In several studies with retrovirus based vectors, a significant group of patients acquired severe diseases as a side effect; often this was leukaemia (Strauss and Costanzi-Strauss, 2007; Haviernik and Bunting, 2004). These side effects are normally caused by non-directed integration into the genome at critical locations in respect to cancer development. In addition, studies using other viral systems such as adenovirus based vectors often resulted in the death of the patient, as seen in the case of Jesse Gelsinger who died from an immune response after being injected with an adenoviral gene therapy vector (Teichler Zallen, 2000). In comparison to viralvectors in gene therapy, the use of non-viralvectors does not seem to involve such a high risk factor, however, they have also been much less efficient till date. Several non-viral transfection procedures use endocytosis to take up the vector into the cell (See chapter 1.3). A known bottle neck for endocytosis mediated non-viralgene therapy is the efficiency of uptake of the vector into a living cell without degradation during transfection procedure. Further problems are the liberation of the endocytosed DNA from endocytic/lysosomal compartments, the transport of the vector to the nucleus, the import into the nucleus and the integration into the genome, preferable at a defined location. There are two possible approaches to optimize a process such as transfection, a non-systematic screening or a systematic approach based on logic advisements.
Furthermore, many non-viralvectors – particularly polyplexes – exhibit a net positive surface charge due to the excess of polycation required for high gene transfer activity (35). During their circulation in the patient’s blood, these positively charged particles can interact with negatively charged physiological compounds which results in significant toxicity and/or poor efficiency, due to binding to plasma proteins or blood cells, aggregation and complement activation (36, 37). Surface neutralization by PEGylation, i.e. the modification with the hydrophilic polymer poly(ethylene glycol) (PEG) prevents polyplexes from these undesired interactions as well as from rapid elimination by the reticuloendothelial system (RES) and thus decreases toxicity and extends circulation time (38). However, covering the polyplex surface with neutral polymers not only reduces undesired interactions, but also the desired binding to the negatively charged cell membrane of the target cells (39). Following the viral concept, the resulting reduced association and polyplex uptake into the cell can be overcome by the attachment of a targeting ligand to the polyplex for receptor- mediated uptake. PEI polyplexes containing PEG and cell-targeting ligands have been successfully applied in vitro and in vivo (38-44). While the uptake problem of PEG shielded particles can be solved satisfactorily, there still remain difficulties concerning the intracellular fate of PEGylated polyplexes. Extensive PEGylation, though being beneficial for systemic circulation, appears to negatively affect endosomal release of the delivered nucleic acid and leads to a remarkable reduction of transfection efficiency.
One goal of in vivo gene delivery is efficient and specific transfection of distant tissues, such as tumors or disseminated metastasis. Viralvectors, although producing high levels of transduction, have many disadvantages in this direction: Viralvectors are difficult to prepare, lack the ability to be targeted to distinct tissues or cell species, and have draw- backs regarding cytotoxicity and immunogenicity (1, 2). Therefore, non-viral vector are regarded as a promising alternative (3, 4). Polyethylenimine, one of the most cited poly- cations for non-viralgene delivery, was not only successfully utilized as transfection reagent in vitro but also proofed to be efficient after intravenous injection in vivo (5-7). However, it was repeatedly documented that intravenous injection of unmodified PEI polyplexes results in accumulation of the vector mainly in the lung and liver (8-10). A possible solution for this problem is to modify PEI with hydrophilic polymers such as polyethylenglycol (PEG) or poly(N-(2-hydroxypropyl)methacrylamide) (pHPMA) (11, 12). It was shown for liposomes that PEGylation strongly increased blood circulation times due to decreased unspecific interactions (13). However, grafting of PEI with PEG did not in every case result in increased blood levels of polyplexes (8, 14). It was hy- pothesized by de Wolf et al. that polyplexes prepared with PEG-PEI copolymers are still prone to efficient opsonization and therefore prolongation of blood residence time is relatively small compared to results found with PEG-liposomes (10). In addition to suf- ficiently long circulation times in blood, polyplexes need to specifically bind to the de- sired target tissue. Antibody conjugation was shown to be exceptionally suitable to tar- get PEI polyplexes to a certain tissue (5). Furthermore, it was recently shown by us that antibody modification not only potentially increases specificity of vectors but also re- sulted in increased area under the curve of intravenously injected polyplexes (see chap- ter 3). Therefore, we sought to study targeting of a subcutaneously growing, HER2 overexpressing tumor with conjugates characterized in chapter 3.
Gene therapy is a young, emerging field of molecular medicine. It is based on the administration of therapeutic DNA sequences into the target cell to repair genetic defects. In 1990 the first gene therapy treatment utilizing a viral vector was performed in a SCID (severe combined immunodeficiency) patient with adenosine deaminase deficiency. This novel approach caused worldwide attention and a boost for gene therapy. Until 1999, the database “Gene Therapy Clinical Trials World Wide” recorded more than 400 clinical trials (Wiley 2010). However, the death of Jesse Gelsinger in 1999 after being given a high dose of adenoviral particles meant a drawback for the popularity of gene therapy (Raper et al., 2003). Furthermore an additional phase 2 study treating 27 SCID-X1 patients with a retrovirus based vector could indeed achieve a significant improvement of the disease, but 5 patients developed leukemias after several years, which was due to insertional mutagenesis after provirus integration. Both studies emphasize the importance of a detailed characterization and continuous improvement of gene therapy vehicles to prevent negative side effects. Since 1990 until to date more than 1,500 clinical trials with various viral and non-viralvectors have been performed with adenovirus, retrovirus and plasmid DNA being the predominantly used vector types (Figure 1.1) and especially over the past few years, very promising results were obtained. The treatment of patients suffering from liver congenital amaurosis with an adeno-associated virus could partially restore the ability to see in nearly blind patients without inducing toxicity (Bainbridge et al., 2008; Cideciyan et al., 2009; Wright et al., 2010). Furthermore a phenotypic correction of up to 30 month in two X- linked adrenoleukodystrophie patients could be demonstrated after ex-vivo administration of a lentiviral vector encoding for the defective ABCD1 gene (Cartier et al., 2009). Last but not least in a very recent publication the phenotypic correction of a ß-thallasemia patient for 22 month after treatment with a lentiviral vector was reported (Cavazzana-Calvo et al., 2010).
Because of their well-defined particle size and shape, dendrimers are of particular interest for gene transfer, but also for drug delivery and imaging applications (12, 94). Starburst polyamidoamine (PAMAM) is the most commonly used dendrimer in gene delivery and is commercially available as generation 4 “fractured” PAMAM (88) from Qiagen (SuperFect®) or in a variety of generations as cationic or half anionic dendrimers (96) from Dendritech. The heat-degraded structures are believed to feature higher chain flexibility which is drawn on for the increased transfection efficiency at low generation (88). PAMAM is also believed to interact with DNA like non-acetylated histones (89) in a way that initiation of transcription in the complexed DNA is inhibited. Due to their high buffer capacity, PAMAM dendrimers are thought to owe their transfection efficiency to the “proton sponge” effect first described for PEI (46). In an attempt to use poly (L-lysine) (PLL) modified PAMAM as transfection reagent, it was shown that PLL selectively condensed DNA, whereas PAMAM maintained endosomal release (90). While high transfection efficiencies of non-modified PAMAM are achieved only at relatively high generations of 5-10 (91), synthetic efforts were made to reduce the toxicity of PAMAM. This was successfully achieved by quarternization of the internal amines and by neutralizing the surface with a hydroxyl periphery, which on the other hand also reduced transfection efficiency (92). A year later, the same group reported another approach where they maintained high transfection efficiency in spite of low toxicity by synthesis of a PAMAM-PEG-PAMAM triblock copolymer (93). As already described for PEI, several modifications of PAMAM were synthetized to improve in vitro and in vivo parameters. Besides targeting strategies like the attachment of mannose (94) and galactose (95), further modifications include more amphiphilic vectors, such as alkyl chain- or phenylalanine- modified PAMAMs (96), other amino acid-modified PAMAMs (101), and cyclodextrin attachment to the surface (97).
The comparison between two different vector types for their gene transfer activities is not at all trivial, since lentiviral and AAV vector constructs differ in many aspects such as their structural organization, mode of gene expression and entry receptors used. Additionally, LVs integrate the transgene into the host cell’s genome allowing stable transgene expression, while AAV vectors are present in the cell episomally. Therefore, the normalization of lentiviral and AAV vector particles for a convincing side by side comparison is critical. Normalization can be based on physical, genomic or functional titers. Genomic titers include vector particles containing vector genomes, but comprise also non-functional particles that contain genetic information, but are not able to transduce target cells or lead to transgene expression. Quantification of genomic titers is performed by quantitative real-time PCR. While the DNA from AAV vectors can be isolated and used for qPCR directly, LV RNA has to be reverse transcribed into cDNA for qPCR first. Physical titers can be determined by ELISA and include empty and non-functional particles. Functional titers are determined by transduction of cells expressing the entry receptor of the respective vector. Thus, comparing different vectors that do not use the identical entry receptor bias the quantification of functional titers. While VSVG-pseudotyped LVs can transduce a variety of cells, since the VSV receptor is expressed on most of the mammalian cells, the transduction ability of recombinant AAV vectors depends on the natural tropism of the AAV serotype used for capsid formation. Therefore, comparison of different AAV vector serotypes is already challenging. Comparing different vector types, such as lentiviral and AAV vectors, is even more difficult. Nevertheless, several studies previously compared the performance of lentiviral and AAV vectors in various gene transfer approaches in terms of efficacy, safety and biodistribution.
Gene therapy is a controversial topic due to the variety of problems that come along with its application, However, it might be the future hope and cure of many diseases, ranging from cancer, to autoimmune diseases and genetic disorders. One challenge in the field is the development of safe and efficient administration vehicles for transgenes, so-called genevectors. But after years of intense research scientists are still dealing with basic questions of this therapeutic approach like e.g. safety, administration and long- term stable and consistent transgene expression. Most currently used genevectors have a viral background, mainly for the reasons of efficient administration into the target cell and stable transgene expression. Adenovirus-based vectors and vectors based on adeno- associated virus (AAV) that have a depleted integrative potential, make up roughly one quarter of applied vectors for gene therapeutic approaches 1 and offer the advantage of a non-integrating character. However, because adenovirus is a common human virus the application of adenovirus-based vectors can easily lead to immunological reactions of the patient. AAV-based vectors display low immunogenicity but this vector type has only a low transgene capacity up to a maximum of 4.7 kbp (Daya and Berns, 2008). Other vectors, like e.g. retro- and lentivirus-based vectors bring a big disadvantage along - integration into the host’s genome. The main problem of integration is
siRNA delivery to the lung represents a promising non-invasive approach to treating lung cancer or lung diseases like acute lung injury or disorders like influenza [1-3]. Local delivery of siRNAs relevant to lung diseases via the airways is advantageous for gene therapy since the target organ is directly accessible. The large respiratory surface area provides improved transfection efficiency, with reduced systemic side-effects. Successful siRNA delivery using non-viral vector systems to the target cells or tissue is mainly dependent on well balanced electrostatic interactions between a positively charged polymer or liposome and a negatively charged phosphate backbone of the nucleic acid [4-6]. A wide range of polymers with different architectures and functionalities has been engineered to further optimize and to increase targeted delivery, biocompatibility and prolonged efficiency . Non-viral vector systems provide an attractive alternative to recombinant viralvectors due to reduced pathogenicity and immunostimulation [8, 9]. However, cationic transfection reagents used for siRNA delivery often exhibited severe cytotoxicity, which precludes their clinical applications. Pulmonary delivery of plasmid DNA (pDNA) using poly(ethylene imine) (PEI) - based nanocarriers was frequently described in the literature [10-12], but only scant information is available for PEI mediated delivery of siRNA to the lungs. For instance, pulmonary siRNA application was investigated to differentiate off-target effects (caused by the polymers or siRNA sequence) from specific knockdown effects, and to explore the feasibility of cell specific RNA-targeting [8, 13].
Sleeping Beauty and CAR T cells: clinical proof-of-concept obtained
The potential to use SB transposition to integrate the genetic information of the CAR into T cells has first been explored by the group of Cooper et al. (Singh et al., 2008 ). It was demonstrated that SB transposase can be provided either as plasmid DNA or mRNA in combination with a plasmid-encoded CAR transposon and introduced into T cells by electroporation to yield functional CD19-CAR T cells. Consistent with observa- tions in other mammalian cell types, the use of SB11 and hyperactive SB100X accomplished higher rates of gene transfer than first-generation SB transposase (Jin et al., 2011 ; Singh et al., 2013 ). The same group has also provided the successful clinical debut of SB-engineered CD19-CAR T cells, and recently reported results of two pilot clinical trials in 26 patients with ALL and NHL who had undergone autologous (n ¼ 7, ClinicalTrials.gov Identifier 00968760) or allogeneic (n ¼ 19, ClinicalTrials.gov Identifier NCT01497184) HSC transplantation (HSCT) prior to CAR T-cell therapy (Kebriaei et al., 2016 ). From these clinical studies, it was concluded that the administration of SB-engineered CD19-CAR T cells is safe, and may provide additional tumor control in patients after HSCT. The SB transpos- ition strategy pursued in this trial was relatively “basic”, and comprised the nucleofection of plasmid-encoded SB11 transposase and a plasmid-encoded, pT-based CAR transposon, followed by ex vivo propagation of CAR-modified T cells for 28 days (four stimulation cycles with artificial antigen presenting cells). These studies are the first CAR T cell clinical trials that rely on non-viral SB-based gene transfer, and provide proof-of- concept for the utility of SB transposition in CAR T-cell engineering. However, because CD19-CAR T cells were administered as an adjuvant therapy after HSCT, the presented data are somewhat less spectacular than the dramatic and durable anti-tumor responses with some- times year-long persistence of CD19-CAR T cells that have been reported in other clinical trials that used viralgene transfer vectors and administered CD19-CAR T cells outside the HSCT setting (Maude et al., 2014 ; Turtle et al., 2016a , 2016b ). We would like to point out, how- ever, that comparisons between clinical trials of CAR
Delivering genes to specific cells is a process including many obstacles. Nucleic acids are negatively charged, hydrophilic and large macromolecules that cannot penetrate through lipid membranes. Additionally they are quickly eliminated in the organism by nucleases, mainly before they reach their target cell. These properties are not very suitable for their use as therapeutics. In vivo experiments already showed that the intravenous application of naked DNA does not lead to high transfection levels. If naked DNA is injected locally it has to be in large amounts to be efficient. Therefore genevectors that are useful to protect DNA and that are able to deliver it successfully to its target cell are examined intensely. There are many different gene delivery approaches existing in the viral as well in the non-viral field. For example cationic lipids (2) and cationic polymers (PLL, PEI) that both complex DNA and result in so called lipo- or polyplexes or liposomes that envelope DNA solutions with a lipid bilayer system or combinations of both. Further on different viralvectors, like adeno-, retro- or lentiviral vectors are employed. Following, two of the main important systems are listed that are additionally directly related to this work.
Recombinant adenoviral vectors (rAdV) are a promising gene therapeutic tool for efficient transduction and long-term transgene expression with limited side effects. We and others (Kreppel et al. 2002; Ehrhardt and Kay 2002; Brunetti-Pierri et al. 2007; Jozkowicz and Dulak 2005) showed that after injection of a single dose of gene-deleted adenoviral vector (GD AdV) particles the viral genome is maintained. Even after induction of cell cycling of the host cell in murine liver the GD AdV DNA molecule is more stable than non-viral DNA (Ehrhardt et al. 2003). Therefore, one part of this study was to reveal the mechanism of persistence of rAdVs. Several mechanisms that potentially could lead to this persistence have been discussed (Jager and Ehrhardt 2007; Kreppel et al. 2002; Ehrhardt et al. 2003). The rAdV genome may (i) replicate episomally, (ii) form concatemers that were shown to be predominantly responsible for persistent transgene expression in the context of rAAV. Furthermore, (iii) circularisation of rAdV DNA might improve vector DNA maintenance, which was shown for example for members of the Herpesviridae. Herpes simplex virus 1 persists in circular conformation in non-dividing host cells without replication (Roizman and Sears 1996). The Epstein-Barr-Virus genome on the other hand, is maintained in a circular conformation in dividing lymphocytes with simultaneous replication (Kieff 1996).
Several adenoviral serotypes are known, with serotypes 2 and 5 being the types most extensively used for vector construction because their molecular composition is well characterised (Van Ormondt, et al., 1984; Chroboczek et al., 1992). The serotype 5 vector system is based on bacterial plasmids containing the adenovirus genome with deletions of the E1 and E3 genes. Deletion of E1 renders the virus replication defective. In addition, all or part of the E3 region, which is not essential for virus function, is deleted in order to accommodate the genes of interest, and the plasmid vector can then be grown in bacterial culture. The purified plasmid DNA subsequently is transfected into the 293 line of human embryonic kidney cells. This cell line was derived following transformation of 293 cells (Graham et al., 1977) and can thus transcomplement the E1-deficient viral genome. The virus can be isolated from 293 cell media and purified using the limiting dilution plaque assay (Graham and Prevek, 1991). The problems associated with the use of recombinant adenoviruses in gene therapy are mainly due to the hosts cellular and humoral immune response. A second generation of recombinant adenovirus vectors has been generated to overcome this problem. These vectors additionally use an E2a temperature sensitive mutant (which at non-permissive temperatures fails to express late gene products even when E1 in expressed in trans) (Engelhardt et al., 1994) or an E4 deletion (Armentano et al., 1997). The most recent "gutless" vectors contain only the inverted repeats (ITRs) and packaging sequence around the transgene with all the necessary viral genes being provided in trans by a helper virus (Chen et al., 1997).
Thyroid cancer, even in advanced metastatic disease, can be effectively treated by radioiodine therapy, due to thyroidal expression of NIS (Dai et al., 1996; Smanik et al., 1996; Spitzweg and Morris, 2002a). NIS expressing thyroid cancer metastases can be detected and treated by administration of radioiodine, while avoiding adverse effects of ionising radiation to other organs, which do not express NIS and thus do not concentrate radioiodine. NIS therefore represents one of the oldest and most successful targets for molecular imaging and targeted radionuclide therapy. Cloning and characterization of the NIS gene has therefore allowed the development of the NIS gene therapy concept based on NIS gene transfer into nonthyroidal tumor cells, followed by diagnostic and therapeutic application of radioiodine (Dai et al., 1996; Smanik et al., 1996; Hingorani et al., 2010a). One of the major challenges on the way to efficient application of the NIS gene therapy concept in the clinical setting of metastatic cancer is optimal tumor targeting in the presence of low toxicity and sufficiently high transduction efficiency after systemic administration of gene delivery vectors. Only a limited number of studies have investigated systemic NIS gene delivery approaches with the aim of NIS-targeted radionuclide therapy of metastatic disease using an oncolytic measles virus or vesicular stomatitis virus encoding human NIS in multiple myeloma mouse models (Dingli et al., 2004; Goel et al., 2007; Liu et al., 2010). In a recent study we have utilized a promising non-viralgene delivery system for tumor- targeted NIS gene transfer in the syngeneic Neuro2A neuroblastoma mouse model. Branched polycations based on OEI-grafted polypropylenimine dendrimers (G2-HD-OEI) have recently been characterized as biodegradable synthetic gene delivery vectors with high
Up to date about 70% of the gene therapy trials have used viralvectors derived from naturally infecting viruses. The adenovirus-based vectors are the most commonly used vectors (25% of all the gene therapy trials). They are efficient vehicles with high transduction efficiency and high level of gene expression. However, their genetic material does not integrate into cell genome and is not replicated at cell division. Therefore, usage of these viruses is not appropriate for therapies which require transmission of the transgene also into the derivative cells (Evans et al., 2001). Additionally, the adenoviral vectors are known to elicit strong immunogenic reactions, an undesired side effect which has promoted further improvement of the system. The most common strategy is based on deletion of all coding non-essential viral regions apart from the packaging signal and the inverted terminal repeats which are needed for genome packaging and replication. To generate viral particles of these so-called helper-dependent or gutless adenoviruses, all viral genes and the proteins needed for its genome replication, packaging and capside formation are supplied in trans by coinfecting the virus producing cells with helper virus. Although this system has reduced the immunogenicity and toxicity of these vectors and has increased the packaging capacity of transgenic DNA up to 36 kb, technical limitations make it extremely difficult to obtain vector stocks devoid of replication-competent particles. The replication-competent particles account for 0.1–1% of all of the viral particles present. This drawback may hinder their general use in clinical applications especially when high doses need to be administrated. These stocks may therefore be strongly immunogenic (Dai et al., 1995; Schiedner et al., 1998; Raper et al., 2003). Nowadays, new systems are being developed to reduce or even avoid the packaging of the helper vector by flanking the packaging region of the helper virus with loxP sites or attP/attB sites, which are then induced to recombine in coinfected recombinase-expressing cells. Thus, the packaging region of the helper virus is eliminated avoiding packaging of the helper adenovirus (Ng et al., 1999; Alba et al., 2007).
In order to find out whether the observed variable gene expression was due to differentially regulated gene expression or due to a more efficient delivery of viral genomes, adenovirus genomes were quantified by real time PCR in transduced cells 48h post transduction (Fig.19). Care was taken to eliminate residual surface-bound viral particles before DNA extraction by applying several washing steps. Adenovirus DNA was isolated as previously described and equal amounts of DNA used in each PCR reaction. The determined genomic EGFP copy number (equivalent to virus genome number) correlated well with transgene expression in cells for both vectors (fig.19 and table 5). It may be concluded, that differences in transgene expression are attributed to differences occurring during virus uptake and intracellular transport. Two additional parameters, addressing the transduction efficiency (% of initially applied viral genomes detected at 48h), and EGFP expression activity (particles required per relative fluorescence unit [RFU]) were considered. Most of the values were in accordance with the expected results, though derivations occurred especially for the highest Ad19aEGFP vector dose used. Those derivations are highlighted in red in table 5. First parameter analyzed: In theory, the percentage of vector genomes detected 48h post transduction reflects the transduction efficiency and should be constant for all titers used, assuming vector binding and uptake under non-saturating conditions. In agreement with low EGFP expression, low amounts of vector genomes were found in primary mouse myoblasts and L6 cells. The amount of detected vector genomes increased proportionally to EGFP expression in pig, ape, and human myoblasts. However, it should be emphasized that very high amounts of initial Ad19aEGFP vector genomes ranging from 3.94% to 6.59% were detected in human myoblasts, whereas only 0.096% to 0.132% of the initial vector genomes could be detected in FHMs transduced with Ad5EGFP. This result reflects the extremely efficient uptake of Ad19aEGFP virus by primary human myoblasts. Two prominent derivations were observed in ape and pig myoblasts transduced with 12500 vg/cell Ad19aEGFP. Here, a disproportionate increase of cell-associated vector genomes was observed, which did not correlate with transgene expression. This disproportionate increase in vector genomes may be attributed to either inefficient downstream events occurring after vector endocytosis (e.g. mislocalization of the vector), or to species-dependent differential promoter regulation.
Gene transfer activity of viralvectors also strongly depends on the type of virus used for transduction. Murine retroviral vectors, for example, are very useful for ex vivo gene transfer because of their high efficiency to integrate into the host cell genome. In particular, they have been optimized for gene transfer into hematopoietic stem cells. The gene therapy of human SCID-X1 can be regarded as the first success story in gene therapy (3). All ten children treated in this trial were cured of the immunodeficiency symptoms. The therapy, however, was associated with the incidence of leukemia in two of the treated children, and subsequent chemotherapy was required. Insertional oncogenesis by the retroviral vector was identified as the reason for this severe side effect, highlighting a serious risk of such stably but randomly integrating viral vector systems that may cause fatal long term changes in gene expression in addition to the desired therapeutic effect.
Many diseases including immunodeficiency disorders, haemophilia or cystic fibrosis are caused by abnormalities in the genome. Patients suffering from one of these genetic conditions would greatly benefit from correction or transfer of the respective genes. Since the relationship between gene defects and diseases has been identified several decades ago, the idea of gene therapy is not very new. Already in 1972, Friedman & Roblin proposed the potential therapeutic use as well as scientific and ethical problems arising from gene therapy (Friedmann & Roblin, 1972). Nowadays, a lot of improvement in gene transfer techniques has been made and many promising gene therapy trials are ongoing. Among the gene delivery vehicles used, vectors derived from retroviruses are probably the most suitable tool for efficient gene transfer and stable integration of the transgene. Especially the development of lentiviral vectors (LV) has gained a lot of attention in the expanding and competitive field of vectorology (Naldini et al, 1996; Schambach et al, 2013). Compared to other retrovirus- derived vectors, as for example γ-retroviral vectors, LVs stand out due to their capacity to transduce non-dividing cells and decreased genotoxicity. The potential of lentiviral vectors is also reflected by the fact that they are used in 4.2% of all vector gene therapy clinical trials (http://www.wiley.co.uk/genmed/clinical). Recently, two studies reported promising results from ex vivo clinical trials treating patients with rare hematopoietic diseases using improved lentiviral vectors. Remarkably, patients suffering from metachromatic leukodystrophy or Wiskott-Aldrich syndrome benefitted from treatment with gene-corrected hematopoietic stem cells without any signs of oncogene activation which was a major drawback of previous gene therapy trials using γ-retroviral vectors (Biffi et al, 2013; Aiuti et al, 2013).
We have observed faster onset of viralgene expression in coculture infection containing cell- to-cell spread of HIV relative to cell-free HIV infection. The earlier onset of viralgene expres- sion in coculture was lost when target cells were separated from donor cells by a transwell membrane. A faster virus cycle in cell-to-cell spread relative to the non-directed, cell-free mode of infection has been previously observed directly [ 2 , 13 , 41 ] and inferred through modelling of infection dynamics [ 39 , 40 ]. Here we used time-lapse microscopy of HIV infection to directly quantify and investigate the mechanism behind the faster onset of viralgene expression. We minimized possible differences between cell-to-cell spread and cell-free infection in the extra- cellular transit time from donor to target cell by limiting the time window of transmission to 2 hours. We have also minimized any contribution of virus sequence to different viralgene expression dynamics by using viruses with identical sequences derived from a molecular clone. Hence, variability in gene expression is a result of the interaction of the virus with the host cell. After exclusion of donor-target cell fusions, we found a minimum time for early viral protein expression in both infection modes, corresponding to a period of intracellular delay indicative of true infection [ 53 – 55 ]. We found that we could recapitulate the faster onset of viralgene expression by increasing the MOI of cell-free virus, and that there was no evidence for coopera- tivity or interference between co-infecting viruses. There was also no evidence for trans-accel- eration of HIV gene expression onset from the surrounding infected cells.
27 genes encoding cellular proteins not previously impli- cated as HIV-1 restriction factors were amplified from primary human blood cells and cloned into a bi- cistronic vector coexpressing the gene candidate BFP. Over-expression of a striking proportion of candidate genes (16 of 27; 59%) reduced the production of infec- tious wt HIV-1 particles by more than 75% without causing cytotoxic effects and this number was even higher in the absence of intact vpr, vpu and nef genes (20 of 27; 74%). For example, genes of the apolipoprotein L gene family (APOL1, APOL3, APOL6) were generally very potent inhibitors. All members of this gene family have rapidly evolved in primates , and APOL1 has been shown to confer immunity to Trypanosoma brucei that can be countered by a trypanosome-encoded antag- onist . Members of the TNF-receptor superfamily (TNFRSF10A and 10D) were also powerful inhibitors. Notably, TNFRSF10A was also identified as one of the most potent anti-HIV-1 factors among 380 type I interferon-stimulated genes in a previous study . An- other potent inhibitor of HIV-1 was CD164 (also known as endolyn) a cell adhesion molecule that interacts with CXCR4 . Furthermore, the high activity of OAS1 is noteworthy because this interferon-induced factor is known to generate 2',5'-oligoadenylates which activate RNase L thus causing viral RNA degradation and inhib- ition of viral replication . Notably, OAS1 affected the percentage of HIV-1 infected (eGFP+) cells less severely than the mean fluorescence intensity of eGFP expres- sion, which would be expected in the case of RNase L mediated viral RNA degradation. A similar phenotype was observed for IFI16, which is involved in transcrip- tional regulation and plays a role in the sensing of intra- cellular viral DNA [28,29]. More recently, IFI16 has been reported as a key player in inducing pyroptosis in abortively infected quiescent CD4+ T cells, thereby con- tributing to the massive depletion of CD4+ T cells ob- served during HIV infection [30,31]. Accumulating evidence suggest that restriction factors, such as TRIM5α  and BST2/tetherin [33-35] may also act as immune sensors and direct anti-HIV effects of IFI16 warrant fur- ther study. Conversely, some genes that displayed anti- HIV activity are known to be involved in signaling, both in various cancers (CD164, TNFRSF10A/D, ZWINT, MT1X, IL3) and immune or T cell activation (GBP5, CD1A). Thus, these genes may not inhibit HIV-1 directly but could also induce cellular responses associated with the expression of other innate immunity factors. This possibil- ity is in agreement with our observation that some candi- date factors inhibited various viral promoters. It is also noteworthy, however, that 293T cells lack many intact immune signaling pathways and did not express detect- able levels of interferon after transfection with constructs expressing the various candidate factors. Further studies
New therapeutic approaches require an increasing amount of recombinant proteins. Presently available expression systems such as bacterial or yeast fermentation, insect cell cultures or mammalian tissue culture often are expensive, potentially hazardous or cannot perform post- translational modifications. Plants could offer an inexpensive alternative to produce correctly processed, functional recombinant proteins free of potential hazards in a large-scale. Here, transient expression using plant virus-based genevectors could be an approach to shorten cycle times and increase yields of recombinant product. Being a very recent development, plant virus-based vectors are not currently used for commercial production of recombinant protein. However, environmental safety and upscalability have already been demonstrated, and pilot facilities have been established (Biosource, Vacaville, CA). As viruses are selected by evolution for most efficient genomes, inserted foreign genes that do not represent an advantage for the virus, so that most might be recombined out within a few generations as suggested by data obtained in growthrooms and greenhouses. This makes field-use of virus vectors safe, as mainly wt-like virus results from mutation events. Selection markers such as antibiotic resistance that are necessary for development of transgenic plants can be omitted in viral expression vectors, thereby reducing possibility of hazardous gene transfer into environment. Potentially high yields of recombinant protein expressed by virus vectors could even allow exclusive production in greenhouses, which would further reduce possible risk of recombinant constructs. These inherent safety features could be important to reassure public concerns that have been presently raised against widespread use of transgenic plants. Furthermore, the application of established plant virus-based vectors requires relatively simple technology that could be easily applied in less developed countries, allowing the inexpensive and safe production of therapeutics such as vaccines at the location they are most needed. Detailed knowledge about plant viruses and their host interactions might lead to vectors with specifically designed viral functions, host range and temporal stability.