Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Plasmin proteolytic activity has been pathophysiologically implicated in several biologic processes, including fibrinolysis, embryogenesis, cancer cell invasion, and wound healing, specifically the abnormal fibroproliferative response to lung injury (1). The proteolytic activation of the ubiquitous zymogen plasminogen to the active enzyme, plasmin by tissue-type plasminogen activator (t-PA), is rapid and efficient. Furthermore, the action of t-PA on fibrin-bound plasminogen is enhanced approximately 1,000-fold in the presence of fibrin, thereby localizing the plasmin activity to sites of fibrin formation (8). t-PA and plasminogen may also bind to cell surface receptors in numerous cell types, further localizing plasmin activity to specific sites on the cell surface (9, 10). Given the potential importance of this protease on diverse biologic processes, we sought to develop vectors useful for study of its effects and its directed overexpression.

To this end, human adenoviral vectors have been successfully utilized to induce foreign gene expression into murine lungs for a number of genes, including the cystic fibrosis transmembrane conductance regulator, interleukin-1 receptor antagonist, transforming growth factor , and soluble tumor necrosis factor receptor (11). The E1-deleted, E1 foreign gene-inserted, serotype 5 adenoviral vector carries the advantages of replication deficiency and respiratory epithelial cell tropism. Although the mouse is not the natural host for this virus, there is a host response to intrapulmonary adenovirus (Ad) that is characterized by an alveolar peribronchial and perivascular infiltration of inflammatory cells as well as a humoral immune response (16). This host response may place limits on both the maximal tolerable dose of virus, the duration of foreign gene expression, as well as result in lung injury. However, the relatively short time course of numerous experimental lung injury/fibrosis models (< 1 mo) that parallels that of human acute lung injury provide a clinically relevant system with which to test the physiologic effects of transient foreign gene expression.

An adenoviral, E1-deleted, E1-inserted, cytomegalovirus (CMV)-driven, recombinant human (rh) t-PA vector was successfully generated using homologous recombination in trans-complementing 293 cells. Using our vector, we found a 100-fold, dose-dependent, bronchoalveolar compartment-specific increase of functional rht-PA expression in murine lungs. This expression persisted for at least 2 wk after a single intratracheal instillation of 10⁸ plaque-forming units (PFU) of Ad-CMV-t-PA. These results allow for the further development of the therapeutic potential of bronchoalveolar compartment t-PA expression using in vivo experimental model systems.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Materials

The human tissue-type plasminogen activator full-length cDNA was kindly provided by Dr. F. Booyse (University of Alabama at Birmingham). Cell culture media (DMEM/ Ham's F-12) was obtained from the UAB Comprehensive Cancer Center Media Preparation Facility. Meyer's hematoxylin, fetal bovine serum, glutaraldehyde, ferricyanide, and ferrocyanide were obtained from Sigma Chemical Co. (St. Louis, MO). Opti-mem media was obtained from Life Sciences Technologies (Gaithersburg, MD). Taq polymerase was obtained from Perkin Elmer Corp. (Norwalk, CT), and oligonucleotide primers for polymerase chain reaction were obtained from DNA International (Lake Oswego, OR). Sephadex G-25 columns were obtained from Pharmacia Biotech (Uppsala, Sweden). X-gal and DOTAP liposome transfection reagent were obtained from Boehringer Mannheim (Indianapolis, IN). Glu-plasminogen and cyanogen bromide digests of fibrinogen were obtained from American Diagnostica (New Haven, CT), and the plasmin-sensitive chromogenic substrate, S-2251, was obtained from Kabi (Franklin, OH). Tissue preservation media was obtained from Triangle Biomedical Sciences (Durham, NC). Avertin (2,2,2,tri-bromoethanol/3'-amyl alcohol) was obtained from Aldrich Chemical Co. (Milwaukee, WI). Goat -human t-PA IgG was obtained from Biopool (Ventura, CA), and biotin-tagged rabbit -goat IgG, streptavidin-peroxidase conjugate, and the peroxidase-substrate aminoethyl carbazole were obtained from Zymed Immunochemicals (S. San Francisco, CA). Low-temperature-melting agarose was obtained from FMC Bioproducts (Rockland, ME).

Preparation, Amplification, and Purification of the Adenoviral-t-PA Construct

The full-length cDNA of rht-PA was ligated to the polylinker region of the adenoviral vector pACCMVpLpARS(+) to generate the construct pAdCMV-tPA (Figure 1). pAdCMVt-PA (2 µg) and pJM17 (3 µg [19-21]) were co-transfected into and allowed to recombine in subconfluent E1A trans-complementing 293 cells (5 × 10⁵ cells/60-mm-diam dish) as previously described (Figure 1) (21). Confirmation of the generation of the rht-PA encoding adenovirus was done by performing the polymerase chain reaction on the crude cell lysates using adenoviral tail fiber-specific primers (5'-TGGGGCTATACTACTGAATGAAAAATGAC-3', 5'-GGGACAGTTCAAAGTGCTCATAT-3') and t-PA-specific primers (5'-CCATGATCCTGATAGGCAAG-3', 5'-ACAGTCTAGCATGAGCCTCC-3') (36 cycles; 92°C dissociation, 60°C annealing, 72°C extension). The virus was amplified in 293 cells, purified by two CsCl gradient separations, and desalted over a G-25 sizing column. Resultant viral titers, in plaque-forming units, were determined by directly counting the number of viral plaques present in 293 cell monolayers, as described (22). The recombinant adenovirus Ad-CMV--galactosidase (Ad-CMV--Gal) was kindly provided by Dr. D. C. Tang (University of Alabama at Birmingham) and has been described previously (23).

View larger version (20K): [in this window] [in a new window]   Figure 1.   Construction and validation of rht-PA-encoding adenovirus (rAdCMVtPA). pAd5CMV-htPA (rht-PA-encoding shuttle adenovirus vector) and pJM17 (packaging-deficient adenovirus, serotype 5 genome) were co-transfected into E1 trans-complementing 293 cells. Intracellular, homologous recombination of the two plasmids yielded rAdCMVtPA (Ad-CMV-t-PA), an E1-deleted, E1 rht-PA-inserted, CMV promoter-driven rht-PA- encoding adenoviral construct. Standard map units (mu) are indicated.

Animal Protocol

All animal interventions have been approved by the AAALAC accredited Animal Resources Facility at the University of Alabama at Birmingham and were performed under the guidelines for animal care recommended by the American Physiological Society. All animals were pathogen-free, housed in micro-isolator cages, and fed with autoclaved food and water. The animals (C57Bl/6 female mice, 18 to 20 g body weight) were anesthetized with intraperitoneally administered Avertin (0.013 ml of a 2.5% solution/g body weight), the trachea was dissected free under sterile conditions, and a 0.25-mm (outer diameter) intratracheal catheter was inserted to the level of the carina. The recombinant adenoviral constructs (in 100 µl sterile saline/0.1% sterile endotoxin-free bovine serum albumin [BSA]) or vehicle (100 µl sterile saline/0.1% BSA) were slowly instilled into the lungs over a period of 20 min. The skin wound was closed with absorbable sutures, and the animals were allowed to recover. Before tissue harvest, the animals were euthanized and exsanguinated, and the lungs were perfused via the pulmonary artery with 10 ml of phosphate-buffered saline at 4°C through the right ventricle during lung inflation. Tissues were either minced and snap-frozen in liquid nitrogen for biochemical analysis, or the lungs were fixed in inflation with tissue-preservation media for histologic and immunohistochemical analysis, as described by the investigator (3, 24).

Biochemical Analysis of t-PA Expression

Antigenic t-PA and functional t-PA levels were measured in tissue homogenate supernatants. Snap-frozen tissue was homogenized in 10 vol of cold 0.25% TX-100/Tris-buffered saline (TBS) (pH 7.4) in a polytron (Brinkmann Instruments, Westbury, NY) and extracted by centrifugation (14,000 × g for 20 min at 4°C).

t-PA antigen in lung lysates and from cell-conditioned media was characterized by Western blotting using a polyclonal goat anti-human t-PA IgG by methods described previously (24). Human t-PA antigen levels in tissue extracts and cell-conditioned media were quantified using a commercially available, double antibody sandwich, enzyme-linked immunosorbent assay (ELISA) method (Biopool) (2, 25). Plate wells are precoated with polyclonal goat -human t-PA IgG/nonimmune goat IgG and samples or standards (0 to 40 ng/ml human t-PA) are incubated in triplicate for 1 h at 21°C while rotating the plate. The horseradish peroxidase-conjugated secondary antibody (HRP--t-PA Fab fragments) is added, and the wells are washed, followed by application of the colorimetric substrate 1,2 phenylenediamine dihydrochloride substrate in 0.2% H₂O₂. The reaction is stopped after 15 min at 21°C by acidification, and the absorbance is read at 490 nm on a microplate reader (MR5000; Dynatech, Chantilly, VA). The unknown sample values are determined by comparison with the known standards in the same buffer. In preliminary experiments, it was determined that this assay predominantly measures rht-PA based on a threshold for detection of rht-PA of 1.5 ng/ml but a lack of detection of up to 400 ng/ml of recombinant murine t-PA.

Plasminogen activator activity in lung lysates and from cell-conditioned media was characterized by fibrin autography using methods described previously (2). Briefly, samples were subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) under nonreducing conditions. The polyacrylamide gel was then placed on an agarose indicator film containing fibrinogen (2.5 mg/ml), plasminogen (10 µg/ml), and thrombin (0.4 NIH units). Incubation of the film at 37°C for 4 to 8 h results in the development of a transparent proteolytic zone in the otherwise opaque indicator film. This zone reflects plasminogen activator activity in a location corresponding to the apparent M_r of the activator.

t-PA activity was measured using a plasmin-based, chromogenic substrate assay described previously (2, 25). The samples were acidified to inhibit -2 antiplasmin followed by a standard incubation with Glu-plasminogen, cyanogen bromide digests of fibrinogen, and the plasmin-sensitive chromogenic substrate S-2251 (37°C for 6 to 12 h). The absorbance of the samples at 405 nm is compared with that of a known amount of human single-chain t-PA, calibrated against the NIBSAC international standard (Lot No. 86/ 670) that has undergone the same incubations. For selected samples, the t-PA activity was measured as above but in the presence of neutralizing -t-PA antibodies (10 and 50 µg/ml; American Diagnostica, Greenwich, CT) or the urokinase specific inhibitor amiloride (5 mM) (26).

Histochemical Analysis of -Galactosidase Expression and Immunohistochemical Analysis of t-PA Expression

For determination of -galactosidase (-Gal) expression by histochemical techniques, lungs were fixed in situ in inflation with 1.5% glutaraldehyde (10 min at 4°C), lung tissue slices were reproducibly selected from each lobe, and 1- to 2-mm lung slices were stained overnight at 37°C in high-pH, X-gal staining solution (1.0 mg/ml 5-bromo-4-chloro-3-indoyl--D-galactoside [X-gal], 0.02% Nonidet P-40, 1 mM MgCl₂, 10 mM potassium ferricyanide, 10 mM potassium ferrocyanide in Tris-HCl [pH 9.5]). The tissue was embedded in tissue-preservation media, sectioned (5 µm), mounted on glass slides, and examined under the light microscope.

For determination of the distribution of t-PA by immunohistochemical techniques, lung tissue was frozen in inflation by intratracheal instillation of tissue-preservation media, and 5-µm-thick cryostat sections were mounted on Vectabond (Vector Laboratories, Burlingame, CA)-coated slides. After blocking endogenous peroxidase activity (3% H₂O₂ in methanol), nonspecific protein binding was blocked with 10% heat-inactivated rabbit serum. This was followed by a primary antibody incubation (5 µg/ml goat -human t-PA IgG; Biopool) for 2 h at room temperature, by washing, and by application of secondary antibody (biotin-tagged rabbit -goat IgG [10 µg/ml]) and streptavidin- horseradish peroxidase conjugate. The peroxidase-substrate aminoethyl carbazole was applied, and the sections were counterstained with filtered Meyer's hematoxylin. For all histologic analyses, 10 randomly selected fields (magnification ×200) per lobe on four separate tissue sections were examined.

Statistical Analysis

For interval data (e.g., t-PA antigen, microplate t-PA activity), values from vehicle alone and adenoviral groups were compared by means of analysis of variance (ANOVA) (27). Where differences were detected, they were analyzed by the multiple-comparison procedure Newman-Keuls test (27). If data were found to have an inhomogenous variance, they were analyzed by a Kruskal-Wallis ANOVA based on ranks and by a multiple-comparison procedure using Dunn's method. Statistical significance was accepted at the P < 0.05 level for all analyses.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Expression of Functional t-PA Mediated by Ad-CMV-t-PA Construct In Vitro

The Ad-CMV-t-PA vector was constructed by homologous recombination in E1 trans-complementing 293 cells, validated using t-PA-specific primers (Figure 1), and tested for its capacity to express functional t-PA in vitro. A size and functional characterization of the expressed t-PA in vitro by immunoblotting and by fibrin autography indicated that the protein migrates at the expected M_r (60 kD), is capable of activating plasminogen, and will form complexes with murine PAI-1 (Figures 2A and 2B). rAd-CMV-t-PA-infected 293 cell supernatants contained higher t-PA antigen (400-fold) and t-PA activity (800-fold) levels (antigen: 42,995 ± 25,842 ng/ml versus 107 ± 17 ng/ml; P = 0.01; activity: 12,556 ± 2,494 IU/ml versus 16.05 ± 2.4 IU/ml; mean ± SD; P = 0.00006) as compared with that of Ad-CMV--gal virally infected cells or noninfected, which were no different from each other. Thus, rAd-CMV-t-PA will drive expression of high levels of functional rht-PA in vitro.

View larger version (21K): [in this window] [in a new window]   Figure 2.   Functional and immunologic characterization of adenoviral rht-PA expressed in vitro. Conditioned media were harvested from virally infected trans-complementing 293 cells as described in MATERIALS AND METHODS. Equal volumes of the media were immunoblotted for t-PA (panel A) or were separated by SDS-PAGE followed by fibrin autography (panel B) as described in MATERIALS AND METHODS. Note that a 1:200 dilution of the media from Ad-CMV-t-PA-infected cells was used for the fibrin autography assay (panel B). Molecular mass markers are indicated (in kilodaltons) at the left margin. (Panel A) Lane 1: Ad-CMV-t-PA-infected cell media (5 µl ); lane 2: Ad-CMV--Gal-infected cell media (5 µl); lane 3: noninfected cell media (5 µl); lane 4: rht-PA standard (5 ng). (Panel B) Lane 1: Ad-CMV-t-PA-infected cell media (15 µl of a 1:200 dilution); lane 2: Ad-CMV--Gal- infected cell media (15 µl); lane 3: noninfected cell media (15 µl); lane 4: human t-PA standard (0.5 IU).

Dose Response and Time Course of Adenovirally Mediated Expression of Functional t-PA In Vivo in Murine Lungs

To determine the dose response and time course of adenovirally mediated t-PA expression after a single intratracheal instillation in intact mice, antigenic and functional t-PA was measured by several complementary techniques. Concordant with the in vitro findings, the rht-PA protein expressed in whole lung lysates in vivo migrates at the appropriate M_r (60 kD) and is functionally active as qualitatively assessed by Western blot and by fibrin autography (Figures 3A and 3B). The additional 90-kD, t-PA-immunoreactive band in panel A, lane 3 was also immunoreactive to PAI-1 and co-migrated with t-PA-PAI-1 complexes formed in vitro (Figure 3, lane 4); therefore, it likely represents the degraded form of t-PA-PAI-1 complexes. Ad-CMV-t-PA instillation induced a dramatic, dose-dependent (10⁶ to 10⁸ PFU) increase of lung t-PA antigen. We noted a maximum 100-fold increase of t-PA antigen at 10⁸ PFU/ animal, as compared with vehicle alone (HBS/0.1%BSA) (544.6 ± 132 ng/lung set, Ad-CMV-t-PA versus 4.96 ± 3.2 ng/lung set, vehicle; P = 0.00003) at 3 d after intratracheal instillation (Figure 4, left panel). In contrast, Ad-CMV--Gal instillation did not induce t-PA expression (Figure 4, both panels, open bars). The Ad-CMV-t-PA (10⁷ PFU)- induced expression was tissue specific, as 94.9% of the total t-PA antigen detected was extracted from the small airways and lung parenchyma, 0.3% was extracted from the trachea (0.15 ng ± 0.1 ng; mean ± SD), and the remaining 4.8% was extracted from the heart (1.33 ng ± 1.8 ng; mean ± SD), kidney (1.58 ng ± 0.07 ng; mean ± SD), and liver (not detectable) combined. Plasma t-PA was undetectable in vehicle-instilled or Ad-CMV-t-PA-instilled mice.

View larger version (21K): [in this window] [in a new window]   Figure 3.   Functional and immunologic characterization of adenoviral rht-PA expressed in vivo. Animals were instilled intratracheally with 108 PFU of either Ad-CMV-t-PA, Ad-CMV--Gal, or vehicle alone (HBS/0.1% BSA), and lung tissue was harvested 3 d later and extracted with TBS/0.2% TX-100/20 mM EDTA as described in MATERIALS AND METHODS. Equal volumes were separated by SDS-PAGE followed either by immunoblotting for t-PA (panel A) or by fibrin autography (panel B) as described in MATERIALS AND METHODS. Molecular mass markers (in kilodaltons) are indicated. (Panel A) Lane 1: Ad-CMV--Gal-instilled animals' lung extract (10 µl); lane 2: vehicle-instilled animals' lung extract (10 µl); lane 3: Ad-CMV-t-PA-instilled animals' lung extract (10 µl); lane 4: rht-PA-murine PAI-1 complexes (56 ng); lane 5: rht-PA standard (5 ng). (Panel B) Lane 1: vehicle-instilled animals' lung extract (15 µl); lane 2: Ad-CMV--Gal-instilled animals' lung extract (15 µl); lane 3: Ad-CMV-t-PA-instilled animals' lung extract (15 µl).

View larger version (16K): [in this window] [in a new window]   View larger version (14K): [in this window] [in a new window]   Figure 4.   t-PA content of Ad-CMV-tPA, Ad-CMV--Gal, and vehicle-instilled murine lungs. Lungs from animals that were instilled with either indicated vector (106 to 108 PFU of Ad-CMV-t-PA or 106 to 1010 PFU of Ad-CMV--Gal) or vehicle 3 d previously were homogenized in TBS/0.2% TX-100/20 mM EDTA as described in MATERIALS AND METHODS. t-PA antigen levels in lung lysates were determined by ELISA (left panel), and t-PA activity levels were determined by a plasmin-based chromogenic substrate assay (right panel) as described in MATERIALS AND METHODS. Vehicle is HBS/0.1% BSA. Open bars indicate t-PA antigen values from vehicle (bar 1)-instilled or pooled 106 to 1010 PFU of Ad-CMV--Gal (bar 2) (n = 6/dose)-instilled animals. Filled bars (bars 3 through 6) denote Ad-CMV-t-PA at the indicated plaque-forming units (n = 6/dose). (Left panel) t-PA antigen values. (Right panel) t-PA activity values. Data are plotted as mean ± SE. *P < 0.05 compared with vehicle-instilled animals' lungs.

A quantitative assay of plasminogen activator activity indicates that the expressed t-PA is functionally active in vivo with a maximal 4-fold induction above that of control levels (122 ± 82 IU/lung set, Ad-CMV-t-PA versus 29.3 ± 5.8 IU/lung set, vehicle; P = 0.005) 3 d after instillation of 10⁸ PFU of Ad-CMV-t-PA (Figure 4, right panel ). Curiously, the fibrin autography revealed an increase in the murine urokinase-type plasminogen activator (u-PA) ( 45 kD) lytic zone in the Ad-CMV-t-PA-instilled animals (Figure 3B). The increased u-PA zone on fibrin autography is consistent with our observation that 11 ± 2% of the total plasminogen activator activity of Ad-CMV-t-PA instilled lungs was suppressible with the u-PA specific inhibitor amiloride whereas no amiloride-suppressible activity was detected in vehicle-instilled lungs. The remainder (91 ± 3%) of the total plasminogen activator activity in Ad-CMV-t-PA instilled lungs was inhibited by neutralizing t-PA antibodies (50 µg/ml), while only 70 ± 11% was inhibited by t-PA antibodies in vehicle-instilled lungs. An examination of the time course of expression reveals that after a single intratracheal instillation of 10⁸ PFU of Ad-CMV-t-PA/animal, t-PA antigen was expressed in the lung at 50-fold over control levels for at least 14 d (Figure 5).

View larger version (14K): [in this window] [in a new window]   Figure 5.   Time course of pulmonary t-PA expression after a single intratracheal (IT) instillation of Ad-CMV-t-PA (108 PFU). Mice (n = 6/time point) were intratracheally instilled with 108 PFU of Ad-CMV-t-PA. Lung tissue was homogenized in TBS/ 0.2% TX-100/20 mM EDTA at the indicated times thereafter, and lysate t-PA antigen and activity levels were measured by ELISA and plasmin-based microplate assay, respectively, as described in MATERIALS AND METHODS. Solid bars denote t-PA antigen values, and open bars denote t-PA activity values. For comparison, values from the vehicle-instilled animals are from the 3 day time point. Data are plotted as mean ± SE. *P < 0.05 compared with vehicle-instilled animals' lungs.

Localization of Adenovirally Mediated t-PA Expression in Murine Lungs

Expression of adenovirally mediated rht-PA in intratracheally instilled murine lung tissue sections was localized by immunohistochemical techniques using the same primary antibody as used for the Western blots. Three days after intratracheal instillation of 10⁸ PFU Ad-CMV-t-PA, t-PA antigen expression was localized to the bronchial epithelium and bronchiolar epithelium, and to peribronchiolar alveolar cells (Figure 6A). Of 134 bronchial and bronchiolar regions counted, expression of t-PA antigen was noted in 81% of the airways, and of these > 80% demonstrated peribronchiolar alveolar cell t-PA expression. Serial sections stained with nonimmune goat IgG as well as vehicle-instilled animals' sections stained with anti-t-PA antibody or nonimmune IgG were negative in corresponding areas (Figures 6B through 6D). The distribution of -Gal activity 3 d after instillation of Ad-CMV--Gal was identical to that of adenovirally induced t-PA (data not shown). Thus, intratracheal AdCMV-t-PA instillation achieved bronchiolar and peribronchial-alveolar cell localized expression.

View larger version (97K): [in this window] [in a new window]   Figure 6.   Localization of Ad-CMV-t-PA-driven t-PA antigen in murine lungs. Mice (n = 3/condition) were instilled intratracheally with 108 PFU of Ad-CMV-t-PA or vehicle alone (HBS/0.1% BSA) 3 d before lung tissue harvest. Lung tissue was fixed in inflation, immunohistochemically stained with either goat anti-human t-PA antibody (panels A and C) or nonimmune goat IgG (panels B and D), and counterstained with hematoxylin as described in MATERIALS AND METHODS. A red-brown precipitate denotes positive staining. The symbol b marks bronchiolar lumens, and the symbol v marks vascular structures. Original magnification: ×200. Panels A and B. Ad-CMV-t-PA-instilled animals' lung. Panels C and D. Vehicle-instilled animals' lungs.

Histologic Analysis of Ad-CMV-t-PA Effects at 3 Days

As a measure of inflammation, total extractable protein levels and alveolar cell counts (10 × 160-magnification fields/tissue) were performed. In the Ad-CMV-t-PA group (10⁸ PFU) at 3 d, there was no increase in total extractable protein (6.7 ± 2.1 mg/lung set for vehicle-instilled animals versus 7.0 ± 1.3 mg/lung set in Ad-CMV-t-PA-instilled animals) and no increase in alveolar cell counts (262 ± 33 cells/grid at ×160 magnification for Ad-CMV-t-PA, 262 ± 37 cells/grid at ×160 magnification for vehicle) nor was there perivascular or peribronchial cell accumulation compared with that of vehicle alone (representative photomicrographs in Figures 6A through 6D). As a positive control, animals given 10¹⁰ PFU Ad-CMV--Gal 3 d previously demonstrated an increase in total extractable protein of 40% (to 9.6 ± 0.8 mg/lung set), as well as obvious collections of inflammatory cells surrounding vascular structures (data not shown).

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

The major finding in this report is the successful expression of functional rht-PA in murine lungs using an E1- inserted, replication-defective adenoviral-rht-PA construct. This expression was dramatically dose dependent, with a greater than 100-fold increase in t-PA antigen at 10⁸ PFU Ad-CMV-t-PA. It was tissue specific, with > 94% of the t-PA found within the lung parenchyma, and was maintained for at least 2 wk at levels 50-fold greater than that of controls. Although no obvious signs of inflammation were detected at 3 d, our measurements of inflammation were crude and were performed relatively early in the expected time course for adenovirus-induced lung inflammation.

The selective expression of rht-PA in cells or animals exposed to rAd-CMV-t-PA is supported by the concordant observations of the 400-fold increase in t-PA antigen in vitro; a dose-, time-, and tissue-specific increase in t-PA antigen in vivo (maximal 100-fold) (Figure 4, left panel ); the formation of immunoreactive t-PA-plasminogen activator inhibitor type 1 (PAI-1) complexes in vitro and in vivo (Figures 2A and 3A); the presence of a zone of plasminogen activation and fibrinolysis at the expected M_r for rht-PA (Figures 2B and 3B); and the concomitant increase in t-PA-specific activity (Figures 3B and 4B). Taken together, these data unequivocally demonstrate successful adenovirally mediated delivery and expression of active rht-PA to murine lungs.

The increased t-PA detected after instillation of Ad-CMV-t-PA is not due to systematic differences in extraction as total extractable protein levels (5,000 µg/ml) were not different between animals. Furthermore, a dose-dependent induction of murine t-PA by the adenovirus itself or its gene products is highly unlikely, as there was no dose response of t-PA to a five log range of Ad-CMV--Gal instillation. Although our ELISA and immunoblot analyses found that a 200-fold molar excess (400 ng/ml) of murine recombinant t-PA relative to human recombinant t-PA (2 ng/ml) was not measurable, wild-type murine t-PA was detected in vehicle-instilled and sham-treated lungs. This is most likely explained by the presence of antigenic epitopes in mammalian post-translationally modified t-PA molecules that are not present in the prokaryotically expressed, recombinant t-PA molecules.

The calculated "effective specific activity" (PA activity/ t-PA antigen) of the adenovirally expressed rht-PA in vitro (290,000 IU/mg) and in vivo (190,000 IU/mg) is similar, indicating that post-translational processing of the rht-PA in a murine system does not result in a loss of functional activity. Furthermore, despite a 100-fold increase in t-PA antigen in vivo, complexes with PAI-1 were the only t-PA- inhibitor complexes detected by immunoblotting, suggesting that other inhibitors for human t-PA in murine lung are not quantitatively important. In further support of this notion, free PAI-1 protein levels were, on average, 2-fold higher in t-PA-overexpressing animals with low t-PA activity relative to those with high t-PA activity by immunoblotting in selected samples (data not shown). However, we have not excluded the possibility that other significant non-PAI-1 inhibitors for murine t-PA exist in murine lungs from this analysis. Although the expected M_r for t-PA-PAI-1 complexes is approximately 100 to 110 kD, we observed a band at 90 kD that was immunoreactive to both t-PA and PAI-1 and that co-migrated with that of human t-PA-murine PAI-1 complexes formed in vitro (Figure 3A). Based on these concordant observations, the 90 kD t-PA-PAI-1 complex likely represents adenovirally expressed rht-PA bound to murine PAI-1 in its cleaved conformation. Alternatively, it may reflect rht-PA-PAI-1 complexes with a proteolytically degraded form of rht-PA or some other unique cross-species t-PA-PAI-1 interaction.

It is also interesting to note the increase in u-PA fibrinolytic activity in the Ad-CMV-t-PA-instilled animals as evidenced by the 45 kD band in the fibrin autography (Figure 3B) and its 11% suppressibility with amiloride in the microplate assay (26). We and others have shown that u-PA is present in the parenchymal compartment of normal murine lungs in the single chain, pro-enzyme form (3, 28). The generation of active plasmin in vivo, by the overexpressed t-PA, may activate this reserve of pro-u-PA by cleavage to its active two-chain form, thereby resulting in an observed increase in u-PA activity.

We know of one other published study where t-PA gene transfer was attempted. In that study, a 5-fold increase in t-PA antigen and activity was noted in human saphenous vein endothelial cell primary cultures 14 d after incubation with 10⁶ PFU of an LTR promoter-driven human t-PA using a retroviral vector system (29). Our vector system, where 100-fold and 400-fold increases in t-PA antigen were noted in vivo and in vitro, respectively, compares favorably to the prior study. Potential uses for our vector include the study of in vivo t-PA overexpression in rat or mouse lung injury and/or inhalational injury models. If administered via the intraperitoneal or intravenous route or through an occluded vessel, it is potentially effective in generating hepatic, pulmonary vascular, or arterial wall cell t-PA expression, as seen with marker gene-adenoviral constructs (12, 16, 30). Thus, the vector would be potentially useful in studying the role of t-PA in experimental models of hepatic disease, of vascular disease, or of pulmonary parenchymal disease.