A Receptor-targeted Nanocomplex Vector System Optimized for Respiratory Gene Transfer

Abstract

Synthetic vectors for cystic fibrosis (CF) gene therapy are required that efficiently and safely transfect airway epithelial cells, rather than alveolar epithelial cells or macrophages, and that are nonimmunogenic, thus allowing for repeated delivery. We have compared several vector systems against these criteria including GL67, polyethylenimine (PEI) 22 and 25 kd and two new, synthetic vector formulations, comprising a cationic, receptor-targeting peptide K16GACSERSMNFCG (E), and the cationic liposomes (L) DHDTMA/DOPE or DOSEP3/DOPE. The lipid and peptide formulations self assemble into receptor-targeted nanocomplexes (RTNs) LED-1 and LED-2, respectively, on mixing with plasmid (D). LED-1 transfected airway epithelium efficiently, while LED-2 and GL67 preferentially transfected alveolar cells. PEI transfected airway epithelial cells with high efficiency, but was more toxic to the mice than the other formulations. On repeat dosing, LED-1 was equally as effective as the single dose, while GL67 was 30% less effective and PEI 22 kd displayed a 90% reduction of efficiency on repeated delivery. LED-1 thus was the only formulation that fulfilled the criteria for a CF gene therapy vector while GL67 and LED-2 may be appropriate for other respiratory diseases. Opportunities for PEI depend on a solution to its toxicity problems. LED-1 formulations were stable to nebulization, the most appropriate delivery method for CF. Synthetic vectors for cystic fibrosis (CF) gene therapy are required that efficiently and safely transfect airway epithelial cells, rather than alveolar epithelial cells or macrophages, and that are nonimmunogenic, thus allowing for repeated delivery. We have compared several vector systems against these criteria including GL67, polyethylenimine (PEI) 22 and 25 kd and two new, synthetic vector formulations, comprising a cationic, receptor-targeting peptide K16GACSERSMNFCG (E), and the cationic liposomes (L) DHDTMA/DOPE or DOSEP3/DOPE. The lipid and peptide formulations self assemble into receptor-targeted nanocomplexes (RTNs) LED-1 and LED-2, respectively, on mixing with plasmid (D). LED-1 transfected airway epithelium efficiently, while LED-2 and GL67 preferentially transfected alveolar cells. PEI transfected airway epithelial cells with high efficiency, but was more toxic to the mice than the other formulations. On repeat dosing, LED-1 was equally as effective as the single dose, while GL67 was 30% less effective and PEI 22 kd displayed a 90% reduction of efficiency on repeated delivery. LED-1 thus was the only formulation that fulfilled the criteria for a CF gene therapy vector while GL67 and LED-2 may be appropriate for other respiratory diseases. Opportunities for PEI depend on a solution to its toxicity problems. LED-1 formulations were stable to nebulization, the most appropriate delivery method for CF. IntroductionCystic fibrosis (CF) is one of the most common autosomal genetic diseases, affecting 60,000 individuals worldwide.1Gibson RL Burns JL Ramsey BW Pathophysiology and management of pulmonary infections in cystic fibrosis.Am J Respir Crit Care Med. 2003; 168: 918-951Crossref PubMed Scopus (1294) Google Scholar It is caused by mutations in the gene encoding the CF transmembrane conductance regulator (CFTR), a cyclic adenosine monophosphate–activated chloride channel. Mutations in this gene lead to imbalanced water and ion movement across the airway epithelium, resulting in thickened mucus, chronic bacterial infection and inflammation, with progressive loss of pulmonary function and ultimately death.2Gill DR Davies LA Pringle IA Hyde SC The development of gene therapy for diseases of the lung.Cell Mol Life Sci. 2004; 61: 355-368Crossref PubMed Scopus (47) Google Scholar At present, no clinically effective drug is available to correct the CF genotype or CF lung disease manifestations. Several clinical trials have been performed to date assessing the potential of gene therapy to limit the progression of CF lung disease, but, so far, a clinically relevant treatment has yet to emerge.3Alton EW Stern M Farley R Jaffe A Chadwick SL Phillips J et al.Cationic lipid-mediated CFTR gene transfer to the lungs and nose of patients with cystic fibrosis: a double-blind placebo-controlled trial.Lancet. 1999; 353: 947-954Abstract Full Text Full Text PDF PubMed Scopus (379) Google Scholar,4Porteous DJ Dorin JR McLachlan G Davidson-Smith H Davidson H Stevenson BJ et al.Evidence for safety and efficacy of DOTAP cationic liposome mediated CFTR gene transfer to the nasal epithelium of patients with cystic fibrosis.Gene Ther. 1997; 4: 210-218Crossref PubMed Scopus (288) Google Scholar,5Harvey BG Leopold PL Hackett NR Grasso TM Williams PM Tucker AL et al.Airway epithelial CFTR mRNA expression in cystic fibrosis patients after repetitive administration of a recombinant adenovirus.J Clin Invest. 1999; 104: 1245-1255Crossref PubMed Scopus (202) Google Scholar,6Crystal RG McElvaney NG Rosenfeld MA Chu CS Mastrangeli A Hay JG et al.Administration of an adenovirus containing the human CFTR cDNA to the respiratory tract of individuals with cystic fibrosis.Nat Genet. 1994; 8: 42-51Crossref PubMed Scopus (790) Google Scholar,7Bellon G Michel-Calemard L Thouvenot D Jagneaux V Poitevin F Malcus C et al.Aerosol administration of a recombinant adenovirus expressing CFTR to cystic fibrosis patients: a phase I clinical trial.Hum Gene Ther. 1997; 8: 15-25Crossref PubMed Scopus (178) Google Scholar,8Flotte TR Zeitlin PL Reynolds TC Heald AE Pedersen P Beck S et al.Phase I trial of intranasal and endobronchial administration of a recombinant adeno-associated virus serotype 2 (rAAV2)-CFTR vector in adult cystic fibrosis patients: a two-part clinical study.Hum Gene Ther. 2003; 14: 1079-1088Crossref PubMed Scopus (194) Google Scholar,9Wagner JA Nepomuceno IB Messner AH Moran ML Batson EP Dimiceli S et al.A phase II, double-blind, randomized, placebo-controlled clinical trial of tgAAVCF using maxillary sinus delivery in patients with cystic fibrosis with antrostomies.Hum Gene Ther. 2002; 13: 1349-1359Crossref PubMed Scopus (217) Google ScholarEfficient synthetic vectors for gene transfer to airway epithelial cells, that are well tolerated in vivo, would provide useful tools, both for research in the respiratory field, and in the development of therapeutic applications such as gene therapy for CF. Liposomal and other nonviral vector systems are potentially useful for CF gene therapy as they are less immunogenic than viral vectors such as adenovirus and adeno-associated virus, allowing repeated and regular delivery in vivo.10Glover DJ Lipps HJ Jans DA Towards safe, non-viral therapeutic gene expression in humans.Nat Rev Genet. 2005; 6: 299-310Crossref PubMed Scopus (499) Google Scholar,11Li S Huang L Nonviral gene therapy: promises and challenges.Gene Ther. 2000; 7: 31-34Crossref PubMed Scopus (534) Google Scholar However, generally, nonviral vectors produce lower levels of transfection in airways in vivo compared to their viral counterparts due in large part to their lack of targeting and binding to apical receptors and surfaces, and their inability to transfect nondividing cells.We have described previously a modular, self-assembling vector system, LID, which is a formulation of Lipofectin [L, a cationic liposome, itself comprising a 1:1 (by weight) mixture of DOTMA and DOPE], an integrin receptor-targeting/DNA-binding peptide (I;K16GACRRETAWACG) and plasmid DNA (D).12Hart SL Arancibia-Carcamo CV Wolfert MA Mailhos C O'Reilly NJ Ali RR et al.Lipid-mediated enhancement of transfection by a nonviral integrin-targeting vector.Hum Gene Ther. 1998; 9: 575-585Crossref PubMed Scopus (178) Google Scholar,13Jenkins RG Herrick SE Meng QH Kinnon C Laurent GJ McAnulty RJ et al.An integrin-targeted non-viral vector for pulmonary gene therapy.Gene Ther. 2000; 7: 393-400Crossref PubMed Scopus (78) Google Scholar,14Jenkins G Hart SL Hodges RJ Meng QH Kinnon C Laurent GJ et al.Cyclooxygenase-2 overexpression, using an integrin-targeted gene delivery system (the LID vector), inhibits fibroblast proliferation in vitro and leads to increased prostaglandin E-2 in the lung.Chest. 2002; 121: 102S-104SAbstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar,15Cunningham S Meng QH Klein N McAnulty RJ Hart SL Evaluation of a porcine model for pulmonary gene transfer using a novel synthetic vector.J Gene Med. 2002; 4: 438-446Crossref PubMed Scopus (47) Google Scholar,16Parkes R Meng QH Siapati KE Mcewan JR Hart SL High efficiency transfection of porcine vascular cells in vitro with a synthetic vector system.J Gene Med. 2002; 4: 292-299Crossref PubMed Scopus (25) Google Scholar,17Jenkins RG Meng QH Hodges RJ Lee LK Bottoms SE Laurent GJ et al.Formation of LID vector complexes in water alters physicochemical properties and enhances pulmonary gene expression in vivo.Gene Ther. 2003; 10: 1026-1034Crossref PubMed Scopus (26) Google Scholar,18Meng QH Robinson D Jenkins RG McAnulty RJ Hart SL Efficient transfection of non-proliferating human airway epithelial cells with a synthetic vector system.J Gene Med. 2004; 6: 210-221Crossref PubMed Scopus (23) Google Scholar This formulation displayed receptor-targeted transfection mediated by the peptide, with endosomal release of DNA to the cytoplasm enhanced by the liposome component, while the peptide also conferred the capacity to transfect nondividing epithelial cells.12Hart SL Arancibia-Carcamo CV Wolfert MA Mailhos C O'Reilly NJ Ali RR et al.Lipid-mediated enhancement of transfection by a nonviral integrin-targeting vector.Hum Gene Ther. 1998; 9: 575-585Crossref PubMed Scopus (178) Google Scholar,18Meng QH Robinson D Jenkins RG McAnulty RJ Hart SL Efficient transfection of non-proliferating human airway epithelial cells with a synthetic vector system.J Gene Med. 2004; 6: 210-221Crossref PubMed Scopus (23) Google Scholar This formulation has the potential to overcome some of the obstacles to gene transfer. Here we have evaluated novel formulations of cationic liposomes and targeting peptides, which retain these beneficial properties but with lipid and peptide components optimized specifically for transfection of airway epithelial cells, in vivo and in vitro.In this study we describe the LED-1 formulation comprising a peptide with an epithelial targeting sequence, SERSMNF, identified by phage display technology that targets intercellular adhesion molecule-1 (ICAM-1), a receptor for the respiratory pathogen rhinovirus.19Writer MJ Marshall B Pilkington-Miksa MA Barker SE Jacobsen M Kritz A et al.Targeted gene delivery to human airway epithelial cells with synthetic vectors incorporating novel targeting peptides selected by phage display.J Drug Target. 2004; 12: 185-193Crossref PubMed Scopus (42) Google Scholar The integrin receptors targeted by the LID complexes are predominantly basolateral, whereas ICAM is located in the apical membrane and thus makes an attractive target for receptor-mediated uptake. Cationic lipids were optimized for epithelial cells based on structural variants of DOTMA, differing most notably by shorter alkyl chains.20Writer M Hurley CA Sarkar S Copeman DM Wong JB Odlyha M et al.Analysis and optimization of the cationic lipid component of a lipid/peptide vector formulation for enhanced transfection in vitro and in vivo.J Liposome Res. 2006; 16: 373-389Crossref PubMed Scopus (22) Google Scholar In addition, we have described LED-2, which contains the same peptide as LED-1 but a different liposome component, DOSEP3/DOPE in a 1:1 weight ratio, instead of DHDTMA/DOPE. The cationic lipid DOSEP3 contains an additional polyethylene glycol (PEG) sequence and was included to test whether PEGylation, which is known to shield cationic complexes and reduce binding of serum proteins, might help to reduce inflammation in the lung compared to LED-1. The aim of the study was to evaluate the suitability of LED-1 for gene therapy of CF, with regard to transfection efficiency of human airway epithelial cells in vitro and in murine lung airway epithelium in vivo, and its inflammatory potential and cytotoxicity, compared to vectors GL67 and polyethylenimine (PEI). In addition we have assessed the capacity of LED-1 for delivery by nebulization.ResultsDeveloping a second generation RTN vectorA series of cationic lipids that varied in alkyl chain tail length from 12 (C12) to 18 (C18) carbon atoms, and in saturation were formulated as liposomes with DOPE in a 1:1 weight ratio, then into receptor-targeted nanocomplexes (RTNs) with peptide K16GACSERSMNFCG (peptide E) and plasmid DNA. An unsaturated lipid with a C16 alkyl chain and with the double bond at the C11 position (DHDTMA; Figure 1) was optimal in transfections performed in vitro (data not shown). The formulation of this lipid (L) with peptide E (E) and plasmid DNA (D) was termed LED-1. The RTN formulation LED-2 comprised the same peptide as LED-1 but formulated with a PEGylated cationic lipid, DOSEP3 (Figure 1) and DOPE in a 1:1 weight ratio.Complexes were prepared at a lipid:peptide:DNA ratio of 2:4:1 as this ratio was found to be optimal by in vitro tests (data not shown). pCILux is ∼6 kilobase (kb) in length which equates to 0.25 pmol of plasmid per microgram. One microgram of plasmid pCILux, equivalent to 0.25 pmol of DNA, is formulated with 4 μg gross of peptide E which is ∼0.8 nmol net peptide. This equates to ∼3,200 peptide ligands per plasmid, and therefore per complex (assuming one plasmid per complex).Luciferase reporter gene expression from LED-1 in both 1HAEo [472,218 ± 66,418 relative light unit (RLU)/mg protein] and CFBE41o– (477,414 ± 103,898 RLU/mg protein) cells was significantly higher (P < 0.05) for both cell lines than the LID formulation (177,649 ± 29,247 and 306,042 ± 30,363 RLU/mg protein, respectively) (Supplementary Figure S1). Similarly, when tested for green fluorescent protein (GFP) expression by flow cytometry, again, LED-1 transfected with higher efficiency than LID both CFBE41o– (∼1.5-fold; 20% versus 12.9%) and 1HAEo– cells (∼2-fold; 42.16% versus 22.01%). LED-1 and LID efficiencies were also compared in transfections in vivo in which mice were administered by intratracheal instillation with a single dose of 16 μg of pCILux plasmid DNA in 50 μl water. LED-1 mediated luciferase expression (median value 750 RLU/mg protein) was approximately fourfold higher (P < 0.05) than with LID (median value 196.5 RLU/mg protein) (Supplementary Figure S2).Physicochemical properties of RTN particlesLED-1 formulations were prepared for biophysical characterization in water at lipid:peptide:DNA weight ratios of 2:4:1, with a DNA concentration in the formulations of 320 μg/ml, as used for in vivo experiments. Particles were homogeneous, displayed a low value of polydispersity (data not shown), with an average size of <150 nm, and were stable for at least 4 hours (Supplementary Figure S3). Similar results were obtained from two independent experiments. The zeta potential of the particles in the LED-1 formulation was+48 mV, i.e., the vector particles were strongly cationic.In vitro comparison of transfection efficienciesTransfections with LED-1, LED-2, GL67, PEI 25 kd, and PEI 22 kd were compared in four different human airway epithelial cell lines including two normal cell types (16HBEo– and 1HAEo–) and two CF cell lines (CFBE41o– and CFTE29o–). LED-1 transfected each of the four cell types with the highest efficiencies. Compared to PEI 22 kd, the next best formulation, LED-1 transfection efficiencies were 6-fold (P < 0.01) higher for 16HBEo– cells, 11-fold for CFBE41o– cells (P < 0.001), 123-fold for 1HAEo– cells (P < 0.001), and 5-fold for CFTE29o– cells (P < 0.001) (Figure 2a). When tested for GFP expression by flow cytometry, again, LED-1 transfected all cell lines significantly better (P < 0.05) than the other reagents, with the exception of 16HBEo– cells where there was no significant difference between LED-1, PEI 22 kd, and PEI 25 kd (Figure 2b).Figure 2In vitro transfections reveal that LED-1 is more efficient at transfecting four different epithelial cell lines. (a) Transfections with LED-1, GL67, polyethylenimine (PEI) 22 kd, and PEI 25 kd complexed with pCILux were performed in two normal (16HBEo– and 1HAEo–) and two cystic fibrosis (CFBE41o– and CFTE29o–) human airway epithelial cell lines. The N/P ratio for both PEI 22 kd and PEI 25 kd was 10:1. Inall transfections 0.25 μg of DNA/well was used. Twenty-four hours after transfection a luciferase assay was performed followed by a Bradford protein assay. The experiment was repeated on two occasions. Each result is the mean of six values and error bars represent the SD about the mean (mean ± SD). The graph was plotted using a logarithmic scale for the, y-axis. (b) Transfections with LED-1, GL67, PEI 22 kd, and PEI 25 kd complexed with pEGFP-N1 were also performed in these cell lines. The N/P ratio for both PEI 22 kd and PEI 25 kd was 10:1. In all transfections 0.25 μg of DNA/well was used. Twenty-four hours after transfection wells were photographed and then fixed and assayed for green fluorescent protein expression by flow cytometry. Each result is the mean of three values and error bars represent the SD about the mean (mean ± SD).View Large Image Figure ViewerDownload Hi-res image Download (PPT)In vivo comparison of lung transfection efficiencies after single and repeated deliveryA dose–response study was performed in vivo following intratracheal installation of LED-1 with 8 or 16 μg of pCILux plasmid DNA in 50 μl water. The higher dose (median value 1,134 RLU/mg protein) was selected as it resulted in more than a twofold higher luciferase expression than the lower dose (median value 499 RLU/mg protein) (Supplementary Figure S4). Vector formulations of LED-1, LED-2, GL67, PEI 22 kd, and PEI 25 kd were then prepared and administered in doses containing 16 μg of pCILux plasmid DNA in equal volumes of 50 μl. In luciferase assays of lung extracts prepared 24 hours after transfection from a single dose of vector formulation, PEI 22 kd was the most efficient vector, producing at least a 14-fold higher expression than GL67 (P < 0.01), LED-1 (P < 0.01), LED-2 (P < 0.01), or PEI 25 kd (P < 0.001) (data not shown for PEI 25 kd) (Figure 3). GL67 and LED-1 vectors displayed similar activity levels, both producing two- to threefold higher expression than either PEI 25 kd or LED-2 (P < 0.01). However, PEI 22 kd, even at the relatively low dose used here, was highly toxic with a mortality rate of ∼25%.Figure 3Comparison of in vivo transfection efficacy of several nonviral vectors after a single or triple dose regime. Luciferase activity in whole-lung lysates 24 hours following instillation after 1 or 3 weekly administrations of four different vectors. Luciferase activity is significantly reduced following the third instillation for both polyethylenimine (PEI) 22 kd (P < 0.01) and GL67 (P < 0.05) but not with the LED-1 or the LED-2 vector. The graph was plotted using a logarithmic scale for the y-axis. RLU, relative light unit.View Large Image Figure ViewerDownload Hi-res image Download (PPT)A repeat dosing study was performed with mice receiving two weekly administrations of each formulation containing control plasmid (pCI), followed by a third dose with the luciferase plasmid, pCILux. Luciferase activity levels were assessed in samples taken 24 hours after the administration of pCILux (Figure 3). LED-1 and LED-2 displayed luciferase activity levels similar to that obtained with the initial single dose, whereas both PEI 22 kd (96% reduction; P < 0.01) and GL67 (67% reduction; P < 0.05) showed significantly reduced luciferase activity levels.Localization of luciferase expression in murine lungLung structures were unaffected by a single intratracheal instillation of empty plasmid (pCI) complexed with LED-1, LED-2, GL67, or PEI 22 kd and were negative for luciferase immunostaining (Figure 4a, f, k, and p). Following a single instillation of pCILux complexed with LED-1 and PEI 22 kd vectors, immunostaining for luciferase in lung sections was strong and predominantly associated with bronchial epithelial cells, although alveolar macrophage staining was also evident (Figure 4b, c, q, and r). PEI 22 kd appeared to penetrate deeper than LED-1 into the structures of the lung with staining for luciferase also apparent in some basal cells, fibroblasts and smooth muscle cells in the airway wall (data not shown). By contrast, following instillation of pCILux with LED-2 or GL67, there was minimal staining of the bronchial epithelium and luciferase expression was associated mainly with macrophages although for GL67, staining was also apparent in areas of the alveolar epithelium (Figure 4g, h, l, and m). Following instillation of pCILux complexed with LED-1, 99 ± 1% of airways showed evidence of bronchial epithelial cell transfection. By contrast, 73 ± 7%, 49 ± 13%, and 38 ± 8% of airways showed epithelial expression with PEI 22 kd, LED-2, and GL67, respectively. Furthermore, semi-quantitative scoring of the proportion of airway epithelial cells staining for luciferase (Supplementary Figure S5) showed higher levels of cell staining in LED-1 transfected animals as compared to all other groups (P < 0.001 in all cases).Figure 4Immunohistochemical localization of luciferase following transfection with various vectors. Representative sections are shown from animals transfected with LED-1 (a–e), LED-2 (f–j), GL67 (k–o) and polyethylenimine (PEI) 22 kd (p–t) complexed with pCI (a,f,k,p) or pCILux (b,c,g,h,l,m,q,r) following a single instillation or following 3 weekly administrations (d,e,i,j,n,o,s,t) at low (b,d,g,i,l,n,q,s) and high (a,c,e,f,h,j,k,m,o,p,r,t) magnification. Sections from animals transfected with pCI acted as negative controls and showed no positive staining. Transfection following a single instillation of vectors with pCILux showed brown staining, indicative of luciferase expression, which was predominantly localized to the bronchial epithelium (denoted by *low, and Ep in high, magnification images) with LED-1 (b,c) and PEI 22 kd (q,r). In contrast, for LED-2 (h) and GL67 (m) staining was predominantly associated with macrophages (Mac). Following 3 weekly instillations, with LED-1 (d,e) and PEI 22 kd (s,t) staining was still predominantly associated with the bronchial epithelial. In contrast to the single instillation, moderate staining of the bronchial epithelium was apparent following repeated transfection with LED-2 (i,j) and GL67 (n,o). Four or five animals were analyzed per group and staining was consistent between animals in each group. The bar scale of 50 μm applies to all high power images, whereas the bar scale of 400 μm applies to all low power images. Ep, epithelial cells.View Large Image Figure ViewerDownload Hi-res image Download (PPT)In animals instilled on three separate occasions, 7 days apart, bronchial epithelial staining for luciferase was strongest following transfection with LED-1 (Figure 4d and e). In contrast to the single transfection studies, moderate staining of the bronchial epithelium was also present in animals following repeated transfection with GL67 (Figure 4n and o). Luciferase staining was again associated with macrophages for all four vectors (Figure 4e, j, o, and t). Semi-quantitative scoring (Supplementary Figure S5) showed no significant difference in the proportion of bronchial epithelial cells staining following repeat transfection of LED-1 or LED-2 as compared with single instillations. As with the single instillations, repeat instillations of LED-1 showed the greatest proportion of bronchial epithelial cells staining for luciferase compared with the other groups (P < 0.01 in all cases). A reduction in bronchial epithelial cell staining was observed following repeated instillation, as compared to that from a single instillation of PEI 22 kd (P < 0.001). By contrast, the proportion of bronchial epithelial cells staining for luciferase following repeated instillation with GL67 was increased when compared with that for a single transfection (P < 0.001).Host response to vector instillation by histologyA single intratracheal instillation of the control plasmid pCI complexed with GL67, LED-1, LED-2, or PEI 22 kd had no distinguishable effect on the histological appearance of the lung when compared with untreated controls (Supplementary Figure S6a). A single instillation of the luciferase expressing plasmid, pCILux, with all vectors induced more inflammation than pCI with mild peribronchial mixed inflammatory cell infiltration, which with LED-1 affected between 1 and 5% of the lung (Supplementary Figure S6b).Animals given 2 weekly instillations of pCI with GL67 vectors followed by a third instillation of vector with pCILux showed evidence of mild inflammation similar in extent to that following a single instillation (data not shown) while a similar triple instillation strategy with LED-1 or PEI 22 kd induced inflammation that affected 5–10% and 15–30% of the lung respectively, together with small areas of tissue disruption and remodeling that was more severe with PEI 22 kd (Supplementary Figure S6c and d).Analysis of cytokines in bronchoalveolar lavage fluidTwenty-four hours after a single vector instillation lavage fluid was collected and immunofluorescent analysis performed in a Luminex assay for a panel of common mouse inflammatory cytokines including interferon-γ, interleukin-2 (IL-2), IL-5, IL-6, IL-10, IL-12, and tumor necrosis factor-α. The vectors, LED-1, PEI 22 kd, and GL67, were compared both with pCILux and with pCI control plasmid. All vectors induced an inflammatory response (Figure 5 and Supplementary Table S1) with both plasmids pCI and pCILux, when compared to noninstilled controls, with all inflammatory cytokines being elevated. IL-10 levels were not above background for all the different vectors tested (data not shown). Specifically vector LED-1 with pCILux caused a greater increase in cytokine IL-5 when compared to GL67 with pCILux (P < 0.001), while LED-1 with pCI caused a greater increase in cytokines IL-2 (P < 0.05) and IL-4 (P < 0.01) compared to that caused by GL67 with pCI. PEI 22 kd induced a greater increase in most cytokines, especially IL-4 and IL-5 compared with both GL67 (P < 0.001) and LED-1 (P < 0.001) except for IL-5 levels from LED-1 with pCILux, which were not significantly different from PEI. Interestingly, PEI 22 kd induced the smallest increase in IL-12 with levels in GL67/pCI-transfected lungs significantly higher than with PEI 22 kd/pCI (P < 0.01).Figure 5Inflammatory mediator profile in bronchoalveolar lavage fluid following intratracheal instillation of three different vector formulations (with pCILux or pCI). All vectors induced an inflammatory response with both plasmids pCI and pCILux, with all inflammatory cytokines elevated. In all cases interleukin-12 (IL-12) was elevated with the exception of polyethylenimine (PEI) 22 kd, which showed though the highest levels for cytokines IL-1b, IL-2, IL-4, IL-5, and IL-6. Each value represents the mean ± SEM for at least six animals and is expressed as picogram per 100 μl.View Large Image Figure ViewerDownload Hi-res image Download (PPT)TUNEL and caspase 3 staining for apoptosisThe extent of apoptosis in the lung was investigated for PEI, GL67, and LED-1 complexed with pCI or pCILux after single and repeat dosing. Twenty-four hours after the single or the third instillation, mice were culled and lung sections were stained immunohistochemically for active caspase 3 and terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) stained. Both patterns of staining were similar for each vector formulation, with slightly increased numbers of cells positive for active caspase 3 compared to TUNEL staining, due possibly to earlier and more protracted activation of caspase 3 during the apoptotic process. Untreated control lung was generally negative for both markers with only occasional bronchial epithelial cells and macrophages staining positively (Figure 6a), which is consistent with the low levels of turnover of these cells in the normal lung. The lungs of animals instilled with pCI complexed with GL67 showed similar numbers of positive bronchial epithelial cells to untreated controls, although there were increased numbers of apoptotic macrophages and occasional small areas of alveolar epithelial cell apoptosis apparent (Figure 6b). A similar result was observed with GL67 complexed with pCILux. LED-2 complexed with pCI or pCILux gave very similar results to GL67 (data not shown). The lungs of animals instilled with LED-1 complexed with pCI were similar to those instilled with GL67, but there was a slight increase in the number of apoptotic bronchial epithelial cells (Figure 6c). Plasmid pCILux, compared to pCI, when complexed with LED-1 showed a small increase in bronchial epithelial cell apoptosis as well as alveolar epithelial cells and macrophages in areas of inflammation (data not shown). By contrast, instillation of PEI 22 kd complexed with pCI or pCILux (data not shown) induced a marked increase in apoptotic bronchial epithelial cells, macrophages, and alveolar epithelial cells (Figure 6d). In a further group of animals, ins