infants. These infections can be associated with serious morbidity and mortality, especially in children less than 6 yr of age and children with comorbid chronic illnesses (1). This is reflected in the approximately 90,000 hospitalizations and 4,500 deaths in infants and young children caused by RSV each year in the USA (4).

The pulmonary histopathology of RSV infection has been described in studies using human tracheal organ cultures and autopsy specimens from infants and young children who succumb acutely from RSV infection. These studies demonstrate that RSV replicates in ciliated epithelial cells of the airway. They also demonstrate that RSV infection of the lower airway and lung is associated with necrosis of the bronchiolar epithelium, a mononuclear inflammatory response, and the plugging of the smaller bronchioles with mucus, cellular debris, fibrin strands, and DNA-like materials (3). Submucosal, adventitial, and interstitial edema are also prominent features of RSV bronchiolitis and pneumonia (3). It has been assumed that the edema is due, at least in part, to the virus-induced epithelial necrosis that is characteristic of these lesions. Recent studies have, however, demonstrated that cytokines such as vascular endothelial cell growth factor (VEGF)/vascular permeability factor (VPF) can also directly induce blood vessel permeability alterations (5). The degree to which VEGF/VPF contributes to the alterations seen in RSV-infected lungs and the ability of viruses, other than human immunodeficiency virus (6), to induce VEGF/VPF elaboration have not been investigated.

VEGF/VPF, the prototype of the VEGF family of proteins (6), is a pleiotropic dimeric glycoprotein. It was first described as a tumor cell-derived molecule that increased vascular permeability. This response is now known to be mediated via the induction of blood vessel fenestrations and the formation of vesiculo-vacuolar organelles that form channels through which blood-borne proteins such as fibrin can extravasate (5). VEGF/VPF has also been extensively studied because of its ability to stimulate processes involved in angiogenesis, including endothelial cell proliferation, endothelial migration, matrix remodeling, vasodilation, and the inhibition of endothelial cell apoptosis (reviewed in Reference 6). These studies have demonstrated that VEGF/ VPF plays a crucial role in physiologic angiogenesis, such as that seen in reproduction and wound healing, and in pathologic angiogenic responses, such as those seen with tumor growth, diabetic retinopathy, and chronic inflammatory and fibrotic disorders (6, 9). VEGF/VPF can be produced by a variety of cells including epithelial cells, monocytes, macrophages, T cells, keratinocytes, fibroblasts, granulocytes, eosinophils, and smooth-muscle cells (6, 12, 13). In highly vascular tissues such as the lung, epithelial cells appear to be a particularly important source of this molecule (11, 14). The respiratory epithelium is a major site of RSV infection and replication, and RSV-induced epithelial cytokine production is felt to play an important role in the pathogenesis of RSV-induced inflammation and injury (3, 15). Despite the importance of edema, paradigms of wound healing, and epithelial infection in the pathogenesis of RSV infection, the ability of RSV to regulate epithelial cell VEGF/VPF production has not been investigated and the levels of VEGF/VPF during the course of human RSV infection have not been reported.

To address these issues we determined whether human alveolar and airway epithelial cells elaborated VEGF/ VPF after RSV infection and characterized the levels of VEGF/VPF in the nasal washes of patients with RSV infections. These studies demonstrate that RSV stimulates the production and elaboration of the 165- and 121-amino acid isoforms of VEGF/VPF by human respiratory epithelial cells and that this stimulation is mediated by a novel messenger RNA (mRNA)-independent mechanism. They also highlight the relevance of these in vitro observations by demonstrating that patients with RSV infection have significant levels of the 165- and 121-amino acid isoforms of VEGF/VPF in their nasal washings.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Viral Stock Preparation

RSV (A2 strain) and RV type 14 (RV 14) were obtained from the American Type Culture Collection (Rockville, MD) and viral stocks were prepared as previously described by our laboratory (15). In brief, sensitive cell systems were infected with a low-input multiplicity of infection (MOI). Hep-2 cells were used for RSV and MRC-5 cells were used for RV. When infection was advanced, cell supernatants were harvested, cells were disrupted by freezing and thawing, and debris was pelleted by low-speed centrifugation. Aliquots of clarified supernatants were frozen at 70°C until used. Titers of infectivity of stock viruses were determined by inoculation of serial dilutions into sensitive cell systems and quantification of plaque formation. Interleukin (IL)-1, tumor necrosis factor-, IL-8, and endotoxin were not detected in these stock preparations. In selected experiments, ultraviolet (UV)-inactivated virus was employed. This was accomplished by exposing viral stocks to ambient light or to UV light as previously described (15).

Cells and Cell Culture

A549 alveolar epithelial-like cells, A431 epidermoid carcinoma cells, and normal human bronchial epithelial (NHBE) cells were used in these studies. The A549 and A431 cells were obtained from ATCC. The NHBE cells and serum-free bronchial epithelial growth medium with supplements were purchased from Clonetics (San Diego, CA). The A549 cells and A431 cells were grown to confluence in Dulbecco's medium supplemented with penicillin, streptomycin, L-glutamine, and 10% fetal calf serum (DMEM). The NHBE cells were grown and subcultured following the supplier's instructions. On the day of infection, the medium bathing the cells was aspirated and the cultures were inoculated with virus stock at an MOI of 3.0. After adsorption at 37°C for 90 min the viral solution was removed, the cells were washed with phosphate-buffered saline (PBS), bronchial epithelial cell growth medium or DMEM was added, and the cells were incubated for the desired periods of time. The supernatants were then removed, clarified by low-speed centrifugation, and stored at 70°C until analyzed. The cell monolayers were rinsed with PBS and harvested for quantification of mRNA as described later.

Quantification of VEGF/VPF

The levels of immunoreactive VEGF/VPF in the cell supernatants and nasal samples were quantitated by enzyme-linked immunosorbent assay (ELISA) using kits obtained from R&D Systems (Minneapolis, MN) according to the manufacturer's protocol. These assays can detect as little as 5 to 15 pg/ml of the noted moieties.

Reverse Transcriptase/Polymerase Chain Reaction Analysis of VEGF/VPF mRNA

The total cellular RNA from the uninfected and virus-infected cells was extracted using Trizol reagent (Life Technologies, Grand Island, NY) according to the manufacturer's instructions. Reverse transcription and polymerase chain reaction (PCR) were performed using the Access RT-PCR kit purchased from Promega (Madison, WI) according to the manufacturer's instructions. The protocol and reverse transcriptase (RT)-PCR primers used were those described by Burchardt and colleagues (19). All primers were synthesized at the Yale Oligonucleotide Synthesis Laboratory (New Haven, CT). To amplify all of the isoforms of VEGF/ VPF, including those with alternatively spliced exons 6 and 7, we used a sense oligonucleotide (5'-TGCACCCATGGCAGAAGGAGG-3') from exon 1 and antisense oligonucleotide (5'-TCACCGCCTCGGCTTGTTCACA-3') from exon 8. With these primers, VEGF₁₂₁, VEGF₁₆₅, VEGF₁₈₉, and VEGF₂₀₆ yield products with expected sizes of 360, 497, 564, and 615 base pairs, respectively. For -actin, the following primers were used; sense 5'-GCGCTCGTCGTCGACAACGG-3' and antisense 5'-GATAGACAACGTACATGGCTG-3'.

Analysis of RT-PCR Products

The RT-PCR reaction products were fractionated on 1.5% agarose gels and the ethidium bromide (Eth Br) bands were visualized and photographed under UV light. Southern blot analysis was also used. In these assays, the RT-PCR reaction product was transferred to a nylon membrane and evaluated by hybridizing using enhanced chemiluminescence (ECL)-labeled oligonucleotide probes. Oligonucleotide labeling and signal detection were achieved using the ECL 3'-Oligolabeling and Detection System (Amersham Life Science, Buckinghamshire, UK) with protocols provided by the company. Internal oligonucleotide probes were used that either detected all isoforms of VEGF/VPF or were exon-specific and allowed for the selective detection of the 121-, 165-, or 189-amino acid isoforms of the cytokine as described by Ballaun and associates (20).

RNA Isolation and Northern Analysis

Total cellular RNA was extracted from cell monolayers at desired time points using an acid guanidinium thiocyanate-phenol- chloroform extraction protocol. Equal amounts (10 µg) of RNA from each experimental condition were size-fractionated by electrophoresis through 1% agarose/17% formaldehyde gels, transferred to nylon membranes, and hybridized with [³²P]-labeled complementary DNA (cDNA) probes. The cDNA that contains coding region for VEGF₁₂₁ and hybridizes with all VEGF/VPF isoforms was obtained by RT-PCR as described earlier. This cDNA was isolated and labeled to a high-specific activity using a random primer method. Hybridization was assessed after washing under conditions of increasing stringency and quantitated via autoradiography. The adequacy of gel loading was routinely assessed by Eth Br staining and by stripping the membrane and reprobing with a cDNA encoding -actin.

Western Blot Analysis

Uninfected and virus-infected cells were incubated as noted earlier. The supernatants were then removed, a protease inhibitor mixture was added, and the supernatants were concentrated 10-fold using Centricon 10 concentrators (Amicon, Beverly, MA). The human nasal aspirates were obtained as described later and concentrated 5-fold using the same concentrators.

Equal volumes of test and control materials were fractionated on 12.5% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) gels under reducing conditions and transferred to membranes (Millipore, Bedford, MA) that were incubated overnight in blocking buffer (5% wt/vol nonfat dry milk in PBS/0.1% Tween) at 4°C. The membranes were then incubated for 2 h at room temperature with the primary antiserum, antihuman VEGF/VPF (R&D Systems), which recognizes all known isoforms of VEGF/VPF. The membranes were then washed three times in PBS/0.1% Tween and incubated for 2 h at room temperature with mouse antigoat immunoglobulin (Ig) G (Pierce, Rockford, IL). Immunoreactive VEGF/VPF was detected using a chemiluminescent procedure (ECL plus Western Blotting Detection System, Amersham Life Science) according to the manufacturer's instructions.

In all cases, studies were undertaken to determine whether bands that were detected were specifically reacting with our anti-VEGF/VPF primary antibody. This was done by running identical gels in parallel to the ones described earlier. The only difference between the detection gels and the control gels was the absence of primary antisera during Western blot development of the control membranes. Bands were felt to represent VEGF/VPF moieties when present in Western blots developed using primary anti-VEGF/VPF antibody and absent in the absence of this antibody.

Two positive controls were used in these studies. The first is the recombinant 165-amino acid isoform of VEGF (VEGF/VPF₁₆₅) (R&D Systems). The second consisted of supernatants from stimulated A431, cells which are known to produce significant quantities of the 165- and 121-amino acid isoforms of VEGF/VPF (20, 21).

Assessment of Cell Viability

Cell viability was assessed by Trypan blue dye exclusion and lactate dehydrogenase (LDH) release assay. The release of intracellular LDH was determined using a LDH assay kit purchased from Sigma (St. Louis, MO) according to the manufacturer's instructions.

Effects of Cycloheximide/Actinomycin D

Confluent A549 cells and NHBE cells were grown to confluence, incubated with virus for 90 min, and washed as described earlier. They were then incubated in complete medium in the presence or absence of cycloheximide (Sigma) or actinomycin D (Sigma) for up to 24 h. The levels of VEGF/VPF in the resulting supernatants were evaluated by ELISA as described.

Metabolic Labeling and Radioimmunoprecipitation of VEGF

To determine whether RSV-infected epithelial cells released newly formed or exclusively preformed VEGF/VPF, confluent A549 cells in 100-mm culture dishes were inoculated for 90 min with RSV as described earlier. The cells were then washed twice with methionine- and cysteine-free DMEM (CELLect; ICN, Irvine, CA), incubated in methionine- and cysteine-free DMEM, and labeled with [³⁵S] translabel (25 µci/ml; ICN). This labels newly formed, but not preformed, proteins. The incorporation of radiolabel was determined by collecting supernatants at designated time points (4, 24, and 48 h) after viral inoculation which were used in immunoprecipitation assays. In these assays, the radiolabeled supernatants (10 ml) were centrifuged at 4,500 rpm to remove residual cells and were precleared with goat IgG (Santa Cruz Biotechnology, Santa Cruz, CA) and protein A/G PLUS-agarose conjugates (Santa Cruz Biotechnology). After repeat centrifugation, the supernatants were transferred to new 15-ml tubes, incubated with agarose-conjugated anti-VEGF primary antibody (Santa Cruz Biotechnology) for 16 h at 4°C, and centrifuged at 2,500 rpm for 5 min. The resulting immunoprecipitates were washed twice with cold PBS and once with ice-cold buffer solution (50 mM Tris-HCl, pH 7.0; and 0.3 M NaCl) and analyzed using 12.5% SDS-PAGE gels under reducing conditions. After fixation in 10% acetic acid, the gels were soaked in fluorographic solution (Autofluor; National Diagnostics, Atlanta, GA), dried, and exposed to X-ray films (Hyperfilm-MP; Amersham Life Science).

Nasal Aspirates

To determine whether VEGF/VPF could be detected in vivo, we quantitated the levels of this cytokine in nasal aspirates from virus-infected and virus-free children. The aspirates were obtained in a prospective fashion over the fall and winter months of the years 1997 to 2000 from children presenting to Yale-New Haven Hospital. As previously described (16), they were submitted to the Clinical Virology Laboratory by the patient's primary care physician on the basis of the clinical suspicion of virus infection or hospital epidemiologic surveillance. Virus infection was assessed using direct immunofluorescent assays of viral antigens (influenza A and B, parainfluenza 1-3, adenovirus, RSV) and viral culture as described by our laboratories (22). Comparisons were made of aspirates from patients with RSV infection (RSV antigen-positive or culture-positive), influenza infection (influenza antigen-positive or culture-positive) or patients without documentable viral infection (antigen-negative and culture-negative).

Statistical Analysis

Data that could not be assumed to be normally distributed are expressed as medians and interquartile ranges and analyzed with Wilcoxon 2 Sample Test when comparing two variables. Normally distributed data are expressed as means ± standard error of the mean (SEM) and assessed for significance with Student's t test or analysis of variance, as appropriate.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

RSV Stimulation of VEGF/VPF In Vitro

Because epithelial cells are the major site of RSV infection (3), in vitro studies were undertaken to determine whether alveolar and/or airway epithelial cells produced VEGF/ VPF before or after infection with this virus. Under basal conditions A549 cells and NHBE cells produced modest levels of VEGF/VPF which increased over time in culture (Figure 1). However, RSV infection increased VEGF/VPF elaboration at all time points. This induction was time-dependent with an interesting kinetic. Significant increases in supernatant VEGF could be detected within 2 h of RSV infection and continued to be appreciated 48 h after RSV infection (Figure 1) (P < 0.05 at all time points). In contrast, IL-11 was barely detectable in the absence of virus infection and was released in significant quantities 16 to 48 h after virus infection (data not shown) (15, 17). In all cases, the induction of VEGF appeared to be virus-mediated because it was decreased by > 90% when the RSV stock was exposed to UV light before monolayer incubation (Table 1). Thus, RSV is a potent stimulator of epithelial cell VEGF/VPF elaboration.

View larger version (16K): [in this window] [in a new window]   Figure 1.   VEGF/VPF elaboration by RSV-infected epithelial cells. Confluent A549 cells (A) and NHBE cells (B) were incubated in medium in the presence and absence of RSV infection as noted in MATERIALS AND METHODS. The levels of VEGF/VPF in the cell supernatants were evaluated at the noted time points by ELISA. Data are presented as means ± SEM from triplicate determinations at each time point. One experiment of three is shown (*P < 0.05).

Specificity of In Vitro Effects of RSV

To determine whether the ability to stimulate epithelial cell VEGF/VPF elaboration is a general property of respiratory viruses, studies were undertaken to determine whether RV had similar effects. In contrast to RSV, RV did not stimulate VEGF/VPF elaboration by A549 cells or NHBE cells (data not shown). It was, however, an equipotent stimulator of IL-6, IL-8, and IL-11 elaboration by these cells under identical conditions (data not shown) (15). Thus, the ability to stimulate VEGF/VPF elaboration is at least partially specific for RSV.

Isoforms of VEGF Produced by Lung Epithelial Cells

Multiple VEGF/VPF isoforms are generated as a result of extensive alternate splicing of the single VEGF/VPF gene (6, 19, 20, 23). Thus, to further understand the effects of RSV on lung epithelial cells, studies were undertaken to characterize the isoforms of VEGF/VPF produced by uninfected and RSV-infected epithelial cells. As shown in Figure 2, supernatants from uninfected A549 cells contained a moiety that migrated with the 165-amino acid VEGF/VPF isoform elaborated by A431 cells. A moiety compatible with the 121-amino acid VEGF/VPF isoform was also detectable with long periods of supernatant conditioning. At all time points RSV infection increased the elaboration of these moieties (Figure 2). Supernatants from uninfected NHBE cells also contained 165- and 121- amino acid VEGF/VPF isoforms, and RSV infection augmented the elaboration of these moieties (Figure 2). For both cell lines, bands compatible with larger VEGF/VPF moieties were not appreciated. Interestingly, supernatants from NHBE cells also contained a smaller moiety that reacted with our antiserum. It, however, was produced in a constitutive fashion and its elaboration was not regulated by RSV (data not shown). The significance of this protein is not known at the present time. Overall, these studies demonstrate that RSV infection augments the release of the 165- and 121-amino acid isoforms of VEGF/VPF by lung epithelial cells.

View larger version (31K): [in this window] [in a new window]   Figure 2.   Isoforms of VEGF/VPF produced by uninfected and RSV-infected epithelial cells. Confluent A549 cells (A) and NHBE cells (B) were incubated for the noted periods of time in medium alone () or in the presence of RSV (+). The isoforms of VEGF/VPF in the resulting supernatants was assessed by Western blot analysis described in MATERIALS AND METHODS. The left lane contains the recombinant human 165-amino acid VEGF/ VPF isoform (165). The right lane shows the 165- and 121-amino acid VEGF/VPF isoforms produced by A431 cells (A431).

Role of Protein Synthesis in Epithelial Cell VEGF/VPF Elaboration

VEGF/VPF has been reported to exist in a preformed pool in some cells (12). Thus, the rapid kinetic of VEGF/ VPF elaboration by RSV-infected epithelial cells raised the possibility that RSV might interact with and induce the elaboration of preformed cytokine. To address this issue, the effects of the protein synthesis inhibitor cycloheximide on RSV-induced VEGF/VPF elaboration were assessed. These studies demonstrated that protein biosynthesis was required because RSV stimulation of VEGF/VPF elaboration was abrogated in the presence of 10 µg/ml cycloheximide (Figure 3). This suggests that cycloheximide directly blocks VEGF/VPF production or blocks the production of an intermediate protein that is involved in the secretion of preformed VEGF/VPF cytokine. To differentiate among these options and reinforce our findings with cycloheximide, epithelial cells were grown to confluence and infected with RSV in the presence of [³⁵S]-methionine and cysteine to label newly translated proteins. The VEGF/ VPF in the resulting supernatants was then analyzed via immunoprecipitation and autoradiography. As can be seen in Figure 4, VEGF/VPF was produced in this setting inasmuch as bands compatible with the 121- and 165- amino acid isoforms of the cytokine were readily appreciated. These studies demonstrate that RSV-induced VEGF/ VPF elaboration is mediated, at least in part, by virus-induced VEGF/VPF protein biosynthesis.

View larger version (13K): [in this window] [in a new window]   Figure 3.   Effects of cycloheximide on RSV stimulation of VEGF/ VPF elaboration. Confluent A549 cells were incubated in the presence and absence of virus for 90 min, washed, and incubated for 24 h in the presence or absence of cycloheximide (10 µg/ml). The levels of VEGF/VPF in the resulting supernatants were assessed by ELISA. Data are presented as means ± SEM of three separate experiments (*P < 0.01).

View larger version (22K): [in this window] [in a new window]   Figure 4.   35S labeling of VEGF/VPF induced by RSV. A549 cells were grown to confluence and then incubated for 90 min with medium alone () or with RSV (+). The cells were then washed and incubated in methionine- and cysteine-free DMEM supplemented with 35S-translabel for the noted periods of time. At the end of the incubation period the supernatants were removed and the VEGF/VPF isoforms that they contained were evaluated via immunoprecipitation and autoradiography as described in MATERIALS AND METHODS. The locations of the 165- and 121-amino acid isoforms of VEGF/VPF are noted.

Mechanism of RSV Regulation of VEGF/VPF Elaboration

To further understand the mechanism(s) by which RSV stimulates epithelial cell VEGF/VPF elaboration, we quantitated the levels of VEGF/VPF mRNA in A549 cells and NHBE cells before and after virus infection using RT-PCR analysis. mRNA encoding the 165-, 121-, and 189-amino acid isoforms of VEGF/VPF were readily detected in unstimulated cells (Figure 5). Interestingly, RSV infection did not cause impressive alterations in the levels of these mRNA transcripts when evaluated from 30 min to 48 h after virus administration (Figure 5 and data not shown). A similar lack of induction was noted with Northern blot analysis (data not shown). In accord with these observations, actinomycin D (0.02 to 2 µg/ml) did not abrogate the ability of RSV to induce VEGF/VPF elaboration (data not shown). The mechanism of RSV induction of VEGF/VPF also did not appear to involve significant cell toxicity because the increase in VEGF/VPF elaboration by NHBE and A549 cells was not associated with significant LDH release or alterations in cell viability (data not shown). When viewed in combination, these studies demonstrate that RSV infection increases the elaboration of VEGF/VPF by A549 and NHBE cells, at least in part, via a noncytotoxic translational and/or post-translational mechanism.

View larger version (57K): [in this window] [in a new window]   Figure 5.   VEGF/VPF mRNA in uninfected and RSV-infected epithelial cells. Uninfected () and RSV infected (+) A549 cells and NHBE cells were incubated for 4 or 24 h as noted and the mRNA encoding VEGF/VPF in these cells was evaluated via RT-PCR analysis as described in MATERIALS AND METHODS. The top panel illustrates the Eth Br gels after RT-PCR with primers that can detect all isoforms of VEGF/VPF. The middle panel illustrates the same RNA samples using RT-PCR primers specific for -actin. In the bottom panel, RT-PCR was performed using the primers in the top panel and, after agarose gel electrophoresis, the RT-PCR products were transferred and the amount of each transcript was evaluated via Southern analysis using a VEGF/VPF DNA probe that hybridizes with all isoforms of the cytokine. The locations of the molecular weight (MW) markers and the mRNA encoding the 189-, 165-, and 121-amino acid isoforms of VEGF/VPF are indicated.

Nasal VEGF in RSV Infections

To determine whether respiratory virus stimulation of VEGF/VPF elaboration in vitro is relevant to the in vivo state, studies were undertaken to determine whether VEGF/ VPF could be detected in the nasal secretions of patients with RSV respiratory tract infections. Significant levels of VEGF/VPF were noted in aspirates from patients with RSV infections (range 16.4 to 2266.3 pg/ml). The median level of nasal VEGF/VPF for this group was 413.7 pg/ml (interquartile range 125.4 to 870.7) (Figure 6). Interestingly, these levels were significantly higher than the levels in secretions from patients with documented influenza infection (range 18.6 to 816.8; median 173.7; interquartile range 125.4 to 870.7 pg/ml) (P = 0.008 versus RSV-infected patients) and the patients without documentable virus (range 21.2 to 841.2; median 117.8; interquartile range 30.8 to 305.8 pg/ml) (P = 0.0012 versus RSV-infected patients) (Figure 6).

View larger version (12K): [in this window] [in a new window]   Figure 6.   VEGF/VPF levels in human nasal samples. Nasal aspirates were obtained from patients with documented influenza infection (Influenza; n = 25), patients with documented RSV infection (RSV; n = 47), and patients without documentable virus infection (Virus []; n = 21) as described in MATERIALS AND METHODS. The levels of VEGF/VPF were evaluated by ELISA. Individual values and medians and interquartile ranges for the different groups are illustrated.

Isoforms of VEGF in Nasal Fluids

To further characterize the nasal VEGF/VPF, Western blot analysis was used to define the VEGF/VPF isoforms in the nasal fluids from patients with RSV respiratory tract infections. As illustrated in Figure 7, the 165- and 121- amino acid isoforms of VEGF/VPF were detected in the nasal secretions of these individuals. Appropriately specific immunoreactive moieties compatible with larger VEGF/VPF isoforms could not be appreciated. These studies demonstrate that the 165- and 121-amino acid isoforms are the major VEGF/VPF moieties found in nasal secretions in the setting of RSV infection.

View larger version (26K): [in this window] [in a new window]   Figure 7.   Isoforms of VEGF/ VPF in nasal secretions from patients with RSV infection. Western blot analysis was used to define the VEGF/ VPF isoforms in nasal secretions from two representative RSV-infected patients (A and B). The isoforms of VEGF/VPF in the secretions from these individuals are compared with recombinant 165-amino acid VEGF/ VPF (165) and the 165- and 121-amino acid isoforms of VEGF/ VPF produced by stimulated A431 cells (A431).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In these studies we addressed the hypothesis that the vascular alterations contributing to the wound healing and tissue edema in RSV respiratory infections could be mediated, at least in part, by vascular-active substances that are elaborated during the course of the disease(s). VEGF/ VPF was chosen for these studies on the basis of its documented ability to induce transient vascular leakage, endothelial cell proliferation, and other endothelial alterations characteristic of the angiogenic state (6). Our findings support this hypothesis. They demonstrate that airway and alveolar epithelial and epithelial-like cells elaborate VEGF/ VPF after RSV infection, that the 165- and 121-amino acid isoforms of VEGF/VPF are elaborated in this setting, and that the elaboration of these moieties is mediated, at least in part, by a translational and/or post-translational, noncytotoxic, biosynthetic process. These studies also provide evidence that these in vitro observations are relevant to the in vivo state because VEGF/VPF can be detected in nasal washings from patients with RSV respiratory tract infections. These are the first studies to demonstrate that any respiratory virus can induce VEGF/VPF elaboration in vivo or in vitro and that virus-infected epithelial cells produce this important cytokine. When viewed in combination, these studies raise the novel possibility that the tissue edema in RSV-infected tissues is due, at least in part, to the effects of virus-induced VEGF/VPF on local vascular elements.

Inflammation is a cardinal feature of RSV lower respiratory tract infections. Increased microvascular permeability is an obligatory accompaniment of inflammation and is, in turn, causally related to local tissue edema (24). In RSV infections, plasma exudes into the interstitium, airway walls, alveoli, and airway lumen. This generates airway wall thickening and contributes to the formation of viscous mucus and airway mucus plugging (24, 25). The exuding plasma proteins and growth factors may also contribute to the generation of airway fibrosis and smooth-muscle hypertrophy and/or hyperplasia (25, 26). Wheezing and airway mucus plugging are common clinical features in childhood RSV infections. It is thus easy to see how airway permeability alterations could contribute to the airways obstruction, structural alterations, and physiologic dysregulation characteristic of severe lower respiratory tract RSV infections.

At least three mechanisms have been proposed to contribute to the pathogenesis of inflammation-induced tissue edema. They include: (1) local vascular injury and cell cytotoxicity; (2) contraction of postcapillary venular endothelium causing gaps between otherwise tightly associated cells; and (3) the elaboration of vasoactive substances such as platelet-activating factor, leukotrienes, and histamine (24, 25, 27). Surprisingly, little is known about the role of each of these processes in the pathogenesis of the edema seen in common disorders such as asthma. Even less is known about their role in lower respiratory tract viral infections. Our studies demonstrate that VEGF/VPF is elaborated by RSV-infected epithelial cells and found in the nasal secretions of patients with RSV infections. These studies suggest that RSV-induced VEGF/VPF elaboration may play an important role in the pathogenesis of the edema and other features of RSV pneumonias and/or bronchiolitis. They also suggest that respiratory epithelial cells are a major source of VEGF/VPF in this setting. This observation is in accord with a variety of studies demonstrating that alveolar and airway epithelial cells express VEGF/ VPF at baseline and in a prominent fashion during recovery from injury (11, 23, 28). It is important to point out, however, that these studies do not prove that the many other cells capable of elaborating VEGF/VPF (fibroblasts, smooth-muscle cells, granulocytes, macrophages, and eosinophils) (12, 13) are not involved in the VEGF/VPF elaboration noted in the in vivo setting.

Angiogenesis is defined as the formation of new capillary blood vessels from existent microvessels. It is one of the most pervasive and essential biologic events encountered in mammalian organisms (29). Under normal conditions, angiogenesis occurs infrequently. It can, however, be induced in a rapid fashion in response to diverse physiologic stimuli. Among the most extensively studied angiogenesis-dependent stimuli are growth and wound repair (29). VEGF/VPF plays a major role in these responses by stimulating blood vessel formation and endothelial cell proliferation, inhibiting endothelial cell apoptosis, and stabilizing newly formed vessels (6, 11, 23, 29). Exaggerated levels of VEGF/VPF have also been documented in states characterized by tissue fibrosis, and VEGF/VPF can stimulate the proliferation of some epithelial tissues (28, 32). RSV lower respiratory tract infections are characterized by inflammation and necrosis (1). It is thus tempting to speculate that RSV-induced VEGF/VPF elaboration, in addition to contributing to the generation of tissue edema, also acts to foster and/or promote healing and the re-epithelialization of denuded basement membranes. Additional investigation will be required to test this hypothesis. These experiments will need to be done with approaches other than the use of mice with targeted genetic disruptions of VEGF/VPF because even animals that lack one of the two VEGF/VPF alleles die before birth because of defects in the development of the cardiovascular system (33).

At least five VEGF/VPF isoforms have been described that are produced by the alternative splicing of the mRNA from the single VEGF/VPF gene (6, 19, 20, 23). An important biologic property that distinguishes the different VEGF/VPF isoforms is their heparin and heparan-sulfate binding ability. The 121- and 165-amino acid isoforms of VEGF/VPF do not bind and do bind weakly, respectively. As a result, they are soluble moieties in cell supernatants and biologic fluids. In contrast, the 189- and 206-amino acid isoforms bind with a high affinity and remain cell- and/or matrix-associated in vitro and in vivo. The isoforms of VEGF/VPF also differ in their biologic functions, the use of neuropillin-1 as a coreceptor, and their expression and regulation at sites of injury and repair (6, 11, 23). Cells that synthesize VEGF/VPF usually produce more than one isoform (6, 13, 20), with the 165- and 121-amino acid isoforms dominating in most cells and tissues. Our studies characterize, for the first time, the VEGF/VPF isoforms elaborated by alveolar and airway epithelial cells. In accord with other tissues, the 121- and 165-amino acid isoforms were the major products of NHBE cells. In contrast, A549 cells produced the 165-amino acid VEGF/VPF isoform in much greater quantities than the 121-amino acid moiety. The implications of these findings are not clear. In addition, it is impossible to determine whether the differences between A549 and NHBE cells represent true differences between airway and alveolar epithelial cells or differences associated with the malignant transformation of the A549 cell line.

RSV-induced epithelial cell cytokine production plays an important role in the pathogenesis of RSV respiratory tract infections. Studies of these virus-epithelial interactions have demonstrated that RSV can stimulate epithelial cell cytokine production by altering cytokine gene transcription (34, 35) or mRNA stability (36). Our studies demonstrate, for the first time, that RSV stimulates epithelial cell VEGF/VPF elaboration, at least in part, via a translational and/or post-translational mechanism. This is the first demonstration of a translational and/or post-translational site of RSV regulation of epithelial cell cytokine production.

In summary, these studies demonstrate that airway and alveolar epithelial and epithelial-like cells elaborate VEGF/ VPF after RSV infection and highlight the virus and isoform specificity of this induction and the unique mechanism mediating this response. They also demonstrate that RSV infections in vivo are associated with the accumulation of the 165- and 121-amino acid isoforms of VEGF/VPF. These findings suggest that VEGF/VPF plays an important role in the pathogenesis of RSV-induced pathologies and that regulation of VEGF/VPF elaboration and/or effector function may be a useful way to modulate RSV- induced respiratory responses. They also highlight the complexity of RSV-epithelial cell interactions at sites of RSV infection.