Genetic Modification of the Lung Directed Toward Treatment of Human Disease

Retroviral and lentiviral vectors

The RNA-based vectors have been most useful in ex vivo lung gene transfer applications. The first vector to be employed for lung gene therapy, which was directed toward ex vivo gene therapy for α1AT deficiency in rodents, was derived from the Maloney murine leukemia virus, a murine retrovirus. ¹⁴ The retrovirus and lentivirus genomes are comprised of two long terminal repeats at either end of the genome and a packaging sequence. A transgene cassette up to 8 kb can be inserted in the space created by the removal of the (gag, pol, and env) native viral genes. During the virus life cycle, the RNA genome is converted into double-stranded DNA that then integrates into the host chromosomal DNA in a random fashion. Due to this property, the retroviral and lentiviral vectors have been explored as an option in gene therapy strategies for proliferating cells in order to achieve stable transduction of the target cell with transmission of the therapeutic gene to all daughter cells. ²

The requirement for the target cells to be actively proliferating at the time of infection is less useful for in vivo transduction of lung cells where the rate of cellular proliferation is low, thereby limiting the application of retroviruses for in vivo gene therapy. This issue has been overcome in part by lentiviral vectors that are capable of transfecting post-mitotic cells. ⁴²

Two types of lentiviral vectors have been tested for in vivo lung gene delivery: vectors derived from human immunodeficiency virus and feline immunodeficiency virus. ^{43 –45} However, the use of lentiviruses for in vivo lung gene therapy is still hampered by the absence of suitable cellular receptors on the more accessible apical surface of lung epithelial cells. ^5,46 This problem may be partly overcome with pseudotyped lentiviral vectors, such as those pseudotyped with modified viral envelopes from diverse origins such as Ebola, baculovirus, Sendai virus (SeV), influenza, and parainfluenza. ^{5,47 –51} Additionally, because the RNA virus integrates the transgene into the target cell genome, there is the risk of random integration, which could possibly cause oncogene activation or tumor-suppressor gene inactivation. ^52,53 This was the likely cause of T-cell leukemia in children treated for severe combined immunodeficiency with retrovirus-mediated gene therapy. However, the insertional mutagenesis risk has been mostly overcome in the design of currently used lentivirus vectors. ^{54 –58} Specifically, studies comparing genotoxicity of retroviral and lentiviral vectors have demonstrated that retroviruses can insert into growth-control genes, which could lead to tumor formation, but lentiviruses do not tend to do so. ⁵⁹ In addition, lentiviruses lead to efficient and persistent gene expression and can be re-administered if necessary without the development of an inhibitory immune response. ^16,60,61 Even with these significant improvements, a continuing challenge that has limited the use of RNA virus vectors for lung gene therapy is the difficulty in manufacturing and purifying them in large quantities. ^2,5

Adenovirus vectors

The first example of successfully using a viral vector for in vivo gene transfer used a replication deficient Ad to express human α1AT in the lung of cotton rats. ⁶² Recombinant Ad vectors, which have a natural tropism for the respiratory tract, are comprised of a double-stranded genome. ² The E1 region is deleted to render the vectors replication deficient, and the E3 region is deleted to make room for the expression cassette. ⁶³ Ad vectors result in highly efficient gene transfer in vivo and are capable of infecting both proliferating and non-proliferating cells. ⁶⁴ They are also relatively easy to produce and purify in high quantities. However, Ad vectors are highly immunogenic and evoke host immunity to the vector, limiting gene expression to 2–3 weeks. ^{63 –67}

In an attempt to extend expression mediated by Ad vectors, “second-generation” Ad vectors were engineered in order to attempt to minimize the expression of viral antigens. Typically, these vectors have the E2 or E4 early viral genes deleted or mutated, in addition to the E1 and E3 deletions present in the first-generation Ad vectors. ^66,68,69 In experimental animals, these vectors were able to reduce but not eliminate anti-vector-induced inflammation likely mediated by the antigenic capsid and low level expression of late viral genes. ⁶⁹ Further attempts to reduce cellular immunity against the vector led to the design of “gutted/helper-dependent” Ad viruses, where all the viral coding sequences have been deleted. These have been shown in some experimental animals models to mediate longer-term transgene expression, but a major hurdle for this vector system is production and contamination of the vector preparation with replication competent Ad. ^{70 –74}

With Ad-based vectors, the Ad genome is not integrated into the host genome and persists in the nucleus as an episome. Thus, persistence of expression of the therapeutic gene depends on the proliferation of the target cell. If it is quiescent and anti-vector immunity is not an issue, the expression will persist. If the cell proliferates, only one of the daughter cells would continue to contain the Ad genome, and expression will be diluted. Ad vectors also elicit a strong humoral immune response against the vector capsid. As a result, effective expression following repeat administration to the lung is prevented, making it unsuitable for diseases that require chronic expression. ^65,75 Another challenge for the use of Ad vectors directed toward the airway epithelium (e.g., to treat CF) is that the CAR cellular receptor for Ad is expressed more highly on the basolateral surface of the human airway epithelial cells, which are protected by tight junctions that make vector–receptor interactions inefficient. ⁷⁶ To circumvent this issue, although not practical for human applications, studies have been carried out to disrupt airway epithelial tight junctions transiently in order to increase the efficiency of Ad-mediated transduction of airway epithelia through the use of detergents, calcium phosphate, or Ca²⁺ chelators in the vector formulating buffers. ^{77 –80} Studies have also explored the potential of pseudotyping the Ad vectors to target other more suitable cell surface receptors. ⁸¹ Despite the issues relating to Ad vectors, there is an extensive safety record of their use in human lung gene therapy. ^{82 –94}

AAV vectors

AAV, a nonpathogenic, single-stranded parvovirus, does not cause any human disease. ⁹⁵ The wild-type AAV consists of two genes—rep (encoding regulatory proteins) and cap (encoding capsid proteins)—bound by two inverted terminal repeats (ITRs). ⁹⁶ When used as a gene therapy vector, the rep and cap genes of the virus are replaced by the cDNA sequence of the transgene of interest and associated regulatory sequences. The length of this insert is limited to ∼4.7 kb but can be up to 5 kb. ⁹⁷ Unlike the wild-type AAV, which integrates into a specific locus on human chromosome 19, ^98,99 recombinant AAV does not integrate but rather persists in the nucleus as an episome. ¹⁰⁰ Production of AAV vectors require “help” from genes from Ad or herpes virus vectors. ^{101 –103}

AAV vectors are able to transduce non-dividing quiescent cells and lead to persistent transgene expression, as long as the target cells are not proliferating. In experimental animals, AAV vectors have been used to transfer genes to the airway epithelium, alveolar epithelium, pulmonary vascular endothelium, and pleural mesothelioma. ^{5,12,104 –113} Initial studies were carried out using the human AAV serotypes 2 and 5, but there are now >100 AAV serotypes, both naturally occurring and created in the laboratory. ^114,115 Most commonly, AAV2 ITRs are used to flank the expression cassette, which is then pseudotyped into capsids of the different serotypes. These variant AAVs have the potential to bypass neutralizing antibodies to human AAV and circumvent lung cell receptor availability limitations, which are an issue for the common AAV2 and AAV5 vectors. ^{107,109,116 –119} Several of these serotypes of AAV have been shown to mediate efficient transduction of airway and alveolar epithelium, pulmonary endothelium, and the pleural mesothelium. ^{4,5,16,105,107,109 –112} The practical disadvantages of using AAV vectors for lung gene transfer include the limited size of the expression cassette that can be accommodated and the lack of the relevant receptors on the target cells. The great advantage of AAV vectors for lung gene transfer is that if their genome can be delivered to the target cell population and the cells are not replicating or are replicating very slowly, the vector genome will persist and continue to express the transgene. ^117,119,120

Other viral vectors

In addition to the commonly used viral vectors already described, a number of other viruses also have been used for lung gene therapy, including polyomaviruses, vaccinia virus, baculovirus, and SeV.

Polyomaviruses, including Simian virus 40 (SV40) and John Cunningham virus (JCV), can transduce both resting and dividing cells and can deliver sustained gene expression by integrating into the host genome. ^121,122 These vectors can also be produced in high titers and are relatively non-immunogenic. The disadvantages of SV40 or JCV are the small expression cassette (∼2.5–5.0 kb) capacity and potential risks of random integration. ^121,122 These polyomavirus vectors have been tested in preclinical experiments for CF ¹²³ and lung cancer. ¹²⁴

Vaccinia virus (VV) is another potential virus being developed as a lung gene therapy vector. VV will infect most cell types and has the capacity for insertion of >25 kb of foreign DNA into the genome. VV produces high levels of protein and can be made to high titers. The challenges in using VV as a gene delivery vector include the cytopathic effect on infected cells, the immunogenicity of VV proteins, and pre-existing immunity to VV in individuals that have received smallpox vaccination. ¹²⁵ Some of these disadvantages can be used favorably in the treatment of cancer by using VV for oncolytic virotherapy ¹²⁵ and have been tested for treatment of both lung cancer and mesothelioma. ^126,127

Another viral vector under development for lung gene therapy is based on the insect-derived baculovirus expression vectors (BEV). These vectors are widely used for recombinant protein expression but also have potential as gene therapy vectors because they can carry large DNA inserts (>38 kb) and can transduce a wide variety of cell types, including stem cells. Other potential advantages of BEV are that they cannot replicate in mammalian cells and there is no pre-existing anti-BEV immunity in humans. ¹²⁸ However, limitations to the use of BEV are the short expression time of the transgene, the necessity for entry through the basolateral membrane for some cells, and the rapid virus inactivation by serum complement. ^128,129 BEV have been utilized in preclinical experiments targeting lung cancer. ¹³⁰

Vectors based on SeV have also been developed for lung gene transfer. SeV naturally infects respiratory epithelial cells, as well as many other cell types. Gene transfer efficiency in the respiratory tract of mice and ferrets in vivo is three to four logs higher than Ad5 or plasmid/liposomes. ¹³¹ However, gene expression is transient, and repeated administration leads to diminished gene expression. ¹³² These constraints might be overcome by pseudotyping the SeV envelope proteins onto a lentivirus, allowing for efficient transfection of lung cells and persistent gene expression. ¹³³ SeV vectors have been employed for therapy of experimental animals relevant to CF ¹³⁴ and pulmonary arterial hypertension. ¹³⁵

Non-viral vectors

In non-viral vectors, either the therapeutic gene is administered to the target cell in the form of “naked” DNA or mRNA, or the nucleic acid is complexed with other macromolecules with the goal of enhancing entry into the cells and transfection to the nucleus. ^{136 –138} Non-viral, plasmid-based vectors overcome the limitations of viral vectors such as limited size of the expression cassette (AAV), immunogenicity (Ad), and insertional mutagenesis (retro and lentivirus). While non-viral vectors possess the potential advantage of a better safety profile and reduced immunogenicity compared with viral vectors, the effectiveness of non-viral vectors is restricted by a low efficiency of in vivo gene transfer. ^{139 –141}

Liposomes combined with plasmids have been widely used to enhance translocation of DNA into the cytosol via membrane fusion or endocytosis, and there is extensive literature in using liposome/plasmid complexes to delivery genes to the lung. ^142,143 Typically, cationic lipids are complexed with anionic plasmid DNA to increase the transfection efficiency. ^140,144 Strategies to improve liposomes have focused on lipid formulations, and antibodies or other ligands have been incorporated into the design to permit targeting to specific cell types. ¹⁴⁴ Nanoparticles consist of the nucleic acid complexed with another material such as lipid or polymers, including peptides or polysaccharides. ¹⁴⁵ Solid lipid nanoparticles, which remain solid at both room and body temperatures, are effective at protecting nucleic acids from nuclease degradation and are often used for the delivery of siRNA. Polymer-based nanoparticles are more stable than lipids or liposomes and consist of natural or synthetic cationic polymers complexed with DNA. ¹⁴⁰ Chitosan is a natural cationic polysaccharide that has muco-adhesive properties, making it a good choice for lung and oral gene therapy. A synthetic polymer, polyethylenimine (PEI) is also commonly used for in vitro and in vivo gene transfer because of its high efficiency. However, PEI can induce cytotoxicity, and studies are ongoing to investigate various methods to improve biocompatibility, potency, and stability. ¹⁴⁵ Another class of non-viral vectors, referred to as “molecular conjugate” vectors, have also been developed to accomplish the delivery of heterologous genes to target cells via the receptor-mediated endocytosis pathway. The basic design of molecular conjugates consists of plasmid DNA complexed to polylysine and a macromolecular receptor ligand, which can be internalized by the target cell. ¹⁴⁶ Non-viral vectors employing liposomes, nanoparticles, or “molecular conjugates” have been used to deliver plasmids for gene therapy to the airways for the treatment of CF, ^{147 –149} lung cancer, ^{150 –164} food allergy, ¹⁶⁵ and infectious disorders. ^{166 –168}

Antisense gene silencing

The strategy of gene silencing is used to inhibit the biological function of specific proteins. ¹⁶⁹ Several RNA-based silencing strategies have been developed that have been used to modify lung gene expression. ¹⁷⁰ Antisense oligonucleotides (ASOs), consisting of RNA complementary to the mRNA of the target genes, were the earliest RNA-based approaches to be tested. ASOs act by blocking access to ribosomes and can be modified to recruit RNase H or stimulate the immune response. ^169,171 The disadvantages of ASOs include non-specific immune stimulation and modest silencing effect. DNAzymes consist of a complementary DNA and a catalytic subunit that can cleave mRNA. ^172,173 The technology of RNA interference (RNAi) includes short interfering RNA (siRNA), short hairpin RNA (shRNA), or micro RNA (miRNA). These RNAs are 21–22 nucleotides in length and are highly optimized for gene silencing by targeting the complementary cellular mRNA for degradation. ¹⁷⁴ Delivery of oligonucleotides is generally inefficient in a naked form. Thus, delivery to cells is enhanced by making liposome complexes or expressing the RNA from viral vectors. ¹⁶⁹ Each of these described gene-silencing strategies has been tested to treat diseases of the lung such as asthma, ^{172,175 –184} allergic rhinitis, ¹⁸⁵ lung cancer, ^{151,152,155,158,186 –201} pulmonary hypertension, ²⁰² and infection. ^{167,168,203 –205}

Gene editing

Gene therapy strategies have focused on expressing an exogenous DNA to correct a disease caused by a genetic mutation. New technologies that are currently being developed may one day allow in vivo editing to repair the mutated gene, as well as the insertion of new genes or deletions of problematic endogenous genes. Gene editing technology relies on homology-directed repair by the cellular machinery of double-stranded breaks made in the genomic DNA. Double-stranded breaks can be introduced in precise locations by employing targeted nucleases such as zinc finger nucleases, transcription activator-like effector nucleases (TALENs), or the CRISPR/Cas9 system. ²⁰⁶ Zinc finger nucleases and TALENs are proteins that both combine a DNA-binding domain that recognizes certain base pair sequences with the FokI endonuclease domain. This combination allows for targeted cleavage of double-stranded DNA. The drawback of these nuclease-based systems is that a new protein must be engineered for each DNA target site of interest. The CRISPR/Cas9 system uses guide RNAs that include sequence complementary to the target site. Heteroduplex formation between the guide RNA and the target DNA allows for cleavage by the Cas9 nuclease. Because the DNA recognition and cleavage functions are decoupled, multiple guide RNAs can be easily used to guide the cleavage of Cas9 at multiple sites. ^206,207 Genome editing has been used in mice to correct the defect in CF ¹⁴⁹ and surfactant protein deficiency ²⁰⁸ with limited success. Recently, an initial test of CRISPR/Cas9 gene editing in the human lung has been performed in a patient with metastatic non-small-cell lung cancer. ²⁰⁹ These genome editing technologies hold tremendous potential, but much work must still be done to ensure the specificity of the cleavage sites to avoid unintended off target effects and also to improve the efficiency of delivery of the nucleases and guide RNAs to the targeted cells.

Preclinical studies

A detailed summary of preclinical studies of CF lung gene therapy can be found in Table 1. Studies have been carried out in wild-type and experimental models of CF. Eleven different mouse models of CF have been developed (reviewed in Lavelle et al. ²³² ), including several knockouts (CFTR^−/−) generated by different methods in several genetic backgrounds. ^{233 –239} These knockout mice express between 0% and 10% of normal levels of CFTR mRNA and have a low survival rate, but they vary in terms of the pathology of disease. Most of the CFTR knockout mouse models have little or no lung disease but do have pronounced intestinal abnormalities that are often more severe than in the human disease. There are also knock-in models for the F508del, ^{240 –242} G551D, ²⁴³ and G480C ²⁴⁴ human mutations. These mouse models have abnormal electrophysiological profiles and sodium hyperabsorption, similar to the human in the upper respiratory tract. No CFTR mouse models develop the spontaneous lung inflammation found in humans, but lung disease can be induced in some models by exposure to bacteria. ²⁴⁵ Following colonization with Pseudomonas aeruginosa on agar beads, CFTR^−/− mice demonstrated defective bacterial clearance and enhanced levels of inflammation in the lung. ^246,247

The first successful demonstration of viral vector transfer of the human CFTR coding sequence to the airway epithelium of experimental animals used intratracheal administration of an E1⁻E3⁻ Ad. ^248,249 The first test of gene therapy relevant to CF in a mouse used cationic liposomes complexed to a non-viral plasmid encoding a Rous sarcoma virus long terminal repeat (RSV-LTR) driving the human CFTR cDNA. ²⁵⁰ Intratracheal administration of this complex to C57Bl/6NCR wild-type mice resulted in expression of human CFTR that was detected in lung homogenates for up to 4 weeks. Subsequently, numerous studies have been conducted to test the ability of various vectors and formulations to transfer the CFTR gene into the airways of small and large animal models (Table 1). Most have focused on mRNA or protein expression, as well as safety and toxicity studies in normal animals.

Studies in CFTR^−/− mice demonstrated that intratracheal administration of an Ad vector expressing CFTR partially corrected the chloride transport defect in the nasal epithelium. ^77,251 Intratracheal or intranasal administration of plasmid CFTR cDNA with cationic liposomes to CFTR^−/− mice resulted in a partial correction of chloride secretion or sodium adsorption in the trachea. ^{252 –256} Improvement in the nasal airway electrophysical function of F508del mice, measured by transepithelial potential difference, was demonstrated using a truncated version of CFTR (CFTRΔR) expressed from either an Ad or recombinant AAV2 vector after intranasal instillation. ^257,258 Long-lasting expression and partial recovery of nasal electrophysical function up to 12 months post-administration was achieved in CFTR^−/− mice by pretreatment with lysophosphatidylcholine, followed by intranasal inoculation of a vesicular stomatitis virus glycoprotein pseudotyped lentivirus vector expressing human CFTR. ^259,260

Several groups have demonstrated that gene therapy can lead to clinically relevant phenotypes in CFTR^−/− mice. Koehler et al. ⁷⁴ showed that intranasal application of a helper-dependent Ad vector expressing CFTR driven by the human cytokeratin 18 (K18) promoter resulted in mRNA and protein expression in bronchioles of CFTR^−/− mice for 28 days. After challenge with Burkholderia cepacia complex, the treated CFTR^−/− mice showed less severe histopathology and similar lung bacteria counts to CFTR^+/+ littermates. ⁷⁴ Sirninger et al. ²⁶¹ treated CFTR^−/− mice with an AAV5 vector expressing a truncated CFTR in the trachea and then challenged the mice with intratracheal P. aeruginosa after 6 or 10 weeks. Treated mice showed significantly less weight loss and lung inflammation than untreated mice did. Similar results were observed with intratracheal application of CFTR expressed by a SV40 vector. ¹²³ Mueller et al. ²⁶² evaluated a CFTR^−/− mouse model sensitized by exposure to Aspergillus fumigatus to mimic the allergic bronchopulmonary aspergillosis, which is observed in ∼15% of CF patients. Intratracheal treatment of these mice with an AAV5 vector expressing a truncated version of CFTR attenuated the hyper-immunoglobulin E (IgE) and cytokine response that are hallmarks of this phenotype. Lastly, although most gene therapy strategies focus on delivering a functional copy of CFTR to airway cells, McNeer et al. ¹⁴⁹ have demonstrated in CFTR F508del mice that it may be possible to correct the defective gene using site-specific gene editing in vivo. After intranasal application of polymer nanoparticles containing triplex-forming peptide nucleic acids and CFTR DNA, gene modification was confirmed by deep sequencing, and correction was observed in 5% of nasal epithelium and 1% of lung cells. Partial correction in nasal cAMP-mediated transepithelial potential difference response was observed, as was a reduction in inflammatory cells in lung epithelial lining fluid.

Porcine models of CFTR with either a null allele lacking production of the CFTR protein (CFTR^−/−) or containing alleles with the CFTR^{ΔF508/ΔF508} mutation have been developed. ^263,264 Both CFTR pig models recapitulate from birth the pulmonary disease observed in humans, including airway wall thickening and obstruction with excess mucous material. ^263,264 The porcine CFTR models have established that defective clearance of bacterial infection leads to inflammation and further pathology in CF airways. Newborn CFTR^−/− pigs lack lung inflammation, but Staphylococcus aureus was more often recovered from bronchoalveolar lavage fluid from both CFTR^−/− and CFTR^{ΔF508/ΔF508} piglets than from wild-type piglets shortly after birth. ^257,263,265 CFTR^−/− pigs have been used to test the efficacy of gene delivery of a helper-dependent Ad vector carrying human CFTR. ²⁶⁶ Intratracheal administration of the vector resulted in expression of human CFTR protein on the apical surface of airway epithelial cells and submucosal glands with no systemic toxicity. In another study, CFTR was administered using a feline immunodeficiency virus-based lentiviral vector pseudotyped with baculovirus GP64 envelope protein by aerosolization to the lungs of newborn CF pigs. This therapy resulted in a significant increase in transepithelial cAMP-stimulated current in tracheal and bronchus tissues, and increased tracheal airway surface liquid pH and bacterial killing in vector-treated pigs. ²⁶⁷ Administration of an AAV2 vector with five point mutations in the capsid to increase tropism for pig airway epithelia carrying the CFTR cDNA led to increased CFTR protein expression in ciliated and non-ciliated cells and improved chloride anion transport and airway surface liquid pH and bacterial killing. ²⁶⁸ These preclinical studies are important because they show improvement not only in anion transport but in functional phenotypic measurements such as bacterial clearance as well.

Recently, a CFTR^−/− ferret model has been developed because of the similarity to the lung anatomy and cell biology of humans. ^269,270 The CFTR^−/− ferrets showed increased vulnerability to lung infections and had poor nutritional status. These animals showed increased levels of bacteria in the lungs and had airway dysfunction from impaired submucosal gland fluid secretion and increased chloride permeability, which is consistent with CF patients. ²⁷⁰ These features, along with the small size and short time to reach adolescence, may make the ferret CFTR^−/− model useful for the future study of CF disease and treatment. However, to date, this model has not been used to test any gene transfer strategies.

Clinical trials

Prior to beginning human clinical trials, AdCFTR was used to demonstrate that it was possible to transfer CFTR into human CF bronchial epithelial cells in vitro. ²⁴⁸ Following this success, a subject with CF was treated with intra-airway administration of AdCFTR in April 1993, representing the first human to undergo in vivo gene transfer with a recombinant viral vector. ²⁷¹ To date, 26 human trials of gene therapy for the pulmonary manifestations of CF have been completed (Table 2). Several methods have been used to deliver the CFTR cDNA to the lung, including Ad and AAV vectors, cationic liposomes, and nanoparticles. The bulk of the clinical studies have focused on safety, molecular evidence of gene transfer, and electrophysiological outcome measures such as correction of transepithelial potential difference. ²⁷² Overall, CFTR gene therapy has been shown to be safe and well tolerated.

Early trials with Ad vectors demonstrated the proof of principle that CFTR can be transferred into the nasal or lung epithelium, albeit with limited efficiency and transient expression. ^{83,84,86 –88,273,274} However, CFTR gene expression mediated with Ad vectors diminishes over time, even after repeated administration, likely due to immune responses to the vector. ^85,88,275 In some studies, an improvement in transepithelial potential or cAMP response has been demonstrated after CFTR gene transfer. ^{83,89,147,276} Two trials delivering liposome-CFTR cDNA complexes by nebulizer to the nasal or lung epithelium have suggested improvement in clinical outcomes. ^277,278 Following a single dose, there were improvements in the pulmonary airway potential difference and chloride efflux, as well as decreased inflammatory cells in sputum and less bacterial adherence in the lung. ²⁷⁸ CF patients given a monthly application of a liposome-CFTR cDNA complex for 1 year showed stabilization of lung function, as measured by forced expiratory volume in 1 s (FEV₁), while the lung function of placebo recipients declined over time. ²⁷⁷

For future studies, both preclinical and clinical, it will be important to understand the identity and the number of cells that need to receive a functioning CFTR in order to elicit a relevant clinical response. Until recently, many studies in mice evaluated CF disease only by a change in electrophysical measurements of ion transport, such as nasal transepithelial potential difference, because a defect similar to that in human CF airways is exhibited. A correction of these measurements, and thus the ion transport defect, was then used to evaluate treatment efficacy. Although this technique is non-invasive and can generally differentiate between wild-type and CF individuals based on low-chloride response, the differences can be smaller, and these measurements may vary significantly between mouse strains and individuals. ²⁷⁹ Additionally, the different composition of olfactory and ciliated epithelium in mouse nasal epithelium (50/50) versus human airway epithelium (97/3) suggests that much of the observed response in the mouse may be from the olfactory cells, which are not the main targets of CFTR correction, that react differently from ciliated cells when stimulated with forskolin or amiloride. ²⁵¹ Future preclinical studies of CFTR should focus on other critical parameters of disease such as bacterial clearance in addition to measuring nasal potential difference to insure an accurate picture of CFTR functional rescue.

Perspective

The ongoing challenges of CF gene therapy in the lung are numerous. There are several lung pathologies related to disease, and the level of protein replacement necessary to correct them and the cell populations to target remain unclear and require further study. Although it has been demonstrated that CFTR can be transferred into the cells of the lung, the level of protein expression and percent of transduced cells was low. Further, the natural turnover of these cells necessitates repeated delivery of the gene or targeting the gene to lung progenitor cell populations to maintain expression. Current methods of delivery, including both viral and non-viral vectors, induce an inflammatory immune response upon repeated administration, and therefore improved methods of delivery or ways to combat inflammation must be sought. Additionally, continued clinical testing of new gene therapies depends on the development of better assays and clinical outcome markers that are more indicative of disease improvement. Furthermore, although pulmonary disease causes the majority of morbidity and mortality for CF patients, the CFTR mutation affects other organs in which replacement by gene therapy may be beneficial.

Preclinical studies

Several viral and non-viral gene delivery systems have been evaluated for treating the α1AT deficiency. The general strategy is to restore the necessary protective levels of α1AT to the serum (11 μM) and lung epithelial lining fluid (1.2 μM) to prevent lung disease ^{9,282,285,296} by delivering the normal human M type α1AT cDNA. ^7,9 Protein augmentation studies have established that levels >80 mg/mL have no adverse effects/clinical sequelae, thereby defining an acceptable and flexible range of therapeutic protein expression. ²⁹⁸ Gene therapy investigations related to α1AT have been primarily tested in small and large wild-type animal models, including mice, rats, sheep, baboons, dogs, rabbits, and nonhuman primates. These studies were recently reviewed by Chiuchiolo et al. ⁹ and are summarized in Table 3.

Currently, there is no animal model that recapitulates the α1AT lung disease. ³⁰⁰ There have been various attempts to create knockout animal models of the disease, but it has been challenging because there are several murine α1AT genes, and targeted deletion of the α1AT gene (SERPIN1A) resulted in embryonic lethality. ^301,302 There are transgenic mouse models, such as the PiZ mice that carry the human α1AT transgene with the Z mutation superimposed on the wild-type murine α1AT genes, which recapitulate some aspects of the liver disease but not the lung disease. ^301,303,304 There is also the pallid mouse, which is a naturally occurring strain of mice with a mutation in the pallid gene that hampers the normal secretion of α1AT. ³⁰⁵ This results in spontaneous emphysema in these mice later in life. These mice have ∼50% of normal levels of circulating α1AT and are therefore not an ideal model for gene therapy studies. ³⁰¹

If the gene can be effectively transferred, one advantage of α1AT deficiency as a gene therapy target is that the protein can be produced by many different cells/organs. ⁹ In an ex vivo strategy, the first experimental animal model to assess α1AT delivery used a retrovirus to modify fibroblasts genetically, which were then transplanted to the peritoneal cavity of nude mice. At 4 weeks, human α1AT was detected in both sera and the epithelial surface of the lung. ¹⁴ The first in vivo approach to α1AT gene therapy used direct intratracheal administration of an Ad vector encoding human α1AT to cotton rats, resulting in the synthesis and secretion of an α1AT protein product of normal size and function, detectable in the lung epithelial lining fluid and bronchoalveolar lavage fluid for at least 1 week. ⁶² The transient expression of the transgene could not be addressed by repeated administration of the vector due to vector-induced immunity. Liposomes were investigated for their ability to deliver the α1AT cDNA directly to the lungs of rabbits by either an aerosol or an intravenous route at weekly intervals for a period of 4 weeks. ^306,307 In this strategy, α1AT expression persisted for the 4-week period studied, with no evidence of associated toxicity, but only low levels of α1AT were achieved.

The current gene therapy approaches for α1AT deficiency focus primarily on AAV vector-based gene therapy, with the goal of maintaining sustained expression of α1AT therapeutic levels. Various AAV serotypes delivered using different methods of administration have been explored. Initial studies were carried out using AAV2 vectors delivered intramuscularly or intravenously. ⁶

Efficiency in delivery to the lung was improved using AAV5 vectors delivered by the intrapleural route. ²¹¹ The pleura presents several structural advantages that make it an attractive, easily accessible site for gene delivery to the lung parenchyma. Foremost, it provides a large surface area for gene transfer. In addition, the parietal (chest-wall surface) pleura has direct opening to the lymphatic system; when AAV vectors are administered to the pleura, a significant amount reaches the systemic circulation, enabling gene transfer to the liver. ²¹¹ Assessment of 25 different AAV serotypes administered to the pleura of wild-type mice demonstrated that AAV serotype rh.10 (AAVrh.10) yielded the highest sustained levels of α1AT (>2.5 times the target level). ¹⁰⁷ Studies in nonhuman primates and mice with AAVrh.10-mediated gene transfer of α1AT demonstrated effectiveness and safety. ³⁰⁸

In a comparison of serotypes AAV1, 2, 3, 4, and 5 via intramuscular delivery, AAV1 mediated the highest muscle transduction efficiency, with sustained protein expression in serum up to 84 weeks, and high serum α1AT levels 100-fold higher compared with AAV2. ³⁰⁹ Follow-up safety studies in mice and rabbits showed this vector to be safe. ³¹⁰

In other studies comparing various AAV serotypes delivered either intratracheally or intranasally, the highest levels of α1AT in lung epithelial lining fluid and lung tissue were achieved with AAV8. ⁴⁵ In a study of expression of α1AT comparing AAV1 to AAV5 by direct aerosol administration via a fiberoptic bronchoscopy in chimpanzees, AAV1 was more potent and less immunogenic than AAV5 was. ³¹¹ Of all the various AAV serotypes tested in animal studies, AAV1, AAV2, and AAVrh.10 vectors encoding α1AT cDNA have moved to the clinic.

Clinical trials

To date, there are only a few clinical trials of gene therapy for α1AT deficiency (Table 4). The preclinical studies by Canonico et al. ^306,307 involving administration of a plasmid encoding human α1AT complexed to cationic liposomes in the lungs of rabbits led to the first clinical study of α1AT gene replacement. Five subjects with α1AT deficiency received this plasmid–liposome complex by instillation in one nostril, with the other nostril serving as a control. α1AT protein levels in nasal lavage fluid increased in the transfected nostril but not in the control, peaking at 5 days post administration to one-third of the normal level. α1AT levels in the nasal lavage were transient, returning to baseline by day 14. ³¹²

AAV2 preclinical studies led to a Phase I trial for the intramuscular delivery of an AAV2 vector expressing the α1AT cDNA. A total of 12 α1AT deficient subjects participated in a dose-ranging study (2.1 × 10¹²–6.9 × 10¹³ vector genomes [vg]). A low, transient elevation of α1AT protein in serum was detected in one subject from the 2.1 × 10¹³ vg dose group on day 30 after vector administration. A second clinical trial with α1AT deficient subjects used intramuscular delivery of an AAV1 vector. This study enrolled nine α1AT deficient subjects who received total doses ranging from 6.9 × 10¹² to 6.9 × 10¹³ vg. There were no significant safety issues, with sustained levels of 0.1% of the serum therapeutic threshold of 11 μM up to 1 year for the subjects receiving the highest dose. All the subjects had an immune response to the capsid at the 14-day timepoint. ³¹³ Following a safety study with the higher doses in mice, a Phase II clinical study with a higher dose was initiated. ³¹⁴ Nine α1AT deficient subjects were enrolled, and they were administered doses up to 6.0 × 10¹¹–6.0 × 10¹² vg/kg. There was a dose-dependent increase in α1AT serum levels that peaked 30 days after vector administration, with serum α1AT mean value of 0.6 μM in the highest dose cohort, declining to 0.2 μM by day 90. ^9,314 The investigators observed a T-cell regulatory response that allowed ongoing transgene expression at 1 year, suggesting that immunomodulatory suppression may not be necessary for this approach to AAV-mediated gene therapy. ³¹⁶

Experimental animal studies demonstrating the safety and efficacy of AAVrh.10 delivered α1AT to the pleural space has led to the approval of a Phase I/II clinical trial. The aim of this trial is to assess the hypothesis that a single intrapleural administration of a serotype AAVrh.10 vector expressing the normal M-type α1AT to individuals with α1AT deficiency is safe and results in persistent therapeutic serum and alveolar epithelial lining fluid levels of α1AT. ^107,308,317

Perspective

Significant advances have been made both in the safety of gene therapy for α1AT and in improved expression levels achieved through alterations in vector design. In spite of these advances, the critical challenge to successful gene therapy for α1AT deficiency is the requirement to maintain serum α1AT levels of 11 μM, the approved target for protein augmentation therapy necessary to protect the lung from the action of neutrophil proteases.

Preclinical studies

Several animal models have been used to mimic human asthma, although specific models that spontaneously develop lung disease with all of the human characteristics are lacking. ³²⁴ Although transgenic mice are the most often used animal model, rats, guinea pigs, ferrets, monkeys, cats, pigs, sheep, and horses have all been used in various gene therapy studies targeted toward key mediators of asthma. ^322,324,325

Mice are most commonly used to model asthma. The animals are first sensitized to a high dose of allergen (generally ovalbumin) with alum adjuvant, leading to the development of specific T-cell and IgE and IgG1 antibody responses. Further exposure of mice to the allergen via inhalation, intranasal, or intratracheal routes allows antigen distribution to the lower airways. This results in both an early acute neutrophilic inflammatory response and a late sustained accumulation of eosinophils and lymphocytes in the lung, an increase in Th2 cytokine levels (IL-4, IL-5, IL-13), goblet cell hyperplasia, mucous production, and increased airway reactivity to a chemical bronchoconstrictor. ^326,327 The use of this sensitization protocol on mice lacking various cytokines, receptors, or immune cell types has led to the identification of a number of mediators of the allergic asthma response.

Models of asthma airway inflammation in the mouse have been used to test a number of candidates for gene therapy (Table 5). Early studies focused on the overexpression of interferon gamma (IFN-γ) to shift the immune response toward the Th1 pathway. Systemic or intratracheal administration of an IFN-γ plasmid coupled with liposomes led to decreased airway hyper-responsiveness and eosinophilia in lung epithelial lining fluid. ^328,329 Systemic expression of a IFN-γ plasmid reduced the levels of serum IgE. ³²⁹ Tests with intranasal application of nanoparticles containing an IFN-γ plasmid also yielded decreased airway hyper-responsiveness, lower IL-5 levels, and decreased cytokine production by CD8+ T cells. ¹⁵⁹

Expression of the inhibitory or immunosuppressive cytokines IL-10, IL-12, or TGF-β individually from plasmids after intratracheal administration resulted in decreased airway hyper-responsiveness and fewer eosinophils and neutrophils in lung epithelial lining fluid. ^330,331 Administration of an Ad vector expressing IL-10 and IL-12 in combination decreased the levels of IL-4, IL-5, and eotaxin in the mouse ovalbumin-induced asthma model. ³³² Overexpression of the transcription factor T-bet, the master regulator of Th1 versus Th2 lineage commitment, has been evaluated in the mouse ovalbumin-induced asthma model using intranasal administration of an AAV vector. T-bet overexpression resulted in reduced levels of IL-4 and IL-5 and increased IFN-γ levels in lung lavage fluid. Reduction of eosinophils in lung lavage fluid, IgE in serum, and bronchial inflammation were also observed. ³³³ Overexpression of galectin-3 from a plasmid after intranasal instillation or diacylglyceral kinase alpha from a plasmid after intramuscular injection were also shown to decrease airway hyper-responsiveness, eosinophilia, and serum IgE levels in the ovalbumin-induced asthma mouse model. ^158,334

Mediators of the inflammatory response in allergic asthma are prime targets for inhibition through gene therapy. IL-5 has been targeted by numerous groups using intratracheal administration of lentivirus vectors to deliver siRNA, or intravenous injection of recombinant AAV to deliver ASOs against this cytokine. These studies demonstrated that inhibiting IL-5 expression reduces eosinophils and eotaxin in lavage fluid and inhibits lung inflammation. ^196,335,336

Another important mediator relevant to asthma is IL-4, which has been targeted with multiple gene therapy strategies. Cao et al. ³³⁷ treated rats with ovalbumin-induced asthma intravenously with an AAV vector expressing ASOs targeting cytokine IL-4 and observed decreased eosinophils and cells expressing TGF-β and reduced airway inflammation. Three studies tested an IL-4 receptor antagonist (IL-4RA) expressed by intratracheal administration of a plasmid, ³³⁸ AAV2 vector, ³³⁹ or moloney murine leukemia retrovirus ³⁴⁰ vector in asthmatic mice, with resulting reduction in airway hyper-responsiveness, eosinophil infiltration, and Th2 cytokine production. Another group employed naked ASOs to the IL-4 receptor via inhalation and observed similar results. ¹⁸²

An additional target for treatment of asthma is GATA-3. Intranasal application of a naked ASOs or lentivirus expressing a shRNA against GATA-3 showed reductions in airway hyper-responsiveness, eosinophils, and Th2 cytokine production in the ovalbumin-induced asthma mouse model. ^178,341 Targeting GATA-3 with a naked DNAzyme by intranasal application helped to reduce airway hyper-responsiveness, inflammation, and mucous production in an ovalbumin-induced mouse model. ¹⁷² Other interesting asthma-related targets for inhibition by gene-silencing techniques include NFκB p65 subunit, ¹⁷⁵ Syk kinase, ¹⁸³ acidic mammalian chitinase (AMCase), ³⁴² the voltage-dependent calcium channel Ca_v1, ³⁴³ CD86, ¹⁷⁶ CD40, ¹⁸⁴ and STAT6¹⁷⁷ (Table 5).

Clinical trials

Only a few clinical trials have been carried out for gene therapy for asthma (Table 6). In one study, mild asthmatic patients were given a mixture of naked ASOs against CCR3 (the eotaxin receptor) and the common β chain of the IL-3/IL-5/GM-CSF receptors ¹⁷⁹ by nebulizer. After allergen challenge, the treated group had lowered early asthmatic response, as measured by FEV₁, and reduced eosinophilia in the airways. A second study with the same therapy showed an additional reduction of the late asthmatic response in treated patients. ¹⁸¹ Recently, a naked DNAzyme against GATA-3 was tested in patients by inhalation and found to be safe and well tolerated, but the efficacy of this treatment has not yet been determined. ¹⁸⁰

Perspective

None of the currently available treatments for asthma represent a long-term solution, as they must be taken regularly to alleviate symptoms. Asthma is a complex syndrome without a single genetic mutation that can easily be targeted. Although many mediators have already been identified, more work is needed to understand the key regulators that should be targeted to best control an aberrant allergic immune response without dampening an effective response against pathogens and other foreign agents. Further understanding of asthma will be aided by the development of better animal models, as the mouse models currently available only represent the acute and transient disease but do not recapitulate the effects of chronic disease. Overcoming these challenges will require further development of novel delivery methods and gene targeting agents.

Preclinical studies

The mouse models for allergic rhinitis are less well developed than models for asthma, but they are built on the same principle of sensitization to an allergen followed by subsequent challenges. ³²⁴ Mice are first sensitized to the allergen (usually ovalbumin) with alum adjuvant by intraperitoneal injection at a low dose. Then, non-anesthetized mice are challenged with the same allergen daily by intranasal inhalation, which confines the antigen in the upper airways. After challenge, sneezing and nasal itching as well as nasal hyper-responsiveness are observed. ³⁴⁷ Increased numbers of eosinophils, basophils, and CD4+ T cells are found in the nasal mucosal tissue, and there are elevated levels of allergen-specific IgE in the serum. This model represents the symptoms of allergic rhinitis in the upper airways without showing any changes in the lower airways. ³⁴⁷

The allergic rhinitis mouse model has been used to study several different types of gene therapy (Table 7). Intranasal instillation of an Epstein–Barr virus (EBV)-based plasmid expressing a decoy TNF-α receptor-IgGFc inhibited the allergic response in this model. ³⁴⁸ The treated group showed decreased sneezing and itching of the nose and less infiltration of eosinophils, mast cells, and IL-5⁺ cells in the nasal mucosa. Similar results were obtained after intranasal administration of an EBV-based plasmid expressing IL-12, which promotes differentiation to the Th1 response. In addition to less eosinophil infiltration in the nasal mucosa, a lower level of IgE was measured in serum. ³⁴⁹ In a different study, intranasal delivery of siRNA with liposomes against STAT6 decreased inflammation, mucous production, sneezing, and nasal rubbing in mice with induced rhinitis. ¹⁸⁵ A lentiviral vector expressing siRNA against CCR3 was tested by intranasal administration in a mouse model of allergic rhinitis. Lower levels of eosinophils were observed in nasal lavage, blood, and bone marrow after treatment. ³⁵⁰ Finally, intranasal administration of an Ad5 vector expressing antibody against inducible co-stimulator, a molecule required for T-cell activation, attenuated nasal inflammation, decreased eosinophil infiltration, and decreased IL-5 expression. ³⁵¹

Clinical trials

To date, there have been no clinical trials of gene therapy for allergic rhinitis.

Perspective

Gene therapy for allergic rhinitis faces many similar challenges to the treatment of asthma, including the lack of a single obvious target, the incomplete understanding of the key mediators of disease, and the need to tune the immune response carefully in order to avoid an allergic response without affecting the response to pathogens. As with gene therapy for other airway diseases, transferring the gene to the correct cells for long-term expression is a challenge. This problem is exacerbated in allergic rhinitis because of the loss of epithelial cells due to inflammation and increased mucous production that acts as a barrier to vectors. As allergic rhinitis is generally controlled with currently available medications and treatments, further interest in gene therapy treatment likely would require significant advances in vector design and delivery that would allow for a single-dose long-term effective treatment.

Preclinical studies

Mouse strains that readily develop Th2 responses, such as C3H/HeJ and BALB/c, and Brown Norway rats have been used to establish small animal models for food allergy. These animals are able to produce IgE antibodies in response to sensitization with food allergens. ^324,356 Sensitization involves repeatedly exposing the animal to small amounts of allergen by oral, nasal, intraperitoneal, or cutaneous routes daily or weekly followed by a larger dose oral challenge. ³⁵⁷ Large animal models, such as dogs, pigs, and sheep, have also been useful for understanding food allergy because their physiology is more similar to humans, and dogs spontaneously develop allergic reactions to many of the same food allergens as people. ^324,356

Two preclinical experiments testing gene therapy for food allergy have been conducted in mouse models of peanut allergy (Table 8). Roy et al. ¹⁶⁵ assessed the treatment of peanut allergy in mice by oral immunization with nanoparticles containing plasmid DNA expressing the dominant peanut allergen gene Arah2. Mice receiving the prophylactic vaccine showed reduced levels of IgE, plasma histamine, and vascular leakage in response to peanut challenge. Recently, Pagovich et al. ³⁵⁸ used a humanized mouse model of peanut allergy and delivered an AAVrh.10 vector expressing anti-IgE antibody as either a prophylactic or therapeutic treatment. Mice treated with anti-IgE gene therapy after peanut sensitization had lower free serum IgE, histamine, and anaphylaxis score upon peanut challenge. The treated mice showed no clinical signs of allergy and had significantly improved long-term survival compared with untreated mice.

Clinical trials

To date, no clinical trials of gene therapy for food allergy have been initiated.

Perspective

Food allergy is a severe disease that is currently only controlled by avoidance of triggering foods and treatment of anaphylaxis after exposure. Gene therapy could provide the means of increasing tolerance to triggering foods or reducing the IgE response upon exposure.

Preclinical studies

A variety of preclinical experiments have been described for gene therapy for lung cancer. The following are examples; a detailed list can be found in Table 9.

Murine models used to study lung cancer can be categorized into four groups: xenograft, syngeneic, transgenic, and carcinogen-inducible (reviewed in Kellar et al. ³⁶⁶ ). The xenograft model is developed by injecting human cancer cell lines or tumor tissues removed from patients into immunocompromised mice. Xenograft models are useful for therapy studies because they re-create the tumor microenvironment. Syngeneic models are derived by engrafting immunocompetent mice with compatible cancer cells from a mouse tumor, such as the Lewis lung carcinoma. ³⁶⁷ This model has the advantage of maintaining syngeneic immune and toxicity responses. Transgenic models utilize mice that have been genetically engineered with mutations or oncogenes that drive development of tumors in the lungs. ³⁶⁶ These models are ideal for determining the role of genetic abnormalities in tumor formation. One frequently utilized transgenic mouse model harbors oncogenic alleles of the K-ras that increase the susceptibility to developing spontaneous early-onset lung cancer. ³⁶⁸ Carcinogen-inducible mouse models require the introduction of a carcinogen to induce genetic mutations that lead to the development of lung tumors in susceptible mice. For example, intraperitoneal injection of urethane reliably introduces mutations in K-ras and p53 that lead to the development of adenocarcinoma. ³⁶⁹ These inducible mouse models are advantageous in that all stages of tumors can be observed, but the response to the carcinogen for development of tumors may vary, and the incubation time until tumor development can be long. ³⁶⁶

The earliest lung cancer gene therapy in animal models focused on delivering the herpes simplex virus (HSV) thymidine kinase (tk) “suicide gene” or the tumor suppressor p53. The suicide gene approach delivers the gene for an enzyme that induces sensitivity of the tumor cells to another agent, such as the pro-drug ganciclovir, which in itself is harmless until processed by the enzyme. Delivery of the HSVtk gene as a plasmid with liposomes ³⁷⁰ or via Ad, ^371,372 JC polyoma virus, ¹²⁴ or lentiviral ³⁷³ vector into non–small cell lung cancer (NSCLC) mouse models followed by ganciclovir treatment was shown to be effective at killing tumor cells. Gene transfer of p53 by a number of different methods, including Ad vectors ³⁷⁴ and plasmid delivery by liposomes or nanoparticles ^{161,164,375 –380} into various mouse models of lung cancer resulted in cell apoptosis, tumor growth suppression, and enhanced survival. Other models have used Ad vector or nanoparticle delivered plasmid HSVtk or p53 gene therapy in combination with other therapies such as traditional chemotherapy ^162,381 or additional gene therapy with genes such as IL-2³⁸² or microRNA-125b. ³⁸³

The Akt (protein kinase B or PKB) signaling pathway is activated in many tumor types, including 90% of NSCLC, which makes this an attractive lung gene therapy target. Two approaches have been tested in animal models that resulted in lower Akt signaling and decreased lung tumorigenesis. One approach used plasmid/liposome complexes or Ad5 vector to deliver the regulatory protein phosphatase and tensin homolog (PTEN) ^157,384,385 or lentivirus expressing carboxyl-terminal modulator protein (CTMP) ^386,387 to downregulate Akt signaling. Another approach used nanoparticles to deliver shRNA or ASOs to downregulate the expression of Akt directly. ^{150 –152} A dual expression plasmid expressing Akt shRNA and the gene for programmed cell death protein 4 (PDCD4) showed a synergistic effect further decreasing lung tumorigenesis. ¹⁹⁵

A variety of gene therapy vectors, including Ad, ^388,389 AAV, ³⁹⁰ and baculovirus, ¹³⁰ or plasmid/liposome complexes, ^160,391,392 have been used to overexpress immunostimulatory cytokines such as IFN-β, IL-12, and IL-15 in lung tumor models, with consequent cytotoxic immune responses and inhibition of tumor growth. The immune system can also be stimulated by introduction of ligands to stimulate antigen-presenting cells. Expression of CD40 ligand from an Ad vector suppressed tumor growth and increased survival in a mouse lung metastasis model. ³⁹³

Angiogenesis plays a key role in tumor formation, and this process has been targeted with gene therapy to deliver inhibitor proteins to block blood-vessel formation in lung tumors. The intravenous administration of Ad vectors expressing endostatin ^{394 –396} or flt-1, ^159,397 AAV2 vectors expressing vasostatin or plasminogen, ^398,399 lentivirus expressing kallistatin, ⁴⁰⁰ or intratumoral injection of AAV or Ad expressing human pigment epithelial-derived factor (PEDF) ^401,402 or plasmid expressing canstatin ⁴⁰¹ all decreased the number of blood vessels, slowed tumor growth, and in some cases suppressed metastasis or enhanced survival. AAV expression of an anti-VEGF antibody yielded long-term antibody expression in the lung up to 40 weeks and significantly suppressed metastatic lung tumor growth and blood-vessel number and increased survival. ¹¹¹

Another group of proteins targeted for gene therapy for lung cancer are those involved in apoptosis. Expressing pro-apoptotic proteins such as TNF-related apoptosis-inducing ligand, (TRAIL) from an Ad5 vector or plasmid ^{404 –406} or PDCD4 from plasmid nanoparticles ^153,154 in lung tumors induced apoptosis and promoted tumor regression in mouse models. In vivo RNAi silencing by lentivirus expressed shRNA of survivin, an apoptosis inhibitor upregulated in numerous tumor types, markedly inhibited tumor growth in a mouse NSCLC model. ⁴⁰⁷

Clinical trials

A number of gene therapy strategies have been tested in the clinic for the treatment of lung cancer (Table 10).

Gene transfer–mediated addition of tumor suppressor genes has been attempted with p53 in trials in patients with NSCLC. Intratumoral injection of up to 5 × 10¹¹ particles of Ad vector expressing p53 achieved gene expression, was minimally toxic, and elicited tumor stabilization or regression in some patients. ^{408 –413} For patients with advanced lung cancer or malignant pleural effusion induced by lung cancer, the combination of intratumoral or intracavitary Ad vector mediated p53 gene therapy with the platinum-based chemotherapy drug, cisplatin was found to be more effective than chemotherapy alone. ^{414 –416} A clinical trial for metastatic lung cancer tested systemic treatment with intravenous injection of liposome nanoparticles containing plasmid DNA expressing the gene for tumor suppressor candidate 2 (TUSC2), a tumor suppressor gene commonly deleted in lung cancer cells. Tumors removed from patients post treatment showed TUSC2 protein expression and induction of apoptotic genes in tumor cells, and 20% of patients showed stabilization of disease. ¹⁶³

Tumor growth factors contributing to tumor expansion and progression are attractive targets for downregulation by gene therapy with ASOs. However, clinical trials employing naked ASOs administered by intravenous infusion directed against protein kinase C-α, ^197,200 Raf-1, ^{191,192,199,417} or Bcl-2^193,194,198 in lung cancer patients have been performed but showed varying levels of toxicity and no clinical benefit.

Another approach to treating lung cancer with gene therapy is to modify immune cells genetically ex vivo and then deliver the modified cells to the patient to stimulate an immune response against the tumor cells. In the first trial of this type, lymphocytes from patients with advanced lung cancer with pleural effusions were transduced with a retrovirus expressing IL-2 and then reinfused into the chest cavity. This treatment was found to be safe, and pleural effusions were resolved for at least 4 weeks. A reduction in tumor size was noted in 1/10 patients. ⁴¹⁸ Cells from NSCLC tumors were infected with an Ad5 expressing GM-CSF, irradiated, and re-administered to the patient intradermally at biweekly intervals. An immune response was elicited in 70% of patients, including tumor infiltrating T cells, and long-term disease stabilization was achieved for 25% of patients. ⁴¹⁹ Further studies showed that longer survival was correlated with higher expression of GM-CSF. ⁴²⁰ Another target for this approach is the downregulation of the immunosuppressive protein TGF-β2 to enhance antitumor immunogenicity. NSCLC patient tumor cells transfected with a plasmid expressing ASOs to TGF-β2 were injected intradermally into patients. Increased cytokine production and increased survival time were observed and correlated with the dose of cells injected. ⁴²¹ However, a Phase III study found no difference in overall survival for treatment with TGF-β2 gene therapy after chemotherapy than for chemotherapy alone. ⁴²² In another study, NSCLC patients were given an intradermal injection of an irradiated adenocarcinoma cell line transfected with CD80, a cytokine important for T-cell activation, and HLA-A1 or -A2 (matched to patient HLA type). Nearly all patients (94%) had a measurable CD8 T-cell response after three immunizations that persisted for up to 3 years in surviving patients (30%). ⁴²³

Lastly, viral vectors have been under development to “vaccinate” lung cancer patients with tumor antigens to stimulate the antitumor immune response. Trials for this approach have been conducted using intramuscular administration of an Ad vector expressing L523S lung cancer antigen, which is expressed on 80% of NSCLC cells ⁴²⁴ or by subcutaneous injection of a vaccinia virus expressing IL-2 and MUC1, which are expressed in excess by tumor cells. ¹²⁷ However, neither strategy showed clinical efficacy.

Perspective

Among all malignancies, lung cancer has one of the lowest survival rates due to the lack of efficacy of the treatment options available and the late stage of disease diagnosis. However, the genes and pathways involved in the development of lung cancer and subsequent new therapies continue to be identified and developed. Early detection and a personalized approach to treatment by pinpointing the exact mutations present in an individual patient may make targeted treatments more effective. Novel new strategies to target mutated or overexpressed genes such as the CRISPR-Cas9 system may allow genes to be edited in vivo and revolutionize treatment of certain tumors. The first trial using this gene editing system for lung cancer treatment was performed in China in November 2016. ²⁰⁹ The combination of traditional treatments such as surgery and chemotherapy, along with various new types of gene therapy, may prove to be effective approach for combatting lung cancer.

Preclinical studies

The most commonly used model for mesothelioma are subcutaneous or orthotopic xenografts of human mesothelioma tumor cells in mice or rats. ⁴²⁵ The majority of preclinical experiments in rodent models have been performed with Ad vectors, and many of the gene targets are similar to those identified for lung cancer (Table 11). Intraperitoneal or intrapleural administration of an Ad5 vector expressing HSVtk followed by ganciclovir effectively regressed mesothelioma tumor progression. ^{371,426 –430} Intratumoral treatment with a retrovirus expressing the suicide gene cytosine deaminase followed by 5-fluorocytosine treatment also reduced mesothelioma tumors in mouse models. ^431,432

Using gene therapy to stimulate the immune response by intratumoral treatment with Ad vector expressing IFN-β decreased tumor size and increased the time to tumor recurrence and disease-free survival. ⁴³³ The influx of CD4 and CD8 cells to the peritoneal space stimulated by the IFN-β expression mediated tumor regression. ^434,435 Intratumoral recombinant E1/E3-deleted Ad vector expression of CD40 ligand increased infiltration of intratumoral CD8 T cells and induced tumor regression. ⁴³⁶ Similar decreases in tumor growth and increased survival were observed upon treatment with a combination of Ad expressed angiogenesis inhibitor proteins. ⁴³⁷

Clinical trials

Ad-mediated gene therapy has been the basis for the majority of clinical trials for mesothelioma (Table 12). Ad5 expressing HSVtk, a classic suicide vector, was introduced into the pleura of patients with mesothelioma followed by treatment with ganciclovir. ⁴³⁸ Expression of HSVtk made the transduced cells susceptible to the drug causing apoptosis. ⁴³⁹ This treatment was found to be safe, and gene transfer was achieved in 60% of patients into multiple layers of tumor cells. ⁴³⁸ Approximately 40% of these treated patients showed tumor stabilization or regression, and three patients survived >7 years after treatment. ^438,440

Another gene therapy strategy for mesothelioma is to introduce a gene for an immunostimulatory cytokine to stimulate an antitumor immune response. ⁴³⁹ Intratumoral injection of a vaccinia virus expressing IL-2 was tested weekly for 12 weeks in mesothelioma patients. IL-2 expression was detected up to 3 weeks after injection, despite an IgG response to the vector. ¹²⁶ Intrapleural administration of the cytokine IFN-β expressed with an E1/E3-deleted Ad vector has also been evaluated in mesothelioma patients. Gene transfer and antitumoral immune response was detected in 70% of patients, and 40% of them showed tumor stabilization or regression with several surviving >18 months. ^441,442 Finally, a recent trial of intrapleural administration of an Ad5 vector expressing IFN-α showed disease stabilization in 62% of patients and a partial response in 25% of patients. ⁴⁴³

Perspective

Mesothelioma is an aggressive tumor that is rarely treated effectively with therapies currently available. Like lung cancer, new research in target identification and pinpointing the exact mutations in an individual tumor may allow for the development of more effective gene therapy strategies that can be used in combination with surgery and chemotherapy. Further, designing gene therapy vectors to target cancer stem cells may help to eliminate tumor recurrence, but much research is still needed to define the properties of these stem cells and the genes to target for delivery and therapy.

Preclinical studies

There have been several targets for investigating gene therapy strategies for the treatment of PAH (Table 13). Studies have shown that there is a link between the hereditary forms of PAH and mutations in bone morphogenetic protein receptor 2 (BMPR2). Decreases in levels of BMPR2 have been correlated in PAH patients, and BMPR2 mutations are present in >70% of idiopathic PAH cases. ^445,446 Investigators have observed that BMPR2 inhibits smooth-muscle cell proliferation, promotes the survival of pulmonary arterial endothelial cells, and prevents pulmonary damage. Mutations in BMPR2 lead to vascular smooth-muscle proliferation, PAH, and right-side heart failure. As a result, BMPR2 gene therapy has been considered as a possible therapeutic option, with the aim of targeting pulmonary arterial endothelial cells either by the vascular route or via inhaled gene therapy.

In one study, intravenous delivery of an Ad5 vector encoding BMPR2 was targeted to pulmonary endothelium by linking the vector capsid to a bispecific antibody that targets the virus to angiotensin-converting enzyme (ACE), a membrane bound protease that is highly expressed on pulmonary endothelial cells. This was tested in a rat model of hypoxic pulmonary hypertension and substantially reduced the pulmonary hypertensive response to chronic hypoxia. ⁴⁴⁷ Positive results were obtained in a subsequent study where two models of PAH—the hypoxia rat model and the monocrotaline (MCT) rat model—were administered the Ad5 vector encoding BMPR2 combined with a pulmonary targeting conjugate. ⁴⁴⁸ This resulted in amelioration of the decrease in BMPR2 expression induced by hypoxia or MCT challenge, amelioration of the increase in TGF-β expression seen in the MCT model, and reduced development of pulmonary hypertension and associated vascular remodeling. ⁴⁴⁸ In a recent study, a transgenic mouse model of PAH based on the overexpression of a BMPR2 dominant negative mutant was administered intravenously an Ad5 vector encoding the BMPR2 gene combined with a pulmonary targeting conjugate (similar to that described above where the Ad is linked to a bispecific antibody). This led to the mitigation of PAH symptoms in the mutant mice, including the reversal of the increase in the systolic pressure of the right ventricle and of right ventricle hypertrophy. ⁴⁴⁹ Similar results were observed in the MCT rat model, where additional studies were carried out to elucidate the PAH related changes in the endothelial cell signaling. ⁴⁵⁰

Connective tissue growth factor (CTGF) has also been associated with the pathogenesis of PAH. A plasmid-based CTGF-specific shRNA complexed with a transfection reagent, PEI (ExGen 500), was tested in the MCT model of PAH. The results demonstrated that intratracheal administration of this plasmid suppressed the pulmonary vascular remodeling that is induced in the rat model by monocrotaline. ²⁰²

Another target for a gene therapeutic approach for PAH is the C-type natriuretic peptide/natriuretic peptide receptor 2 (CNP/NPR2) signaling pathway, which is linked to elevations of intracellular cyclic guanosine monophosphate (cGMP). Elevations in cGMP have been suggested to suppress proliferation effectively and induce apoptosis in the pulmonary arterial smooth-muscle cells. A SeV carrying a constitutively active mutant of NPR2 was administered directly into the left pulmonary artery of a hypoxia rat model of PAH. This treatment resulted in a significant reduction in the right ventricular systolic pressure 2 weeks after treatment, and there was also histological improvement in the lungs of the treated animals. ¹³⁵

Thickening and narrowing of the pulmonary vasculature is triggered by abnormal calcium levels within the vascular cells. The sarcoplasmic reticulum calcium ATPase pump (SERCA2a) regulates intracellular calcium in vascular cells and prevents the cells in the vessel wall from proliferating. When SERCA2a is downregulated, calcium stays longer in the cells and induces pathways that lead to overgrowth of new and enlarged cells. Based on this concept, the delivery of SERCA2a could lead to overexpression of SERCA protein, which could help lung cells restore their proper use of calcium. To assess this, an aerosolized AAV1 vector expressing SERCA2 was administered through the intratracheal route and was shown to reverse PAH symptoms in the MCT rat model of the disease. ⁴⁵¹ This was replicated in Yorkshire swine, a large animal model that more closely resembles PAH in humans. ⁴⁵² The findings of these studies were that both the heart and lung function had improved post treatment and the abnormal cellular changes that cause PAH were reduced.

Clinical trials

To date, there have been no clinical trials of gene therapy for pulmonary arterial hypertension.

Perspective

PAH is a complex disease. Its pathogenesis is driven by a combination of genetic factors, a myriad of environmental factors and triggers such as inflammation, and the fact that it is a multi-organ disease. Over the past two decades, while several experimental gene therapies have been shown to have beneficial effects in preclinical studies, they have not yet led to clinical studies, as no one strategy has been able to address the multifactorial and multi-organ nature of the disease effectively. It is likely that a future therapeutic for this disease will require a combinatorial approach with multiple treatment strategies.

Preclinical studies

Gene therapy mediated delivery of IL-10 has been assessed in animal models of lung transplant rejection (Table 14). Ad5-mediated intratracheal delivery of IL-10 to lungs before removal from the donor has resulted in improved outcomes for subsequent lung transplants in rats and pigs, reducing ischemia-reperfusion injury and improving graft function. ^{453,457 –459} By keeping the donor lungs at body temperature, IL-10 expression was improved, helping to prevent the lungs from deteriorating and improving the success of transplants in experimental Yorkshire pigs. Delivery via a bronchoscopy of an E1/E3 deleted Ad5 encoding the IL-10 cDNA into pig and human donor lungs significantly improved the suitability of donor lungs for transplant. The higher levels of IL-10 persisted in the lungs for 30 days. ⁴⁶⁰ As a next step, the researchers demonstrated that any Ad-mediated inflammation could be reduced if the vector encoding the IL-10 gene was delivered to the lungs during the acellular normothermic ex vivo lung perfusion process in a pig model. Administration would be decoupled from the host immune system, thereby improving IL-10 expression level and duration. This was compared with the in vivo delivery model described above, demonstrating that the ex vivo method was superior. ⁴⁶¹ In another study, bone marrow–derived mesenchymal cells were transduced with a retrovirus encoding the IL-10 cDNA, and the cells were delivered intravenously to an ischemia rat model. The results showed evidence of preventing ischemia-reperfusion injury in the setting of lung transplantation. ⁴⁶²

Clinical trials

To date, there have been no clinical trials of gene therapy for preventing lung transplant rejection.

Perspective

Proper optimization of lungs for organ donation can reduce both early and late post-transplant morbidity and mortality while increasing their usability, and this is becoming increasingly important as waiting lists grow. IL-10 gene therapy holds promise because even its transient expression at time of reperfusion is able to reduce ischemia-reperfusion injury, often the reason for non-usability of a donor lung. While promising, this has not yet led to a clinical trial, and future studies could include improved vector systems for better efficacy or immunologic benefit.

Preclinical studies

Oxidants are involved in the pathogenesis of many lung diseases, resulting from increased production of activated oxygen species that overwhelm the antioxidant defenses in the lung. ⁴⁶⁵ This provides a strong rationale for considering the use of a free-radical scavenging enzyme such as catalase or superoxide dismutase (SOD) to protect against this damage. Using Ad-mediated delivery of catalase or SOD to the lung via intratracheal delivery, these antioxidant therapies were assessed in protection against two oxidant stress scenarios, including 100% hypoxia and ischemia-reperfusion syndrome. ⁴⁶⁵ The results demonstrated that after exposure to 100% O₂, survival was improved in rats that were administered SOD and/or catalase Ad vectors compared with controls, although a similar beneficial impact was not observed in the ischemia-reperfusion model. ⁴⁶⁵ In another study, heme oxygenase-1 (HO-1) was tested for its ability to prevent hyperoxia-mediated injury when delivered by an Ad5 vector intratracheally to the rat lung. Rats treated with HO-1 cDNA exhibited reduced lung injury and increased survival in response to hyperoxia than controls. ⁴⁶⁶ When Ad-mediated gene therapy with HO-1 was tested in a mouse model of ALI (resulting from exposure to aerosolized lipopolysaccharide), the inflammatory impact was attenuated, and there was a concomitant increase in the levels of protective IL-10, an anti-inflammatory cytokine. ⁴⁶⁷ Ad-mediated intranasal delivery of HO-1 also provided therapeutic benefit in an ALI model induced by type A influenza virus. ⁴⁶⁸ Ad5-mediated expression of another antioxidant gene, 1-cys peroxiredoxin (which reduces peroxides) when delivered intranasally, protected the mouse lung against hyperoxic injury. ⁴⁶⁹ In another gene transfer study, transthoracic electroporation-mediated transfer of plasmids encoding for the β1 unit of Na⁺, K⁺-ATPase protected mice from LPS-induced lung injury by improving the clearance of alveolar fluid, which accumulates in patients with ALI/ARDS. ⁴⁷⁰

ALI/ARDS is linked to vascular leak in the alveoli. Based on the knowledge that vascular endothelial growth factor (VEGF) can mediate vascular leak, Kaner et al. ⁴⁷¹ created an in vivo model of VEGF overexpression by Ad5 mediated delivery of VEGF165 cDNA to the respiratory epithelium of the lung by the intratracheal route to C57Bl/6 mice. These mice showed increased expression of VEGF mRNA and VEGF protein in the lung and a dose-dependent increase in lung wet/dry weight ratios over time. At necropsy, lung histology showed widespread intra-alveolar edema, and pulmonary capillary permeability was significantly increased. In order to confirm the specificity of these findings, mice were pretreated with intranasal administration of an Ad vector expressing a truncated soluble form of the VEGF receptor flt-1 (Adsflt), which eliminated the increased lung wet/dry weight ratio, whereas an identical Ad vector with an irrelevant transgene had no effect upon subsequent AdVEGF165-induced pulmonary edema.

Lung damage can also be induced by ionizing radiation, resulting in acute damage and subsequent fibrosis. Overexpression of the cDNA for human manganese SOD delivered intratracheally via either a plasmid/liposome or an Ad5 to the lungs of mice prior to whole lung radiation reduced its deleterious impact. ⁴⁷² In another study, mesenchymal stem cells modified by Ad were utilized as ex vivo gene therapy vehicles to deliver hepatocyte growth factor to enhance repair of radiation-induced lung injury. This therapy reduced expression of pro-inflammatory cytokines, increased expression of the protective anti-inflammatory IL-10, and inhibited lung fibrosis. ⁴⁷³

TGF-β plays a key role in development of lung fibrosis. Based on the knowledge that TGF-β expression can be modulated with Smad7, an antagonist of TGF-β cellular signaling, delivery of an Ad vector expressing the Smad7 cDNA intratracheally to mice with bleomycin-induced lung fibrosis prevented the development of fibrotic changes. ⁴⁷⁴ VEGF is also thought to have a pro-inflammatory role, as it relates to lung injury/fibrosis. In a bleomycin model of pulmonary fibrosis, a plasmid expressing soluble flt-1 (sflt-1) demonstrated therapeutic benefit following administration to the lung. ^471,475

Clinical trials

To date, there have been no clinical trials of gene therapy relating to lung injury.

Perspective

Extensive research on lung injury, including acute lung injury, acute respiratory distress syndrome, and pulmonary fibrosis, has revealed a variety of molecular mechanisms that contribute to the pathogenies of these diseases. Therefore, the potential for the use of gene therapy is great, although to date no one approach has been the “magic bullet,” and it is likely that combinations of these approaches will be needed to lead to a successful treatment. ⁴⁶⁴

RSV infection

RSV is a significant pathogen of infants and young children, infecting nearly all children by 2 years of age worldwide. Although most children resolve the disease within 3 weeks of onset, the remainder develop severe symptoms that require hospitalization, and up to 200,000 children die yearly from RSV infections worldwide. ⁴⁷⁶ The virus also causes complications in elderly adults, with prolonged hospital stays and high mortality rates.

RSV causes a lower respiratory tract infection presenting as acute bronchitis, bronchiolitis, or pneumonia. ⁴⁷⁶ RSV is also associated with the development of recurrent wheeze and asthma in adulthood. ⁴⁷⁷ Current treatment for RSV focuses on alleviating disease symptoms through the use of supplemental oxygen or intravenous fluids. High-risk children may be treated with palivizumab, a humanized monoclonal antibody that targets an RSV envelope protein. ⁴⁷⁷ Currently, no vaccine is available to prevent RSV infection, but an antiviral fusion inhibitor is in clinical trials. ⁴⁷⁸ RSV belongs to the paramyxovirus family and is transmitted by contact with respiratory secretions or aerosolized droplets. The virus initially replicates in the nasopharynx and then spreads to the epithelium of the lower airways. Virus shedding occurs for 7 days in adults and up to 2 weeks in children. RSV infection results in secretion of inflammatory cytokines and recruitment of neutrophils. Lung histopathology from fatal cases has shown airway narrowing caused by submucosal edema, as well as plugs of mucous and cellular debris from infiltrating leukocytes and dead bronchial epithelial cells. ⁴⁷⁹

Preclinical studies

Cotton rats and mice are both semi-permissive for RSV infection, and develop airway obstruction and hyper-responsiveness associated with increased local production of proinflammatory cytokines. RSV-infected mice have bronchiolar mononuclear cell infiltration and interstitial pneumonia similar to humans. ⁴⁷⁹

Several gene modification strategies for controlling RSV infection have been tested in preclinical animal experiments (Table 16). Kumar et al. ⁴⁸⁰ showed that treating mice intranasally with a plasmid expressing IFN-γ resulted in reduced RSV replication and lung inflammation. Prophylactic treatment of RSV-infected mice with nanospheres containing a mixture of plasmid DNA encoding RSV antigens reduced viral titers in the lung and induced an immune response, including neutralizing antibody and cytotoxic T lymphocytes directed toward RSV. ¹⁶⁶ Intramuscular administration of a nonhuman primate Ad vector expressing the RSV fusion glycoprotein induced an immune response consisting of neutralizing antibody and RSV-specific CD8+ and CD4+ T cells that protected cotton rats and mice from infection after challenge. ⁴⁸¹ More recently, several groups have targeted viral proteins using siRNA as a therapeutic antiviral approach. Intranasal application of liposomes and siRNA targeting the viral phosphoprotein, ⁴⁸² non-structural protein 1, ^167,168 or nucleocapsid protein ⁴⁸³ for silencing inhibited RSV infection and prevented RSV-associated lung pathology. Intrapleural administration of an AAVrh.10 vector to express this anti-RSV antibody in mice resulted in long-term expression of antibody and protection from RSV challenge up to 21 weeks after administration. ⁴⁸⁴

Clinical trials

A single RNAi-based therapy has been tested for RSV in humans (Table 17). ALN-RSV01 is a siRNA directed against the RSV nucleocapsid protein that was administered by nasal spray to healthy individuals 3 days before and after RSV challenge. Among treated individuals, there was a 38% reduction in the number of infections. ^203,204

Influenza

Influenza, one of the most common respiratory infections, causes high morbidity and mortality, particularly among infants and the elderly. Seasonal influenza causes ∼200,000 hospitalizations with 36,000 deaths each year. ⁴⁸⁵ Additionally, pandemic influenza viruses emerge periodically, infecting up to 50% of the population. Influenza-caused disease consists of high fever, headache, fatigue, and inflammation of the upper respiratory tract and trachea, as well as more severe complications such as bronchitis or pneumonia in high-risk groups. ⁴⁸⁶ Moderately effective seasonal influenza live attenuated and inactivated vaccines are available as well as antiviral drugs for therapeutic treatment. However, immunity from the vaccine persists only for months, as there is continuous virus evolution, making annual revaccination necessary, and virus resistance to drugs develops rapidly. ⁴⁸⁶

Influenza virus belongs to the orthomyxovirus family and consists of three types (A, B, and C), with influenza A being the most widespread. Viruses are further categorized into subtypes by their hemagglutinin (H) and neuraminidase (N) proteins. The constant accumulation of mutations in these two envelope proteins allows the virus to evade preexisting immunity to previously acquired virus infection. ⁴⁸⁶ Influenza transmission occurs mainly by inhalation of infectious droplets. Virus replication occurs in both the upper and lower respiratory tract, with the nasal mucosa epithelial cells and alveolar epithelial cells being the main targets for infection. In influenza viral pneumonia, both the inflammatory immune response and viral cytopathic effect result in cell apoptosis, leading to edema and respiratory distress. ⁴⁸⁷

Preclinical experiments

Pigs, ferrets, nonhuman primates, and mice have all been used to model influenza infection in animals. ⁴⁸⁶ Several gene therapy–based vaccination strategies have been tested against influenza virus (Table 16). Expression of the influenza nucleoprotein from Ad5 or AAV12 vectors after intranasal vaccination elicited anti-nucleoprotein antibody responses and T-cell responses to nucleoprotein epitopes. ^488,489 Intranasal vaccination with PanAd3, an adenovirus derived from bonobo, expressing both nucleoprotein and matrix 1 protein induced strong antibody and T-cell responses and protected mice from a lethal dose of influenza H1N1. ⁴⁹⁰ Another combination vaccine of an Ad5 vector expressing both H5 hemagglutinin and matrix 2 protein administered intranasally induced a strong antibody response and protected against a heterotypic influenza challenge. ⁴⁹¹ Another strategy used an AAV9 vector expressing the anti-influenza neutralizing antibody FI6. Intranasal vaccination reduced viral load and conferred complete protection to mice and ferrets against challenge from the H5N1 and H1N1 influenza strains ⁴⁹² and partial protection against the H7N9 strain. ⁴⁹³ Young, old, and immunodeficient SCID mice were completely protected from an H1N1 influenza challenge after intranasal vaccination. ⁴⁹⁴ Intramuscular injection of an AAV2/8 vector encoding a broadly neutralizing monoclonal antibody offered protection from infection with diverse strains of H1N1 influenza for 11 months in mice. ⁴⁹⁵ Finally, delivering miRNAs directed against the matrix 1, matrix 2, and nucleoprotein of influenza expressed by an AdC68 vector completely protected against homologous influenza challenge and partially protected against a heterotypic challenge. ²⁰⁵

Clinical trials

Early phase trials of an Ad5 vector encoding avian influenza A hemagglutinin and a Toll-like receptor 3 ligand administered by oral capsule showed a positive safety profile and induction of antigen specific cytotoxic and interferon responses ^496,497 (Table 17).

P. aeruginosa

P. aeruginosa causes a range of severe opportunistic infections in patients with compromised immune systems, including nosocomial infections and ventilator-associated pneumonia. ⁴⁹⁸ P. aeruginosa is the predominant pathogen associated with chronic lung disease in CF patients and is found in 80% of patients by 18 years of age. ⁴⁹⁹ P. aeruginosa is a gram-negative bacterium that primarily infects the lower respiratory tract. Transmission of the bacteria occurs by direct contact with human carriers or environmental reservoirs. Strains that establish early infections are able to penetrate the mucous layer. If not treated during the early stage, P. aeruginosa can adapt to grow in the lung environment, resulting in a chronic infection. In patients with CF, if the infection remains untreated, it can lead to decreased lung function and death at a young age. ⁴⁹⁹ Once the infection proceeds to the chronic stage, the bacteria form a biofilm that protects the pathogen from antibiotics and clearance by phagocytic immune cells. ⁴⁹⁹

Preclinical experiments

Strategies for vaccination to protect from P. aeruginosa infection have been tested in mice (Table 16). For example, transferring dendritic cells pulsed with P. aeruginosa in vitro prolonged the survival of mice after intrapulmonary challenge with P. aeruginosa. ⁵⁰⁰ Insertion of an epitope from the P. aeruginosa outer membrane protein F (OprF) protein into the hexon of Ad5 has also been tested. Intramuscular vaccination with this vector induced a serum humoral immune response in mice and protected against P. aeruginosa challenge. ⁵⁰¹ The addition of a RGD sequence to the fiber of the Ad5-OprF vector yielded similar humoral responses and an increased CD4 and CD8 IFN-γ T-cell response and increased survival after P. aeruginosa intrapulmonary challenge. ⁵⁰² Using the nonhuman primate–derived AdC7 hexon instead of Ad5 as the vector offered more robust immunity toward P. aeruginosa infection. ^503,504 Gene therapy to augment the immune response has also been tested against P. aeruginosa. In a mouse model of chronic P. aeruginosa infection, intratracheal inoculation with AAV5 expressing IL10 was shown to decrease pro-inflammatory cytokines and neutrophil infiltration. ⁵⁰⁵ In another trial, an Ad5 vector expressing IFN-γ was administered to rats prior to intratracheal bacterial challenge. The lung lavage contained elevated levels of IFN-γ from 3 to 28 days after administration, and bacterial clearance was enhanced. ⁵⁰⁶

Clinical trials

There have been no clinical trials for P. aeruginosa vaccines.

Bioterrorism agents

Infectious bioterrorism agents are pathogens that can be easily disseminated and transmitted between people and that cause high mortality. ⁵⁰⁷ Category A agents that manifest disease in the lungs and can be readily transmitted by aerosol include anthrax, plague, and tularemia.

Anthrax

Disease from anthrax dates back to ancient times and was found in both livestock and people. The development of a vaccine for livestock in the 1930s greatly reduced the number of cases derived from environmental exposures. ⁵⁰⁸ The first definite use as a bioterrorism weapon was in 2001, resulting in several confirmed cases of inhalation anthrax. ⁵⁰⁸

Anthrax disease is caused by the gram-positive bacterium Bacillus anthracis. The bacterium can persist as dormant spores in the environment that are resistant to temperature change and can survive for long periods of time. After inhalation, the spores germinate to produce active bacteria that rapidly divide and produce lethal toxin and edema toxin. Clinical manifestations of inhalation anthrax begin with non-specific flu-like symptoms, but rapid replication of the bacteria in the bloodstream eventually leads to sepsis and vascular collapse. ⁵⁰⁸ An anthrax vaccine is available and consists of the bacterial protein known as protective antigen, the component common to both the lethal and edema toxins. Protective antigen is highly immunogenic, and anti-protective antigen antibodies correlate with survival. ⁵⁰⁸

Preclinical experiments

Gene therapy–based vaccination has been tested in mice as a method to induce immunity rapidly against anthrax toxin (Table 16). Both Ad5 and AdC7 vectors have been used to express the B. anthracis protective antigen protein. Intramuscular administration of either vector resulted in rapid anti-protective antigen antibody production and increased protection against B. anthracis lethal toxin challenge. ^509,510 Administration of the AdC7-protective antigen vector yielded 100% survival in mice after lethal toxin challenge, even in the presence of pre-existing Ad immunity. ⁵¹⁰ Several groups have also tried genetic passive immunotherapy approaches. An Ad vector expressing a single-chain anti-protective antigen antibody conferred neutralizing activity for 2 weeks and a survival advantage if administered intravenously up to 14 days prior to lethal toxin challenge. ⁵¹¹ Combination treatment with intravenous Ad5 and intrapleural AAVrh.10 vectors expressing anti-protective antigen monoclonal antibody was able to protect mice from lethal toxin challenge from 1 day to 6 months after administration. ⁵¹² Intravenous injection of an Ad5 vector expressing an anti-protective antigen camelid heavy chain only antibody was able to protect mice against anthrax toxin challenge and spore infection. ⁵¹³

Clinical trials

A Phase I study for an anthrax vaccine based on an Ad4 vector expressing protective antigen is currently active (https://clinicaltrials.gov/ct2/show/NCT01979406?term=anthrax&rank=1).

Plague

Plague has been a human disease for at least 5,000 years and has caused three major worldwide pandemics, leading to the deaths of >50% of the population for each. ⁵¹⁴ Corpses of plague victims were used as an early form of bioweapon in the mid-1300s, and this bacterium continues to be of high interest as a bioterrorism agent today. ⁵¹⁵ Pneumonic plague is the most severe form of the disease, caused by inhalation of the bacterium Yersinia pestis from respiratory droplets. Symptoms include cough with bloody sputum, headache, fever, nausea, and vomiting. Mortality is close to 100% if not treated with antibiotics within 24 h of symptom onset. There is currently no licensed vaccine against plague. ⁵¹⁴

Preclinical experiments

Eliciting protection against aerosolized Y. pestis infection has been tested using an Ad5 vector displaying either the V antigen or F1 capsular antigen on the Ad capsid (Table 16). This intramuscular Ad5-based vaccination stimulated better protection from a Y. pestis lethal respiratory tract challenge in mice than an equivalent administration of recombinant protein and adjuvant. ⁵¹⁶ Intravenous treatment with an Ad5 vector expressing a high affinity anti-V antigen IgG enhanced survival of mice after a lethal intranasal challenge 3 days after immunization. ⁵¹⁷

Clinical trials

There have been no clinical trials for gene therapy for plague vaccines.

Tularemia

Tularemia is an infection with multiple manifestations caused by the bacterium Francisella tularensis. Only 100–200 natural cases are reported yearly in the United States. Transmission can occur through contact with animals, arthropod vectors, or by inhalation of infected aerosolized droplets. The clinical manifestation of inhalation is pneumonia, including high fever, cough, and pleuritic chest pain. Antibiotics, particularly streptomycin, can be used to treat tularemia with a high success rate. A live, attenuated vaccine has gone through development but has not been fully approved by the Food and Drug Administration because of concerns about the effectiveness. ⁵¹⁸

Preclinical experiments

Intramuscular vaccination of an Ad5 vector expressing the Tul4 membrane protein from F. tularensis elicited a robust antibody response against Tul4 and provided 60% protection from lethal challenge in mice after three doses ⁵¹⁹ (Table 16).

Clinical trials

No clinical trials for tularemia gene therapy have been reported.

Perspective

Gene therapy for infectious agents may provide a method to invoke immunity or combat disease for pathogens for which traditional vaccination or treatment strategies have failed to provide an adequate response either because of the complexity of the pathogen or susceptibility of certain groups of individuals.