Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The delivery of exogenous genetic material to a specific vascular bed is a major challenge in the successful application of gene therapy strategies in cardiovascular disease (1). Significant in vivo gene transfer to either the endothelial or smooth-muscle cell (SMC) layers of the vessel wall, whether by angioplasty, intraluminal instillation, or direct injection, has been possible only by the use of either adenoviral or retroviral vector-based systems. Using these techniques, several groups have demonstrated effective delivery of an exogenous gene either through reporter-gene transfection studies or by inducing biologic changes in the vessel by functional gene transfer (2). However, inflammation provoked by first-generation adenoviral vectors has limited the duration of gene expression in animal experiments (6, 7), and the clinical sequelae to this immune response has limited their utility in human trials (8, 9). The delivery of retroviral vectors results in the incorporation of the foreign DNA into the host genome with the potential for disruption of an important host gene or spontaneous viral replication, raising concerns for human somatic gene therapy (10). In addition, the intravascular delivery of viral vectors or liposome-complexed plasmid DNA results in unwanted gene transfer in distant (nontargeted) vascular beds, particularly the liver.

Cell-based methods of gene transfer rely upon the ex vivo transfection of a specific cell or cell line and the reintroduction of that transfected cell population into either the original host or an immunologically identical or immunosuppressed individual. This strategy has already been effective in the clinical treatment of certain human diseases (e.g., familial hypercholesterolemia) (13, 14), and if a nonviral method of ex vivo gene transfer is used, the immune complications related to viral protein expression can be completely avoided. Although cell-based gene transfer has been used in the systemic circulation, no investigations have been performed in the pulmonary vasculature where the low-pressure system and natural filtering function of the pulmonary microvasculature could provide ideal conditions for this method of delivery. In addition, the delivery of genetically engineered cells into the pulmonary circulation offers the promise of selective pulmonary vascular gene transfer, thus avoiding unwanted systemic effects.

Among the various pulmonary vascular pathologies that could potentially be amenable to this form of gene transfer, primary pulmonary hypertension (PPH) is an obvious target. This condition predominantly affects females in the third decade of life, and has a mean survival of less than three years from the time of diagnosis (15, 16). At present, few satisfactory long-term treatment options other than lung transplantation (17, 18) are available; and the principal therapy, continuous intravenous infusion of prostacyclin derivatives (19), is associated with significant morbidity (22). The underlying etiology of PPH remains unclear, however several groups have suggested that endothelial dysfunction associated with a decrease in expression or biologic activity of the vasodilator nitric oxide synthase (NOS), either alone or in combination with the overexpression of the potent vasoconstrictor endothelin (25), plays a critical role. Therefore, we hypothesized that the delivery of vascular cells engineered to overexpress NOS to the lung could restore the balance of vasoconstrictor and vasodilator endothelial factors and inhibit the development of pulmonary hypertension.

The purpose of this study was to determine the in vivo efficiency, the duration of transgene expression in the pulmonary vasculature, and the potential biologic efficacy of a cell-based gene-transfer approach. In the present report, we describe a method of nonviral in vitro transfection and the optimal conditions for transfection of primary pulmonary artery SMCs using this system. We have studied transplanted cell localization and survival using a fluorescent method of cell labeling, and have determined the time course of transgene expression in the pulmonary vasculature and parenchyma using cells transfected ex vivo with the reporter gene -galactosidase (Gal). Finally, we determined that the delivery of endothelial NOS (eNOS)- transfected cells in the monocrotaline model of pulmonary hypertension significantly reduced the development of right ventricular hypertension and adverse remodeling 4 wk after cell-based gene transfer and monocrotaline injection. These results suggest that a cell-based approach may be a useful strategy for achieving selective therapeutic gene transfer to the pulmonary vasculature.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Pulmonary Artery Explant Culture

Fisher 344 rats (Charles River Co., St. Constant, PQ, Canada) were obtained at 21 d of age and were killed by overdose with ketamine and xylazine. The main pulmonary artery was excised and transferred immediately into a phosphate-buffered saline (PBS) solution containing 2% penicillin and streptomycin (GIBCO BRL, Burlington, ON, Canada). The specimens were transfered to a sterile tissue-culture area where the adventitia was carefully removed with sterile forceps, the artery opened longitudinally, and the endothelium removed by abrasion of the intimal surface with a scalpel. The vessel was cut into approximately 4-mm-square pieces, which were placed intimal-surface down on individual fibronectin-coated (Sigma Chemical Co., Mississauga, ON, Canada) tissue-culture plates (Falcon; Becton Dickinson Canada, Mississauga, ON, Canada). The explants were then grown in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal calf serum (FCS) and 2% penicillamine and streptomycin (all from GIBCO BRL), in a humidified environment with 95% O₂ and 5% CO₂ at 37°C. Explants were passaged using 0.05% trypsin/ethylenediaminetetraacetic acid (GIB-CO BRL) once many cells of a thin, fusiform SMC phenotype could clearly be seen growing from the pulmonary artery segment, at which time the remaining explanted tissue was removed. The cells were then grown in DMEM with 10% FCS and 2% penicillamine and streptomycin until they were to be used in further experiments.

-Actin and von Willebrand Factor Fluorescent Staining

To confirm their SMC identity and rule out endothelial cell contamination, cells at the third passage were plated onto coverslips and grown until 70% confluent, at which time they were fixed in acetone at room temperature (RT) for 10 min. The cells were incubated with FCS for 30 min at 37°C to block nonspecific binding sites, and then with a monoclonal anti--actin antibody (5 µg/ml) (Boehringer Mannheim, Laval, PQ, Canada) and a rabbit-derived polyclonal anti-von Willebrand factor (anti-vWF) antibody (1:200 dilution) (Sigma) for 60 min at 37°C in a covered humidified chamber. Negative control coverslips were incubated with PBS for the same duration of time. The coverslips were then washed in PBS and incubated for 60 min at RT in a PBS solution containing a Cy3-conjugated donkey antimouse immunoglobulin (Ig)G antibody (1:200 dilution) (Jackson ImmunoResearch Laboratories, Mississauga, ON, Canada), a fluorescein isothiocyanate-conjugated goat antirabbit IgG antibody (1:200) (Jackson ImmunoResearch Laboratories), and Hoescht 33258 (Sigma), a fluorescent nuclear counterstain. The coverslips were again washed with PBS and mounted using a 1:1 solution of PBS and gycerol. Slides were examined using an Olympus BX50 epifluorescent microscope with standard fluorescein, rhodamine, and autofluorescent emission and excitation filters. For each coverslip, the immunofluorescence for actin, vWF, and the nuclear counterstain Hoescht was indicated as positive or negative. All of the explant-derived cultures were found to be at least 97% pure SMC with very rare endothelial contamination. This could be attributed to the vigorous debridement of the endothelial lining during the initiation of the explant and to early passaging with removal of the residual explant material.

Fluorescent Cell Labeling

Cells between the fifth and ninth passages were grown until 80% confluent and were then labeled with the viable fluorophore chloromethyl trimethyl rhodamine (CMTMR; Molecular Probes, Inc., Eugene, OR). CMTMR affords an accurate method of detecting ex vivo-labeled cells because the molecule undergoes irreversible esterification and glucoronidation after passing into the cytoplasm of a cell and thereby generates a membrane-impermeable final product. This active fluorophore is then unable to diffuse from the original labeled cell into adjacent cells or structures, and may be detected in vivo for several months. The fluorescent probe was prepared by dissolving the lyophilized product in dimethyl sulfoxide to a concentration of 10 mM. This solution was stored at 20°C and diluted to a final concentration of 25 µM in serum-free DMEM immediately before use. Cells were exposed to the labeling agent for 45 min, and were then washed with PBS twice and the regular medium (DMEM with 10% FCS and 2% penicillin and streptomycin) was replaced. The cells were grown overnight and harvested 24 h later for injection into the internal jugular vein of recipient Fisher 344 rats. A series of in vitro experiments was also performed by plating the cells on coverslips and incubating them with the fluorophore to determine the quality and duration of fluorescence over time.

Ex Vivo Cell Transfection with Either pCMV-Gal, NOS, or pcDNA3.1

The vector CMV-Gal (Clontech, Inc., Palo Alto, CA), which contains the Gal gene under the control of the cytomegalovirus (CMV) promoter/enhancer sequence, was used as a reporter gene to follow the course of in vivo transgene expression. The full-length coding sequence of human endothelial NOS was obtained as a generous gift from Dr. P. Marsden (University of Toronto), and was cloned into the mammalian expression vector pcDNA3.1 (Invitrogen, Carlsbad, CA) at the EcoR1 restriction site. The insert-deficient vector (pcDNA3.1) was used as a control for the monocrotaline experiments. All plasmid DNA was introduced into a JM109 strain of Escherichia coli via the heat-shock method of transformation, and bacteria were cultured overnight in luria broth media containing 100 µg/ml of ampicillin. The plasmids were then purified using an endotoxin-free purification kit according to the manufacturer's instructions (Qiagen Endotoxin-Free Maxi Kit; Qiagen, Inc., Mississauga, ON, Canada), producing plasmid DNA with an A₂₆₀/A₂₈₀ ratio of greater than 1.75 and a concentration of at least 1.0 µg/µl. SMCs between the fifth and ninth passages were transfected using Superfect (Qiagen). This product is composed of charged polycations around which the plasmid DNA coils in a manner similiar to histone-genomic DNA interactions. This Superfect-DNA complex then interacts with cell-surface receptors and is actively transported into the cytoplasm, after which the plasmid DNA can translocate to the nucleus. This technique allows the transfection reaction to be performed in the presence of serum and produces no toxic metabolites, unlike similiar lipid-based systems.

Cells between the fifth and ninth passages were trypsinized the day before transfection to obtain a density of 5 × 10⁵ cells/dish. The following day, 5 µg of plasmid DNA was mixed with 300 µl of serum-free DMEM in a sterile microcentrifuge tube. The plasmid-medium solution was then vortexed with 50 µl of Superfect transfection agent, after which the tubes were incubated for 10 min at RT. The transfection mixture was then combined with 3 ml of DMEM with 10% FCS and 2% penicillin and streptomycin and applied to the culture dishes after the cells had been washed with PBS. The solution was allowed to incubate at 37°C for 2 h, and the cells were then washed with PBS twice and the standard medium was replaced. The transfected cells were allowed to grow overnight and were then harvested 24 h later for animal injection. For every series of transfection reactions that was performed, one 100-mm dish of pulmonary artery SMCs was stained in vitro to provide an estimate of the transfection efficiency of the total series.

Animal Surgery

All animal procedures were approved by the Animal Care Committee of St. Michael's Hospital. Six-wk-old Fisher 344 rats (Charles River Co.) were anesthetized by intraperitoneal injection of xylazine (4.6 mg/kg) and ketamine (70 mg/kg), and the cervical area was shaved and cleaned with iodine and ethanol. A midcervical incision was made with a scalpel and the right internal, external, and common jugular veins were identified. Plastic tubing of 1 mm external diameter was connected to a 23-gauge needle and flushed with sterile saline. This tubing was then used to cannulate the external jugular vein and was introduced approximately 5 cm into the vein to what was estimated to be the superior vena caval level; rapid venous blood return was used to confirm the catheter location.

To determine the approximate number of cells to inject, measurement of right ventricular pressure after intrajugular cell delivery was performed. In these animals, both the left and right common jugular veins were identified. In the left jugular system, the plastic tubing was inserted to the level of the brachiocephalic vein. In the right jugular system, a 2 French microtip catheter (Millar Instruments, Inc., Houston, TX) was inserted and connected to a physiologic recorder (PPG Biomedical Systems Divisions, Pleasantville, NY) and a chart recorder (Astro-Med, Inc., West Warwick, RI). The catheter was passed through the superior vena cava and right atrium and into the right ventricle, where the systolic and diastolic pressures were recorded. The animals were then injected either with 1 ml of PBS (n = 3) or with 1 million pulmonary artery SMCs resuspended in 1 ml of PBS (n = 4) via the left internal jugular vein and the right ventricular pressure was measured at 5, 15, and 30 min. The catheters were then removed and the incision closed with 3-0 absorbable suture. The animals were allowed to recover for 24 h, at which time they were reanesthetized and the microtip catheter was reinserted via the right internal jugular vein into the right ventricle. The systolic and diastolic pressures were again recorded, and the animals were then killed. The systolic and diastolic pressures over time were compared between the PBS and cell-injected groups.

For experiments to determine the time course of cell survival and transgene expression in the lung, pulmonary artery SMCs that had been labeled with the fluorophore CMTMR or transfected with the plasmid vector CMV-Gal were trypsinized and centrifuged at 850 rpm for 5 min. The excess media was removed and the pellet of cells was resuspended in a total volume of 2 ml of PBS. A 50-µl aliquot of these resuspended cells was taken and counted on a hemocytometer grid to determine the total number of cells present per milliliter of PBS. The solution was then divided into 1-ml aliquots of approximately 500,000 cells and transfered in a sterile manner to the animal care facility. These cells were then resuspended by gentle vortexing and injected into the animals via the external jugular vein catheter. The solution was infused slowly over 1 to 2 min and the catheter was then flushed again with sterile saline before removal. The external jugular vein was ligated, the incision was closed with 3-0 interrupted absorbable sutures, and the animals were allowed to recover from surgery.

To determine whether cell-based gene transfer of eNOS would be capable of inhibiting the development of pulmonary hypertension in an animal model of disease, pulmonary artery SMCs that had been transfected with either eNOS under the control of the CMV enhancer/promoter (pNOS) or with pcDNA3.1 were trypsinized and divided into aliquots of 500,000 cells. Six-wk-old Fisher 344 rats were then anesthetized and injected subcutaneously with 80 mg/kg of monocrotaline (Aldrich Chemical Co., Milwaukee, WI), and then injected via a catheter in the external jugular vein with either 500,000 pNOS- or pcDNA3.1-transfected cells. The vein was tied off, the incision was closed in the normal fashion, and the animals were allowed to recover. At 28 d after injection, animals were reanesthetized and a Millar microtip catheter was inserted via the right internal jugular vein into the right ventricle. The right ventricular systolic pressures (RVSPs) were recorded, and the catheter was then inserted into the ascending aorta and the systemic arterial pressures recorded. The animals were then killed and the hearts excised. The right ventricular-to-left ventricular plus septal weight ratios (RV/LV ratios) were determined as an indicator of hypertrophic response to long-standing pulmonary hypertension. The systolic pressures and RV/LV ratios were compared between the pNOS and pcDNA3.1 groups.

Detection of Fluorescently Labeled Cells in Tissue

At 15 min, 48 h, 7 d, or 14 d after delivery of labeled cells (n = 5 for each time point except 15 min, where n = 4) or saline injection (negative control, n = 6), the animals were killed by anesthetic overdose and the chest cavity was opened. The pulmonary artery and trachea were flushed with saline, and the right and left lungs excised. Transverse slices were taken from the basal, medial, and apical segments of both lungs, and specimens obtained from the livers, spleens, kidneys, and gastrocnemius muscles. Tissue specimens were embedded in OCT compound (Sakura Finetek U.S.A., Inc., Torrance, CA) en face and then flash-frozen in liquid nitrogen. Ten-micron sections were cut from these frozen blocks at two different tissue levels separated by at least 200 µ, and these sections were then examined under a fluorescent microscope, using a rhodamine filter, and the number of intensely fluorescing cells were counted in each en face tissue specimen.

To estimate the total number of labeled cells present within the entire lung, the number of fluorescent cells was counted in each lung section and averaged over the number of sections counted. A mathematic approximation could be made of the total number of cells present within the lung by utilizing Simpson's rule for the volume of a truncated cone. This equation bases the total volume of a cone on the relative areas of three different sections such that:

The height of the lung was measured after organ harvesting, and the area of each transverse section was determined by planimetry. The average number of cells present in the three sections, divided by the total volume of these sections, yielded an estimate of the cell number per unit volume. By multiplying this number by the total lung volume, an estimate of the total number of cells within the lung could be obtained. To correct for the appearance of a single cell in multiple adjacent lung sections, rats were injected with 500,000 CMTMR-labeled cells and killed. The lungs were prepared, harvested, and embedded in the usual manner, and 20 serial sections, each 5 µ thick, were taken through the lung parenchyma. Each section was examined using a rhodamine filter, and distinct individual cells were identified and their presence was determined on adjacent sections. The number of 5-µ sections in which a single cell could be identified was counted and the average dimensions of a pulmonary artery SMC in vivo was obtained. The average diameter observed was 16.4 ± 1.22 µ. Therefore, the total number of cells calculated using Simpson's formula was multiplied by 0.61 to correct for the presence of one cell in, on average, each 1.64 10-µ section.

Detection of Gal Expression in Tissue

Animals were killed at three time points after cell-based gene transfer (48 h, 7 d, and 14 d, n = 7 for each time point), the pulmonary arteries were flushed with saline, and the tracheas were cannulated and flushed with 2% paraformaldehyde until the lungs were well inflated. Transverse slices were taken from the basal, medial, and apical segments of both lungs, and specimens obtained from the livers, spleens, kidneys, and gastrocnemius muscles of certain animals. The specimens were incubated in 2% paraformaldehyde with 0.2% glutaraldehyde for 1 h, then rinsed in PBS. The tissue was then incubated for 18 h at 37°C with a chromogen solution containing 0.2% 5-bromo-4-chloro-3-indolyl--D-galactoside (X-Gal; Boeh-ringer Mannheim), 5 mM potassium ferrocyanide, 5 mM potassium ferricyanide, and 2 mM magnesium chloride, all dissolved in PBS. The specimens were then rinsed in PBS, embedded in OCT compound, cut into 10-µ sections, and counterstained with neutral red.

The en face sections were examined microscopically, and the number of intensely blue staining cells was determined. Because one dish of cells was used for in vitro staining to determine the transfection efficiency for each reaction series, an estimate of the percentage of cells that were transfected with the reporter gene plasmid pCMV-Gal could be made for every animal. Using this information and the mathematic calculation described for approximating the number of fluorescent cells present, an estimate could be made of the total number of transfected cells remaining at the time of animal death.

Statistical Analysis

Data are presented as means ± standard error of the mean. Unpaired t tests were used to compare differences in right ventricular pressures between the saline and cell-injected groups, or right ventricular pressures and RV/LV ratios in the pNOS- and pcDNA3.1-transfected animals. Differences in the number of fluorescently labeled cells or transfected cells over time were assessed by means of an analysis of variance, with a post hoc analysis using Fisher's Protected Least Significant Difference test. A value of P < 0.05 was accepted to denote statistical significance.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Right Ventricular Pressure after Intravenous Cell Delivery

In preliminary experiments it was determined that administration of 5 × 10⁵ cells was well tolerated and resulted in little change in basal vascular resistance. However, 5 min after injection of 1 million SMCs into the internal jugular vein of recipient animals, the RVSP rose from a baseline of 24 ± 1 to 37 ± 1 mm Hg, as compared with the PBS- injected animals, where no significant increase in pressure occurred (P < 0.01 for cell-injected as compared with PBS-injected). After 15 min, the cell-injected animals demonstrated a progressive rise in RVSP to 46 ± 5 mm Hg (P < 0.02 for cell-injected as compared with PBS). The increase in RVSP plateaued approximately 30 min after injection at 51 ± 5 mm Hg in the cell-injected group as compared with 24 mm Hg in the PBS group (P < 0.01). These findings suggest that the intravenous injection of 1 million cells into the rat elevates pulmonary vascular resistance, likely by the physical obstruction of microvascular channels. At 24 h after injection of 1 × 10⁶ cells, there was a small but significant residual elevation in RVSP (29 ± 1 versus 23 ± 3 mm Hg in the saline-injected, P < 0.05), consistent with a decrease in the number of occluded pulmonary microvessels due to either transplanted cell loss or transmigration of the cells out of the vascular lumen. Because of the persistent rise in pulmonary pressure observed after injection of 1 × 10⁶ cells, we selected a dose of 500,000 cells per animal for the subsequent experiments.

Fluorescent Cell Labeling and In Vivo Detection of Fluorescently Labeled Cells

Immediately after incubation with the fluorophore CMTMR at a concentration of 25 µM, 100% of cultured cells were found to fluoresce intensely when examined under a rhodamine filter (Figure 1A). Cells were also examined 48 h and 7 d after labeling, and despite numerous cell divisions, 100% of the cells present on the coverslip continued to fluoresce brightly (data not shown).

View larger version (50K): [in this window] [in a new window]   Figure 1.   In vitro 100% of pulmonary artery SMCs fluoresce intensely immediately after incubation with the viable fluorophore CMTMR (A). Multiple cell-shaped fluorescent signals could be observed within the lung 15 min after jugular injection (B; identified by white arrows). A decrease in cell number and change in position to a perivascular location was seen at 48 h (C; identified by white arrows). The white scale bar is 50 µ in length.

Approximately 57 ± 5% of the labeled cells could be identified within the lung 15 min after intravenous delivery (Figure 1B). Most of these cells appeared to be lodged in the capillary circulation at the alveolar level. By 48 h after cell delivery, a significant decrease in the total number of fluorescent cells identified was noted (34 ± 7%, P < 0.01 as compared with the 15-min time point), and the location of the cells also appeared to have changed. Many bright fluorescent signals were now identified within the pulmonary parenchyma, or were lodged within the wall of small vascular structures (Figure 1C). At 7 and 14 d after injection, a further decrease in cell number was noted (16 ± 3 and 15 ± 5%, respectively, both P < 0.001 as compared with the 15-min time point), however the cells appeared to remain in approximately the same location (see Figure 2). No brightly fluorescent signals were seen in any of the lungs injected with nonlabeled SMCs.

View larger version (24K): [in this window] [in a new window]   Figure 2.   Percentage of cells detected after gene transfer. The percentage of CMTMR-labeled or Gal-transfected cells identified is plotted over time after internal jugular vein injection. Approximately 56% of CMTMR-labeled cells were seen 15 min after cell transplantation, decreasing to 34% at 48 h, 16% at 7 d, and 15% at 14 d. A similar decrease in cell number was observed after Gal-transfected cell injection, with 36% of cells being identified at 48 h, 28% at 7 d, and 25% at 14 d.

In the spleen, liver, and skeletal muscle tissue, no fluorescent signals were identified. In two out of four kidneys examined at 48 h after injection, irregular fluorescent signals could be identified. None of these appeared to conform to the shape of a whole cell, and they were presumed to represent those cells that were sheared or destroyed during cell injection or shortly thereafter. In addition, no fluorescent signals were identified in any organ outside of the lung 7 d after injection.

In Vivo Detection of Gal Expression

In a total of 15 separate transfection reactions using the pCMV-Gal plasmid, an average transfection efficiency of 13 ± 0.5% was obtained with the primary pulmonary artery SMCs in vitro (Figure 3A). No staining was seen in mock-transfected cultures.

View larger version (96K): [in this window] [in a new window]   Figure 3.   An in vitro transfection efficiency of approximately 15% in the primary pulmonary SMCs could be obtained using a nonviral technique, as seen here with the reporter gene Gal (A). At 48 h after cell-based gene transfer, multiple intense blue staining cells could be seen throughout the lung, predominantly located in alveolar septae adjacent to small vessels (B; identified by black arrows). At 14 d after injection, a decrease in cell number and in intensity of staining could be seen; however, those cells that remained appeared located within the pulmonary parenchyma (C; identified by black arrows). The black scale bar is 50 µ in length.

After incubation with the X-Gal chromogen solution, microscopic evidence of cell-based transgene expression could clearly be seen at 48 h after injection of pCMV-Gal-transfected SMCs into the internal jugular vein (n = 7), with multiple intense blue staining cells being seen throughout the lung (Figure 3B), representing approximately 36 ± 6% of the original transfected cells that were injected (see Figure 2). As with the fluorescently labeled cells, most of the Gal-expressing cells appeared to be lodged within the distal microvasculature. By 7 d after injection (n = 4), a decline in the number of Gal-positive cells was noted (28 ± 6%), and the intensity of staining also appeared to decrease. Again, the cells appeared to have migrated into either the pulmonary parenchyma or vascular wall. At 14 d (n = 6) after cell-based gene transfer, no further decrease in the number of cells identified was noted but the intensity of Gal staining of each cell had decreased further (Figure 3C). No evidence of Gal expression was detected in any of the lungs from animals (n = 4, 3 at 7 d and 1 at 14 d) injected with nontransfected SMCs. At all three time points, no evidence of pulmonary pathologyas determined by the presence of an abnormal polymorphonuclear or lymphocytic infiltrate, septal thickening, or alveolar destructioncould be detected.

In the spleens and skeletal muscles of animals injected with transfected or nontransfected SMCs, no blue staining cells could be identified. Livers and renal specimens from animals injected with either transfected (n = 5) or nontransfected (n = 3) SMCs occasionally showed faint blue staining across the cut edge of the tissue (n = 2 for each group), but no intense staining was seen at any time point and no staining was seen further than one high-power field into the tissue.

Effect of eNOS Gene Transfer in Monocrotaline-Induced Pulmonary Hypertension

At 4 wk after monocrotaline injection and delivery of cells transfected with the control vector pcDNA3.1 (n = 6), the RVSP was dramatically increased to 50 ± 4 mm Hg. However, in those animals which received the pNOS-transfected cells (n = 7), the elevation in right ventricular pressure was reduced to 33 ± 3 mm Hg (P < 0.005) (Figure 4). The RV/LV ratio was also significantly reduced in those animals injected with pNOS as compared with those injected with pcDNA3.1 (0.35 ± 0.02 versus 0.26 ± 0.02 in pcDNA3.1 versus pNOS, respectively, P < 0.05) (Figure 4). No difference in aortic pressure was noted.

View larger version (54K): [in this window] [in a new window]   Figure 4.   Right ventricular systolic pressure and RV/LV ratio 4 wk after monocrotaline injection and cell-based gene transfer. The RVSP is graphed in the left panel for the control-transfected animals (pcDNA3.1) and the animals injected with NOS-transfected SMCs (pNOS). At 4 wk after injection of the pulmonary endothelial toxin monocrotaline and transfected cells, the RVSP was increased to 50 mm Hg in the pcDNA3.1 group, but was significantly decreased to 33 mm Hg in the pNOS-transfected animals. Similarly, in the right panel, the RV/LV ratio, a measure of long-standing pulmonary and right ventricular hypertension, was significantly elevated to 0.35 in the pcDNA3.1 group, but was decreased to 0.26 mm Hg in the pNOS-transfected animals.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

This investigation represents the first evidence of successful cell-based gene transfer to the pulmonary vasculature, and provides one of the first demonstrations of non-viral based gene transfer to the lung. This method of delivery was associated with a high percentage of cells being retained within the lung at 48 h, as determined by both the fluorescent labeling technique and by the reporter gene studies using Gal, and with moderate but persistent gene expression over 14 d. These results roughly parallel what has previously been demonstrated with a viral-based method of intravascular gene delivery to the pulmonary vasculature (34, 35); however this cell-based technique avoids the use of a potentially immunogenic viral construct, was not associated with any significant pulmonary or systemic inflammation, and permits more selective transgene expression within the pulmonary microvasculature.

The present report addresses several key questions related to the feasibility of a cell-based gene transfer approach for the pulmonary circulation, including the survival of genetically engineered cells and the selectivity of their localization and transgene expression within the lungs. As demonstrated by our examination of distant organs, implanted cells were efficiently retained by the lungs. Nevertheless, a significant cell loss did occur over the 2-wk study period. Of the total number of cells retained, there was a decrease of ~ 50% over two distinct time periods, the first occurring during the cell injection procedure itself, and the second over the 2 to 7 d after cell implantation. Several explanations can be offered for both these early and late results. First, the ex vivo manipulation of the cells during trypsinization, centrifugation, and resuspension is likely to cause the destruction of a significant number. Second, while the shear forces within the venous circulation are low some cells may be unable to survive the transit through the right heart and into the pulmonary vascular tree, and subsequently may be unable to adhere to the endothelium and reestablish the necessary adhesion contacts. Because there was good agreement between both the fluorescent method of cell labeling used simply to assess cell survival and the reporter gene studies used to measure transgene expression, it seems unlikely that the decrease in number of Gal-positive cells at later time points can be attributed entirely to loss of the reporter plasmid from the cells, although this may be a contributory factor. A delayed cell loss occurred after 48 h, by which time the cells had lodged in the pulmonary vessels and likely had transmigrated out of the vascular space. The cause of this late cell dropout is entirely speculative; however, it is possible that apoptosis plays an important role. It is known that failure of cells to establish appropriate matrix integrin interactions can induce programmed cell death, and the change from an in vitro to an in vivo cellular environment, with the associated withdrawal of high serum and growth-factor levels, and the transition to a more differentiated phenotype can also trigger signals for apoptosis (36, 37).

In addition to the decrease in number of transplanted cells, there was a decrease in the intensity of reporter-gene staining at the individual cell level. This may be related to the CMV promoter used in this construct. This promoter is known to have several repeated sequences that bind nuclear factor (NF)-B and consequently cause increased RNA polymerase II binding. Therefore, when a transgene is under the control of the CMV promoter it is likely to be expressed at higher levels in tissues that overexpress this factor. The activation of NF-B is associated mainly with inflammatory cytokines (38), and no evidence of inflammation was detected with this syngeneic method of cell-based gene transfer. If no proinflammatory stimulus exists after the initial stress of transit through the vascular space, the activity of the CMV promoter may be decreased, the level of transgene transcription may fall, and the intensity of Gal staining may decrease (39). The gradual loss of plasmids from these rapidly dividing cells in the process of normal cell division, in combination with the progressive destruction of closed covalent plasmid DNA in the cell due to proteolytic activity, could also account for the decrease in staining intensity seen at 7 and 14 d.

The finding that most of the cells appeared to lodge within small pulmonary arterioles is consistent with the normal physiologic role the lung plays as an anatomical filter, and thus it would be expected that relatively large particles, such as resuspended cells, would become lodged within the pulmonary microvasculature. However, this "targeting" of cells to the precapillary resistance vessel bed in a highly selective manner may prove very useful in the treatment for certain pulmonary vascular disorders. The overexpression of a vasoactive gene, such as NOS, at the distal arteriolar level could provide a highly localized effect in a vascular region critical in the control of pulmonary vascular resistance and could amplify the biologic consequences of gene transfer. Indeed, significant reduction in RVSP was seen in monocrotaline-treated animals receiving eNOS-transfected vascular cells. This approach may therefore offer significant advantages over other pulmonary selective gene transfer strategies such as endotracheal gene delivery, which results in predominantly bronchial overexpression, or catheter-based pulmonary vascular gene transfer, which produces diffuse macrovascular and systemic overexpression (35, 40).

In the monocrotaline model, cell-based gene transfer of eNOS was surprisingly effective in minimizing the development of pulmonary hypertension. This significant effect occurred despite a relatively low overall mass of organ-specific transfection, and was likely due to the fact that the transfected cells were targeted, based on their size, to the precapillary pulmonary resistance vessels that play a critical role in controlling pulmonary pressure. This method of pulmonary vascular gene transfer may have benefits over existing techniques by minimizing the overall "load" of foreign transgene that is delivered to the body and may thereby theoretically reduce the incidence of undesired side effects. As well, the "transplantation" of syngeneic cells does not induce inflammation; thus, with certain modifications, such as the use of stably transfected cell lines, this approach may have the potential to provide long-term transgene expression.

This is the first description of eNOS gene transfer successfully reducing pulmonary pressure in a model of chronic pulmonary hypertension. Previous experiments have demonstrated that viral transfer of eNOS is effective in reducing the acute vasoconstrictive response to short-term hypoxic exposure (41). This is consistent with the role of reduced nitric oxide production or bioavailability in the pathogenesis of pulmonary hypertension (26, 28), at least in the monocrotaline model. Whether eNOS overexpression will be effective in preventing the development of chronic pulmonary hypertension in other animal models (i.e., hypoxia) or in reversing established pulmonary hypertension remains to be tested.

In conclusion, a cell-based method of gene transfer to the pulmonary vasculature provides an effective means of overexpressing a reporter gene for up to 2 wk in the pulmonary microcirculation with no evidence of local or systemic inflammation. This method of delivery may provide an effective nonviral form of gene therapy for certain pulmonary vascular disorders, such as primary pulmonary hypertension. By generating recombinant constructs with tissue-specific or even inducible promoter sequences, this method of gene transfer may also enable the lung to be used as a manufacturing plant for systemic protein therapy.