Cover Page
The handle http://hdl.handle.net/1887/43807 holds various files of this Leiden University dissertation
Author: Hoyng, Stefan
Title: Gene therapy and nerve repair Issue Date: 2016-11-01
4
Developing a potentially immunologically inert tetracycline-regulatable
viral vector for gene therapy in the peripheral nerve
Stefan A. Hoyng1,2, Sara Gnavi1,3, Fred de Winter1,2, Ruben Eggers1,
Takeaki Ozawa4, Arnaud Zaldumbide5, Rob C Hoeben5, Martijn J.A. Malessy1,2, Joost Verhaagen1,6
Gene Therapy. 2014 Jun;21(6):549-57.
PMID: 24694534
1. Department of Neuroregeneration, Netherlands Institute for Neuroscience, Royal Netherlands Academy of Arts and Sciences (KNAW), Amsterdam, the Netherlands
2. Department of Neurosurgery, Leiden University Medical Center, Leiden, the Netherlands 3. Neuroscience Institute of the Cavalieri Ottolenghi Foundation (NICO) & Department of Clinical
and Biological Sciences, University of Turin, Turin, Italy
4. Department of Chemistry, School of Science, the University of Tokyo, Tokyo, Japan
5. Department of Molecular Cell Biology, Leiden University Medical Center, Leiden, the Netherlands 6. Center for Neurogenomics and Cognitive Research, Vrije Universiteit Amsterdam, Amsterdam,
the Netherlands
ABSTRACT
Viral vector mediated gene transfer of neurotrophic factors is an emerging and prom- ising strategy to promote the regeneration of injured peripheral nerves. Unfortunately the chronic exposure to neurotrophic factors results in local trapping of regenerating axons or other unwanted side effects. Therefore, tight control of therapeutic gene expres- sion is required. The tetracycline/doxycycline-inducible system is considered to be one of the most promising systems for regulating heterologous gene expression. However, an immune response directed against the transactivator protein rtTA hampers fur- ther translational studies. Immunogenic proteins fused with the Gly-Ala repeat of the Epstein-Barr virus Nuclear Antigen-1 protein have been shown to successfully evade the immune system. In this article we used this strategy to demonstrate that a chimeric transactivator, created by fusing the Gly-Ala repeat with rtTA and embedded in a lentivi- ral vector (i) retained its transactivator function in vitro, in muscle explants, and in vivo following injection into the rat peripheral nerve, (ii) exhibited a reduced leaky expression and (iii) had an immune-evasive advantage over rtTA as shown in a novel bioassay for human antigen presentation. The current findings are an important step towards creat- ing a clinically applicable tetracycline-regulatable viral vector system.
4
INTRODUCTION
Traumatic injuries to the human peripheral nervous system (PNS) often lead to severe motor and sensory deficits with a limited perspective of spontaneous axonal regenera- tion and functional recovery. Over the past decades viral vector-mediated gene transfer of neurotrophic factors, axon guidance cues and transcription factors capable of promot- ing and guiding axonal regrowth as well as increasing survival of remaining neurons have emerged as a promising treatment strategy.1,2 Glial cell-line derived neurotrophic factor (GDNF) has, through its effects on outgrowth and survival of motoneurons, emerged as an important candidate to enhance peripheral nerve regeneration.3,4 Lenti- viral vector-mediated overexpression of GDNF during peripheral nerve regeneration has shown promising results but its uncontrolled expression leads to trapping of regenerat- ing axons also known as the “candy store” effect.5–7 The development of a viral vector capable of regulating GDNF expression is therefore essential and could possibly further enhance axonal regeneration and functional outcome.
To achieve this, a regulatable vector should possess several essential characteristics. First, it should allow a rapid and efficient on-and-off switch of GDNF expression. Second, the activation of GDNF expression should solely depend on the presence of an inducer drug which is safe and well tolerated in humans. Finally, a regulatable vector system should exhibit a low potential of eliciting an immune response in the PNS. Over the past years four main systems for in vivo transgene regulation have emerged from animal studies:
the tetracycline8,9 and rapamycin-inducible systems10, the mammalian steroid receptor (tamoxifen and mifepristone)11,12 and the insect steroid receptor (ecdysteroid) system.13 Of these, the tetracycline inducible system has been the most commonly studied and is, at this moment considered the most potent and clinically relevant.14 Since its original development,8 where transgene expression was repressed in the presence of the inducer drug tetracycline (so called tet-off), several major improvement in the tetracycline inducible system have readied this system for clinical translation. These include (i) the development of a more relevant tet-on system9, in which transgene expression is depend- ent on the binding of a tetracycline/doxycycline (Dox) activated transactivator rtTA to a tet-operator controlled promoter; (ii) the development of a transactivator with increased sensitivity to the inducer drug Dox, and (iii) reduced activity in the absence of Dox caus- ing so called “leaky expression”.15 Importantly, the inducing drug Dox is a well tolerated antibiotic drug that has been administered orally and intravenously in humans for over 30 years and achieves excellent tissue penetration. Multiple in vivo studies have shown promising results regulating a large selection of transgenes including GDNF in a variety of tissues using the tet-on system.14,16–18 Unfortunately preclinical studies in non-human primates revealed that long term regulation of transgene expression was hampered by an active loss of genetically modified cells.19–21 This loss resulted from a rapid cellular and humoral response targeted against the cells that expressed the transactivator rtTA. These findings currently exclude in vivo applicability of the tetracycline inducible system for the human PNS.
During evolution many viruses have developed effective immune-evasive mecha- nisms.22,23 The Epstein-Barr virus (EBV), also called human herpes virus 4 requires the Epstein-Barr Nuclear-Antigen-1 protein (EBNA-1) to maintain stability of its genome during latency. To evade cytotoxic T-cell mediated destruction EBNA-1 possesses a long Gly-Ala repeat (GAr) that has been shown to prevent antigenic peptide presentation to major histocompatibility complex (MHC) class I molecules. Previously, in a proof-of- concept study, we showed that fusing the GAr to the N-terminus of rtTA could possibly create a functional and immunologically inert transactivator (GArrtTA).24
In this study we embedded our GArrtTA-based system in a lentiviral vector and investi- gated the effects of GAr fusion on GArrtTA-mRNA and protein biosynthesis. We evalu- ated the capacity of GArrtTA to regulate transgene expression in vitro, ex vivo using muscle explants and in vivo following application in the rat peripheral nerve. We show that the GArrtTA-based system retains its transactivator function in vitro and in vivo in the peripheral nerve, displays reduced leaky expression, and has an immune-evasive advantage over rtTA in a novel bioassay for human antigen presentation.
RESULTS
The effects of GAr on rtTA biosynthesis
In order to examine the possible effects resulting from the fusion of the Gly-Ala repeat to rtTA on protein biosynthesis we performed an in vitro side by side comparison of rtTA and GArrtTA expression. We first created three independent lentiviral vector batches of rtTA and GArrtTA and titered each by measuring genomic integration in HEK293T cells(Figure 1). We then performed a titer matched serial dilution in HEK293T in which we measured viral genomic integration, (GAr) rtTA mRNA and protein synthesis. As expected following titer matched infection we observed similar integration of all six viruses (p=NS) (Figure 2a). Subsequently we measured (GAr) rtTA mRNA expression and observed a 7-fold lower expression in GArrtTA compared to rtTA expressing cells (p<0.01) (Figure 2b). Interestingly, western blot analysis of (GAr) rtTA protein expres- sion showed an inverted 5-fold increase in GArrtTA compared to rtTA expressing cells (p<0.01) (Figure 3c and d). In summary, fusion of the Gly-Ala repeat to rtTA led to a reduction in mRNA expression while paradoxically increasing protein levels.
The effects of GAr on rtTA function
We next evaluated the effects of the fusion of GAr to rtTA on its transactivating capacity.
To assess (GAr) rtTA function, we created a HEK293T cell line expressing firefly lucif- erase under the control of a tet-operator controlled promoter (TRE-Luc). In this cell-line, at equal genomic integration and in the presence of the maximum non-toxic dose of Dox (10µM),25 both rtTA and GArrtTA successfully induced luciferase expression. There was however an approximately 20 fold lower induction by GArrtTA (data not shown).
In light of these results we decided to investigate the possibility to increase the multi-
4
Figure 1. Schematic representation of the constructs and their titers. The classical transactivator rtTA and the potentially non-immunogenic transactivator GArrtTA under the control of the constitutively active CMV promoter were embedded in second generation lentiviral vectors (CMV-rtTA and CMV-GArrtTA respectively).
Additionally lentiviral vectors encoding for firefly luciferase and GDNF under the control of the tetracycline response element were created (TRE-Luc and TRE-GDNF respectively). In the presence of Dox the transactivator protein (rtTA or GArrtTA) can bind to the TRE and transcription of the gene of interest (firefly luciferase or GDNF) is initiated. Three independent batches of CMV-rtTA and CMV-GArrtTA and a single batch of TRE-Luc, TRE-GDNF and CMV-GFP were produced. Transducing units per ml (TU/ml) for the LV-GFP were manually quantified by counting transduction events in HEK293T cells. The genomic integration of viral WPRE of all stocks was measured in HEK293T cells and the ratio between TU/ml and genomic WPRE content of LV-GFP was used to calculate relative TU/ml titers of all stocks.
plicity of infection (MOI) of GArrtTA to correct for this loss of function and compare the overall Dox responsiveness (Figure 3a). In our TRE-Luc cell line we performed a titer matched infection (for GArrtTA and rtTA) and the experimentally determined (20 fold increased) function matched infection for GArrtTA. Cells were stimulated with increas- ing amounts of Dox (0, 0.01, 0.1, 1 and 10µM) for 24hours. In line with our previous observations low Dox concentrations were sufficient to induce near maximum transacti- vation by rtTA but not by GArrtTA.24 By increasing both Dox levels and GArrtTA levels (function matched group) we were however able to induce similar levels of transgene expression for GArrtTA and rtTA (Figure 3a). Interestingly, in the absence of Dox we observed a reduced leaky expression in not only the GArrtTA titer matched group but also the GArrtTA function matched group compared to rtTA. To further investigate this leaky expression we decided to perform a study in our TRE-Luc cell line infected with increasing amounts of the rtTA or GArrtTA virus (MOI of 0.01, 0.1, 1, 10, and 50) in the absence of Dox (Figure 3b). While increasing rtTA concentrations showed a lin- ear increase in leaky expression we were unable to detect any leaky expression in the GArrtTA group (p<0.001). These findings indicate that fusion of the Gly-Ala repeat to rtTA modifies its capacity to transactivate the tet-operator controlled promoter thus effectively reducing its transactivator efficiency and leaky expression.
Regulation of gene expression ex vivo
As an intermediate step towards in vivo GDNF regulation, we evaluated the capacity of our transactivator to regulate luciferase expression in ex vivo rodent tissue explants.
For this we injected muscle explants of the rat gastrocnemius muscle with a two vec- tor approach. Titer matched GArrtTA or rtTA vectors were injected together with a
TRE-Luc vector. An additional function matched group of GArrtTA was added (20 fold increase). Over a total period of 7 weeks we successfully performed 3 full on/off cycles mediated through the addition or removal of Dox, showing first that fusion of GAr to rtTA still allows the release of GArrtTA from the tet-operator controlled promoter and second that a two vector approach is feasible for in vivo translation (Figure 3c and d).
Furthermore as observed in vitro, increasing GArrtTA successfully resulted in a system with comparable efficiency (GArrtTA function match v rtTA; p=NS). These findings sug- gest that at adequate levels GArrtTA can tightly regulate gene expression and that a two vector approach is realistic for in vivo translation.
Regulation of gene expression in vivo
Our next aim was to regulate the expression of a therapeutically relevant protein medi- ated through GArrtTA in vivo. More specifically we wanted to assess whether GArrtTA could (i) induce GDNF expression in the rat peripheral nerve (ii) lead to biologically relevant levels of GDNF expression and (iii) permit long-term regulation in an immune competent environment. To evaluate this, we performed a side by side comparison of the rtTA and GArrtTA based systems in vivo. Using our two vector approach we injected in a 1:1 ratio either CMV-rtTA or CMV-GArrtTA and TRE-GDNF in the left sciatic nerve of Wistar rats (n=48). During a total period of 18 weeks, GDNF expression in the sciatic nerve was measured by ELISA at 5 different time points (week 2,4,6,8 and 18; Figure 4).
Supplementing the regular diet with Dox for a period of two weeks (week 0 to 2) suc- cessfully induced GDNF expression in the peripheral nerve in all groups (week 2). This showed that GArrtTA is a fully functional transactivator in vivo and demonstrated the excellent penetration of Dox in the rat peripheral nerve. A return to the regular diet (week 2 to 4) resulted in a return to baseline (week 4). An additional 2 rounds of Dox including a late time point (week 4 to 6 and 16 to 18) successfully resulted in a renewed increase of GDNF expression (week 8 and 18) albeit at lower levels compared to the first round (week 2). Overall the induced expression levels of GDNF were comparable to previously observed levels achieved with a constitutively active CMV promoter driving expression. (R. Eggers et al, in press) This shows that GArrtTA has the capacity to switch on-and-off biologically relevant expression levels of GDNF. In line with the in vitro find- ings, we observed almost no leaky expression in the GArrtTA group in the absence of Dox (week 4 and 8). In contrast 2 out of 8 animals in the rtTA group had considerable levels of GDNF in the absence of Dox (187pg/cm at week 4 and 227pg/cm at week 8). All together these findings demonstrate for the first time that GArrtTA is a fully functional transactivator with an almost undetectable leaky expression that can efficiently regulate gene expression in vivo.
Unexpectedly, during our 18 week follow-up we did not observe a selective loss of func- tion in the rtTA group. This finding suggests that a vigorous immune response directed against rtTA does not occur in the peripheral nerve of Wistar rats. This therefore pre- cluded the possibility to confirm the immune-evasive advantages of GArrtTA in our in vivo model.
4
Figure 2. The GAr sequence modifies rtTA biosynthesis. Titer matched dilution and transduction of the three independently produced rtTA and GArrtTA lentiviral batches in HEK293T cells were performed to compare mRNA and protein expression of the two transactivators. (a) Quantitative RT-PCR of genomic DNA measuring viral WPRE integration. All six viruses show equal integration and correct titer matching (p=NS). (b) Quantitative RT-PCR of mRNA measuring (GAr) rtTA messenger content. GArrtTA transduced cells have significantly lower rtTA mRNA content than rtTA transduced cells (**p<0.01) (c,d) Western-blot analysis of rtTA and GArrtTA protein expression.
GArrtTA transduced cells have significantly higher protein content than rtTA transduced cells (**p<0.01) (e) Schematic representation of these findings on mRNA and protein biosynthesis and the hypothesized immune evasive strategy of GAr. Graphs represent mean ± SEM of n=3 independently produced viral stocks.
GAr protects from human cytotoxic T-cell mediated destruction
The immunogenicity of rtTA has however been extensively described in non-human pri- mates and remains a putative risk for further human translational studies.19–21 Therefore our final goal was to validate the immune-evasive advantage of the fusion of GAr to rtTA in a human bioassay. In order to achieve this we set up an in vitro assay were we fused the HLA-A2 restricted CMV pp65-derived epitope NLVPMVATV to the C-terminal end of rtTA or GArrtTA (figure 5b). We subsequently monitored the death of rtTA [pp65]
Figure 3. GArrtTA is a functional transactivator. (a) Comparison of the functionality of the transactivators.
Luciferase activity of a stable TRE-Luc HEK293T cell-line infected with three independently produced rtTA and GArrtTA lentiviral batches either titer (rtTA and GArrtTA; MOI 0.5) or function (GArrtTA; MOI 10) matched and stimulated with increasing concentrations of Dox (0 to 10µM). GArrtTA is a functional transactivator but has reduced efficiency and modified Dox responsiveness. (b) Comparison of the leaky expression of the two transactivators. Luciferase activity of a TRE-Luc HEK293T cell-line infected with increasing amounts of rtTA and GArrtTA (MOI 0 to 50) in the absence of Dox. At high MOI rtTA shows a significant higher leaky expression than GArrtTA (at MOI 10, ****p<0.0001). (c,d) Regulating luciferase expression in transduced rodent muscle explant cultures. Luciferase activity in muscle explants transduced with a mix of either titer (rtTA and GArrtTA) or function (GArrtTA) matched and TRE-Luc lentiviral vectors was measured every 3 or 4 days for a total period of 55 days.
The addition of Dox (2µM during periods 3 to 7, 21 to 28 and 42 to 46 days) successfully induced expression at all three independent time points (day 7, 24 and 46) showing reliable Dox-mediated regulation of transgene expression by GArrtTA over time. (c) Representative images of the luciferase signal in a single explant per group following the addition (at day 7, 24 and 46 respectively) or the absence/removal (day 3, 21, 42 respectively) of Dox.
(d) Quantification of the luciferase activity at all time points. Graphs represent mean ± SEM of n=3 independently produced viral stocks or n=4 independently injected muscle explants.
Figure 4. GArrtTA successfully regulates GDNF expression in the rat peripheral nerve. The left sciatic nerves of 48 rats were injected with a mix of rtTA or GArrtTA and TRE-Luc. GDNF expression was measured by ELISA at 2,4,6,8 and 18 weeks post injection of the viral vector in the nerve. Supplementing Dox to the diet successfully induced GDNF expression at three independent times (week 2, 6, 18) showing that a GArrtTA based system is a fully functional system to regulate gene expression in vivo. Bars represent mean ± SEM. n=4 (week 2, 4, 8), 3 (week 6), and 8 (week 18) animals per group.
4
modified HLA-A2 target cells induced by CMV-specific CD8+ T-cells (CTL). To assess CTL induced apoptosis, we used a caspase-3 sensitive firefly luciferase.26 In apoptotic conditions, this non active cyclic luciferase is cleaved by caspase-3 at the tetrapeptide sequence: Asp–Glu–Val–Asp (DEVD) resulting in the return of luciferase into an active form. We created a lentiviral vector in which the expression of the DEVD-cyclic-lucif- erase reporter cassette (cFLucDEVD) is driven by the constitutively active CMV pro- moter.
To determine whether such reporter could be used to monitor CMV CTL-induced apop- tosis, HLA-A2+ HEK293T cells were incubated with synthetic NLVPMVATV peptides derived from the pp65 of CMV and exposed to increasing amounts of CMV NLVPM- VATV/HLA-A2-specific CTLs. Overnight incubation resulted in an up to 4-fold increase in light emission upon incubation with pp65-specific CTL only in the cells loaded with peptide (p<0.0001) (Figure 5a). Of note, in the absence of peptide, no significant increase in luciferase activity was detectable upon CTL addition. Similarly, transfection of rtTA[pp65] in HEK293T cells induced an up to 10-fold increase in luciferase activity upon incubation with pp65-specific CTL, demonstrating efficient processing and pres- entation of the antigenic epitope to CTLs (Figure 5d). In the same experimental set-up, while transfection of GArrtTA[pp65] gave rise to similar protein expression, as shown by western blot analysis (Figure 5c), pp65 presentation to CMV-CTLs was remarkably impaired by the fused Gly-Ala repeat. A 70% reduction of luciferase activity for the high- est T-cell concentration was observed (rtTA v GArrtTA, p<0.0001).
DISCUSSION
In this article, we present a tetracycline-inducible lentiviral vector system based on a potentially non-immunogenic transactivator which allows the controlled expression of a gene of interest in vivo. To achieve this we embedded the classical regulatable system, based on the immunogenic rtTA transactivator, and our system, based on the potentially non-immunogenic GArrtTA transactivator in lentiviral vectors and performed a side by side comparison in vitro in a cell-line, ex vivo in muscle explants and in vivo in the rodent peripheral nerve. Finally we developed a novel human in vitro assay to validate the capacity of GArrtTA to evade human cytotoxic T-cell mediated destruction validat- ing the clinical relevance of our system. We show here that GArrtTA is a fully functional transactivator which possesses significantly enhanced immune-evasive properties that can successfully regulate transgene expression in vivo.
The Epstein–Barr Nuclear-Antigen-1 protein (EBNA-1) permits latency of the Epstein Bar virus (EBV) in humans. It plays an essential role in the episomal maintenance of the EBV genome while successfully evading immune surveillance.27 Unraveling its composi- tion has revealed that its immune evasive characteristics are the result of a long irregular copolymer repeat of Glycine and Alanine (GAr) present at its N-terminal. This repeat successfully prevents EBNA-1 presentation by MHC class I molecules disabling cytotoxic
T cell recognition and destruction.28,29 Fusion of GAr at the N-terminus of different immunogenic proteins successfully resulted in their immune evasion.6,30,31 Previously, in an in vitro proof-of-concept study, we showed that fusion of GAr to rtTA could possibly create a functional and immunologically inert transactivator.24 In this manuscript we embedded the classical rtTA and our transactivator GArrtTA in a lentiviral backbone.
To investigate the effects of GAr on rtTA biosynthesis, the mRNA and protein expres- sion was quantified following equal genomic integration of both vectors. Interestingly, fusion of GAr created an increase in protein expression (7:1 ratio GArrtTA v rtTA) while mRNA availability was paradoxically reduced (1:5 ratio GArrtTA v rtTA). The increased GArrtTA protein availability is in line with the capacity of GAr to increase protein sta- bility and inhibit proteasomal degradation.28,32 Initial reports conferred the immune evasive advantages of GAr to this inhibition hypothesizing that proteasome mediated degradation was necessary for antigen presentation by the MHC class I. Recently, how- ever, it has become clear that the majority of presented antigenic peptides are not derived from proteasomal degradation but rather from early translation initiation events from
Figure 5. GAr protects in-cis fused transgenes from human cytotoxic T-cell mediated recognition and destruction. (a) Validation of our bioassay. H293T cells possessing a caspase-3 sensitive firefly luciferase were exposed to the CMV pp65-derived peptide NLVPMVATV (Pep). Following the exposure to increasing amounts of the HLA-A2-specific CMV-CTLs, apoptosis and luciferase activity was induced (p<0.001). (b) Schematic representation of the constructs that harbor pp65 tagged transactivators. The HLA-A2 restricted CMV pp65-derived epitope containing a KpnI restriction site was fused to the C-terminus of rtTA and GArrtTA by site directed mutagenesis (indicated by the small black box in the construct). (c) Western-blot analysis of rtTA expression in human HEK293T cells transfected with rtTA [pp65] or GArrtTA [pp65] constructs showing correct protein synthesis. d) GAr successfully inhibits human CTL-mediated apoptosis of rtTA expressing cells in vitro. Cells transfected with rtTA[pp65] or GArrtTA[pp65] constructs in the presence of increasing amounts of HLA-A2-specific CMV-CTL. CTL induced apoptosis was significantly reduced in GArrtTA [pp65] transfected cells (****p<0.0001). Graphs represent mean ± SEM. n=3 measurements. NT, not-treated
4
reading frames throughout the mRNA.33 Indeed, not proteasomal inhibition but inhibi- tion of in-cis translation initiation permits the immune evasion of GAr.34,35 In line with this theory, reducing the overall presence of “immunogenic” viral mRNA could be an additional strategy used by EBNA-1 to evade immune presentation. In EBV EBNA-1 has been shown to regulate different viral promoters including its own.36,37 A plausible strategy to create a stable and non-immunogenic setting during EBV latency could be a negative-feedback of EBNA-1 causing the reduction of viral mRNA while maintaining functional levels of this essential and immune evasive protein. This homeostasis seems ideal for a regulatable vector since constant high transactivator levels are desirable while immune-presentation must be minimized.
Fusion of GAr to rtTA results in a 3 times larger protein (248 versus 678 amino acids).
Therefore we investigated the effects of this fusion on the transactivating capacity of rtTA. Fusion of the GAr to rtTA leads to a reduction in transactivation efficiency. For- tunately, increasing transactivator concentrations resulted in a near normal efficiency of the induction of transgene expression. Importantly, under these conditions, leaky expression (binding of the transactivator in the absence of Dox) was not increased. These characteristics highlight that GArrtTA is an effective transactivator with the advantage of showing almost no leaky expression. Over the years the modification of rtTA has cre- ated highly efficient and sensitive transactivators in vitro.15,38,39 The question remains, however if contaminants in food and other comestible products could, in such sensi- tive transactivators, undesirably induce chronic transgene expression in a clinical set- ting. On the other hand, an insensitive transactivator might be impossible to activate or cause unwanted side effects from the required high Dox levels such as nephrotoxicity or photosensitivity.14, It is important to note that the concentrations of Dox used here are approximately 100 fold higher than the recommended dose for clinical use.40 To inves- tigate whether it was possible to regulate transgene expression through the less sensitive GArrtTA and successfully evade immune-mediated destruction of rtTA we performed a comparative study of the two systems in vivo using lentiviral-mediated delivery of the two transactivators to the rat peripheral nerve.
We demonstrated GDNF expression in the rat peripheral nerve following lenti viral transduction through the addition or removal of Dox for a total period of up to 18 weeks.
The oral administration of Dox was sufficient to allow rtTA and GArrtTA to equally induce previously reported biologically relevant levels of GDNF expression in the peripheral nerve. (R. Eggers et al, in press) In line with our in vitro findings we found a reduced leaky expression in GArrtTA. This is clinically relevant because unwanted side effects following the prolonged application of several neurotrophic factors have been documented.2 Unexpectedly, during our 18 week follow-up we did not observe a loss of function in the rtTA group. Earlier reports from our group have shown that in the rodent peripheral nerve cells expressing foreign proteins following lentiviral transduc- tion are rapidly cleared by the immune system.41 Our results apparently suggest that a vigorous response against rtTA is absent in the peripheral nerve of Wistar rats. The immunogenicity of rtTA has, however, been extensively shown in non-human primates
and precludes the use of this system in translational studies in human subjects.19–21 Using a caspase-sensitive firefly luciferase26 in combination with human cytotoxic T cells we demonstrated that GArrtTA displays a significant immune evasive advantage in human cells. Interestingly even though this immune-evasion was highly efficient it was not complete. Combining this assay with our TRE-Luc cell-line would be an ideal screen to discover novel enhanced GArrtTA transactivators in the future. Reducing the size of GAr or modifying the existing GArrtTA through viral evolution, an elegant method described by Das and colleagues38,39, could potentially create a more efficient transacti- vator that can be activated at clinically acceptable dose of Dox(40), while conserving its non-leaky and immune evasive properties.
In conclusion, we have developed a fully functional lentiviral vector system capable of regulating transgene expression in vivo via a tetracycline inducible transactivator that possesses enhanced immune-evasive properties. The findings reported in this article present an important step towards the milestone of creating a clinically applicable tetra- cycline regulatable viral vector system.
MATERIAL AND METHODS
DNA constructs. Second generation lentiviral (LV) transfer plasmids flanked by the con- stitutively active CMV promoter and the woodchuck hepatitis virus posttranscriptional regulatory element (WPRE) were created by replacing the GFP sequence from pRRL- CMV-GFP-WPRE by the coding sequence for rtTA and GArrtTA.24 Transfer plasmids where expression of GDNF and luc2 (pGL4.10, Promega, Leiden, the Netherlands) was driven by the inducible tet responsive element (TRE) for GDNF and luc2 were created by replacing the CMV promoter by seven repeats of the tet operator DNA sequence fused with a minimal CMV promoter (Figure 1).6,42 pRRL-CMV-cFluc-DEVD-WPRE was generated by removing the cFluc-DEVD fragment from pcDNA3.1-cFlucDEVD26 with PmeI/PmeI and ligating it into pRRL-CMV-WPRE opened by SmaI. The pp65 epitope containing constructs were generated by mutagenic PCR using the following primers: 5’- GACATGCTCCCCGGGAACTTGGTACCAATGGTTGCTACTGTTTAACTAAGTA AGGATCCG-3’; 5’- CGGATCCTTACTTAGTTAAACAGTAGCAACCATGGTACCAA GTTCCCGGGGAGCATGT-3’on pRRL-CMV-rtTA-WPRE and pRRL-CMV- GArrtTA- WPRE (Figure 5b). The pp65 (NLVPMVATV) encoding sequence is underlined. The constructs were validated by restriction analysis using the KpnI site in the mutant con- structs shown in bold in the primers sequences.
Lentiviral production and titering. LV stocks were generated as previously described31,43,44 in three independent rounds of production. Briefly, for each batch of LV two 15cm Petri dishes containing 12.5x106 HEK293T in Iscoves modified Dulbeccos medium (IMDM) containing 10% fetal calf serum (FCS), 1% Penicillin/Streptomycin (P/S) and Glutamax (Invitrogen) were prepared. All cells were maintained at 37°C in a humidified atmos-
4
phere of 5% CO2. Using branched polyethylenimine (Sigma, St Louis, MO) a triple trans- fection with the LV transfer, packaging (pCMVdeltaR8.74) and envelope (pMD.G.2) plasmid was performed (ratio 3:2:1, total DNA 90μg/plate). After 14 hours, the medium was replaced by IMDM containing 2%FCS, 1%P/S and Glutamax. After 24 hours, the medium was harvested, filtered through a 0.22uM filter and concentrated by ultracen- trifugation at 20.000 rpm for 2.5 hours in a SW32Ti rotor (Beckman Coulter BV, the Netherlands). Viral pellets were resuspended in PBS pH7.4, aliquoted and stored at -80°C until further use.
Serial dilutions (10-2,-3,-4 and 10-2 to -7 for LV-GFP) of all viral stocks were used to infect 2 x 105 HEK293T in IMDM 2% FCS, 1% P/S and Glutamax seeded in poly-L-lysine (PLL) coated 24-well culture plates. After 48 hours 1) the number of transducing units per ml (TU/ml) for the LV-GFP stock was manually quantified by counting transduction events in the LV-GFP transduced cells using a fluorescence microscope (10-5,-6,-7), 2) genomic DNA (gDNA) of all samples was extracted and measured for viral integrating events by quantitative PCR (10-2,-3,-4). Briefly, cells were harvested and gDNA was extracted (DNeasy Blood & Tissue Kit, Qiagen, Venlo, the Netherlands). Viral mediated transgene integration was measured using primers directed against the lentiviral WPRE on an ABI 7900HT detection system (Applied Biosystems) using the SYBR green PCR kit (Applied Biosystems). WPRE primers were as follows: 5’- TTCCCGTATGGCTTTCATTT-3’
and 5’- GAGACAGCAACCAGGATTTA-3’. All expression values were normal- ized to that of the reference gene GAPDH. GAPDH primers were as follows:
5’- TGCACCACCAACTGCTTAGC-3’ and 5’-CGCATGGACTGTGGTCATGA-3’. The ratio between the TU/ml and gDNA WPRE content of the LV-GFP stock was used to calculate relative TU/ml titers for all stocks on the basis of their gDNA WPRE content.
In-vitro biosynthesis comparison. Titer matched serial dilutions of the 3 independently produced viral batches of rtTA and GArrtTA were performed in 3 (one for genomic DNA, one for mRNA and one for protein extraction) 24-well culture plates pre-coated with PLL and seeded with 2 x 105 HEK293T in IMDM 2% FCS, 1% P/S and Glutamax. All cells were harvested 48 hours post transduction. To quantify genomic integration, gDNA from three dilutions (MOI 0.5, 5, 50) was extracted and analysed for viral WPRE integrat- ing events as described above. To determine rtTA mRNA expression, RNA from three dilutions (MOI 0.2, 1, 5) was extracted using 250 µl TRIzol reagent (Invitrogen, Carlsbad, CA) per well. RNA was treated with DNase I (Invitrogen) and 1µg was reverse transcribed using a QuantiTect Reverse Transcription Kit (Qiagen) according to manufacturer’s protocol. Quantitative PCR was performed as described above with primers specific for rtTA: 5’-CACCTACCACCGATTCTATG-3’ and 5’- ACAGCTCAATTGCTTGTTTC-3’
and GAPDH. All expression values were normalized to that of GAPDH.
Western blot. Cells were treated with RIPA lysis buffer (50 mM Tris/HCl pH 7.5, 150 mM NaCl, SDS 0.1%, 0.5% DOC, 1% NP40 and protease inhibitors). Proteins were seper- ated by electrophoresis on an acrylamide/bisacrylamide SDS page gel transferred to
Immobilon-P [Immobilon-P transfer membrane (PVDF); Millipore, Etten-Leur, the Netherlands] and imunnoblotted for rtTA which was detected with anti-TetO2 anti- body(1:1,000; MoBiTec) and the loading control used was anti-actin (1:5,000, clone C4;
ICN Biomedicals, Inc., Zoetermeer, the Netherlands). Labeled proteins were detected by using Western Lightning Plus–ECL (Perkin-Elmer, Groningen, the Netherlands). All blots were captured with Amersham Hyperfilm ECL (GE Healthcare Europe, Diegem, Belgium). Band intensities were measured with ImageJ analysis software (U.S. National Institutes of Health, Bethesda, Maryland).
Creation of a TRE-Luc cell line. In a 96-well culture plate a single well was seeded with 1 x 103 HEK293T in IMDM 2% FCS, 1% P/S and Glutamax. 24 hours post seeding the cells were transduced with a TRE-Luc lentiviral vector at an MOI of 100. 24 hours later the medium was refreshed and we transduced the cells an additional time at an equal MOI of 100. The next day we collected 5 single cells and grew them to confluency in unique wells in a 96-well culture plate in IMDM 10% FCS, 1% P/S and Glutamax. Once confluent each unique well was reseeded in a new well in two separate 24-well culture plates. Genomic DNA was extracted from each well from one plate and TRE-Luc inte- gration was measured as described above using primers directed against the Luc2 gene.
Luc2 primers were as follows: 5’- CACATATCGAGGTGGACATT-3’ and 5’- GCATGAA- GAACTGCAAGCTA-3’. All expression values were normalized to that of the reference gene GAPDH. All 5 clones efficiently integrated TRE-Luc at similar levels. Clone #2 was selected and kept in culture for all experiments.
Luciferase activity. To measure luciferase activity D-luciferin sodium salt (Regis Tech- nologies, Morton Grove, Illinois) diluted in PBS was added to each well to obtain a final concentration of 375 µg/ml. Using an IVIS 200 imaging system (Caliper Live sciences, Hopkinton, Massachusetts) the average radiance for each well was measured. The fol- lowing settings were used: binning at 8, camera lens aperture size at 1 and a field of view of 19.5x19.5cm. The average radiance during a 10 second exposure time was measured approximately 10 minutes post addition of D-luciferin.
Doxycycline responsiveness and leakiness assay. A total of four 24-well culture plates pre-coated with PLL were seeded with 2 x 105 TRE-Luc HEK293T in IMDM 2% FCS, 1% P/S and Glutamax. For the Dox responsiveness three plates were used. Each single plate was transduced with three independently produced viral batches of either rtTA (MOI 0.5), GArrtTA (titer matched, MOI 0.5) or GArrtTA (function matched, MOI 10).
24 hours post transduction the medium was refreshed and supplemented with different concentrations of Dox (0, 10 -3, -2, -1, 0, 1 µM). 48 hours post transduction luciferase activity was measured. For the leakiness assay a titer matched serial dilution of the 3 indepen- dently produced viral batches of rtTA and GArrtTA (MOI 10-3, -2, -1, 0, 1 and 50) was per- formed in the remaining plate. 24 hours post transduction the medium was refreshed but no Dox was added. 48 hours post transduction luciferase activity was measured.
4
Experimental animals. A total of 48 female Wistar rats (200-250 g; Harlan, Horst, The Netherlands) were used. Animals were housed under standard conditions at a 12:12 h light/dark cycle with regular (week 3, 4, 7 to 16 weeks) or Dox (6g/kg) enriched food (TD.09282, Harlan) (week 1, 2, 5, 6 and 16 to 18) and water ad libitum. Experimental procedures were performed in accordance with the European guidelines for the care and use of laboratory animals (86\609\EEC) and the local committee for laboratory animal welfare and experimentation.
Ex vivo assay. A total of 16 muscle explants of rat gastrocnemius muscle were harvested from female Wistar rats (approximately 0.8cm3), divided in groups groups (n=4) and kept in culture in two 12 well plates in IMDM 2% FCS, 1% P/S and Glutamax for a total period of 44 days. Using a two vector approach a titer-matched CMV-rtTA (2.3E+09 TU/ml) and CMV GArrtTA (1.0E+10 TU/ml) and a function matched CMV-GArrtTA (20 fold increase) diluted in PBS pH7.4 to create a total volume of 9µL were mixed with 1µL of a TRE-Luc vector (2.0E+10 TU/ml). An additional PBS injected control group was added.
The explants were injected using an operating microscope with 10µl of each viral vec- tor solution. The injection was performed with a glass needle fixed to a 10 µl Hamilton syringe. Fast green (Sigma, Zwijndrecht, the Netherlands) was added to the viral vector solution to aid in the visualisation of the injection procedure. Every 3 to 4 days medium was refreshed and luciferase expression was measured as described above except that the ROI was reduced to the muscle explant and the best exposure time was determined automatically. The medium was supplemented with 2µM of Dox a total of 3 times (day 7-10, 24-28 and 41-44) and 2 extra medium washes of the explants were performed upon return to regular medium.
Surgical procedures. All animals were anesthetized using isoflurane (Isoflo, Abbott, Hoofddorp, the Netherlands). Following the splitting of the left gluteal muscle the sciatic nerve was exposed. Starting 5 mm distal to the internal obturator tendon the sciatic nerve was injected using a glass needle as described above through an operating microscope with 2µl of a viral vector solution. The viral spread was subsequently marked proximally with a 10/0 epineural suture (Ethilon, Johnson & Johnson, the Netherlands). The left sci- atic was injected with a mix of either 1µl CMV-rtTA (2.3E+09 TU/ml) or CMV-GArrtTA (1.0E+10 TU/ml) and 1µl TRE-GDNF (2.3E+10 TU/ml). The wound was closed and a sin- gle dose of Buprenorphine 0.03 ml/100 g body weight (Temgesic, Schering-Plough B.V., Maarssen, the Netherlands) was injected for post-operative analgesia. Animals were kept at 37°C until fully recovered.
GDNF ELISA. At 2,4,6,8 (n=8) and 18 (n=16) weeks post-injection animals were eutha- nized using pentobarbital (sodium pentobarbital; 0.11ml/100g, Sanofi Sante, Maassluis, the Netherlands). The sciatic nerve was re-exposed by splitting the gluteal muscle and the epineurial marking suture was localized. A 1.5cm segment distal to this suture was dis- sected. This segment was snap frozen on dry-ice. To quantify the amount of GDNF, the
nerve segments were homogenized in a mortar cooled with dry ice and containing liquid nitrogen and resuspended in 250 ml lysis buffer (137 mM NaCl, 20 mM Tris/HCL, pH 8.0, 1% Nonidet P40, 10% glycerol, 0.1% Polysorbate 20, 0.5 mM sodium othonovadate and 1 tablet / 50 ml of Roche total protease inhibitor). The concentration of GDNF was meas- ured with an ELISA kit (Emax #g7620, Promega, Madison, Wisconsin) on high binding ELISA plates (Nunc-Immuno Maxisorp #439454). The procedure was performed follow- ing the manufacturer’s instructions and the final GDNF concentration was expressed in pg/cm of nerve segment.
Human cytotoxic T lymphocytes. For maintenance and expansion, pp65CMV-directed CTLs were cultured for two weeks with irradiated allogeneic PBMCs, irradiated CMVpp65 peptide-pulsed HLA-A2-expressing EBV-LCL in IMDM supplemented with 10% human serum, 0.5% LeucoA, 0.1 ng/ml recombinant human (rh)-IL12, 10 ng/ml rh-IL7, 25 U/ml rh-IL2 and 5 ng/ml rh-IL15. Cells were frozen in a solution of 20% human pooled serum and 10% DMSO and kept in liquid nitrogen until use. Upon thawing, cells were allowed to rest in IMDM supplemented with 10% human serum, IL-2 (50U/ml) and IL-15 (0.1ng/ml).
In-vitro apoptosis assay. To measure apoptosis in target cells we used a caspase-3 sen- sitive luciferase reporter as previously described.26 cFluc-DEVD transduced HEK293T cells were transfected in suspension using Polyethylenimine (PEI) with 1µg rtTA [pp65]
or 1µg GArrtTA [pp65] and plated in a 96-well plate in IMDM 10% FCS, 1% P/S and Glutamax. Twenty four hours post transfection, the medium was removed and modified cells were incubated with increasing amounts of HLA-A2-specific CTLs directed against pp65 (ranging from ~1:1 to 10:1 effector: target ratio). After 24h coculture, cells were lysed with a luciferase lysis buffer (25mmol/l Tris-HCl pH=7.8, 2mmol/l CDTA, 2mmol/l DTT, 10% glycerol, 1% triton X-100). Luciferase activity, reflecting the caspase-3 activity, was measured by luminometry (Luciferase Assay System, Promega). Experiments were performed in triplicate for every effector: target ratio used. Luciferase activity in the absence of CTLs was set to 1 and results are shown as relative luciferase activity.
Statistical analysis. All data were presented as mean ± SEM and subjected to nonpara- metric statistical analysis using two-way analysis of variance test with Bonferroni cor- rection using GraphPad Prism soſtware (GraphPad, San Diego, CA). A value of p < 0.05 was considered statistically significant.
ACKNOWLEDGEMENTS
The authors thank Kasper C. Roet, for expert technical assistance in setting up the bio- luminescence assay. This study was partly funded by a grant from the International Spinal Research Trust (TRI004). The authors declare no conflict of interest.
4
REFERENCES
1 Mason MRJ, Tannemaat MR, Malessy MJA, Verhaagen J. Gene therapy for the periph- eral nervous system: a strategy to repair the injured nerve? Curr Gene Ther 2011; 11:
75–89.
2 Hoyng SA, Tannemaat MR, De Winter F, Verhaagen J, Malessy MJA. Nerve surgery and gene therapy: a neurobiological and clinical perspective. J Hand Surg Eur Vol 2011; 36: 735–46.
3 Li L, Wu W, Lin LF, Lei M, Oppenheim RW, Houenou LJ. Rescue of adult mouse moto- neurons from injury-induced cell death by glial cell line-derived neurotrophic factor.
Proc Natl Acad Sci U S A 1995; 92: 9771–5.
4 Henderson CE, Phillips HS, Pollock RA, Davies AM, Lemeulle C, Armanini M, et al. GDNF: a potent survival factor for moto- neurons present in peripheral nerve and muscle. Science 1994; 266: 1062–4.
5 Blits B, Carlstedt TP, Ruitenberg MJ, de Winter F, Hermens WTJMC, Dijkhuizen PA, et al. Rescue and sprouting of moto- neurons following ventral root avulsion and reimplantation combined with intraspinal adeno-associated viral vector-mediated expression of glial cell line-derived neuro- trophic factor or brain-derived neurotrophic factor. Exp Neurol 2004; 189: 303–16.
6 Eggers R, Hendriks WTJ, Tannemaat MR, van Heerikhuize JJ, Pool CW, Carlstedt TP, et al. Neuroregenerative effects of lentiviral vector-mediated GDNF expression in reim- planted ventral roots. Mol Cell Neurosci 2008; 39: 105–17.
7 Tannemaat MR, Eggers R, Hendriks WT, De Ruiter GCW, Van Heerikhuize JJ, Pool CW, et al. Differential effects of lentiviral vector- mediated overexpression of nerve growth factor and glial cell line-derived neuro-
trophic factor on regenerating sensory and motor axons in the transected peripheral nerve. Eur J Neurosci 2008; 28: 1467–79.
8 Gossen M, Bujard H. Tight control of gene expression in mammalian cells by tetracy- cline-responsive promoters. Proc Natl Acad Sci U S A 1992; 89: 5547–51.
9 Gossen M, Freundlieb S, Bender G, Müller G, Hillen W, Bujard H. Transcriptional acti- vation by tetracyclines in mammalian cells.
Science 1995; 268: 1766–9.
10 Rivera VM, Clackson T, Natesan S, Pollock R, Amara JF, Keenan T, et al. A humanized system for pharmacologic control of gene expression. Nat Med 1996; 2: 1028–32.
11 Vegeto E, Allan GF, Schrader WT, Tsai MJ, McDonnell DP, O’Malley BW. The mecha- nism of RU486 antagonism is dependent on the conformation of the carboxy-terminal tail of the human progesterone receptor.
Cell 1992; 69: 703–13.
12 Roscilli G, Rinaudo CD, Cimino M, Sporeno E, Lamartina S, Ciliberto G, et al. Long- term and tight control of gene expression in mouse skeletal muscle by a new hybrid human transcription factor. Mol Ther 2002;
6: 653–63.
13 Karns LR, Kisielewski A, Gulding KM, Seraj JM, Theodorescu D. Manipulation of gene expression by an ecdysone-inducible gene switch in tumor xenografts. BMC Biotech- nol 2001; 1: 11.
14 Stieger K, Belbellaa B, Le Guiner C, Moullier P, Rolling F. In vivo gene regulation using tetracycline-regulatable systems. Adv Drug Deliv Rev 2009; 61: 527–41.
15 Urlinger S, Baron U, Thellmann M, Hasan MT, Bujard H, Hillen W. Exploring the sequence space for tetracycline-dependent transcriptional activators: novel mutations
yield expanded range and sensitivity. Proc Natl Acad Sci U S A 2000; 97: 7963–8.
16 Georgievska B, Jakobsson J, Persson E, Ericson C, Kirik D, Lundberg C. Regulated delivery of glial cell line-derived neuro- trophic factor into rat striatum, using a tetracycline-dependent lentiviral vector.
Hum Gene Ther 2004; 15: 934–44.
17 Chtarto a, Yang X, Bockstael O, Melas C, Blum D, Lehtonen E, et al. Controlled deliv- ery of glial cell line-derived neurotrophic factor by a single tetracycline-inducible AAV vector. Exp Neurol 2007; 204: 387–99.
18 Yang X, Mertens B, Lehtonen E, Vercammen L, Bockstael O, Chtarto A, et al. Revers- ible neurochemical changes mediated by delayed intrastriatal glial cell line-derived neurotrophic factor gene delivery in a partial Parkinson’s disease rat model. J Gene Med 2009; 11: 899–912.
19 Chenuaud P, Larcher T, Rabinowitz JE, Provost N, Joussemet B, Bujard H, et al.
Optimal design of a single recombinant adeno-associated virus derived from serotypes 1 and 2 to achieve more tightly regulated transgene expression from nonhuman primate muscle. Mol Ther 2004;
9: 410–8.
20 Latta-Mahieu M, Rolland M, Caillet C, Wang M, Kennel P, Mahfouz I, et al. Gene transfer of a chimeric trans-activator is immunogenic and results in short-lived transgene expression. Hum Gene Ther 2002; 13: 1611–20.
21 Favre D, Blouin V, Provost N, Spisek R, Porrot F, Bohl D, et al. Lack of an immune response against the tetracycline-depend- ent transactivator correlates with long-term doxycycline-regulated transgene expression in nonhuman primates after intramuscular injection of recombinant adeno-associated virus. J Virol 2002; 76: 11605–11.
22 Zaldumbide a, Hoeben RC. How not to be seen: immune-evasion strategies in gene therapy. Gene Ther 2008; 15: 239–46.
23 Horst D, Verweij MC, Davison AJ, Ress- ing ME, Wiertz EJHJ. Viral evasion of T cell immunity: ancient mechanisms offer- ing new applications. Curr Opin Immunol 2011; 23: 96–103.
24 Zaldumbide A, Weening S, Cramer SJ, Rabelink MJWE, Verhaagen J, Hoeben RC.
A potentially immunologically inert deriva- tive of the reverse tetracycline-controlled transactivator. Biotechnol Lett 2010; 32:
749–54.
25 Ermak G, Cancasci VJ, Davies KJA.
Cytotoxic effect of doxycycline and its implications for tet-on gene expression systems. Anal Biochem 2003; 318: 152–4.
26 Kanno A, Yamanaka Y, Hirano H, Umezawa Y, Ozawa T. Cyclic luciferase for real-time sensing of caspase-3 activities in living mammals. Angew Chem Int Ed Engl 2007; 46: 7595–9.
27 Young LS, Rickinson AB. Epstein-Barr virus: 40 years on. Nat Rev Cancer 2004; 4:
757–68.
28 Levitskaya J, Sharipo A, Leonchiks A, Ciechanover A, Masucci MG. Inhibition of ubiquitin/proteasome-dependent protein degradation by the Gly-Ala repeat domain of the Epstein-Barr virus nuclear antigen 1. Proc Natl Acad Sci U S A 1997; 94: 12616–
21.
29 Blake N, Lee S, Redchenko I, Thomas W, Steven N, Leese A, et al. Human CD8+ T cell responses to EBV EBNA1: HLA class I presentation of the (Gly-Ala)-containing protein requires exogenous processing.
Immunity 1997; 7: 791–802.
30 Ossevoort M, Zaldumbide A, Cramer SJ, van der Voort EIH, Toes REM, Hoeben RC.
Characterization of an immuno “stealth”
4
derivative of the herpes simplex virus thymidine-kinase gene. Cancer Gene Ther 2006; 13: 584–91.
31 Hendriks WTJ, Eggers R, Verhaagen J, Boer GJ. Gene transfer to the spinal cord neural scar with lentiviral vectors: predominant transgene expression in astrocytes but not in meningeal cells. J Neurosci Res 2007; 85:
3041–52.
32 Levitskaya J, Coram M, Levitsky V, Imreh S, Steigerwald-Mullen PM, Klein G, et al.
Inhibition of antigen processing by the internal repeat region of the Epstein-Barr virus nuclear antigen-1. Nature 1995; 375:
685–8.
33 Apcher S, Daskalogianni C, Lejeune F, Manoury B, Imhoos G, Heslop L, et al.
Major source of antigenic peptides for the MHC class I pathway is produced during the pioneer round of mRNA transla- tion. Proc Natl Acad Sci U S A 2011; 108:
11572–7.
34 Apcher S, Komarova A, Daskalogianni C, Yin Y, Malbert-Colas L, Fåhraeus R. mRNA translation regulation by the Gly-Ala repeat of Epstein-Barr virus nuclear antigen 1. J Virol 2009; 83:
1289–98.
35 Apcher S, Daskalogianni C, Manoury B, Fåhraeus R. Epstein Barr virus-encoded EBNA1 interference with MHC class I anti- gen presentation reveals a close correlation between mRNA translation initiation and antigen presentation. PLoS Pathog 2010; 6:
e1001151.
36 Lu F, Wikramasinghe P, Norseen J, Tsai K, Wang P, Showe L, et al. Genome-wide analysis of host-chromosome binding sites for Epstein-Barr Virus Nuclear Antigen 1 (EBNA1). Virol J 2010; 7: 262.
37 Dresang LR, Vereide DT, Sugden B. Iden- tifying sites bound by Epstein-Barr virus
nuclear antigen 1 (EBNA1) in the human genome: defining a position-weighted matrix to predict sites bound by EBNA1 in viral genomes. J Virol 2009; 83: 2930–40.
38 Zhou X, Vink M, Klaver B, Berkhout B, Das a T. Optimization of the Tet-On system for regulated gene expression through viral evolution. Gene Ther 2006; 13: 1382–90.
39 Das AT, Zhou X, Vink M, Klaver B, Verhoef K, Marzio G, et al. Viral evolution as a tool to improve the tetracycline-regulated gene expression system. J Biol Chem 2004; 279:
18776–82.
40 Ariza J, Bosilkovski M, Cascio A, Colme- nero JD, Corbel MJ, Falagas ME, et al.
Perspectives for the treatment of brucellosis in the 21st century: the Ioannina recom- mendations. PLoS Med 2007; 4: e317.
41 Hendriks WTJ, Eggers R, Carlstedt TP, Zaldumbide A, Tannemaat MR, Fallaux FJ, et al. Lentiviral vector-mediated reporter gene expression in avulsed spinal ventral root is short-term, but is prolonged using an immune “stealth” transgene. Restor Neurol Neurosci 2007; 25: 585–99.
42 Baron U, Bujard H. Tet repressor-based system for regulated gene expression in eukaryotic cells: principles and advances.
Methods Enzymol 2000; 327: 401–21.
43 Dull T, Zufferey R, Kelly M, Mandel RJ, Nguyen M, Trono D, et al. A third-gener- ation lentivirus vector with a conditional packaging system. J Virol 1998; 72: 8463–
71.
44 Naldini L, Blömer U, Gallay P, Ory D, Mulligan R, Gage FH, et al. In vivo gene delivery and stable transduction of nondi- viding cells by a lentiviral vector. Science 1996; 272: 263–7.
