* Corresponding author : Suresh K. Mittal, Department of Comparative Pathobiology, College of Veterinary Medicine, Purdue University, 725 Harrison St. West Lafayette, IN 47907-2027, USA; ude.eudrup@lattim
The publisher's final edited version of this article is available at Methods Mol BiolVarious adenovirus (AdV) vector systems have proven to be lucrative options for gene delivery. They can serve as potential vaccine candidates for prevention of several common infectious diseases and hold the promise for gene therapy, especially for cancer. Several AdV vector-based therapies are currently at various stages of clinical trials worldwide, which make an immense interest of both the clinicians and researchers. Since these vectors are easy to manipulate, have broad tropism, and have the capability to yield high titers, this delivery system has a wide range of applications for different clinical settings. This chapter emphasizes on some of the current usages of AdV vectors and their production methods.
Keywords: Adenovirus vector, gene delivery system, gene therapy, vector design, vector production, recombinant vaccines
Adenoviruses (AdVs) are non-enveloped double-stranded DNA viruses containing genomes of approximately 34-44 kilobase pairs (kbp). Initially, a human AdV was first isolated in 1953 from the adenoid tissues [1] and hence was named AdV [2]. They are known to cause inapparent or symptomatic infections of the upper or lower respiratory tract, gastrointestinal tract or eyes, which are usually self-limiting in healthy individuals. Although AdVs were known for a long time, their therapeutic potential as a gene delivery vehicle was realized only with the advent of recombinant DNA technology. With continual advancement in the biology of AdV, it became the first viral gene delivery vector to be used in humans. More than 60 types of human AdVs have been described, of which the vector backbone of human AdV type C5 has been used extensively for gene delivery [3].
AdV vectors have several advantages which make them ideal for gene delivery. The AdV biology has been deciphered, which makes the molecular manipulation of its genome easier. Moreover, several AdVs have low or no virulence in humans and have high transduction efficiency for both replicating as well as non-replicating cell types. The vector can also be grown and purified in very high titers and large quantities at a reasonable cost. Furthermore, AdV vectors possess the minimal risk of insertional mutagenesis because of their inability to integrate into the host genome. The transient transgene expression by AdV vectors have been harnessed for oncolytic therapy and also for expression of vaccine antigens [4-9].
However, the transient nature of transgene expression by AdV vectors sometimes limits their use where continuous transgene expression is necessary for a desired therapeutic effect. Apart from this limitation, AdV vectors are known to activate innate immunity, which can lead to severe toxicity at a very high vector dose. One such evidence is the death of a patient enrolled in the ornithine transcarbamylase (OTC) deficiency clinical trial due to high vector dosage leading to multiple organs failure [10].
Due to the high prevalence of human AdVs, nearly 80% of human population is exposed to one or more AdV types multiple times in their lives [11-13], thereby developing AdV neutralizing antibodies popularly known as ‘pre-existing vector immunity’ [14]. The issue of pre-existing vector immunity can be rectified to some extent by increasing the vector dosage without increasing toxicity [13,15,16]. Alternatively, pre-existing vector immunity can also be circumvented using nonhuman AdV vectors and heterologous prime-boost approaches. Innovation in vector engineering strategies and the use of different immunosuppressive agents can also be used to overcome some of these limitations in the existing vector systems [17-19].
Nonhuman AdV vector systems based on bovine AdV, simian AdV, porcine AdV, ovine AdV, canine AdV, avian AdV and murine AdV [20,21] were developed in search of safe and efficient gene delivery vehicles to overcome the shortcomings of human AdV vector systems, especially the concern of pre-existing vector immunity. For example, bovine AdV vectors are not neutralized by human AdV-specific neutralizing antibodies, and the prevalence of bovine AdV cross-neutralizing antibodies was not detected in human serum samples [22,23]. Moreover, various nonhuman AdV vectors use different receptors for internalization thereby broadening the range of cell types that can be targeted [24,25].
Initially, when the therapeutic potential of AdV vectors was realized, they were evaluated for a broad range of medical conditions including genetic diseases and metabolic disorders. However, soon it was realized that transgene expression is usually for a short duration [26,27]. This limits the use of AdV vectors to conditions where transient transgene expression is required for the desired effects, such as recombinant vaccines and cancer therapeutics.
With the advancement in the field of viral vectored vaccines, various AdV vectors have been tested both in pre-clinical as well as clinical studies [28,6,29] for different infectious diseases. This is due to the fact that AdV vector-based vaccines induce a balanced humoral and cell-mediated immune (CMI) responses [30,31] by stimulating innate immunity through pathways that are both Toll-like receptor (TLR)-dependent and TLR-independent [32,33]. AdV vectors expressing different antigens of influenza virus have been tested in different animal models and have shown high protection efficiency against homologous and heterologous influenza viruses [34-39]. A vaccine construct expressing hemagglutinin (HA) of an H5N1 influenza virus provided cross-protection in mice following challenge with different strains of highly pathogenic H5N1 influenza viruses [40]. Similarly, in a clinical trial, immunization with an AdV vector encoding the HA gene of influenza virus increased hemagglutination inhibition (HI) titers in more than 75% of the participants [41].
With time, several modifications were incorporated in the AdV vector system to overcome the existing limitations of pre-existing vector immunity and inadequate antigen-specific immunogenicity. For this purpose, other AdV types from both human and nonhuman AdVs have been evaluated. An AdV35 vector-based HIV vaccine was assessed in a clinical trial and was found very effective and safe [42]. AdV26, another less common type of human AdV, has been recently evaluated for Ebola vaccine in a clinical trial and it elicited a favorable antibody response [43]. In addition to less prevalent human AdVs, several nonhuman AdVs, in particular chimpanzee AdV (ChAdV) vectors have shown very encouraging results in clinical trials for malaria [44], leishmanial [45] and Ebola [46]. Recently, ChAdV vector, ChAd3-EBOZ, encoding for the Ebola G glycoprotein gene of the Zaire strain showed robust antibody and T cell responses in Phase I and II clinical trials [47]. Another approach which has been adopted to improve transgene immunogenicity is the use of single cycle AdV vectors having the deletion of pIIIa protein-coding gene. A single cycle AdV vector encoding influenza HA was assessed for immunogenicity in both cotton rats and hamsters leading to enhanced immune responses at a low dosage [48].
The oncolytic nature of AdV has been utilized to combat various forms of cancer. To achieve effective oncolysis, the virus should infect and replicate within the cancer cells. Most of human AdVs require Coxsackievirus and Adenovirus Receptor (CAR) for virus internalization, but in many forms of cancer, there is a marked downregulation or complete absence of CAR [49], the reason for a marked reduction in cell transduction with many AdV vectors. To increase the interaction between the virus and cancer cell surface molecules, an introduction of a motif, like RGD, in the knob region of AdV fiber improves the interaction with integrins which are expressed on the cancer cell surface [50]. In some cases, a complete swapping of fiber is done for its preferred interaction with a cell surface molecule such as desmoglein 2, which is expressed in large number on cancer cells [51-54]. It seems very assuring that only tumor cells can be lysed by these oncolytic AdV vectors since their replication competency is dependent on the presence of a specific tumor antigen. A successful oncolytic AdV therapy also requires some other vector modifications to overcome the immunological as well as structural barriers of the tumor microenvironment. Oncolytic AdV vector expressing relaxin facilitates better vector spread in the dense extracellular matrix (ECM) [55]. VCN-01 is another armed oncolytic AdV vector which expresses hyaluronidase and is currently being tested in Phase 1 clinical trials [56]. Recently, oncolytic AdV vector ONCOS 102 expressing GM-CSF demonstrated a potent therapeutic effect with minimal side effects [57]. Many other molecules like interferon alpha, tumor necrosis factor alpha, and other interleukins are also being investigated as delivery molecules with oncolytic AdV vectors. It seems that targeting the tumor microenvironment, in addition to the tumor cell lysis, is a better approach for cancer therapeutics using oncolytic AdV vectors.
Several changes have been made in the AdV vector design methodology to improve vector recovery, transgene expression, and safety. AdV vectors can be broadly classified into three types based on the deletions of the viral genes.
First generation AdV vectors contain the deletion of the early (E) region 1 (E1) or E1 & E3 regions of the viral genome. Deletion of the E1 region results in a replication-incompetent vector, and it also serves the purpose of increasing the capacity of the foreign gene cassette for insertion [58]. E1-deleted AdV vectors can only be grown in a cell line (e.g., HEK 293) that constitutively expresses E1 proteins [59]. However, anchorage-dependent cell lines can be used only for the small-scale production of vector preparations. To achieve scalability and batch-to-batch consistency, a suspension cell culture bioreactor system is used with a variant of HEK 293 cell line capable of growing cultures in suspension without serum. A bioreactor with 10,000 L capacity is projected to yield 10 9 −10 10 viral particle/milliliter (VP/mL) [60]. However, the usage of HEK 293 cells can result in the production of contaminating replication-competent AdV due to homologous recombination. The PER.C6 [61] and SL0036 cell lines [62] have been developed with a minimal E1 region to eliminate the possibility of homologous recombination. The major drawback of the first generation AdV vector system is high immunogenicity in the host, which raises safety concern in situations where a very high vector dose is required for desired effects.
Second generation AdV vectors were created to minimize the shortcomings of first-generation vectors. Second generation AdV vectors were designed with deletion of two more gene regions, E2 and/or E4, along with E1 and E3 deletions. The idea was to reduce the vector immunogenicity by minimizing the leaky expression of viral genes [63]. Apart from this, the deletion/s also increase/s the transgene carrying capacity of the vector. However, these multiple regions deleted vectors require an appropriate complimentary cell line for their propagation.
Third generation AdV vectors include the helper-dependent vectors, also known as gutless vectors, which are designed by removing all the AdV genes. These vectors retain only the AdV packaging signal along with the inverted terminal repeat (ITR) sequences of the viral genome [64-67]. Due to the complete absence of protein-coding regions in the helper-dependent vectors, a significant reduction in vector immunogenicity with improved safety occurs when these vectors are used in patients. Moreover, the transgene carrying capacity of helper-dependent vector system can be up to 36 kbp. Production of these vectors requires a helper virus – a first generation empty AdV vector. Initially, the low yield of helper-dependent vector and the contamination with the helper virus were two major concerns with this vector system [64]. Both of these concerns were addressed by replacing the helper virus with the AdV vector in which the packaging sequences are flanked with a site-specific recombination sequence, e.g., loxP. The helper-dependent vector is grown with this novel helper vector in a cell line that expresses an appropriate recombinase, e.g., Cre recombinase for loxP sites [68,67]. The loxP-Cre recombinase or an equivalent system will results in the generation of novel helper virus genomes without the packaging sequences thereby allowing efficient packaging of the helper-dependent vector genomes. Third generation AdV vectors have shown promising results in different animal models with minimal adverse effects [69].
Several techniques were developed to construct AdV vectors which can be broadly divided into two approaches: 1) direct insertion of the foreign gene into the viral genomic DNA, in a plasmid form, using unique restriction enzymes [70,30]; and 2) recombination between two plasmids through homologous recombination either in bacteria or in a permissive cell line [71-75]. The two plasmids system includes a genomic plasmid that contains nearly the complete AdV genome with appropriate deletion/s, and a shuttle plasmid carrying the foreign gene cassette and AdV sequences that are essential for homologous recombination and generation of an infectious AdV vector. There are two commonly used recombination techniques for generating AdV vectors: 1) Homologous recombination in bacteria, and 2) Cre/lox recombination in mammalian cells. The detailed protocols to create and purify AdV vectors using these two recombination approaches are described below.