SV/luc is a T-antigen replacement vector carrying the firefly luciferase (luc) reporter gene 33. Preparation of the SV/luc vector was performed as previously described 34. Recombinant E1, E3-deleted adenoviral vectors (Ad/luc) were propagated in HEK293T cells, by commonly used methods35.

Animal procedures were approved by the Institutional Animal Care Ethical Committee. Under Ketamine/Xylasine/Isoflurane anesthesia, severe sepsis was induced in Sprague-Dawley rats using cecal ligation double puncture (2CLP) 14. 1.25 × 10⁸ IU/ml (infectious units) of SV/luc in 300 μl PBS were administered via a tracheal catheter to 2CLP and sham operated (SO) rats immediately after the procedure and to unoperated (UO) rats. In the positive control group, 10⁹ IU/Ad/luc in 300 μl PBS were administered in the same way to 2CLP animals 15. Negative control animals received PBS only.

The reporter gene used encodes the luciferase protein, an enzyme that converts luciferin in presence of oxygen and ATP to a bioluminescent substance. In vivo light detection was performed using the Roper Chemiluminescence Imaging System, (CCCD-cooled coupled charged camera) model LN/CCD-1300EB equipped with ST-133 controller and a 50 mm Nikon lens (Roper Scientific, Princeton Instrument, Trenton, NJ). The system enables detection of an internal light signal emerging from mammalian tissue. The measurement is the sum of the integrated light signal subtracted the background light emission of an area of equal size 36. A pseudo color image represents light intensity spectrum (from blue – least intense, to red – most intense).

Forty eight hours after vector administration, rats were reanesthetized. 125 mg/kg Beetle Luciferin (Promega, Madison, WI) was administered intraperitoneally, and luciferase activity was measured by photographing the animals first in light (to obtain the animal's image) and then in the dark (to measure light emission). A composite photograph was obtained by superimposing the two images. The average amount of fluorescence measured in light units per area, as detected by the CCCD camera represents signal intensity.

After the light emission measurement, at 48 hrs, the animals were sacrificed and internal organs harvested. One lung was removed, preserved in formalin, embedded in paraffin, cut at 5 μm thickness and stained with hematoxylin and eosin. H&E sections were evaluated for lung injury, degree of injury and its distribution within the lungs 14. Part of the spleen, liver, heart and one kidney were also harvested and prepared in the same way. The remaining part of the spleen, heart, liver, the second kidney and lung were frozen in fluid nitrogen and preserved at -80°C for RNA.

Immunostaining was performed using the procedure described by Lavon et al 37. For luciferase we used a primary rabbit anti-mouse polyclonal antibody (Cortex, San Leandro, CA) at 1:50 dilution, followed by a secondary goat anti-rabbit IgG antibody (biotin conjugated, Zymed, San Francisco, CA). VP1 capsid protein detection was performed using a rabbit anti-SV40 polyclonal primary antiserum 38 followed by FITC-conjugated goat anti-rabbit IgG secondary antiserum (Zymed). DAPI (blue) was used as nuclear counterstain. Double immunostaining for surfactant protein C (ProSP-C) and VP1 was performed in order to detect internalization of VP1 viral capsid protein into the alveolar cells. Serial lung sections were immunostained for VP1 using the same rabbit polyclonal antibody as above followed by goat anti-rabbit Cyte 5, and for ProSP-C using rabbit anti-ProSP-C polyclonal primary antiserum (Alomone Inc, Israel) followed by FITC-conjugated goat anti-mouse IgG. Immunohistochemical detection of Lymphocyte (CD3+ T cells) infiltration was performed using the immunoperoxidase avidin biotin methodology. A primary CD3+ T cell antibody (monoclonal mouse anti-rabbit, affinity purified, Biosource, Carlsbad, California) at 1:100 dilution was used, followed by a secondary goat anti-rabbit IgG antibody (biotin conjugated, Zymed).

Total RNAs from the frozen tissues were extracted using Trireagent (Sigma, Saint Louis, MO) according to the manufacture's instructions. A quantity of 2.5 μg of extracted RNA (for each sample) was subjected to reverse transcription using 400 u/μl RNAse, 0.5 μg/μl oligo dT, 200 u/μl M-MLV RT and 2.5 mM dNTPs. 2 μl of the reacting cDNA products were used as template for PCR. PCR was performed using PCR primers specific for Luciferase: 5'-TGGTCTGCCTAAAGGTGTCG-3' (forward) and 5'-ATGTAGTCTCAGTGAGCCC-3' (reverse), and carried out for 35 cycles. pGL3 Luciferase reporter vector (Promega) served as a positive control. GAPDH was used as a house keeping control gene, using PCR primers specific for GAPDH: 5'-ACCACAGTCCATGCCATCAC-3' and 5'-TCCACCACCCTGTTGCTGTA-3'.

In the present experiment, ARDS was induced secondary to intra-abdominal sepsis (2CLP). Starting 12 hours after 2CLP, and more prominently at 24 and 48 hours, 2CLP rats displayed the typical signs of sepsis: intense pallor/cyanosis of the mucous membranes, dirty fur with erected hairs, distended abdomen, diarrhea, tachypnea (40–60 breath per minute). Animal behavior was grossly abnormal: periods of agitation followed by periods of sleepiness along with limited movements and inability to feed.

Histological examination of the lung sections from 2CLP animals showed changes consistent with ARDS. Macroscopically: lungs were less aerated, covered with white fibrin patches and pleural fluid was found in different quantities. H&E stained sections depicted: alveoli filled with proteinaceous fluid, septal thickening, and interstitial neutrophilic infiltration (Figure 1). Mortality was present only in the septic (2CLP) animals. The 48 hrs mortality rate following 2CLP was similar to previously published numbers in sepsis induced ARDS 131415.

SV/luc or Ad/luc vectors were directly administered into the trachea of 2CLP and SO rats, immediately following the procedure. Using this route we achieved maximal vector concentration and distribution to the lung, limiting the systemic spread. Histopathologically there was no difference between the lungs of the septic rats given PBS or SV/luc (Figure 1). Mortality rate was similar in 2CLP animals given either PBS or SV/luc, suggesting that the vector itself did not add to the sepsis/ARDS induced mortality.

At 24 and 48 hours, all the animals were reanesthetized and photographed with the CCCD camera to detect light emission. At 48 hours all animals were sacrificed and the lungs were harvested.

At 24 hours, Luciferase activity was not detected in any animal groups. We explain this finding by the time needed for the reporter gene to express. However, at 48 hours, luc activity (detected as luminescence by the CCCD camera) was seen over the lung areas in 2CLP animals given SV/luc (Figure 2a). Low levels of luc activity were detected in the tracheostomy region, confirming a degree of vector affinity for airway epithelium (Figure 2b). No luc activity was detected in rats given PBS. No significant in vivo luc activity was detected in the heart, liver, spleen or kidneys of the animals given SV/luc.

Fluorescence over the lungs of SO rats was significantly lower than that seen in 2CLP animals (Figure 2a). Intratracheal administration results in little vector uptake in the normal lungs of SO rats, while changes in the intracellular structure occurring during ARDS may be responsible for enhanced viral expression. Similar findings were observed with adenoviruses, and explained by an ARDS-induced exposure/expression of the CAR and integrin receptors that mediate adenoviral entry into the alveolar cells 14. This mechanism may also apply to the SV40 vector, although different receptors or increased endocytosis may be involved. 394041.

Based on our previous findings 1415, a comparison of luc activity between SV/luc and Ad/luc vectors was performed. As expected, luc expression was lower in the 2CLP-SV/luc than in the 2CLP-Ad/luc group (Figure 2c, 2d), because of the natural tropism of adenoviral vectors for the pulmonary epithelium 14. Another factor may have been the higher titer of the adenoviruses used, a total of 10⁹ IU/ml adenoviral vector as compared to 1,25 × 10⁸ IU/ml of the SV40 vector.

Luc immunostaining was nonspecific, involving both alveolar type I type II cells. It was moderate in septic animals and low in SO animals (Figure 3). This may be due to moderate infectivity of the SV40 vector in the lungs or weak activity of the SV40 promoter in alveolar cells. This finding is also consistent with the literature, describing less transgene expression of some non – mammalian, non – vertebrate encoded proteins (e.g. luciferase, GFP, lac Z) commonly used as markers for transduction in SV40 vectors, than is usually seen with other vector systems (e.g., adenovirus) 18.

Immunostaining for luciferase in the liver, kidney, spleen and the heart was negative in both 2CLP and SO animals groups given SV/luc, suggesting that direct intratracheal administration of the vector results in negligible systemic spread despite alterations in membrane permeability and capillary leak seen in ARDS.

In order to test if the low level of Luc immunostaining in lung tissue is due to a low vector penetration into the alveolar epithelium or to low expression of the reporter gene, we tested for the presence of the VP1 capsid protein by fluorescent immunostaining. A significantly higher number of alveolar cells contained viral capsid proteins compared with the number of luc positive cells (Figure 4a, 4b, and compare with Figure 3). This suggests that vector penetration into lung epithelial cells is efficient, but transgene expression in many of the cells is below detection level. It is possible that the SV40 promoter used in the present vector is weak and therefore gene expression is lower than gene delivery. A stronger, lung specific promoter may significantly improve expression of a therapeutic gene in alveolar cells.

Alveolar type II cells participate in the regenerative process of the lung. Therefore their transduction is critical for successful ARDS gene therapy.

To examine this we performed co-immunostaining of lung tissue for surfactant protein C (Pro SP-C) a marker for alveolar type II cells and VP1 viral capsid protein (Figure 5). The figure depicts co-localization of Pro SP-C and VP1. These findings support our conclusion that the vector is internalized into alveolar type II cells. The ability to transduce these cells is essential because of their importance in alveolar epithelial repair processes and lung recovery in ARDS 10.

RT-PCR was performed on frozen tissues of the lung, kidney, liver, spleen and heart. High levels of mRNA were detected in the lungs of 2CLP animals, similar to those found in the plasmid pGL3 used as positive control for luciferase. Minimal levels were present in the liver, kidney spleen and none at all in the heart, confirming minimal systemic spread of the vector after direct intratracheal instillation. Interestingly, in T0 and SO controls, minimal expression was noticed in the lungs, in spite of the same route of vector administration, suggesting that the pathological processes taking place in the sepsis-induced ARDS lungs (as in 2CLP animals) enhances vector penetration and expression by an unknown mechanism. Regarding the kidney, liver, spleen and heart of T0 and SO animals, the same minimal expression was observed as in the septic animals (Figure 6).

Direct intratracheal administration of SV/luc induced no immunological response as measured by lymphocytic and neutrophilic infiltration of the H&E stained lung sections compared to PBS treated septic lung (Figure 1). To further confirm this we performed immunohistochemical detection for lymphocytic (CD3+ T cells) infiltration (Figure 7a, 7b). As expected, inflammatory cell infiltration was higher in the ARDS compared to the SO lungs. However lymphocytic infiltration in the SV/luc and PBS treated septic rats was similar, indicating that the SV40 vector does not elicite an excessive cellular immune response. These findings support our previous results which demonstrated that there was no cellular immune response against the vector or the transgene following liver transduction by the SV/luc vector, measured by lymphocyte proliferation assay at 84–110 days following vector administration 30.

Moreover, literature data reported that neutralizing antibody activity against an SV40 vector was not detected in the serum of treated mice, even after 8 consecutive intraperitoneal or subcutaneous inoculations 29. In our previous study we detected marginal humoral immune response against the vector only at very high vector concentration 30. Two factors appear responsible of this behavior: the uncommon cell entry route (SV40 vectors evade immune surveillance most likely by entering cells via caveolar endocytosis, followed by vesicular transport directly into the endoplasmic reticulum, bypassing the endosomal-lysosomal pathway) 27, and deletion of T antigen of the viral genome (renders the vector nonimmunogenic) 32.