Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

In acute lung injury, alveolar type I cells are selectively destroyed, and restoration of the alveolar epithelium occurs by proliferation of alveolar type II cells. Alveolar type II cell hyperplasia is a prominent histologic feature after acute lung injury and in interstitial lung disease. Although type II cells proliferate in vitro in response to acidic fibroblast growth factor (aFGF), keratinocyte growth factor (KGF), hepatocyte growth factor, transforming growth factor (TGF)-, and heparin-binding epidermal growth factor (1, 2), the growth factors responsible for proliferation in vivo are not established. Furthermore, the role of hyperplastic type II cells on the underlying fibroproliferative process in acute lung injury is not defined. There is evidence that type II cells inhibit fibroblast proliferation (3). However, hyperplastic type II cells have also been suggested to foster the fibrotic reaction by producing TGF-, TGF-, and platelet-derived growth factor (PDGF). To date, it has been difficult to produce extensive type II cell hyperplasia in rodents to determine whether type II cell hyperplasia by itself produces a fibrotic reaction or modifies the underlying mesenchyme.

KGF (FGF-7) is a member of the fibroblast growth factor (FGF) family and is specific for epithelial cells and microvascular endothelial cells (4, 5). KGF has been shown to be a potent growth factor for type II cells in vitro (6). KGF also increases the expression of surfactant protein (SP)-A, SP-B, and SP-D in cultured alveolar type II cells (7). In vivo, intratracheal instillation of KGF causes a transient type II cell proliferation (10). Type II cell proliferation is maximal on Days 2 and 3 after instillation, and the lung returns to normal by Day 7. Recently, intratracheal instillation of KGF has been shown to prevent lung injury induced by hyperoxia, acid instillation, bleomycin, and -naphthylthiourea (6, 13). However, the KGF has had to be given before the injury in order to be effective. A more sustained delivery of KGF or expression of KGF might allow for successful post-treatment therapy.

Human adenoviruses were initially evaluated as vectors to treat cystic fibrosis because of their ability to infect nondividing pulmonary epithelial cells (16). Recombinant adenoviruses have been used to transfer the cystic fibrosis transmembrane conductance regulator protein complementary DNA (cDNA), as well as other genes and cDNAs, to the respiratory epithelium (17, 18). These viruses contain a 36-kb genome that consists of a series of early genes, which encode regulatory proteins, and late genes, which encode structural proteins. Adenovirus is particularly attractive as a vector because large amounts of virus can be produced, and recombinant virus efficiently infects differentiated, nondividing cells. Deletion of early region 1 (E1) of adenovirus type 5 produces a highly infectious, but replication-defective, virus that can efficiently transfer genes to respiratory epithelial cells in vivo and in vitro (16, 19).

In this study, we constructed a replication-deficient, recombinant human type 5 adenovirus vector expressing the recombinant human KGF gene (Ad5-KGF) and evaluated the effects of Ad5-KGF on the proliferation of rat alveolar type II cells. We demonstrated that Ad5-KGF stimulates the production and secretion of KGF by alveolar epithelial cells in vitro. The expressed protein was purified from type II cell culture medium infected with Ad5-KGF by chromatography on heparin sepharose and shown to be the estimated size for glycosylated KGF. In vivo, we instilled Ad5- KGF to Fischer 344 rats and demonstrated that Ad5-KGF produced extensive type II cell hyperplasia that peaked from Days 3 to 7 after instillation.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Line and Viruses

The human embryo kidney cell line 293 was obtained from the American Type Culture Collection (stock no. CRL1573; Rockville, MD) and used to propagate the adenoviruses with E1 deletions. Adenovirus 5 expressing interleukin (IL)-4, TGF-1 (active), or TGF-1 (latent) were obtained from Jack Gauldie (Hamilton, ON, Canada).

Generation of a Recombinant Adenovirus Encoding Human KGF

The human KGF (hKGF) coding region was amplified by polymerase chain reaction (PCR) using as template the plasmid pCEV9 (obtained from Jeffrey Rubin, National Cancer Institute, Bethesda, MD). This plasmid contains the complete KGF coding sequence isolated from a M426 human embryonic lung fibroblast cDNA library (20). The sense and antisense primers had, respectively, the sequences 5'-CCTAGATCTGCCACCATGCACAAATGGATACTGAC-3' and 5'-CCTCTCGAGTTAAGTTATTGCCATAGGAAG-3'. The sense primer is homologous to nucleotides 1 to 20 in the hKGF open reading frame and contains a consensus (GCCACC) Kozak sequence immediately upstream of the initiation codon. In addition, a BglII site was inserted at the 5'-terminus to facilitate cloning. The antisense primer is homologous to nucleotides 565 to 585 and contains an XhoI recognition sequence at its 5' terminus. These primers allowed amplification of a 609-bp cDNA fragment. The PCR contained 20 mM Tris-HCl, pH 8.8, 10 mM KCl, 2 mM MgSO₄, 0.1% Triton X-100, 0.1 mg/ml bovine serum albumin, 0.2 mM deoxynucleotide triphosphates, 100 pmol of each primer, 100 ng pCEV9 DNA, and 5 U Pfu polymerase (Stratagene, La Jolla, CA) in a 200-µl volume. Before cycling, the mixture was divided into four 50-µl reactions. A temperature profile of 30 s at 94°C, 30 s at 55°C and 1 min at 72°C was used for 25 cycles followed by a final extension of 10 min at 72°C. After purification of the PCR product by gel electrophoresis, it was ligated as a blunt end fragment into the vector pCR-TRAP, which had been cleaved with StuI to form the plasmid pCR-hKGF. The 0.6-kb cloned fragment was recovered from the vector by cleavage with BglII and XhoI, and purified by gel electrophoresis.

The replication-defective, recombinant virus expressing the hKGF transgene was created by first subcloning the 0.6-kb PCR fragment into the shuttle vector pMH5, cleaved with BamHI and SalI (Figure 1). This vector, created by M. Hitt and F. Graham and obtained from Microbix Biosystems Inc. (Toronto, ON, Canada), contains a segment of the adenovirus 5 genome from bp 22 to bp 342 that includes the packaging signal required to encapsulate viral DNA, the murine cytomegalovirus (mCMV) immediate early gene 1 promoter, a short polylinker containing EcoRI, BamHI and SalI recognition sequences, an SV40 polyA signal, and additional adenoviral sequences from bp 3523 to bp 5790. The remainder of the adenovirus genome required to produce infectious virus is provided by the vector pBHG10 (21). The absence of the adenoviral sequences from 343 to 3522 deletes expression of the viral E1 coding region, rendering the virus defective for replication in normal cells.

View larger version (23K): [in this window] [in a new window]   Figure 1.   Construction of recombinant adenovirus vector Ad5-KGF. The replication-defective recombinant virus expressing the hKGF transgene was created by subcloning the 0.6-kb PCR fragment into the shuttle vector pMH5. The remainder of the adenovirus genome required to produce infectious virus is provided by the vector pBHG10. This vector lacks the packaging signal and, therefore, does not produce infectious virus after a single transfection of 293 cells. pMH5-hKGF and pBHG10 were cotransfected into 293 cells. After plaque purification, individual viral clones were amplified in 293 cells and assayed for transgene expression.

The 293 cell line is a human embryonic kidney epithelial cell line that expresses the adenoviral E1 genes and is therefore able to complement the replication defect of E1-deleted adenoviral vectors (22). Subconfluent (50 to 70%) monolayers of 293 cells, plated 24 h before transfection in 25-cm² flasks, were washed twice with 5 ml of Hanks' balanced salt solution, and then 2.5 ml of OPTI-MEM (GIBCO-BRL, Grand Island, NY) was added. For transfection, 0.8 µg of pMH5hKGF DNA and 1.2 µg of pBHG10 DNA were added to 250 µl of OPTI-MEM. This mixture was then added to 2.5 ml of OPTI-MEM containing 12 µl of lipofectamine (GIBCO-BRL) and incubated for 30 min at room temperature. The lipofectamine-DNA mixture was then added to 293 cells. After an incubation of 5 h at 37°C, the transfection medium was removed and replaced with 8 ml of Dulbecco's modified Eagle's medium (DMEM; GIBCO-BRL) containing 10% calf serum. Transfected cells were collected 10 d later and lysed by three cycles of freezing in a dry ice bath and rapid thawing at 37°C. After three cycles of plaque purification, viral pools were amplified in 293 cells and assayed for KGF transgene expression. Cell-free supernatants were used to infect 293 cell monolayers from which viral plaques were picked.

Construction of the LacZ Recombinant Adenovirus

To construct a control virus vector expressing the lacZ gene of Escherichia coli, the vector pIND/lacZ (Invitrogen, Carlsbad, CA) was cut with HindIII and XhoI. The 3.2-kb fragment containing the lacZ open reading frame was then purified by agarose gel electrophoresis, subcloned into the vector pBluescript II KS+ (Stratagene), and cut with HindIII and XhoI. The recombinant vector was then cut with EcoRI and the 3.2-kb fragment was once again purified by gel electrophoresis and subcloned into the EcoRI site of the polylinker region in the shuttle vector pMH5. Orientation of the lacZ fragment in the vector was determined by restriction enzyme analysis. Transfection of 293 cells and viral plaque purification were performed as described for generation of the KGF recombinant virus.

Growth and Purification of Adenoviruses

The 293 cells were cultured in complete medium (DMEM with 5% heat-inactivated calf serum, 2 mM glutamine, 50 U/ml penicillin, 50 µg/ml streptomycin) in 150-cm² plates at 37°C, 5% CO₂, until 70 to 80% confluent. The infectivity of all virus stocks was determined by plaque titration on 293 cells. For preparation of purified virus, subconfluent monolayers of 293 cells in 150-cm² dishes were infected with a multiplicity of 10 to 100 plaque-forming units (PFU) per cell of virus in a 2-ml inoculum. Cells were harvested by scraping after 48 h incubation. Virus was purified from cell lysates using a stepwise process of equilibrium density gradient centrifugation with a cesium chloride modification of a described technique (23). Cells were lysed by five cycles of freezing and thawing. Cesium chloride at a density of 1.45 g/cm³ was used for the cushion, and a density of 1.35 g/cm³ was used for the equilibrium density gradient. This viral purification procedure was done twice. Cesium chloride was removed from purified virus on a PD-10 gel filtration column (Sephadex G25; Sigma, St. Louis, MO) with 10% glycerol in phosphate-buffered saline (PBS) buffer. Actual PFUs were measured by plaque titration on 293 cells for the in vitro studies, and the number of particles/ml was calculated by optical density 260 nm × 1/0.02 × 10¹² for the in vivo studies.

Adult Rat Alveolar Type II Cell Isolation and Culture

Rat alveolar type II cells were isolated from lungs of adult male, specific pathogen-free Sprague-Dawley rats weighing 180 to 280 g (Bantin and Kingman, Inc., Fremont, CA) by dissociation with porcine pancreatic elastase (Worthington Biochemicals, Freehold, NJ) followed by purification on discontinuous metrizamide density gradients as described (24). Cells were suspended in DMEM containing 10% fetal bovine serum (FBS; Irvine Scientific, Santa Ana, CA), supplemented with 2 mM glutamine, 100 U/ml penicillin G, 100 µg/ml streptomycin, 2.5 µg/ml amphotericin B, and 10 µg/ml gentamicin (Sigma Chemical Co. St. Louis, MO), and incubated in suspension with adenoviruses for 1 h at 37°C. After infection, the cells were plated at a density of 1 × 10⁵ cells/well in 0.5 ml DMEM containing 10% FBS onto 48-well plates (Costar Corp., Cambridge, MA). The cells were maintained at 37°C in a humidified incubator containing 90% air/10% CO₂.

DNA Synthesis and KGF Enzyme-Linked Immunosorbent Assays

DNA synthesis by control and treated type II cells was estimated by [³H]thymidine incorporation as described. aFGF and human recombinant KGF (hrKGF) were added 48 h before cell harvest as positive controls.

To measure KGF in the media, the cells were infected with adenoviruses for 1 h and plated at a density of 1 × 10⁵ cells/well in 0.5 ml DMEM with 10% FBS. After 24 h, the monolayers were rinsed to remove nonadherent cells and cultured for an additional 48 h. The media were then collected, and KGF was measured by enzyme-linked immunosorbent assay (ELISA) using a mouse monoclonal antibody (IG4) to hKGF for the capture antibody and biotinylated monoclonal mouse anti-KGF (A1) for the detection antibody (Amgen, Thousand Oaks, CA). The biotinylated antibody was quantitated with the Vectastain ABC Elite Peroxidase kit (Vector Laboratories, Burlingame, CA). The volume of fluid assayed by ELISA was 50 µl per sample. The lower limit of detection for KGF was 0.3 ng/ml, and the assay was linear up to a concentration of 20 ng/ml.

Identification of KGF in the Medium

The medium infected with Ad5-KGF was collected 72 h after plating and purified by heparin sepharose affinity chromatography. A total of 10 ml of the medium was absorbed onto 0.2 ml heparin sepharose (Pharmacia Biotech, Piscataway, NJ) in a 15-ml tube. The tube was rocked for 30 min at 4°C and then the heparin sepharose was sedimented. The supernatant was removed, and the beads were washed twice with 0.3 M NaCl in 10 mM Tris (pH 7.5). The KGF was eluted three times with 100 µl sample buffer (0.625 M Tris-HCl, pH 6.8, 10% glycerol, 2% sodium dodecyl sulfate, 0.005% bromophenol blue, 5% -mercaptoethanol in deionized water). A total volume of 5 µl of the 250 µl collected was loaded in each lane on 8 to 16% gradient polyacrylamide gel (NOVEX, San Diego, CA). A total of 5 ng hrKGF was loaded as a standard. The amounts of KGF protein loaded were measured by ELISA and were 3.52, 6.46, and 8.50 ng for Ad5-KGF multiplicity of infection (MOI) of 2, 4, and 8, respectively.

Viral Instillation In Vivo

Pathogen-free Fischer 344 rats, weighing 180 to 300 g, were housed in a temperature-controlled room with 12-h light-dark cycle and free access to food and water. Rats were anesthetized with Fluothane in 100% oxygen and then intubated with a 16-gauge intravascular Teflon catheter (Quick-Cath; Baxter, Deerfield, IL). A small polyethylene tube was then positioned into the left lobar bronchus through the tracheal catheter with the rats in a left lateral position, and adenoviruses, saline or PBS containing 10% glycerol were instilled. Ad5-KGF or Ad5-LacZ was instilled at a concentration of 10⁹ PFU in 0.3 ml PBS containing 10% glycerol, and this instillation was followed by 0.2 ml saline and a bolus of air to deliver the virus to the terminal gas exchange units.

Lung Pathology

The lungs were fixed by intratracheal instillation at a pressure of 25 cm H₂O with 4% paraformaldehyde in PBS at 4°C for 2 h. They were then sliced and fixed with 4% paraformaldehyde at 4°C overnight and transferred to 70% ethanol. Slices of lungs were embedded in paraffin, sectioned, and stained with hematoxylin and eosin or processed for immunocytochemistry or in situ hybridization.

Bronchoalveolar Lavage

Lung lavage was performed to obtain total and differential cell counts, and to measure KGF, SP-A, and SP-D protein content. A total volume of 25 ml of PBS delivered in 5-ml aliquots was instilled into the left lung via tracheostomy after ligation of the right main bronchus. The total lavage fluid was centrifuged at 1,200 rpm for 10 min at 4°C, and the supernatant was gently aspirated and stored at 20°C for subsequent assay. Cell pellets were resuspended in PBS for total and differential cell counts. Total counts of mononuclear cells were performed using a standard hemacytometer. Differential counts were made from cytocentrifuge smears stained with the Hemacolor stain set (Em Science, Gibbstown, NJ).

-Galactosidase Stain In Vitro

After alveolar type II cell isolation, the cells were incubated in suspension with Ad5-LacZ at a MOI of 8 PFU/cell for 1 h at 37°C. After infection, the cells were plated at a density of 1 × 10⁵ cells/well in 0.5 ml DMEM containing 10% FBS in a 48-well plate. The nonadherent alveolar type II cells were removed 24 h after plating, and the monolayers cultured for an additional 48 h. Type II cells were then fixed for 1 h with a solution of 2% paraformaldehyde and 2 mM MgCl₂ in PBS and stained overnight at 37°C in a solution of 1 mg/ml X-gal, 2 mM MgCl₂, 4 mM K₄Fe(CN)₆, and 4 mM K₃Fe(CN)₆ in PBS.

-Galactosidase Stain In Vivo

At 24 h after infection, rat lungs were fixed by intratracheal perfusion at a pressure of 25 cm H₂O with 2% paraformaldehyde and 0.2% glutaraldehyde in PBS (pH 7.3) at 4°C for 1 h. The fixative was drained, and the lungs were rinsed twice by PBS, lavaged once with the staining solution (0.5 mg/ml X-gal, 2 mM MgCl₂, 5 mM K₄Fe[CN]₆, and 5 mM K₃Fe[CN]₆ in 20 mM Tris in PBS, pH 8.3), and kept inflated by intratracheal perfusion of the staining solution at 37°C for 4 h (25).

ELISA for SP-A and SP-D

SP-A and SP-D were quantitated by ELISA with rabbit antirat SP-A and antirecombinant SP-D polyclonal antibodies as primary antibodies as reported (26). Native rat SP-A and recombinant rat SP-D expressed in Chinese hamster ovary cells were used as standards.

Immunocytochemistry

Bromodeoxyuridine. To detect cells undergoing DNA synthesis, 2 h before killing and lung fixation, the rats were injected with bromodeoxyuridine (BrdU) 20 mg/100 g body weight intraperitoneally. The lungs were fixed with 4% paraformaldehyde as stated previously. Deparaffinized 4-µm sections were rehydrated through graded alcohols. Sections were permeabilized with 0.5% Triton in PBS for 1 h at room temperature, washed in PBS, and then incubated with 0.2 mg/ml trypsin in 0.1 mM calcium chloride for 30 min at 37°C. The sections were treated with 0.1 N HCl for 10 min at 4°C to remove the histones and then incubated in 2 N HCl in 0.05 M PBS for 30 min at 37°C. After rinsing in 0.1 M borate buffer (pH 8.5) for 10 min, endogenous peroxidase was inhibited by treatment with 0.5% hydrogen peroxide in PBS for 30 min at room temperature. The sections were incubated in PBS containing 3% horse serum in 0.2 M glycine and 0.2 M lysine for 1 h to block nonspecific binding sites. Then the mouse monoclonal antibody (M744) in 3% horse serum was incubated with the tissue sections overnight at 4°C. After washing the section in PBS twice, biotinylated horse antimouse immunoglobulin (IgG) (Vector Laboratories) in 3% horse serum was added and incubated for 30 min at room temperature. After washing in PBS twice, the sections were incubated in streptavidin-biotin-horseradish peroxidase in 3% horse serum for 60 min, washed in PBS twice, and then incubated with diaminobenzidine in 50 mM Tris (pH 7.4) for 15 min. The sections were then washed briefly in water and counterstained with hematoxylin very briefly (10 s), and examined by light microscopy.

Lycopersicon Esculentum Lectin

Lungs were fixed with 4% paraformaldehyde, embedded in paraffin, and processed for immunocytochemistry. Biotinylated Lycopersicon esculentum lectin (tomato) (Vector Laboratories) was used at a concentration of 3 µg/ml for 1 h and detected with streptavidin-horseradish peroxidase and diaminobenzidine for color development. This lectin stains type I cells and some macrophages (27).

SP-C

A rabbit polyclonal antibody to human pro-SP-C (a generous gift of Dr. Jeffrey Whitsett, University of Cincinnati, Cincinnati, OH) was used to detect alveolar epithelial type II cells. The method for immunodetection was performed as described previously (13).

In Situ Hybridization

Expression of KGF in lungs infected with Ad5-KGF was localized by in situ hybridization. Lungs were fixed in freshly prepared 4% paraformaldehyde in RNase-free PBS overnight at 4°C, dehydrated, and embedded in paraffin. In situ hybridization was performed on 4- to 6-µm sections as previously described (28), with the exception that [³³P]uridine triphosphate (2,000 to 4,000 Ci/ mmol; NEN Life Science Products, Boston, MA) was used in the transcription of RNA probes.

Statistical Analysis

One-way analyses of variance were performed to determine differences among groups, with alpha () set at the usual 0.05. Thus, overall differences among groups (F ratios) and mean differences between groups with observed probabilities of type I errors of less than 0.05 (P < 0.05) were considered significant. After overall significance, post-hoc group comparisons were performed using Tukey's honestly significant difference multiple comparison procedure.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effect of Ad5-KGF on Cellular DNA Synthesis

To determine if biologically active KGF was produced in vitro, type II cells were infected with Ad5-KGF and [³H]- thymidine incorporation was measured. After isolation of rat type II cells, cells were incubated with Ad5-KGF at a MOI of 2, 4, and 8 PFU/cell after 24 h of culture. [³H]thymidine was added and incorporation was assayed 48 h later. Ad5-KGF at a MOI of 2, 4, and 8 PFU/cell induced a dose-dependent increase in thymidine incorporation by type II cells that was greater than that measured in parallel cultures with control medium, hrKGF (10 and 100 ng/ml), or aFGF (200 ng/ml) (Figure 2). To determine whether Ad5-KGF stimulation of DNA synthesis was specific for Ad5-KGF, [³H]thymidine incorporation was measured with type II cells infected with adenoviruses expressing -galactosidase (Ad5-LacZ), IL-4 (Ad5-IL4), or active or latent TGF- (Ad5-TGF1). None of these other adenoviruses stimulated DNA synthesis by type II cells (Figure 3).

View larger version (18K): [in this window] [in a new window]   Figure 2.   Thymidine incorporation by type II cells infected with Ad5-KGF. Ad5-KGF at an MOI of 2, 4, and 8 PFU/cell shows a dose-dependent increase of thymidine incorporation that was greater than that measured in parallel cultures stimulated with recombinant human KGF or aFGF. Data represent the means ± standard error of the mean (SEM) of eight independent experiments. * Significantly different from control (P < 0.0001). Significantly different from aFGF 200 ng/ml (P < 0.0001).

View larger version (16K): [in this window] [in a new window]   Figure 3.   Thymidine incorporation by type II cells infected with Ad5-LacZ, TGF-1 (active), TGF-1 (latent), and IL-4. Ad5-LacZ, TGF-1 (active), TGF-1 (latent), and IL-4 do not affect thymidine incorporation (n = 4). Only Ad5-KGF stimulates thymidine incorporation by type II cells. The data of Ad5-KGF are the same as Figure 2. Data of other adenoviruses represent the means ± SEM of five experiments. Ad5-KGF MOI of 2, 4, and 8 are significantly different from Ad5-LacZ, TGF-1 (active), TGF-1 (latent) and IL-4 (P < 0.0001).

KGF Protein Production by Ad5-KGF Infected Type II Cells

Type II cells infected with Ad5-KGF at a MOI of 2, 4, or 8 PFU/cell produced 10.9 ± 1.09, 19.2 ± 1.02, and 26.0 ± 1.00 ng/ml of KGF, respectively, in culture supernatants as measured by ELISA (Figure 4). No KGF was detected in type II cell cultures infected with other viruses (data not shown). Western blotting was done to confirm that the ELISA activity detected in culture supernatants was KGF (Figure 5). KGF in the type II cell culture medium was concentrated and partially purified with heparin sepharose and tested by Western blot analysis with goat polyclonal antibody (R&D Systems Inc., Minneapolis, MN) and rabbit antigoat IgG (Bio-Rad Laboratories, Hercules, CA). The amounts of KGF protein measured by ELISA were 3.52, 6.46, and 8.50 ng for Ad5-KGF MOI of 2, 4, and 8 PFU/cell, respectively. The size of the secreted KGF was approximately 20 kD and hrKGF produced in E. coli is approximately 16 kD, which suggests that the KGF produced by type II cells was glycosylated as expected.

View larger version (25K): [in this window] [in a new window]   Figure 4.   KGF ELISA of type II cell medium infected with Ad5-KGF. KGF concentration in the medium increases in a dose-dependent manner (10.9 ± 1.1 to 26.0 ± 1.0 ng/ml) as the MOI of Ad5-KGF is increased from 2 to 8 PFU/cell. Type II cells infected with Ad5-KGF produce significant amounts of KGF protein measured by KGF ELISA. Data represent the means ± SEM of eight experiments. * Significantly different from Ad5-KGF MOI of 2 (P < 0.0001).  Significantly different from Ad5-KGF MOI of 4 (P < 0.0001).

View larger version (43K): [in this window] [in a new window]   Figure 5.   Western blotting of human KGF protein released into type II cell culture medium infected with Ad5-KGF. The amount of KGF protein in type II cell culture medium increases dose dependently (Ad5-KGF MOI of 2, 4, and 8). The molecular mass of KGF produced by Ad5-KGF is approximately 20 kD and human rKGF is approximately 16 kD. Type II cell media (10 ml) infected with Ad5-KGF were collected 72 h after plating and purified on heparin sepharose. The amounts of KGF loaded were measured by ELISA and were 3.52, 6.46, and 8.50 ng for Ad5-KGF MOI of 2, 4, and 8, respectively. A total of 5 ng human rKGF was loaded as the standard.

Because the thymidine incorporation for type II cells infected with Ad5-KGF was much greater than that observed with cells treated with recombinant KGF, KGF was purified from the culture medium on heparin sepharose, quantitated by ELISA, and assayed on type II cells. This was done to test the hypothesis that glycosylated KGF produced in rat type II cells would be significantly more potent than nonglycosylated KGF produced in E. coli. However, glycosylated KGF and nonglycosylated KGF did not show any significant differences in thymidine incorporation in concentrations ranging from 0.01 to 100 ng/ml.

-Galactosidase Stain In Vitro

To evaluate the extent of the adenovirus infection of type II cells in vitro, we infected type II cells with Ad5-LacZ and detected the infected cells by their expression of -galactosidase. Type II cells were infected with Ad5-LacZ or Ad5-KGF at a MOI of 8 PFU/cell for 24 h after isolation, cultured for an additional 48 h, fixed for 1 h, and then stained with X-gal solution overnight at 37°C. Nearly all cells expressed LacZ (Figure 6A). No staining for -galactosidase was found in type II cells infected with Ad5-KGF.

View larger version (122K): [in this window] [in a new window]   Figure 6.   Morphologic and histochemical effects of Ad5-KGF. (A) -galactosidase stain of type II cells infected with Ad5-LacZ. Type II cells were infected with Ad5- LacZ at an MOI of 8. Staining for -galactosidase demonstrates that after 72 h nearly all the cells express lacZ. Type II cells were fixed for 1 h with fix solution (2% paraformaldehyde/2 mM MgCl2 in PBS) and stained overnight at 37°C with staining solution (1 mg/ ml X-gal, 2 mM MgCl2, 4 mM K4Fe[CN]6, and K3Fe[CN]6 in PBS). Cells infected with Ad5-KGF or uninfected type II cells showed no -galactosidase activity. (B) -galactosidase stain in vivo. The left lung was instilled with Ad-LacZ 109 PFU, removed, and then stained with the X-gal solution 24 h later (right). Rat lungs instilled with PBS containing 10% glycerol were also incubated with the X-gal solution 24 h later and were unstained (left). (C) Histology of rat left lung instilled with PBS containing 10% glycerol at Day 3. Histology showed almost normal architecture consisting of flat and thin alveolar septae in the hematoxylin and eosin section (original magnification: ×200). The bars in sections C through H represent 20 µM. (D) Histology of rat left lung instilled with Ad5-LacZ 109 PFU at Day 3. Ad5-LacZ- instilled lung showed a mild inflammatory cell infiltration in some areas at Day 3, but it was not severe (original magnification: ×200). (E) Histology of rat left lung instilled with Ad5-KGF 109 PFU at Day 3. Intrabronchial instillation of Ad5-KGF produced marked alveolar type II cell hyperplasia in left lung at Day 3. This section was stained with hematoxylin and eosin (original magnification: ×200). (F) Immunochemistry of BrdU. Ad5-KGF caused a marked proliferation of cuboidal cells lining alveolar septae and most of these cells expressed BrdU, a marker of DNA synthesis (original magnification: ×200). (G) Immunochemistry of SP-C. Most of the alveolar lining cells contained immunoreactive SP-C protein, which is a marker specific for the alveolar type II cells (original magnification: ×200). (H) In situ hybridization of hKGF in rat left lung instilled with Ad5-KGF 109 PFU at Day 3. Only a few cells express KGF by in situ hybridization at Day 3 and these cells are designated by arrows (original magnification: ×200).

-Galactosidase Stain In Vivo

To determine the ability to infect the distal lung and define a viral dose to be used in vivo, 10⁹ PFU of Ad5-LacZ was instilled into the left lungs of Fischer 344 rats, and -galactosidase stain was used to detect viral distribution 24 h later. -galactosidase was detected over most of the surface of the lung (Figure 6B, right). A left lung instilled with PBS containing 10% glycerol was also stained with X-gal solution because there is some endogenous -galactosidase in rat lung. No staining for LacZ at pH 8.3 was found in lungs instilled with PBS (Figure 6B, left). This result shows that adenovirus instilled by this method is effective for infection in the distal portions of the rat lung.

Rat Lungs Instilled with Ad5-KGF, Ad5-LacZ, and PBS

To determine the effect of Ad5-KGF in vivo, we instilled 10⁹ PFU of Ad5-KGF into the left lungs of rats. Lungs instilled with PBS containing 10% glycerol showed normal histology consisting of thin alveolar septae (Figure 6C). In contrast, intrabronchial instillation of Ad5-KGF resulted in a marked increase of alveolar type II cells at Day 3 in hematoxylin and eosin stained sections (Figure 6E). About 50% of the left lung showed alveolar type II cell hyperplasia. Most of these cells expressed BrdU, a marker of DNA synthesis (Figure 6F), and contained immunoreactive SP-C protein, which is specific for alveolar type II cells (Figure 6G). Conversely, Ad5-LacZ 10⁹ PFU showed no obvious proliferation of cells along the alveolar septae at Day 3, but there was some inflammation around vessels, especially pulmonary veins (Figure 6D). Figure 7 shows other sections at higher power to provide a better image of the epithelial and interstitial changes. In general, there was significant type II cell proliferation as seen by SP-C immunostaining, routine histology, and electron microscopy (data not shown). The remarkable finding was that on one side of the alveolar wall there was extensive type II cell hyperplasia, whereas on the opposite side there were apparently normal type I cells. Although this can be seen in the histology sections and electron micrographs, it is most clearly seen by staining with the tomato lectin (Lycopersicon esculentum), which stains type I cells but not type II cells.

View larger version (62K): [in this window] [in a new window]   Figure 7.   Histology and staining with Lycopersicon esculentum lectin. Lungs were fixed with 4% paraformalhyde 3 d after instillation of 109 PFU Ad5-KGF. (A) Staining with Lysopersicon esculentum, which stains type I cells and some macrophages. (B) A section stained with hematoxylin and eosin.

KGF in Lavage and In Situ Hybridization

To determine the extent and duration of KGF expression, we measured KGF in lavage from the left lungs of rats infected with Ad5-KGF 10⁹ PFU and performed in situ hybridization. KGF concentrations in lavage fluid at Days 2, 3, and 7 after instillation were 3.60 ± 1.88, 0.67 ± 0.33, and 0.51 ± 0.19 ng/ml, respectively (Figure 8). No KGF was detected in lavage fluid from Ad5-LacZ- and PBS-instilled animals. By in situ hybridization, only a few cells per high powered field showed positive signals for KGF messenger RNA at Day 3 (Figure 6H). In these studies, we cannot define precisely which cells are expressing Ad5-KGF. In other studies with Ad5-LacZ, airway epithelial cells, alveolar epithelial cells, and alveolar macrophages all express LacZ (data not shown). Hybridization with a sense probe did not show any positive signal. We also investigated Ad5-LacZ- and PBS-instilled lungs and did not detect any positive signal for KGF.

View larger version (10K): [in this window] [in a new window]   Figure 8.   KGF protein in lavage fluid KGF was measured in lavage fluid from Days 2 to 28 after instillation of Ad5-KGF by ELISA. KGF concentration in lavage fluid was below the level of detection for the Ad5-LacZ and PBS groups. Data represent the means ± SEM of four or five rats for each day.

Total Number of Cells in Lavage

To evaluate the inflammation of lungs infected with adenoviruses, we measured total number of cells in lavage. Ad5-LacZ produced an increase in inflammatory cells in lavage fluid on Days 3 and 7 (Figure 9). There was no significant increase in total cells in lavage in animals instilled with PBS or Ad5-KGF. Total number of cells in lavage indicated there was more inflammation in the lungs of rats instilled with Ad5-LacZ than with Ad5-KGF, and this result was consistent with the histology.

View larger version (34K): [in this window] [in a new window]   Figure 9.   Total number of cells in BAL fluid. Total number of cells in lavage was increased in Ad5-LacZ-treated animals on Days 3 and 7 compared with the PBS and uninstilled groups on the same day. There were no significant increases at all time points in the PBS or Ad5-KGF groups. Data represent the means ± SEM of four or five rats each day per group. * Significantly different from uninstilled group (P < 0.05).  Significantly different from PBS group on the same day (P < 0.05). # Significantly different from Ad5-LacZ group on the same day (P < 0.05).

SP-A and SP-D in Lavage and SP-D in Serum

KGF has been shown to increase the expression of SP-A and SP-D in cultured alveolar type II cells in vitro (9), so we evaluated SP-A in lavage and SP-D in lavage and serum. Lavage SP-A was increased on Days 2 and 3 after intrabronchial administration of Ad5-KGF (Figure 10A). The increase of SP-A in the Ad5-KGF group reached maximum at Day 3 and gradually decreased after Day 7. In the animals that received Ad5-KGF, the SP-A concentration in lavage fluid was 6.42 ± 0.84, 7.65 ± 1.31, and 4.76 ± 0.75 µg/ml, at 2, 3, and 7 d after instillation, respectively. SP-A level in uninstilled rats was 2.86 ± 0.90 µg/ml, and the SP-A in lavage fluid of PBS-treated rats was similar to that of the uninstilled controls. There was no statistical increase in lavage fluid SP-A after Ad5-LacZ instillation. SP-D in lavage fluid was increased on days 2, 3, and 7 after instillation of Ad5-KGF (Figure 10B). The values on Days 2, 3, and 7 were 380 ± 35.5, 521 ± 64.6, and 469 ± 26.6 ng/ml, respectively, and greater than untreated controls (170 ± 3.46 ng/ ml). These changes in SP-A and SP-D reflected the extent of type II cell hyperplasia observed in histologic sections. In Ad5-LacZ-instilled lungs, lavage SP-D levels were elevated on day 7. We also investigated the lung histology of mice infected with Ad5-LacZ. The virus Ad5-LacZ induced mild to moderate inflammation between Days 3 and 21, and the extent of inflammation was maximum at Day 7. The increase in lavage levels of SP-D in Ad5-LacZ was similar to the extent of inflammation seen in the histologic sections. Lavage fluid SP-D levels in lungs instilled with PBS remained low; the PBS-instilled lungs had near normal histology.

View larger version (41K): [in this window] [in a new window]   View larger version (39K): [in this window] [in a new window]   View larger version (38K): [in this window] [in a new window]   Figure 10.   Lavage SP-A, lavage SP-D, and serum SP-D. (A) SP-A in lavage fluid was increased on Days 2 and 3 after intrabronchial administration of Ad5-KGF. The increase of SP-A in the Ad5-KGF group reached maximum at Day 3 and decreased to the control level by Day 7. PBS- and Ad5-LacZ-treated rats did not differ from the values for the uninstilled rats at all time points. (B) SP-D in lavage fluid in the Ad5-KGF-treated group was increased at Day 2, reached a maximum at Day 3, and remained elevated until after Day 14. In the Ad5-LacZ group, lavage SP-D was increased at Day 7. There was a slight increase in the PBS group on Day 2, but it was not significant. (C) The changes in serum SP-D were similar to the pattern for lavage SP-D. Serum SP-D in Ad5-KGF was increased on Day 2, reached maximum at Day 3, and remained elevated until Day 28. Serum SP-D in the Ad5-LacZ group was increased on Days 14, 21, and 28. In the PBS group, serum SP-D was not increased at any time point. Data represent the means ± SEM of four or five rats each day per group. * Significantly different from uninstilled group (P < 0.05).  Significantly different from Ad5-LacZ group on the same day (P < 0.05). # Significantly different from Ad5-LacZ group on the same day (P < 0.05).

Serum SP-D showed similar changes to lavage fluid SP-D (Figure 10C). Serum SP-D in Ad5-KGF was increased after 2 days, reached maximum at Day 3, and remained elevated until Day 28. Serum SP-D concentration on Days 2, 3, 7, and 14 were 109 ± 13, 168 ± 22, 159 ± 15, and 125 ± 9 ng/ml, respectively, and greater than that seen in untreated controls (23 ± 3 ng/ml). Serum SP-D in the Ad5-LacZ group was elevated on Days 14, 21, and 28. Serum SP-D was not increased in the PBS group compared with uninstilled controls at any time point.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Adenovirus encoding hKGF is effective in stimulating type II cell proliferation in vitro and in vivo. There are many reports that recombinant KGF administration induces transient alveolar type II cell proliferation in vitro and in vivo (6, 10, 11, 13). We created a replication-deficient adenovirus that expresses hKGF under a mCMV promoter. We demonstrated that Ad5-KGF stimulates the production and secretion of KGF by alveolar epithelial cells in vitro. The expressed protein was isolated from Ad5-KGF-infected type II cell cultures on heparin sepharose and shown to be the estimated size for glycosylated KGF. To test that these adenoviruses would infect distal lung epithelial cells, type II cells were infected with Ad5-Lac Z in vitro and demonstrated that nearly all the cells were able to be infected. Ad5-KGF instilled in vivo produced extensive type II cell hyperplasia that was present 2 d after instillation, peaked from Days 3 to 7, and was absent by Day 28. There was no residual fibrosis or chronic inflammation after the type II cell hyperplasia resolved.

Adenoviral vectors efficiently transfer foreign genes into lung tissue in vivo (29, 30). However, adenovirus-mediated gene expression is transient and may be associated with a significant inflammatory response, depending on the viral dose. Yei and coworkers (30) reported that adenovirus-mediated gene transfer to the airway epithelium of cotton rats is accompanied by dose- and time-dependent inflammation. They used doses ranging from 5 × 10⁷ PFU to 5 × 10⁹ PFU. They showed that adenovirus-induced inflammation commenced 3 d after administration, peaked at Day 4, and continued until Day 22. In our experiments, we instilled 10⁹ PFU into the left lungs of Fischer 344 rats. This dose produced extensive type II cell hyperplasia at Day 3 but later produced lung inflammation, although less than a comparable dose of Ad5-LacZ. The concentration of adenovirus to produce maximum gene expression and limited inflammation will likely depend on the strain of rat. In our experiments, 10⁹ PFU Ad5-LacZ induced inflammation at Day 3 with a peak at Day 7. The lungs returned almost to normal by Day 28. The lungs of animals instilled with Ad5-LacZ appeared to have more inflammation and had more total cells and percentage of neutrophils than did those of animals instilled with Ad5-KGF. Xing and colleagues (29) demonstrated that total number of bronchoalveolar lavage (BAL) cells recovered from Sprague-Dawley rats infected with Ad5-LacZ 2 × 10⁸ PFU was 4.8 × 10⁶ at Day 7. Our results showed that 9.5 × 10⁶ cells were recovered from a rat left lung infected with 10⁹ PFU Ad5-LacZ at Day 7 and 5.1 × 10⁶ cells were recovered from that infected with 10⁹ PFU Ad5-KGF at Day 7. Unfortunately, this inflammatory response limits the use of these vectors in treating acute lung injury. The host response is mainly to the virus but is also to the foreign protein LacZ. Yang and associates (31) demonstrated a response to both the virus and the transgene after the instillation of a LacZ expressing virus but that the response to the virus itself was sufficient to account for the inflammatory host response. We cannot exclude the possibility that some of the inflammation associated with the Ad5-LacZ is due to a host response to the transgene.

KGF has been shown to increase the expression of SP-A and SP-D in cultured alveolar type II cells in vitro (9). SP-D is synthesized by both type II and nonciliated bronchiolar cells in the rodent lung. In Clara or nonciliated bronchiolar cells of rodents, SP-D is found in the dense secretory granules (32). Measurements of SP-D in serum appear useful for evaluating parenchymal lung diseases. The serum SP-D concentration significantly increases in patients with pulmonary alveolar proteinosis, idiopathic pulmonary fibrosis, interstitial pneumonia with collagen vascular disease, and acute respiratory distress syndrome (33, 34). In our experiments, Ad5-KGF increased serum and lavage SP-D levels consistent with the time and extent of alveolar type II cell hyperplasia seen in histologic sections. In a separate study, we demonstrated that instillation of KGF increased serum SP-D levels and that SP-D in rat serum can be detected by Western blotting (data not shown). Instillation of Ad-LacZ produced mild to moderate inflammation that was accompanied by an increase in serum levels of SP-D.

This replication-deficient adenovirus expressing KGF was designed for two purposes. The first was to develop a means of producing extensive type II cell hyperplasia in rats. In humans, it is very easy to find rows of hyperplastic type II cells in histopathology sections of patients with interstitial lung disease. However, it is much more difficult to find this extensive type II cell hyperplasia in rodent lungs with a variety of injuries, including bleomycin. Therefore in rodents, it has been difficult to investigate the biologic properties of hyperplastic type II cells and their ability to alter the inflammatory and fibroproliferative response. There is evidence that type II cells produce neutrophil and monocyte chemotactic factors such as macrophage inflammatory protein-2 and monocyte chemoattractant protein-1 and augment the fibrotic reaction by producing PDGF, TGF-, and TGF- (35, 36). However, there is also evidence that type II cells inhibit fibroblast proliferation (3). Ad5-KGF should make it possible to study the biologic properties of hyperplastic type II cells in vivo and permit their isolation for studies in vitro. The second purpose was to have a means of delivering KGF for a sustained period in a local area for the treatment of lung injury. Currently, because of the inflammation induced by adenoviral vectors, therapy of acute lung injury may not be possible. However, there are strategies for minimizing this problem. One is to create adenovirus with diminished amounts of viral proteins. New generations of adenovirus vectors are being developed. Another approach is to decrease the host response to the adenovirus. Three promising approaches are to delete the tumor necrosis factor (TNF)- receptor, to use soluble TNF- receptors to remove TNF from the extracellular fluid, and to overexpress the adenoviral protein E3-14.7K (37). The adenoviral protein E3-14.7 is a nonsecreted cytoplasmic protein that blocks the TNF--induced cytotoxicity during viral infections. Harrod and coworkers (38) have overexpressed this protein under the SP-C promoter in transgenic mice and demonstrated that these mice are relatively resistant to lipopolysaccharide, TNF-, and adenoviral vectors.

In summary, Ad5-KGF induced extensive type II cell hyperplasia in rats in vivo and there was no residual fibrosis. However, the accompanying inflammatory response limits their current use as therapeutic vehicles.