Final Report Summary - BRAINCAV (Non-human adenovirus vectors for gene transfer to the brain)
Executive summary:
Prevention and treatment of diseases that affect the brain will remain formidable challenges for years. Yet, as suggested by recent clinical trials, 'gene therapy' offers substantial potential and reason to believe that we can develop long-term, effective therapies for some brain diseases. Gene therapy is an approach that, instead of using classic methods (i.e. chemicals, surgery, radiation) uses DNA or RNA as drugs. Notably though, this strategy also has its own unique obstacles and risks - in particular the need to address feasibility, efficacy and safety. BRAINCAV's foundation is the ability of a virus called 'CAV-2' to deliver DNA to neurons - the most fascinating cells in the body. To determine if CAV-2 vectors have a clinical use for therapy for brain disorders the consortium used approaches that spanned basic research through pre-clinical model feasibility, efficacy and safety.
Our specific objectives included trying to understand the interaction of CAV-2 with fluid and cells of the brain. Does CAV-2 negatively affect neuron function or modify the genes that are expressed? Does CAV-2 leave the brain and go to other tissues? Does it interact with other cells besides neurons? Why do we primarily see neurons modified? How does it spread throughout the brain? We addressed these questions and have had encouraging results concerning the safety and biology of CAV-2 vectors. Convinced that CAV-2 vector deliver to the brain could be therapeutic, we treated mice and dogs that have a rare disease similar to one found in children (Sly Syndrome). Our results showed correction of many of the anomalies associated with the disease in the brain. On the other end of the scale, we also tried to better understand Parkinson's disease, the second most common inherited brain disease in Europe. We created animals with Parkinson's disease and we can now test therapies on these animals.
Project Context and Objectives:
Our objectives were to bring a novel viral vector 'from bench to bedside'. BRAINCAV tackles gene transfer to the brain by turning a virus into a vector capable of helping us understand fundamental neurobiology questions, disease progression and address therapy for global and focal brain disorders affecting children and the aging population. Our multidisciplinary fundamental and applied research project initiated a European-lead breakthrough to the domain of gene therapy for diseases affecting the brain.
Our scientific objectives are distributed into five complementary and intertwined work packages:
-To understand and define the interaction of canine adenovirus serotype 2 (CAV-2) with fluid and cells of the brain and periphery
-To use a transcriptional analysis to better understand the effects of CAV-2 vector interaction with neurons
-To test a therapy for an orphan neurodegenerative diseases commonly affecting the entire brain of children
-To develop genetic, long-lived Parkinson's disease animals and to treat its focal neurodegeneration
-To develop processes to produce and purify vectors that could bring this vector platform to the doorstep of clinical trials
Each viral vector has its specific advantages and drawbacks such as cloning capacity, memory or induced immunity, toxicity, specificity, safety, titre, or efficacy. If one wanted to genetically modify neurons the vector should preferentially transduce these highly differentiated cells, which make up approximately 10% of the primate brain. The advent of helper-dependent (HD) human adenovirus (HAd) vectors in the late 1990's, led to nontoxic, concentrated, homogenous preparations. HD Ad vectors also poorly integrate into the host genome, yet lead to long-term (at least 3 yrs) expression in many tissues in numerous immunologically naïve animals. The significant cloning capacity (greater than 30 kb) also allows one to subclone an entire gene (versus a cDNA) and include numerous transcriptional regulatory elements. These advantages make HD vectors ideal for long-term gene transfer to post-mitotic cells like neurons. In 2004, we demonstrated that CAV-2 vectors have numerous advantages for clinical gene transfer, and more specifically for treatment of neurodegenerative disorders.
Our proposal was designed to critically assess these advantages, the efficacy of the vector and its safety. Some of the principle characteristics and advances of the CAV-2 vector platform are:
-They preferentially transduce neurons in vitro and in vivo in numerous species
-They circumvent the ubiquitous human anti-Ad memory immune response
-They likely have a minimal effect on neurons
-They are capable of long-term expression
With these characteristics in mind, we define some of the areas in which CAV-2 vectors could be improved. We proposed a work plan to advance HD CAV-2 vectors substantially towards clinical trials.
Some of our basic research questions included how these virions interact with brain cells in relation to cell attachment, internalisation, and the genomic response. For example, 'what cells will be transduced and why' was a simple concept for years. Now we realise that the direct interaction of vectors with cells may be a simple pathway in vitro, but this interaction is significantly more complex in vivo. WP1 was designed to address these questions. The interaction with serum and cerebral spinal fluid (CSF) will dictate and explain the ultimate location of injected particles. Then, once the vector attaches to the target cell (neurons for CAV-2 vectors), we need to better understand how the virion used axon transport to reach the neuron's nucleus.
Many viruses have been selected naturally for their ability to make progeny. Setting the cell up for replication, e.g. perturbing genes involved in cell cycle, is something even vectors initiate. The interaction between the cell and virus also leads to the induction of transcription from a subset of endogenous genes. The cell can, in turn, respond to some of these stimuli by e.g. self-destruction (apoptosis) or inducing an immune response. Who wins this tug-of-war depends on numerous factors. This critic part of gene therapy has led us to propose a global transcriptome analysis to understand the vector-cell interaction at a genetic level.
At the applied research level, we used the characteristics of HD vectors to demonstrate proof-of-efficacy in the healthy and diseased brain of rodents, dogs and primates. We chose two models: a rare monogenetic disorder affecting the entire brain and a relatively common brain disorder that affects a pair of nuclei. We treated the neurodegeneration caused by mucopolysaccharidosis type VII, a lysosomal storage disorder, in a 400 mg mouse brain and a approximately 200-fold larger dog brain. Lysosomal storage disorder is a collection of approximately 60 orphan diseases, primarily affecting children. Most of these disorders result from deficiencies in enzymes normally implicated in the catabolism of macromolecules inside lysosomes. No adequate option is available for the treatment of the debilitating symptoms of neurodegeneration, which is a common pathology in the majority of these orphan diseases. Our challenge was distributing the enzyme throughout the brain. Something CAV-2 vectors were particularly efficient at doing.
We also used HD CAV-2 vectors to better understand, and possibly treat, Parkinson's disease by attempting to create genetic models (in contrast to the acute drug-induced models) that mimic disease progression. Parkinson's disease, one of the most common neurodegenerative disorders, is a chronic, progressive neurodegenerative movement disorder. It primarily affects dopaminergic neurons in substantia nigra (SN) leading to the loss of dopamine in their prime target area, the striatum. Disease modifying therapies are not available, partly, because no adequate animal model exists with chronic degeneration of dopaminergic neurons and a robust behavioural phenotype. In WP4 we will also combine the genetic- and drug-induced models to complement our understanding of Parkinson's disease and its potential therapies.
Project Results:
Results/foregrounds
BRAINCAV was divided into five Science And Tecnology (S&T) work packages. The results are presented and discussed in the format of the respective deliverables.
WP1 - The interaction of CAV-2 with fluid and cells of the brain and periphery
WP1 was designed to further understanding CAV-2 biology in the host and assess pertinent interactions of the virus with the host, for future gene therapy at the preclinical and, ultimately, clinical levels. All deliverables were successfully submitted.
Deliverable (D)1.1: Report describing histological analysis of CAR expression in the primate brain. We compared by immunoblotting the level of CAR expression in the adult primate versus embryonic brain. We found that in addition to a reduced level of CAR expression, CAR ran at different molecular weights suggesting a post-translational modification. We extracted protein from different regions of the primate brain and probed for CAR. Variable levels were found, which is consistent with results in the rodent brain as determined by mRNA content and immunoblotting. We used immunofluorescence (IF) to detect CAR expression in the microcebe brain. CAR was preferentially found on neurons in the soma and axons. We also looked for CAR expression at the primate neuromuscular junction. CAR was detected on the pre-synapse side and not on the muscle as previously reported.
Summary: CAR expression in neurons from the primate CNS was determined using biochemical and histological assays. Compared to data from rodents, CAR expression is similar, but not identical. CAR appears to be primarily expressed on neurons in the brain and not on glia. And while the levels decrease in older animals, CAR was still present and its expression pattern was consistent with CAV-2 tropism.
D1.2 To create a CAV-2 vector containing CAR-ablated knob and D1.3: Report comparing CAV-2 and CAV-2 CAR-ablated and in vivo
The CAV-2 fibre knob (FK) mutated in CAR binding site was tested on primary motor neurons. No specific binding could be detected compared to the native FK, suggesting the tropism of CAV-2 was CAR dependent. We therefore set out to generate a CAV-2 CAR-ablated vector to test the hypothesis that CAV-2 was CAR-dependent. We generated a CAR-binding ablated CAV-2 genome with the mutation R384E in the fibre. Homologous recombination was used to generate CAR-binding ablated CAV-2 vectors. With this aim, we generated shuttle plasmids containing the CAV-2 fibre protein. For the generation of the shuttle plasmid, we amplified a DNA fragment containing the CAV-2 fibre protein. We introduced the modification R384E in the shuttle plasmid and finally, we generated the CAR-ablated CAV-2 virus by homologous recombination between the shuttle plasmid (including the R384E mutation) and the CAV-2 genome. Viral genomes were analysed by sequencing and restriction enzymes. While a wild type CAV-2 genome with a mutation in the CAR binding domain is interesting, it could not be used for in vitro or in vivo analysis without the inclusion of a reporter gene and/or a deletion in the E1 region. It was therefore necessary to make an E1-deleted CAV-2 vector (i.e. not virus) containing a reporter gene (e.g. GFP). Two strategies were used to create CAR-ablated CAV-2 vectors. However, the insertion of the mutated fibre sequence into CAV-2 genome was unsuccessful after numerous attempts. We then tried a second approach using bacterial artificial chromosomes (BACs) and bacterial phage recombinases. Again, after several attempts we didn't succeed.
Corrective action: We therefore created an alternative approach by addressing CAR interaction at the surface of the cell - instead of in the virus. To do this we crossed Nestin-Cre mice with mice that had the second CAR exon flanked by loxP site. This approach created mice that lack CAR in the adult mouse brain. We then generate primary culture cultures of neurons and incubated them with HAd5GFP and CAVRFP (DsRed2). We found a significantly reduced transduction efficiency of CAV-2 especially for the transduction of neurons. HAd5 transduction was modestly impaired and cells with glia morphology were transduced. These data suggested that CAR was largely responsible CAV-2 tropism in vitro. We then injected CAVGFP into the CNS of C57BL/6 and CAR CNS-/- mice. We found a significantly reduced transduction at the site of injection in CAR CNS-/- mice of as well as via axonal transport compared to control C57BL/6 mice, again suggesting that CAR plays a major role in the tropism of CAV-2 vectors in vivo.
Summary: Instead of addressing the role of CAR at the level of the vector, we addressed its role using a genetically induced deletion during development. We tested CAR-depleted neurons in vitro and in vivo and found that CAR plays the most significant role in CAV-2 transduction.
D1.4: Report describing axonal CAV-2 transport in neurons
First, the virus binds to CAR at the cell surface followed by internalization in clathrin-coated pits. Panel A upper part shows electron microscopy of CAV-2 in a coated pit; lower part shows colocalisation of CAV-2 and CAR at the surface of motor neurons. During entry, maturation of CAV-2 containing vesicles occurs, with a shift from Rab5 to Rab7 endosomes that will eventually lead to CAV-2 axonal transport. Panel B shows quantification of colocalisation between CAV-2, Rab5 and Rab7 during internalization. Despite Rab7 maturation, the pH of axonal endosomes stays neutral, and CAV-2 remains inside vesicles during transport. Panel C shows electron microscopy of CAV-2 at late stage of endocytosis, when the transport occurs, and live-cell imaging of co-transport of CAV-2 and the endocytic marker dextran. CAV-2 traffic appears to have a bias for the retrograde direction and both dynein and kinesin are involved. Unexpectedly, CAR was also found in CAV-2 positive vesicles.
Summary: The pathway that CAV-2 uses to enter and be transported in neurons was described at the molecular level.
D1.5 CAV-2 vector biodistribution in primates following stereotaxic injection
CAVGFP was injected into the striata of 7 microcebes. 4 animals were sacrificed 5 h postinjection and 3 animals were sacrificed 5 days postinjection. Brain regions and peripheral tissues were analyzed for vector genome biodistribution analysis and GFP mRNA quantification. At the 5 h very few vector genomes were detected in peripheral tissues compared to brain tissues. There was evidence of retrograde transport of CAV-2, with significant quantities of genomes in areas distant from the injection site (frontal cortex, olfactory bulb and diencephalon). Vector genome levels were highest surrounding the injection site. At the 5 days, few genomes were detected in peripheral tissues. Similar to the animals, in particular at the injection site. Vector genomes were also detected in areas distant from the injection site 5 days after administration. qRTPCR quantification of GFP mRNA expression in brain regions 5 days post-stereotaxic injection showed a broadly similar transduction profile to the vector genome distribution profile. To further assess vector genome distribution and transduction by CAV-2 after stereotaxic injection, we injected a two microcebes with CAVGFP. 3D reconstruction of GFP+ cells quantified in distinct regions of the CNS generated a global view of transduction efficacy.
Conclusion: We assessed the genome biodistribution and transduction profiles of vectors after striatal, stereotaxic administration to microcebes. Highest levels of CAV-2 genomes, GFP mRNA and protein expression were observed in and around the injection site, although vector genomes, GFP mRNA and GFP+ cells were also detected in areas distant from the striatum.
D1.6 Proteomic and biochemical characterization of CAV-2 interactions using human blood and CSF
To complement D1.5 we assessed the interaction of CAV-2 with human serum and CSF components, using HAd5 as a control. SPR analysis of blood coagulation factor X (FX) binding to CAV-2 was carried out. HAd5, which binds FX at nanomolar affinity, was used as a positive control. SPR assays showed clear binding of HAd5 but no binding of CAV-2. In vitro binding experiments were then carried out in SNB19 human glioma cells and DKZeo cells to analyse the effect of blood coagulation factors X or IX or heparin on binding of CAV-2 to the cell surface. Neither FX nor FIX had an effect on binding of CAV-2 to the cell surface, although minimal binding inhibition was observed in the presence of heparin. In vitro cell binding experiments using an enzyme that cleaves HSPGs (Heparinase III) showed that heparin inhibition of CAV-2 binding was due to prevention of binding to HSPGs but not to CAR, which appeared unaffected by Heparinase III pretreatment by FACS analysis. Although these data suggested that there may be CAR-independent interactions of CAV-2 with the cell surface, these are likely to be of very low affinity and may be specific to the cell types that were used for the in vitro assays.
HAd5 interacts with several serum proteins. To identify potential CAV-2 interacting partners in pooled human serum or CSF samples, we carried out magnetocapture pull-down assays, using whole, biotinylated CAV-2 particles to co-precipitate interacting serum and CSF proteins prior to separation by 1D gel electrophoresis and identification by tandem mass spectroscopy. We found high-abundance albumin and IgG proteins. To identify potential low-abundance interacting proteins, samples were processed using an albumin and IgG-depletion kit. We analysed serum and CSF proteins magneto-captured using biotinylated virus as bait. As several putative interacting serum proteins for HAd5 have previously been characterised, HAd5 was used as a positive control for the human serum assays. Several high-abundance proteins were identified across all fractions, at satisfactory levels of confidence, including serotransferrin and ?-1-antitrypsin. In addition, known interacting serum protein partners for HAd5 were identified in co-precipitates, including complement proteins (CO3, CO4A) and transthyretin. Interestingly, more proteins were identified in CAV-2 co-precipitates compared to HAd5 co-precipitates. Moreover, CO3 and CO4A also co-precipitated with CAV-2, suggesting that some adenoviral interactions with serum proteins may be conserved across species.
Summary: CAR is likely the primary functional receptor for CAV-2.
WP 2 High-throughput transcriptional analysis of HD CAV-2 effects on brain cells
Objectives: WP2 was designed to further understanding CAV-2 biology in the host and assess interplay of the virus with the host genome. All deliverables were successfully submitted.
D2.1: hNPCs transcript alterations induced by CAV-2, HAd5, AAV9 and HIV-1 vectors
Human neural precursor cells (hNPCs) were used as a in vitro model to test the molecular effect of HD CAV-2 on the neuron. We were interested in the absolute evaluation of gene expression following vector transduction, to identify its molecular toxicity, and in its relative quantification, to establish how CAV-2 performed compared to other viral vector candidates for neuronal therapy. hNPCs were incuabted with the four vectors, i.e. 2 h at an MOI of 1000 vector genomes/cell with CAV-2, HAd5, AAV-9 and HIV-1.
Affymetrix HG U133 gene chip analysis was performed on RNAs extracted from infected and mock samples at 2 h and 5 days post infection. CAV-2, HAd5, AAV-9 and HIV-1 all modulated the 2D neuronal transcriptome, at 2 h and 5 days posttransduction. To interpret the significance of the modulation, genes that were significantly altered were analyzed with Ingenuity software. The most modulated molecular and cellular functions induced by CAV-2 were evident at 5 days postinfection and include cell cycle, cellular assembly, replication and recombination modulations. Elaboration of IPA data indicated that three major cellular pathways were variably affected: the immune response, the DNA-damage response and the pathways involved in cell trafficking. Particularly notable is the difference between CAV-2 and HIV-1 vectors for what concerns the induced modulation of the DNA damage pathway. The transcriptome variations include 182 commonly modulated genes, with 14% of the genes belonging to the DNA damage pathways modulated in opposite directions. To validate the biological information based on the extended chip analysis we performed qPCR on 24 genes and evaluated the Pearson correlation value (0.72). This indicated the reliability of chip data for the biological evaluation of neuronal response to CAV-2 and to the other viral vectors.
Summary: CAV-2, HAd5, AAV-9 and HIV-1 affect the neuron, independently from the efficiency of GFP expression. All vectors affect immune response, DNA damage and cell trafficking pathways. Particularly evident was the contrasting modulation of the DNA damage response by HIV-1 and CAV-2 vectors.
D2.2: Gene expression alterations induced by CAV-2, HAd5, AAV-9 and HIV-1 vectors in 3D cultures of hNPCs
To test the variations of the CAV-2 effects in 3D neuronal cultures, neurospheres of hNPCs and CAV-2 vector infection in that system were established. A full description of the methodology was published (Brito et al Methods 2012).
Differentiated neurospheres were incubated with CAV-2, HAd5, AAV-9 and HIV-1 at the fixed conditions applied to 2D cultures and described in D2.1. The relative efficiency of transduction in 3D cultures was maintained for CAV-2, while reduced for the other vectors, suggesting that CAV-2 can possibly penetrated the neurospheres more efficiently than other viral particles. RNA was extracted from neurospheres and modulation events were analyzed both by gene chip and by qPCR. After RMA normalization and T-test comparisons (p less than 0.05) 331 and 127 genes were found significantly modulated, respectively at 2 h and 5 days postinfection. Genes were submitted to IPA and the most significantly modulated pathways reported. Compared to the 2D system, CAV-2 effect on neurospheres includes the modulation of the DNA damage response pathway, and has a milder effect on the immune system and a more significant effect on the trafficking pathways. Particularly significant of this last aspect is the modulation of CD44, which has been suggested to contribute to HAd5 entry into cells and that of caveolins.
To evaluate the relative effects of different vectors on neurospheres, specific gene modulations were quantified by qPCR. Concerning the immune system, the analysed genes were not significantly modulated, except for TLR3 and TLR4 at 2 h. Interestingly, DNA damage genes, including RAD51 and BIRC5 have opposite modulation profiles in CAV-2 and HIV-1 transduced neurospheres. This is consistent with the different genome structure, and genome related DNA damage response of the two viruses and with the data obtained in 2D neuronal cultures.
Summary: The 3D system was a valuable tool for the analysis of neuronal function. The analysis of infected neurospheres confirmed specific modulation events observed in the 2D system, including DNA damage response activation. The efficiency of CAV-2 transduction in 3D neurons and the complementary modulation of genes involved in cell trafficking, suggest that this culture method should be exploited for further studying the neuronal trafficking processes following pathogen infections.
D2.3: Report of gene expression alterations induced by CAV-2, HAd5, AAV-5 and HIV-1 vectors in human glia and neurons
A notable effect of CAV-2 on neuronal cultures is the modulation of the DNA damage pathway. We analysed the genes modulated by CAV-2: BIRC5, BRCA2, FANCD2, MRE11, CHEK1, FAS, p53 and RAD51. The induction of RAD51, MAD2L1, BIRC5 and of FANCD2 was highlighted also by qPCR. Together, these data reinforce the principle of the significance of this pathway in the response to Ads, and suggest that, in a clinical perspective, it will have to be evaluated whether and to what extent viral vectors induce, through the activation of the DNA damage response, irreversible cellular toxicity. An aspect that was relevant in the analysis of vector toxicity in a clinical perspective is the evaluation of the immune response. This analysis is also expected to contribute to studies concerning the biology of neurons whose immune function is currently being determined. We thus evaluated the genes modulated by the four vectors in 2D neuronal cultures and identified commonly modulated probes like TLR3 and TLR4, CD44, CASP1 and HAS3. To support the reliability of the results obtained through chip analysis TLR3, TLR4, HAS3, IRF3 and CD44 modulation was quantified by qPCR. These genes were induced by CAV-2 infection suggesting that neurons can activate a response to pathogens, and this appears to occur even when the pathogen, as CAV-2, was attenuated. To evaluate the response of the second component of neurological tissue to viral vectors, we set up differentiation and infection conditions of glial cells. Consistent with poor CAV-2 transduction of the glia, qPCR analysis of single gene modulation showed that this cell type is not significantly affected by CAV-2 vectors under these conditions.
Summary: A detailed analysis of the transcripts modulated by CAV-2 in neurons showed, in the cluster of immune response, the upregulation of TLR 3 and TLR4 at the late time point, along with HAS3 (the enzyme involved in the synthesis of hyaluronan (HA)), CD44, (the HA receptor) and of IRF3. Taken together, these data suggest that CAV-2 induces the activation of gene program of antiviral innate immune response. CAV-2 induces as well a DNA damage pathway with the specific modulation of genes involved in ATM signalling (p21, BIRC5, MAD2L1) and in homologous recombination (FANCD2, RAD51). Glial cells are poorly transduced by CAV-2; consistently, the level of transcript modulation induced by CAV-2 was modest in these cells.
D2.4 Gene expression alterations induced in the primate brain
CAV-2 reaches the cell nucleus via retrograde axonal transport. To better understand the effects induced by the presence of HD-GFP in vivo, a transcriptomic study was performed in the brain of Microcebus murinus.
CAV-2 was introduced in the right caudate nucleus of 8 females Microcebus murinus. 2 periods were investigated for the transcriptome analyse: an acute (24 h) (4 animals) and a delayed time (28 days) postinjection (4 animals). The left caudate nucleus received the PBS vehicle. For controls, 4 animals were naïve for the virus. As the striatum receives synaptic terminals from different brain structures (cortex, SN, thalamus, contro-lateral side), we assayed changes in gene expression induced by HD-GFP in the striatum and in the frontal cortex. We compared the gene expression in the ipsilateral side with the controlateral side, and with non-injected brain samples using transcriptomic approach. Total RNA of each sample, extracted from the striatum and from the frontal cortex samples was prepared for hybridization with Affymetrix HG-133 Plus2 GeneChip. The data were normalized, converted digital intensity values were converted into cell intensity files using the Microarray Affymetrix Software 5.0 (MAS 5). MAS5 labels each transcript as 'present' (P) or 'absent' (A) by taking into account the average difference value calculated between the perfect match (PM) and the mismatch (MM) as detailed in the Affymetrix Genechip procedure (see http://www.Affymetrix.com online). A p-value was associated with the signal intensity. For each gene, the mean signal was calculated for each group. The overall rate of genes with the 'present' detection call was about 18%. Moreover, we noticed a lowest SD (6%) in all the samples studied suggesting a homogeneity of the gene expression. The procedure is described below. We detected 14,400 transcripts to analyze.
We used the significance analysis of microarray (SAM) to identify genes that were differentially expressed when comparing for a brain structure, the three groups together, right side versus left side versus control or when comparing the time after virus infection. Among the genes that presented changes in their expression, the most notable belonged to transcriptional regulation function, the cyotskeleton and synaptic transmission. In the striatum, these genes involved in immune response were detected (e.g. an overexpression for TNFRSF18 at 24 h and for HLAG, IGHM, IGLC2 at 28 days).
Summary: Our results showed that the injection of CAV-2 induced changes in gene expression and that the expression differed. In the acute phase, HD-GFP altered the expression of protein regulating the transcription but also protein involved in the release of neurotransmitter at the presynapse. These gene expression changes could transiently modify the synaptic activity of the neurons. In the chronic phase, CAV-2 induced modifications in some transcripts implicated in the cytoskeleton suggesting a minor modification in cell trafficking. The presence of the HD-GFP-induced an increase of immunoglobulin transcripts suggesting a mild adaptive immune response in the brain.
D2.5: HD CAV-2 vector alterations of gene expression in vitro and in vivo
Neuronal 2D and 3D cell cultures were analysed for the common genes and pathways influenced by HD-GFP transduction. As anticipated in the previous WP2 deliverables, the cell damage/cell cycle and immune response pathways were modulated in both systems. The cell trafficking functions, conversely, were mainly modulated by HD-GFP infection in the 3D neurospheres. When considering the significantly regulated genes together (2D and 3D neuronal results at 5 days post-transduction), 54 probes were commonly modulated in the two systems, corresponding, respectively, to 7.7% of the genes modulated in the 2D system and to 42.5% of the 3D system. The commonly modulated genes analysed and showed that the cell cycle/ DNA damage and cell morphology functional categories were modulated in the 2D and 3D neuronal systems in the same direction (mostly, probe upregulation).
We also studied HD-GFP effect on the neuron by comparing the in vitro (2D and 3D neuronal cultures) and in vivo (Microcebus murinus brain) data. An independent statistical condition applied for the analyses (detailed in the corresponding WP2 deliverable sections), we found no common genes. Given this result, to further develop the comparative study between the in vitro and in vivo systems, we designed a meta-analysis plan. This workflow was applied to the 2D, 3D data and to the Microcebus murinus brain derived chip results and contributed to further define the HD CAV-2 induced signature in primate neurons.
Summary: Taking into account the 2D, 3D and primate brain, we concluded that, at the functional level, HD CAV-2 affected the immune system, the cell cycle/DNA damage response (in 2D and 3D neurospheres), and intracellular trafficking (3D neurospheres). At the single gene level, 54 probes were commonly modulated by HD-GFP between the 2D, 3D, and none among 2D, 3D and primate brain. A meta-analysis based on a common statistical approach for 2D, 3D neurospheres and primate brain was developed.
WP 3 Treating neurodegeneration caused by MPS VII in mice and dogs
Objectives: Assessment of HD-RIGIE efficacy, toxicity and biodistribution in the β-glucuronidase (GUSB)-deficient mouse and dog models.
D3.1: Optimized HD-GUSB dose in the MPS VII mouse brain
The CNS and the skull of MPSVII mice are morphologically different from wild type littermates. We adjusted the injection point to obtain the maximum global transduction. For that reason, CAVGFP was injected into the striatum of 8-weeks old MPS VII mice at the following coordinates: 0,2 mm from bregma; 2 mm lateral; 3.5 mm deep. Results show GFP expression in many cells in the striatum and also in the dopaminergic neurons in the SN. No significant transduction was seen in other areas of the brain. A different cohort of animals was injected more rostrally, at +1,0 mm from bregma, 1,8 mm lateral and 3,5 mm deep . With these coordinates, we detected GFP expression in the striatum, where the virus were injected, and in the SN, corpus callosum and cortex, where we were able to detect transduction in neurons and endothelial cells. Endothelial cells, which overexpress CAR, could have been reached through circulation from the CSF or through microvessels present at the injection point.
Summary: We defined the following coordinates +1.0 mm from bregma, 1.5 mm lateral and 3.0 mm in depth as those that allow for more global transduction of MPSVII mouse brain.
D3.2: Optimized HD-GUSB injection site in the MPS VII mouse brain and D3.3: HD-GUSB effect based on histological analysis of the murine MPS VII brain
A group of 8 transiently immunosupressed animals were injected bilaterally at approximately 9 weeks of age with 2 x 109 pp of HD-RIGIE/striatum. Control littermates (heterozygous and mutant mice) were injected with the same volume (2 μl per hemisphere) of PBS. Mice were sacrificed at 6 weeks and GUSB activity was assayed every 100 nm. No GUSB activity was detected in PBS-treated mutant or even heterozygous mice, showing the low sensitivity of the assay. We detected transgene expression in striatum, several areas of the cortex, corpus callosum, SN and around ventricles. Overall, based on this assay, we estimated a volume transduction about 4 mm around the injection point.
Conclusions:
1. Brain transduction using HD-RIGIE vector is able to drive transgene expression to the whole brain of GUSB mice, 6 and 16 weeks after one single injection of the vector per hemisphere.
2. One injection of HD-RIGIE allowed for short and long-term ß-glu expression up to 60% of heterozygous mice and, more importantly, to complete biochemical recovery of secondary elevation of ß-hex and correction of GAGs in 4 out of the 6 slides of the brain, covering the whole cerebrum with the exception of the cerebellum and the brainstem.
3. Correlating with correction of biochemical parameters, ß-glu expression driven by HD-RIGIE lead to short and long-term clearance of lysosomal storage accumulation at the injection point but also in many distal areas.
D3.4: Behavior modifications following HD-GUSB injection in the murine MPS VII brain
Animals were behaviorally evaluated 6 weeks after treatment in a series of tests used as a screening for behavioral abnormalities and physical condition in mutant mice (Giménez-Llort et al., Eur. J Neurosci. 16: 547550, 2002). The behavioral domains analyzed included physical condition, sensorimotor functions, behavioral and psychological symptoms (anxious-like behavior, exploratory activity, daily life activities) and cognitive function. Experiments were conducted between 10:00 and 17:00 h. In all experiments animals were counterbalanced per genotype and gender. The research was performed in accordance with 86/609/EEC regarding the care and use of animals for experimental procedures. Animals were weighed before each test.
MPSVII mice show poorest physical condition as compared to wt animals, which was improved by HDCAV2GUSB-treatment. Poorest daily life activities were shown in the nesting behavior and seemed to be also improved in HDCAV2GUSB-treated animals. The assessment of animals in an abbreviated version of SHIRPA, wood and wire rod tests and hanger tests evidenced impairment of sensorimotor functions. MPSVII animals showed lower reactivity to handling as well as poorest prehensility in the use of elements of support and strength. Visual and forelimb reflexes, motor coordination and equilibrium were normal.
Summary: MPSVII animals showed impairment in all the domains studied with main affectation on cognition and HDCAV2GUSB was able to successfully improve and/or reverse the major deficits including total learning and memory capacities.
D3.5: HD-GUSB effect based on histological analysis in the dog MPS VII brain
The brains of seventeen dogs were injected with HD CAV-2 vectors. The initial cohort was injected with HD-GFP to determine efficacy, level of retrograde transport, immunogenicity, neuropathology, and optimize doses and coordinates in the healthy and MPS VII dog brains. Initially, stereotactic injection in dogs, and particularly in MPS VII dogs, was challenging because of the variation in the size of pups, the lack of a detailed atlas for 2 to 3-month-old mixed breed dogs, and MPS-associated bone deformation that did not allow a trustworthy location of the bregma or brain structures. Because MRI facilities were not available to determine stereotactic coordinates, injections were targeted to large structures and, when relevant, multiple coordinates along the needle track. The above constraints certainly led to vector delivery into ventricles in some MPS VII dogs, which in this case would lead to loss of extra-chromosomal vector genomes in dividing ependymal cells. At the site of striatal injections, cells with quintessential neuron morphology were predominant. We detected GFP+ cells throughout each hemisphere in particular on the striatum; thalamus; hippocampus; frontal, temporal, occipital cortex and nuclei of Meynert. In dogs injected unilaterally, we detected GFP+ cells throughout the ipsilateral hemispheres and in the contralateral cortex, which has efferent projections to the injected striatum. To distinguish between neurons and glia we stained sections with anti-NeuN and anti-GFAP antibodies. In afferent regions all GFP+ cells were neurons, while at the site of injection the majority of GFP+ cells also expressed NeuN. Combine, these data demonstrated that CAV-2 vectors preferentially transduced neurons in the large dog brain and are capable of efficient axonal transport, which is consistent with the tropism of CAV-2 in rodent brains and in organotypic cultures of human brain biopsies.
Conclusion: we assayed the efficacy of a HD CAV-2 vector containing a human GUSB expression cassette in the canine young MPS VII brain. As little as approximately 2 x 10 8 infectious particles of HD-RIGIE led to histological improvement throughout the MPS VII dog brain. A striking result was the comparable efficacy of HD-RIGIE: the dose injected into the MPS VII dogs was 50 to 200-fold lower than that used for AAV vectors in the brains of Sanfilippo and Hurler syndrome dogs. To put this dose into a more global gene transfer perspective, it would not be surprising to see 8 x 109 physical particles of a HD vector used in the rodent brain. Equally clinically relevant, and in contrast to studies using AAV vectors for MPS therapy, only transient immunosuppression appeared necessary and we found no obvious reduction in efficacy in the MPS VII dogs that were sacrificed 4 versus 1 month postinjection.
D3.6: Behavior modifications following HD-GUSB injection in the dog MPS VII brain
Due to the severity of the joint and cardiac problem associated with MPS VII in these dogs we were unable to assay behaviour modifications associated with brain gene transfer. Eventually this could have been overcome with glu enzyme replacement therapy, but the costs, limited supply of enzyme and minimum time postinjection did not allow this.
WP4: Developing novel genetic Parkinson's disease models and treating neuron death
Objectives:
i) characterize the biological functions of LRRK2;
ii) establish a high-content cell-based array for drug development;
iii) develop models for Parkinson's disease using ectopic expression of wild type and mutant LRRK2; and
iv) test the efficacy of HD CAV-2-mediated delivery of GDNF for neuronal protection against acute and genetically-induced parkinsonism.
D4.1 Report describing HD-LRRK2 vector efficacy in hmNSCs
In the CNS of rodents, dogs and primates, CAV-2 vectors preferentially transduce neurons when injected into the parenchyma. Whether CAV-2 vectors also transduce precursors of human brain cells that maintain the potential to differentiate into neurons or glia was unknown. To address this, we assayed human neural stem cells (hNSCs), which can be derived from specific brain regions and give rise to specific neurons such as dopaminergic neurons, when tissue is derived from fetal midbrain (hmNSCs).
Summary: While HD-LRRK2 transduced MDCK cells efficiently, hmNSCs were transduced with lower efficacy.
D4.2 Report describing the effect of HD-LRRK2 on proliferation and differentiation in hmNSCs
Vector dose of HD-CAV-2 expressing the mutated LRRK2 G2019S gene (HD-CAV-LRRK2) was studied in undifferentiated and differentiated hmNSCs cultures, in comparison to the mock vector HD-GFP. For initial screen of transduction parameters, HD-GFP was used. Different transduction settings were assessed in undifferentiated and differentiated cultures, by combining MOIs and transduction time.
Summary: Overall LRRK2G2019S did not induce significant specific alterations in proliferation apoptotic and neuronal markers in our cellular model.
D4.3: Report describing the cell based HCS system using LRRK2 expression in hmNSCs
Partner 8 did not perform these assays.
D4.4 Report describing HD-LRRK2∗ efficacy in the rat brain
The striatum of 16 rats were injected unilaterally at 2 coordinates with HD-LRRK2 and HD-LRRK2 G2019S. The controls were sham and HD-GFP injected (n = 8).
After the baseline of each individual animal was measured 3 days pre-injection by the amphetamine-induced rotation test, the rats were retested 1-month post-injection vector. We found no significant difference in the rotational behavior depending on the injected virus. At similar times, motor learning was tested using the cylinder test. Again there was no significant difference between groups. Rats were sacrificed 1-month postinjection for histological analyses. Tyrosine hydroxylase (TH)-immunoreactivity in the striatum as a measurement of dopaminergic terminal integrity was not reduced in virus injected animals. The number of TH-positive cells in the SN was also not affected. In addition, there was little GFP fluorescence in injected striata and none in the SN. Thus, we concluded that the amount of injected particles was not sufficient to efficiently transduce dopaminergic neurons in the SN and also that there was no behavioral or histological effect of such injections.
Summary: 16 rats were injected with LRRK2 containing vectors plus controls. We found no changes in the brain. Proposed corrective actions (intranigral injections, inject more particles and/or transiently immunosuppress) were not performed by partner 8.
D4.5: Report describing HD-LRRK2∗ vector efficacy in microcebes
The task was separate into different experiments
1st year: we developed behavioural tests after the injection of 6-OHDA:
1) hourglass: The animal is placed in a cylinder that is turned upside down. We measure the time taken by the microcebe to return to the upright position. Usually, unimpaired animals do it in less than 3 seconds;
2) tower with 7 levels of rungs where the microcebes have to jump from the ground to the 7th level, and where the difficulty increases progressively
3) reward searching in one of 6 aligned tubes; it consisted in reward searching in one of 6 aligned tubes. The animals have to jump on a small platform to reach the reward. We measure the time to catch the reward during 30 trials and the hand used.
4) hill and valley staircase task.
2nd year: we performed pilot studies with HD-LRRK2∗ (G2019S) vectors injected in the caudate putamen of two microcebes (4.5 x 109pp, 3µl injected, coordinates A:5 mm, L: 2mm, P: 8mm, 5 mm, 4mm). The microcebes were sacrificed 15 days after injection. We then investigated the loss of dopaminergic neurons in the SN. We observed a loss of dopaminergic neurons of 24% and 42%. The behavioural study showed that we need to train the animals before injection and follow up them during all the incubation time.
3rd - 4th year: We injected unilaterally at 2 coordinates with 2 x 109 physical particles of HD-LRRK2 and HD-LRRK2∗ G2019S (controls: 6 lemurs HD-GFP, 4 lemurs HD-LRRK2, 8 lemurs HD-LRRK2∗ injected). Prior to vector application, all the animals were trained to perform motor tasks. Then the mouse lemurs were followed up with behavioural tests during 6 months post-injection at the end of the follow up, the animals were sacrificed and the brains were processed for IHC studies.
Behavioral studies
Hourglass test:
G2019S LRRK2 lemurs are considerably impaired from 1 week to 3 weeks after injection in the hourglass test compared to performance before injection and this deficit is observed during the 6 month follow up. The G2019 lemurs did never reach again the baseline (before injection). For wild type LRRK2, the impairment occurs 3 weeks after injection, the animals recover good performances compared to the baseline. Control animals keep baseline performance during the follow up.
Tower test
Control animals show constant performance along the follow up meaning that animals did not encounter any difficulty to jump. G2019S lemurs (particularly 3/8 mouse lemurs) present an important impairment 3 weeks after the injection then recover progressively during 6 weeks and show again motor deficits along the follow up. For wild type animals, important motor deficits are observed 4 weeks after injection until 18 weeks and then the animals are less impaired but never reach again the baseline.
Post-mortem studies
Histological studies were done at the end of the 6 month-follow-up. It is important to note that the behavioural impairment was not maximal at the end of protocol.
Detection of protein expression with immunohistochemistry
We have performed different labelling with various antibodies: anti-LRRK2, anti-Flag, anti-GFP to detect HD-CAV2 transduce neurones, anti-TH to detect dopaminergic neurones, anti-GFAP to detect astrocytic inflammation.
Dosage of dopamine into the striatum of 4 control HD-GFP and 4 HD-LRRK2∗ injected animals
Lemurs brains were rapidly removed and tissue dosages of monoamines (DA, 5-HT) and their metabolites (DOPAC, HVA, 5-HIAA) were performed.
The injection into the striatum of 2 x 109 physical particles of HD-LRRK2 and HD-GFP have clearly an impact of the dopamine level of the striatum whatever the vector injected. The dopamine content is not significantly changed depending on the virus injected into the striatum. However the injection into the ipsilateral side in the G2019S lemurs involved a decrease of the dopamine level in the contralateral side. This decrease is not observed in the control animal suggesting that it is an effect of the LRRK2 expression.
Level of dopamine into the SN of control, WT, and LRRK2∗ lemurs
We have done mapping of TH-immunoreactive neurons into the SN of 4 controls, 4 Wild type and 8 LRRK2∗ (G2019S) and counted the number with the software Mercator (Explora Nova).
We observed a weak loss of TH neurons in the SN in the ipsilateral side whatever the virus injected. But we observed also a decrease in the contralateral side in LRRK2WT and LRRK2∗ compared to control suggesting a role of the overexpression of the LRRK2 gene. This is in agreement with the results observed into the striatum. We need to investigate the number of neurons into the SN in non-injected lemurs to confirm this result.
Morphology of the dopaminergic neurons in the SN
We detected a loss of dopaminergic fibers in G2019S animals compared to control and wild type animals. We observed dystrophic neurites and swollen axons, characteristic of neurodegeneration.
Summary: The injection of HD-CAV-LRRK2∗ into the striatum of mouse lemurs leads to transient motor deficits in prehension, postural stability and rigidity of the body. The loss of dopaminergic neurons is weak but we observe a clear decrease of the number of fibers and apparition of swollen axons and dystrophic neurons, characteristic of neurodegeneration. We observe also a glial activation on the ipsilateral SN. 6 months postinjection into the striatum, this suggested that the HD-LRRK2∗ lemurs present early signs of PD.
D4.6: Report describing HD-LRRK2∗ vector efficacy in Macaca fascicularsi
Injection of HD-LRRK2∗
We performed a pilot study in two Macaques to determine the distribution and axonal transport of CAV-2 into the monkey brain.
Stereotaxic delivery of CAVGFP and CAVCherry was performed into the caudate nucleus, putamen and thalamus of 2 intact animals. Convection enhanced delivery (CED) was used for putamen and thalamus injection while the injection into the caudate nucleus was undertaken by conventional method. GFP and Cherry expression were found throughout the brain. We observed no obvious advantage for CED versus conventional delivery.
Summary:
-CAV-2 was efficiently transported in a retrograde manner in the monkey brain and it showed a neuronal selectivity
-CAV-2 delivery into the brain induced the overexpression of GFP and LRRK2∗ proteins in brain areas projecting to striatum
-HD-LRRK2∗ injection into the putamen induced a loss of striatal 11C-DTBZ uptake and dopaminergic cell loss (VMAT-ir cells) in the homolateral SN
-HD-LRRK2∗ injections into the putamen induced overexpression of LRRK2∗ into the SN cells and provoked damage of the axons and neurites but no parkinsonism
-HD-LRRK2∗ injection into the striatum induced increased levels of phosphorylated Tau protein in several brain structures of the homolateral hemisphere.
-Striatal injections HD-LRRK2∗ induced gliosis and microglial activation in striatum and SN even 6 months after vector injection
-HD-LRRK2 and HD-GFP injection into the SN does not induce any behavioral manifestation 12 months after the surgery
-The decreased density of dopaminergic nigral cells observed after HD-LRRK2 and CAV-2 GFP delivery into the SN may be related to the vector injection itself
WP 5: Production and purification of HD CAV-2 vectors
D5.1 MDCK-E1-Cre recombinase cell line
An MDCK-E1-Cre recombinase cell line for the amplification of CAV-2 derived vectors was established from MDCK from ECACCC (No 84121903, lot 05G029). The cells were maintained and genetically modified in MEM 10% (v/v) FBS, full traceability records of the animal origin reagents used were retained.
Starting with MDCK cells, MDCK-E1-Cre cells were created in a 2-step procedure to constitutively express the CAV-2 E1 region and Cre recombinase. In step 1, MDCK cells were transfected with a plasmid called pCI-neoE1K9 provided by Partner 1-CNRS. pCI-neoE1K9 contains a CAV-2 E1 region expression cassette and a neomycin resistance gene. These cells (MDCK-E1) were selected with G418 and cloned by limiting dilution. E1 expression was functionally screened by the clone's ability to amplify E1-deleted CAV-2 vectors. A high E1-expressing MDCK-E1 clone was then transfected, in step 2, with a plasmid called pZeoCre (provided by Partner 1-CNRS), which expresses an nls-Cre recombinase and Zeocin resistance genes.
MDCK-E1-Cre cells are available and will be tested for production of HD CAV-2 vectors (currently under development by consortium partners CNRS and UAB).
D5.2: Report describing set-up of differential packaging system
The goal of D5.2 was to generate a CAV-2 vector that contained one of the phiC31 recombinase recognition sequences (attB) near the packaging signal and determine if this sequence, as in the case of human adenovirus serotype 5, delayed the production of mature CAV-2 particles.
CNRS designed a 1 kb sequence that contained the inverted terminal repeat (ITR), lox61, attB, a mutated packaging signal (based on CAV4.19 Soudais et al Mol Ther 2001) and lox67. The lox61 and lox67 sequences were chosen to avoid reinsertion of the excised floxed sequence into CAVattBCherry during the final step of HD (HD) vector production This sequence was synthesized, and sub-cloned upstream to a Cherry expression cassette in pTCAV-12VK. The resulting construct, pTCAVattBCherry was recombined with pTG5412, which contains the entire CAV-2 genome, to create a plasmid (pCAVattBCherry) that harboured the E1-deleted vector genome of the CAVattBCherry. CAVattBCherry was removed from the plasmid and transfected into DKZeo cells top produce the vector.
During the amplification it became obvious that CAVattBCherry had a delayed production of mature particles: the amplification cycles were optimal after collecting the cells 4-5 days post-infection instead of the normal 36 hours cycle for CAV-2. CAVattBCherry amplification was not only slower, but also less efficient, likely due to the incorporation of the mutated packaging signal. These characteristics are difficult for production, but optimal for CAVattBCherry use as a helper vector for HD vector production. Post-vector purification, CNRS and UAB tested the packaging delay using CsCl purified vectors via FACS and limiting dilution assays. Both assays showed that CAVattBCherry has a delayed packaging. Restriction digestion of vector DNA purified from infected DKZeo and DKCre cells showed that the genome is also sensitive to Cre recombinase removal of the packaging domain.
Conclusion: D5.2 completed successfully and BRAINCAV has generated a CAV-2 helper vector with characteristics that are optimal for HD vector production.
D5.3 Optimized bioreactor protocol
This technical deliverable described the optimized procedure for the cultivation of MDCK E1 cells adapted to serum-free medium and the production of CAV-2 vectors using a stirred tank bioreactor. The protocol also supports the procedure for production of HD (HD) CAV-2 vectors.
D5.5 Helper and HD master virus bank and GMP master cell bank
Helper and HD virus banks were manufactured in April 2011 and May 2012, respectively. ?E1 helper vector, denominated as JB?5, contains loxP sites flanking packaging domains and a RSV-lacZ expression cassette. For JB?5 master virus bank, DKZeo cells maintained in DMEM supplemented with 10 % (v/v) fetal bovine serum and 1% (v/v) of non-essential amino acids were infected with MOI 5. Viral vectors were harvested 48 hpi and purified by CsCl2 gradients. Thirty vials, 500 μl each, of Helper vector JB?5 in phosphate buffered saline with 10% (v/v) glycerol were produced and stored at -85ºC in the premises of partner 4. HD vector denominated as HD-GFP, contains an GFP expression cassette. HD-GFP bank was prepared after infecting DKCre cells in same culture medium using an MOI ratio between HD-GFP and JB?5 of 5 to 1.
GMP master MDCKE1Cre cell bank:
A GMP-certified Master Cell Bank (MCB) for MDCK-E1 Cre was manufactured in February 2012 using the procedures described in the Batch Production Record PF-012. The manufactured MCB batch number is 00412.
The QC tests used for the approval and release of the GMP-certified MDCK-E1 Cre Master Cell Bank were defined according to the European Pharmacopoeia. 100 vials with, 1 mL each, of the GMP-certified MDCK-E1 Cre Master Cell Bank were produced. They are stored in the vapor phase of liquid nitrogen. Process details are described in the Batch production record (PF-012).
WP 0: Management
Consortium management tasks and achievements
The Project Manager assumed the follow-up of the Grant Agreement signature between the EC and the coordinator CNRS, providing all necessary information/documents to both parties. Initially scheduled on September 2008, the GA was finally signed December 18, 2008. (With the authorization of the EC Project Officer, the project had started October 1st, 2008.) The manager also did a follow up of the GA accession forms to be established by the beneficiaries of the project. The EC funds were transferred to the coordinator end January 2009, and the Project Management team (Coordinator + Manager) processed with the administrative and financial follow-up of the distribution of the EC grant to the beneficiaries. All funds were distributed by beginning February 2009.
Report preparation
PMT has been heavily involved in the preparation of the 3 periodic and the final reports. The Manager provided the templates and sending reminders to the partners. Together with the Coordinator, all elements put on the online reporting tools set up by the European Commission (SESAM and FORCE) were controlled. This has lead so far to a limited number of comments from the European Commission because the major issues were dealt with before submission of the report. In a second step, the PMT ensured that the comments and remarks of the BRAINCAV Scientific Officer were addressed in the best delays.
Grant Agreement amendment
BRAINCAV had to request 4 GA amendments to take in consideration the changes in the consortium. This included the following changes:
-As the coordinator CNRS actually employed personnel from Inserm, it was suggested to adjust the previous budget by diminishing the Inserm budget and increasing the CNRS budget. The manager prepared an amendment that was agreed by the EC on April 2009. Beneficiary 2 (CRUK) asked to receive the whole EC contribution at once, as its work was concentrated on the first period. After formal agreement by the EC project Officer, it was agreed to do so.
-Some beneficiaries to the project disagreed to provide the Project Manager with all information regarding the IP rules internal to their institutions as the deliverable was set as 'PU' (public) when it was seen as strategic. The consortium then agreed to mark the deliverable as 'RE' (restricted access) or 'CO' (confidential).
-Name of Legal Entity Authorized Representative of CNRS has changed from M. Michel Portefaix to Mrs Ghislaine Gibello, since January 1st, 2009.
-To address technically difficult experiments, budget changes were agreed upon between the partners during the ExB meetings and validated by the partners:
-P5, who was responsible for mid-size HD vector production, has not been able to produce HD vectors. To date, P1 is the only partner that can make HD vectors, but did not budget resources to produce vectors for 4 years. After discussing possible options, P5 proposed to reallocate some of its EC contribution to the P1 so that the P1 can continue the production of the vectors. P5 sent an evaluation of the amount that could be reallocate to the P1. As a consequence, P5 EC contribution will be reduced by 17,680 Euros in favour of the P1; the initial EC request will not be otherwise modified. The P1 notes that this sum will not cover any related functioning costs and that it has already devoted extra resources to HD vector production during the first 30 months. At the M36 ExB meeting, P6 also offered to redistribute 20,000 Euros to P1 to ensure the production of the vectors. For similar reasons, the P8 will evaluate if a redistribution of its funding would be appropriate to support this production. The ExB members validated the proposed budget redistributions.
-Significant delays took place in the animal studies as explained in the RTD WP3 and 4 progress reports. This created an insurmountable problem into the ability to deliver on time. In spite of the alternatives explored to maximize the possibilities to stay within the limits of the initial work plan, it was not possible to deliver within 48 months. This was also foreseen as a major worry in the initial proposal. The animal studies were completed thanks to the 6-month extension of the GA. This lead to a 6-month extension, postponing the end of the project to month 54
-Departure of P8-ULEI from the Consortium
-Change in the Coordinator's contact
-Addition of the ICREA third party to P5-UAB
-Several limited changes to the Annexe 1 including the budget redistribution to take into account the departure of P8 and the redistribution of some tasks this partner was taking care of.
All these elements were changed in 4 GA amendments which were prepared for the coordinator by the manager and submitted by the coordinator.
Deviations from the original work plan and proposed corrective actions:
The main difficulties that occurred during the course of the project were solved or addressed by appropriate corrective action. The main issues were linked to the difficulties of working with MPS VII dogs and non-human primates. In addition, we had to deal with the departure of P8, who was WP4 leader and responsible for several tasks in WP4. P8 left the Consortium during the 3rd period and had to be replaced in order to ensure that the work could be finalized in the best conditions by the other BRAINCAV participants. Tasks and budget had to be redistributed to ensure that P8 work could be finished in the best conditions and full respect of the initial workplan.
Potential Impact:
WP6 (WP leader EJ Kremer CNRS) addressed the need to inform scientists, health care professionals, the general public, patient associations, and families about the initiative of the European Union via BRAINCAV.
Health care professionals:
Part-1 CNRS recently presented BRAINCAV'S goals to the University Hospital in Montpellier on period 1.
General public:
Part-1 participated in British Society of Gene Therapy public engagement day (March 29th 2010) to talk about BRAINCAV'S goals in for brain therapies. Part-4-IBET participated in a public engagement day at ITQB.
Patient associations: Part-1 is in regular contact with the French lysosomal storage disorders society (VML) to communicate BRAINCAV'S goal to its members. Part-5 is in regular contact with the Spanish MPS society concerning three MPS VII children in Spain.
Expected results, potential impact and use
Due PD's prevalence, the greatest socio-economic impact BRAINCAV could have is for the understanding and treatment of Parkinson's disease. Over 1 million Europeans suffer from Parkinson's disease with 40% of the victims under the age of 60. Parkinson's disease costs the EU over 50 billion Euros annually in health care, disability, loss of productivity, and long-term care - a cost that will only increase as the population ages. At present approximately 1% of the population greater than 50 years old has Parkinson's disease. This number will increase in concert with the number of older people. Since the submission of BRAINCAV, the use of CAV-2 vectors in the context of PD (fundamental and clinical applications) has attracted significant interest from the scientific community. Numerous colleagues have contacted BRAINCAV members. Recently, there has been a North American led initiative to invest in better PD animal models. BRAINCAV'S WP4 is therefore primed to make a significant contribution in this respect.
List of Websites:
http://www.braincav.eu
Prevention and treatment of diseases that affect the brain will remain formidable challenges for years. Yet, as suggested by recent clinical trials, 'gene therapy' offers substantial potential and reason to believe that we can develop long-term, effective therapies for some brain diseases. Gene therapy is an approach that, instead of using classic methods (i.e. chemicals, surgery, radiation) uses DNA or RNA as drugs. Notably though, this strategy also has its own unique obstacles and risks - in particular the need to address feasibility, efficacy and safety. BRAINCAV's foundation is the ability of a virus called 'CAV-2' to deliver DNA to neurons - the most fascinating cells in the body. To determine if CAV-2 vectors have a clinical use for therapy for brain disorders the consortium used approaches that spanned basic research through pre-clinical model feasibility, efficacy and safety.
Our specific objectives included trying to understand the interaction of CAV-2 with fluid and cells of the brain. Does CAV-2 negatively affect neuron function or modify the genes that are expressed? Does CAV-2 leave the brain and go to other tissues? Does it interact with other cells besides neurons? Why do we primarily see neurons modified? How does it spread throughout the brain? We addressed these questions and have had encouraging results concerning the safety and biology of CAV-2 vectors. Convinced that CAV-2 vector deliver to the brain could be therapeutic, we treated mice and dogs that have a rare disease similar to one found in children (Sly Syndrome). Our results showed correction of many of the anomalies associated with the disease in the brain. On the other end of the scale, we also tried to better understand Parkinson's disease, the second most common inherited brain disease in Europe. We created animals with Parkinson's disease and we can now test therapies on these animals.
Project Context and Objectives:
Our objectives were to bring a novel viral vector 'from bench to bedside'. BRAINCAV tackles gene transfer to the brain by turning a virus into a vector capable of helping us understand fundamental neurobiology questions, disease progression and address therapy for global and focal brain disorders affecting children and the aging population. Our multidisciplinary fundamental and applied research project initiated a European-lead breakthrough to the domain of gene therapy for diseases affecting the brain.
Our scientific objectives are distributed into five complementary and intertwined work packages:
-To understand and define the interaction of canine adenovirus serotype 2 (CAV-2) with fluid and cells of the brain and periphery
-To use a transcriptional analysis to better understand the effects of CAV-2 vector interaction with neurons
-To test a therapy for an orphan neurodegenerative diseases commonly affecting the entire brain of children
-To develop genetic, long-lived Parkinson's disease animals and to treat its focal neurodegeneration
-To develop processes to produce and purify vectors that could bring this vector platform to the doorstep of clinical trials
Each viral vector has its specific advantages and drawbacks such as cloning capacity, memory or induced immunity, toxicity, specificity, safety, titre, or efficacy. If one wanted to genetically modify neurons the vector should preferentially transduce these highly differentiated cells, which make up approximately 10% of the primate brain. The advent of helper-dependent (HD) human adenovirus (HAd) vectors in the late 1990's, led to nontoxic, concentrated, homogenous preparations. HD Ad vectors also poorly integrate into the host genome, yet lead to long-term (at least 3 yrs) expression in many tissues in numerous immunologically naïve animals. The significant cloning capacity (greater than 30 kb) also allows one to subclone an entire gene (versus a cDNA) and include numerous transcriptional regulatory elements. These advantages make HD vectors ideal for long-term gene transfer to post-mitotic cells like neurons. In 2004, we demonstrated that CAV-2 vectors have numerous advantages for clinical gene transfer, and more specifically for treatment of neurodegenerative disorders.
Our proposal was designed to critically assess these advantages, the efficacy of the vector and its safety. Some of the principle characteristics and advances of the CAV-2 vector platform are:
-They preferentially transduce neurons in vitro and in vivo in numerous species
-They circumvent the ubiquitous human anti-Ad memory immune response
-They likely have a minimal effect on neurons
-They are capable of long-term expression
With these characteristics in mind, we define some of the areas in which CAV-2 vectors could be improved. We proposed a work plan to advance HD CAV-2 vectors substantially towards clinical trials.
Some of our basic research questions included how these virions interact with brain cells in relation to cell attachment, internalisation, and the genomic response. For example, 'what cells will be transduced and why' was a simple concept for years. Now we realise that the direct interaction of vectors with cells may be a simple pathway in vitro, but this interaction is significantly more complex in vivo. WP1 was designed to address these questions. The interaction with serum and cerebral spinal fluid (CSF) will dictate and explain the ultimate location of injected particles. Then, once the vector attaches to the target cell (neurons for CAV-2 vectors), we need to better understand how the virion used axon transport to reach the neuron's nucleus.
Many viruses have been selected naturally for their ability to make progeny. Setting the cell up for replication, e.g. perturbing genes involved in cell cycle, is something even vectors initiate. The interaction between the cell and virus also leads to the induction of transcription from a subset of endogenous genes. The cell can, in turn, respond to some of these stimuli by e.g. self-destruction (apoptosis) or inducing an immune response. Who wins this tug-of-war depends on numerous factors. This critic part of gene therapy has led us to propose a global transcriptome analysis to understand the vector-cell interaction at a genetic level.
At the applied research level, we used the characteristics of HD vectors to demonstrate proof-of-efficacy in the healthy and diseased brain of rodents, dogs and primates. We chose two models: a rare monogenetic disorder affecting the entire brain and a relatively common brain disorder that affects a pair of nuclei. We treated the neurodegeneration caused by mucopolysaccharidosis type VII, a lysosomal storage disorder, in a 400 mg mouse brain and a approximately 200-fold larger dog brain. Lysosomal storage disorder is a collection of approximately 60 orphan diseases, primarily affecting children. Most of these disorders result from deficiencies in enzymes normally implicated in the catabolism of macromolecules inside lysosomes. No adequate option is available for the treatment of the debilitating symptoms of neurodegeneration, which is a common pathology in the majority of these orphan diseases. Our challenge was distributing the enzyme throughout the brain. Something CAV-2 vectors were particularly efficient at doing.
We also used HD CAV-2 vectors to better understand, and possibly treat, Parkinson's disease by attempting to create genetic models (in contrast to the acute drug-induced models) that mimic disease progression. Parkinson's disease, one of the most common neurodegenerative disorders, is a chronic, progressive neurodegenerative movement disorder. It primarily affects dopaminergic neurons in substantia nigra (SN) leading to the loss of dopamine in their prime target area, the striatum. Disease modifying therapies are not available, partly, because no adequate animal model exists with chronic degeneration of dopaminergic neurons and a robust behavioural phenotype. In WP4 we will also combine the genetic- and drug-induced models to complement our understanding of Parkinson's disease and its potential therapies.
Project Results:
Results/foregrounds
BRAINCAV was divided into five Science And Tecnology (S&T) work packages. The results are presented and discussed in the format of the respective deliverables.
WP1 - The interaction of CAV-2 with fluid and cells of the brain and periphery
WP1 was designed to further understanding CAV-2 biology in the host and assess pertinent interactions of the virus with the host, for future gene therapy at the preclinical and, ultimately, clinical levels. All deliverables were successfully submitted.
Deliverable (D)1.1: Report describing histological analysis of CAR expression in the primate brain. We compared by immunoblotting the level of CAR expression in the adult primate versus embryonic brain. We found that in addition to a reduced level of CAR expression, CAR ran at different molecular weights suggesting a post-translational modification. We extracted protein from different regions of the primate brain and probed for CAR. Variable levels were found, which is consistent with results in the rodent brain as determined by mRNA content and immunoblotting. We used immunofluorescence (IF) to detect CAR expression in the microcebe brain. CAR was preferentially found on neurons in the soma and axons. We also looked for CAR expression at the primate neuromuscular junction. CAR was detected on the pre-synapse side and not on the muscle as previously reported.
Summary: CAR expression in neurons from the primate CNS was determined using biochemical and histological assays. Compared to data from rodents, CAR expression is similar, but not identical. CAR appears to be primarily expressed on neurons in the brain and not on glia. And while the levels decrease in older animals, CAR was still present and its expression pattern was consistent with CAV-2 tropism.
D1.2 To create a CAV-2 vector containing CAR-ablated knob and D1.3: Report comparing CAV-2 and CAV-2 CAR-ablated and in vivo
The CAV-2 fibre knob (FK) mutated in CAR binding site was tested on primary motor neurons. No specific binding could be detected compared to the native FK, suggesting the tropism of CAV-2 was CAR dependent. We therefore set out to generate a CAV-2 CAR-ablated vector to test the hypothesis that CAV-2 was CAR-dependent. We generated a CAR-binding ablated CAV-2 genome with the mutation R384E in the fibre. Homologous recombination was used to generate CAR-binding ablated CAV-2 vectors. With this aim, we generated shuttle plasmids containing the CAV-2 fibre protein. For the generation of the shuttle plasmid, we amplified a DNA fragment containing the CAV-2 fibre protein. We introduced the modification R384E in the shuttle plasmid and finally, we generated the CAR-ablated CAV-2 virus by homologous recombination between the shuttle plasmid (including the R384E mutation) and the CAV-2 genome. Viral genomes were analysed by sequencing and restriction enzymes. While a wild type CAV-2 genome with a mutation in the CAR binding domain is interesting, it could not be used for in vitro or in vivo analysis without the inclusion of a reporter gene and/or a deletion in the E1 region. It was therefore necessary to make an E1-deleted CAV-2 vector (i.e. not virus) containing a reporter gene (e.g. GFP). Two strategies were used to create CAR-ablated CAV-2 vectors. However, the insertion of the mutated fibre sequence into CAV-2 genome was unsuccessful after numerous attempts. We then tried a second approach using bacterial artificial chromosomes (BACs) and bacterial phage recombinases. Again, after several attempts we didn't succeed.
Corrective action: We therefore created an alternative approach by addressing CAR interaction at the surface of the cell - instead of in the virus. To do this we crossed Nestin-Cre mice with mice that had the second CAR exon flanked by loxP site. This approach created mice that lack CAR in the adult mouse brain. We then generate primary culture cultures of neurons and incubated them with HAd5GFP and CAVRFP (DsRed2). We found a significantly reduced transduction efficiency of CAV-2 especially for the transduction of neurons. HAd5 transduction was modestly impaired and cells with glia morphology were transduced. These data suggested that CAR was largely responsible CAV-2 tropism in vitro. We then injected CAVGFP into the CNS of C57BL/6 and CAR CNS-/- mice. We found a significantly reduced transduction at the site of injection in CAR CNS-/- mice of as well as via axonal transport compared to control C57BL/6 mice, again suggesting that CAR plays a major role in the tropism of CAV-2 vectors in vivo.
Summary: Instead of addressing the role of CAR at the level of the vector, we addressed its role using a genetically induced deletion during development. We tested CAR-depleted neurons in vitro and in vivo and found that CAR plays the most significant role in CAV-2 transduction.
D1.4: Report describing axonal CAV-2 transport in neurons
First, the virus binds to CAR at the cell surface followed by internalization in clathrin-coated pits. Panel A upper part shows electron microscopy of CAV-2 in a coated pit; lower part shows colocalisation of CAV-2 and CAR at the surface of motor neurons. During entry, maturation of CAV-2 containing vesicles occurs, with a shift from Rab5 to Rab7 endosomes that will eventually lead to CAV-2 axonal transport. Panel B shows quantification of colocalisation between CAV-2, Rab5 and Rab7 during internalization. Despite Rab7 maturation, the pH of axonal endosomes stays neutral, and CAV-2 remains inside vesicles during transport. Panel C shows electron microscopy of CAV-2 at late stage of endocytosis, when the transport occurs, and live-cell imaging of co-transport of CAV-2 and the endocytic marker dextran. CAV-2 traffic appears to have a bias for the retrograde direction and both dynein and kinesin are involved. Unexpectedly, CAR was also found in CAV-2 positive vesicles.
Summary: The pathway that CAV-2 uses to enter and be transported in neurons was described at the molecular level.
D1.5 CAV-2 vector biodistribution in primates following stereotaxic injection
CAVGFP was injected into the striata of 7 microcebes. 4 animals were sacrificed 5 h postinjection and 3 animals were sacrificed 5 days postinjection. Brain regions and peripheral tissues were analyzed for vector genome biodistribution analysis and GFP mRNA quantification. At the 5 h very few vector genomes were detected in peripheral tissues compared to brain tissues. There was evidence of retrograde transport of CAV-2, with significant quantities of genomes in areas distant from the injection site (frontal cortex, olfactory bulb and diencephalon). Vector genome levels were highest surrounding the injection site. At the 5 days, few genomes were detected in peripheral tissues. Similar to the animals, in particular at the injection site. Vector genomes were also detected in areas distant from the injection site 5 days after administration. qRTPCR quantification of GFP mRNA expression in brain regions 5 days post-stereotaxic injection showed a broadly similar transduction profile to the vector genome distribution profile. To further assess vector genome distribution and transduction by CAV-2 after stereotaxic injection, we injected a two microcebes with CAVGFP. 3D reconstruction of GFP+ cells quantified in distinct regions of the CNS generated a global view of transduction efficacy.
Conclusion: We assessed the genome biodistribution and transduction profiles of vectors after striatal, stereotaxic administration to microcebes. Highest levels of CAV-2 genomes, GFP mRNA and protein expression were observed in and around the injection site, although vector genomes, GFP mRNA and GFP+ cells were also detected in areas distant from the striatum.
D1.6 Proteomic and biochemical characterization of CAV-2 interactions using human blood and CSF
To complement D1.5 we assessed the interaction of CAV-2 with human serum and CSF components, using HAd5 as a control. SPR analysis of blood coagulation factor X (FX) binding to CAV-2 was carried out. HAd5, which binds FX at nanomolar affinity, was used as a positive control. SPR assays showed clear binding of HAd5 but no binding of CAV-2. In vitro binding experiments were then carried out in SNB19 human glioma cells and DKZeo cells to analyse the effect of blood coagulation factors X or IX or heparin on binding of CAV-2 to the cell surface. Neither FX nor FIX had an effect on binding of CAV-2 to the cell surface, although minimal binding inhibition was observed in the presence of heparin. In vitro cell binding experiments using an enzyme that cleaves HSPGs (Heparinase III) showed that heparin inhibition of CAV-2 binding was due to prevention of binding to HSPGs but not to CAR, which appeared unaffected by Heparinase III pretreatment by FACS analysis. Although these data suggested that there may be CAR-independent interactions of CAV-2 with the cell surface, these are likely to be of very low affinity and may be specific to the cell types that were used for the in vitro assays.
HAd5 interacts with several serum proteins. To identify potential CAV-2 interacting partners in pooled human serum or CSF samples, we carried out magnetocapture pull-down assays, using whole, biotinylated CAV-2 particles to co-precipitate interacting serum and CSF proteins prior to separation by 1D gel electrophoresis and identification by tandem mass spectroscopy. We found high-abundance albumin and IgG proteins. To identify potential low-abundance interacting proteins, samples were processed using an albumin and IgG-depletion kit. We analysed serum and CSF proteins magneto-captured using biotinylated virus as bait. As several putative interacting serum proteins for HAd5 have previously been characterised, HAd5 was used as a positive control for the human serum assays. Several high-abundance proteins were identified across all fractions, at satisfactory levels of confidence, including serotransferrin and ?-1-antitrypsin. In addition, known interacting serum protein partners for HAd5 were identified in co-precipitates, including complement proteins (CO3, CO4A) and transthyretin. Interestingly, more proteins were identified in CAV-2 co-precipitates compared to HAd5 co-precipitates. Moreover, CO3 and CO4A also co-precipitated with CAV-2, suggesting that some adenoviral interactions with serum proteins may be conserved across species.
Summary: CAR is likely the primary functional receptor for CAV-2.
WP 2 High-throughput transcriptional analysis of HD CAV-2 effects on brain cells
Objectives: WP2 was designed to further understanding CAV-2 biology in the host and assess interplay of the virus with the host genome. All deliverables were successfully submitted.
D2.1: hNPCs transcript alterations induced by CAV-2, HAd5, AAV9 and HIV-1 vectors
Human neural precursor cells (hNPCs) were used as a in vitro model to test the molecular effect of HD CAV-2 on the neuron. We were interested in the absolute evaluation of gene expression following vector transduction, to identify its molecular toxicity, and in its relative quantification, to establish how CAV-2 performed compared to other viral vector candidates for neuronal therapy. hNPCs were incuabted with the four vectors, i.e. 2 h at an MOI of 1000 vector genomes/cell with CAV-2, HAd5, AAV-9 and HIV-1.
Affymetrix HG U133 gene chip analysis was performed on RNAs extracted from infected and mock samples at 2 h and 5 days post infection. CAV-2, HAd5, AAV-9 and HIV-1 all modulated the 2D neuronal transcriptome, at 2 h and 5 days posttransduction. To interpret the significance of the modulation, genes that were significantly altered were analyzed with Ingenuity software. The most modulated molecular and cellular functions induced by CAV-2 were evident at 5 days postinfection and include cell cycle, cellular assembly, replication and recombination modulations. Elaboration of IPA data indicated that three major cellular pathways were variably affected: the immune response, the DNA-damage response and the pathways involved in cell trafficking. Particularly notable is the difference between CAV-2 and HIV-1 vectors for what concerns the induced modulation of the DNA damage pathway. The transcriptome variations include 182 commonly modulated genes, with 14% of the genes belonging to the DNA damage pathways modulated in opposite directions. To validate the biological information based on the extended chip analysis we performed qPCR on 24 genes and evaluated the Pearson correlation value (0.72). This indicated the reliability of chip data for the biological evaluation of neuronal response to CAV-2 and to the other viral vectors.
Summary: CAV-2, HAd5, AAV-9 and HIV-1 affect the neuron, independently from the efficiency of GFP expression. All vectors affect immune response, DNA damage and cell trafficking pathways. Particularly evident was the contrasting modulation of the DNA damage response by HIV-1 and CAV-2 vectors.
D2.2: Gene expression alterations induced by CAV-2, HAd5, AAV-9 and HIV-1 vectors in 3D cultures of hNPCs
To test the variations of the CAV-2 effects in 3D neuronal cultures, neurospheres of hNPCs and CAV-2 vector infection in that system were established. A full description of the methodology was published (Brito et al Methods 2012).
Differentiated neurospheres were incubated with CAV-2, HAd5, AAV-9 and HIV-1 at the fixed conditions applied to 2D cultures and described in D2.1. The relative efficiency of transduction in 3D cultures was maintained for CAV-2, while reduced for the other vectors, suggesting that CAV-2 can possibly penetrated the neurospheres more efficiently than other viral particles. RNA was extracted from neurospheres and modulation events were analyzed both by gene chip and by qPCR. After RMA normalization and T-test comparisons (p less than 0.05) 331 and 127 genes were found significantly modulated, respectively at 2 h and 5 days postinfection. Genes were submitted to IPA and the most significantly modulated pathways reported. Compared to the 2D system, CAV-2 effect on neurospheres includes the modulation of the DNA damage response pathway, and has a milder effect on the immune system and a more significant effect on the trafficking pathways. Particularly significant of this last aspect is the modulation of CD44, which has been suggested to contribute to HAd5 entry into cells and that of caveolins.
To evaluate the relative effects of different vectors on neurospheres, specific gene modulations were quantified by qPCR. Concerning the immune system, the analysed genes were not significantly modulated, except for TLR3 and TLR4 at 2 h. Interestingly, DNA damage genes, including RAD51 and BIRC5 have opposite modulation profiles in CAV-2 and HIV-1 transduced neurospheres. This is consistent with the different genome structure, and genome related DNA damage response of the two viruses and with the data obtained in 2D neuronal cultures.
Summary: The 3D system was a valuable tool for the analysis of neuronal function. The analysis of infected neurospheres confirmed specific modulation events observed in the 2D system, including DNA damage response activation. The efficiency of CAV-2 transduction in 3D neurons and the complementary modulation of genes involved in cell trafficking, suggest that this culture method should be exploited for further studying the neuronal trafficking processes following pathogen infections.
D2.3: Report of gene expression alterations induced by CAV-2, HAd5, AAV-5 and HIV-1 vectors in human glia and neurons
A notable effect of CAV-2 on neuronal cultures is the modulation of the DNA damage pathway. We analysed the genes modulated by CAV-2: BIRC5, BRCA2, FANCD2, MRE11, CHEK1, FAS, p53 and RAD51. The induction of RAD51, MAD2L1, BIRC5 and of FANCD2 was highlighted also by qPCR. Together, these data reinforce the principle of the significance of this pathway in the response to Ads, and suggest that, in a clinical perspective, it will have to be evaluated whether and to what extent viral vectors induce, through the activation of the DNA damage response, irreversible cellular toxicity. An aspect that was relevant in the analysis of vector toxicity in a clinical perspective is the evaluation of the immune response. This analysis is also expected to contribute to studies concerning the biology of neurons whose immune function is currently being determined. We thus evaluated the genes modulated by the four vectors in 2D neuronal cultures and identified commonly modulated probes like TLR3 and TLR4, CD44, CASP1 and HAS3. To support the reliability of the results obtained through chip analysis TLR3, TLR4, HAS3, IRF3 and CD44 modulation was quantified by qPCR. These genes were induced by CAV-2 infection suggesting that neurons can activate a response to pathogens, and this appears to occur even when the pathogen, as CAV-2, was attenuated. To evaluate the response of the second component of neurological tissue to viral vectors, we set up differentiation and infection conditions of glial cells. Consistent with poor CAV-2 transduction of the glia, qPCR analysis of single gene modulation showed that this cell type is not significantly affected by CAV-2 vectors under these conditions.
Summary: A detailed analysis of the transcripts modulated by CAV-2 in neurons showed, in the cluster of immune response, the upregulation of TLR 3 and TLR4 at the late time point, along with HAS3 (the enzyme involved in the synthesis of hyaluronan (HA)), CD44, (the HA receptor) and of IRF3. Taken together, these data suggest that CAV-2 induces the activation of gene program of antiviral innate immune response. CAV-2 induces as well a DNA damage pathway with the specific modulation of genes involved in ATM signalling (p21, BIRC5, MAD2L1) and in homologous recombination (FANCD2, RAD51). Glial cells are poorly transduced by CAV-2; consistently, the level of transcript modulation induced by CAV-2 was modest in these cells.
D2.4 Gene expression alterations induced in the primate brain
CAV-2 reaches the cell nucleus via retrograde axonal transport. To better understand the effects induced by the presence of HD-GFP in vivo, a transcriptomic study was performed in the brain of Microcebus murinus.
CAV-2 was introduced in the right caudate nucleus of 8 females Microcebus murinus. 2 periods were investigated for the transcriptome analyse: an acute (24 h) (4 animals) and a delayed time (28 days) postinjection (4 animals). The left caudate nucleus received the PBS vehicle. For controls, 4 animals were naïve for the virus. As the striatum receives synaptic terminals from different brain structures (cortex, SN, thalamus, contro-lateral side), we assayed changes in gene expression induced by HD-GFP in the striatum and in the frontal cortex. We compared the gene expression in the ipsilateral side with the controlateral side, and with non-injected brain samples using transcriptomic approach. Total RNA of each sample, extracted from the striatum and from the frontal cortex samples was prepared for hybridization with Affymetrix HG-133 Plus2 GeneChip. The data were normalized, converted digital intensity values were converted into cell intensity files using the Microarray Affymetrix Software 5.0 (MAS 5). MAS5 labels each transcript as 'present' (P) or 'absent' (A) by taking into account the average difference value calculated between the perfect match (PM) and the mismatch (MM) as detailed in the Affymetrix Genechip procedure (see http://www.Affymetrix.com online). A p-value was associated with the signal intensity. For each gene, the mean signal was calculated for each group. The overall rate of genes with the 'present' detection call was about 18%. Moreover, we noticed a lowest SD (6%) in all the samples studied suggesting a homogeneity of the gene expression. The procedure is described below. We detected 14,400 transcripts to analyze.
We used the significance analysis of microarray (SAM) to identify genes that were differentially expressed when comparing for a brain structure, the three groups together, right side versus left side versus control or when comparing the time after virus infection. Among the genes that presented changes in their expression, the most notable belonged to transcriptional regulation function, the cyotskeleton and synaptic transmission. In the striatum, these genes involved in immune response were detected (e.g. an overexpression for TNFRSF18 at 24 h and for HLAG, IGHM, IGLC2 at 28 days).
Summary: Our results showed that the injection of CAV-2 induced changes in gene expression and that the expression differed. In the acute phase, HD-GFP altered the expression of protein regulating the transcription but also protein involved in the release of neurotransmitter at the presynapse. These gene expression changes could transiently modify the synaptic activity of the neurons. In the chronic phase, CAV-2 induced modifications in some transcripts implicated in the cytoskeleton suggesting a minor modification in cell trafficking. The presence of the HD-GFP-induced an increase of immunoglobulin transcripts suggesting a mild adaptive immune response in the brain.
D2.5: HD CAV-2 vector alterations of gene expression in vitro and in vivo
Neuronal 2D and 3D cell cultures were analysed for the common genes and pathways influenced by HD-GFP transduction. As anticipated in the previous WP2 deliverables, the cell damage/cell cycle and immune response pathways were modulated in both systems. The cell trafficking functions, conversely, were mainly modulated by HD-GFP infection in the 3D neurospheres. When considering the significantly regulated genes together (2D and 3D neuronal results at 5 days post-transduction), 54 probes were commonly modulated in the two systems, corresponding, respectively, to 7.7% of the genes modulated in the 2D system and to 42.5% of the 3D system. The commonly modulated genes analysed and showed that the cell cycle/ DNA damage and cell morphology functional categories were modulated in the 2D and 3D neuronal systems in the same direction (mostly, probe upregulation).
We also studied HD-GFP effect on the neuron by comparing the in vitro (2D and 3D neuronal cultures) and in vivo (Microcebus murinus brain) data. An independent statistical condition applied for the analyses (detailed in the corresponding WP2 deliverable sections), we found no common genes. Given this result, to further develop the comparative study between the in vitro and in vivo systems, we designed a meta-analysis plan. This workflow was applied to the 2D, 3D data and to the Microcebus murinus brain derived chip results and contributed to further define the HD CAV-2 induced signature in primate neurons.
Summary: Taking into account the 2D, 3D and primate brain, we concluded that, at the functional level, HD CAV-2 affected the immune system, the cell cycle/DNA damage response (in 2D and 3D neurospheres), and intracellular trafficking (3D neurospheres). At the single gene level, 54 probes were commonly modulated by HD-GFP between the 2D, 3D, and none among 2D, 3D and primate brain. A meta-analysis based on a common statistical approach for 2D, 3D neurospheres and primate brain was developed.
WP 3 Treating neurodegeneration caused by MPS VII in mice and dogs
Objectives: Assessment of HD-RIGIE efficacy, toxicity and biodistribution in the β-glucuronidase (GUSB)-deficient mouse and dog models.
D3.1: Optimized HD-GUSB dose in the MPS VII mouse brain
The CNS and the skull of MPSVII mice are morphologically different from wild type littermates. We adjusted the injection point to obtain the maximum global transduction. For that reason, CAVGFP was injected into the striatum of 8-weeks old MPS VII mice at the following coordinates: 0,2 mm from bregma; 2 mm lateral; 3.5 mm deep. Results show GFP expression in many cells in the striatum and also in the dopaminergic neurons in the SN. No significant transduction was seen in other areas of the brain. A different cohort of animals was injected more rostrally, at +1,0 mm from bregma, 1,8 mm lateral and 3,5 mm deep . With these coordinates, we detected GFP expression in the striatum, where the virus were injected, and in the SN, corpus callosum and cortex, where we were able to detect transduction in neurons and endothelial cells. Endothelial cells, which overexpress CAR, could have been reached through circulation from the CSF or through microvessels present at the injection point.
Summary: We defined the following coordinates +1.0 mm from bregma, 1.5 mm lateral and 3.0 mm in depth as those that allow for more global transduction of MPSVII mouse brain.
D3.2: Optimized HD-GUSB injection site in the MPS VII mouse brain and D3.3: HD-GUSB effect based on histological analysis of the murine MPS VII brain
A group of 8 transiently immunosupressed animals were injected bilaterally at approximately 9 weeks of age with 2 x 109 pp of HD-RIGIE/striatum. Control littermates (heterozygous and mutant mice) were injected with the same volume (2 μl per hemisphere) of PBS. Mice were sacrificed at 6 weeks and GUSB activity was assayed every 100 nm. No GUSB activity was detected in PBS-treated mutant or even heterozygous mice, showing the low sensitivity of the assay. We detected transgene expression in striatum, several areas of the cortex, corpus callosum, SN and around ventricles. Overall, based on this assay, we estimated a volume transduction about 4 mm around the injection point.
Conclusions:
1. Brain transduction using HD-RIGIE vector is able to drive transgene expression to the whole brain of GUSB mice, 6 and 16 weeks after one single injection of the vector per hemisphere.
2. One injection of HD-RIGIE allowed for short and long-term ß-glu expression up to 60% of heterozygous mice and, more importantly, to complete biochemical recovery of secondary elevation of ß-hex and correction of GAGs in 4 out of the 6 slides of the brain, covering the whole cerebrum with the exception of the cerebellum and the brainstem.
3. Correlating with correction of biochemical parameters, ß-glu expression driven by HD-RIGIE lead to short and long-term clearance of lysosomal storage accumulation at the injection point but also in many distal areas.
D3.4: Behavior modifications following HD-GUSB injection in the murine MPS VII brain
Animals were behaviorally evaluated 6 weeks after treatment in a series of tests used as a screening for behavioral abnormalities and physical condition in mutant mice (Giménez-Llort et al., Eur. J Neurosci. 16: 547550, 2002). The behavioral domains analyzed included physical condition, sensorimotor functions, behavioral and psychological symptoms (anxious-like behavior, exploratory activity, daily life activities) and cognitive function. Experiments were conducted between 10:00 and 17:00 h. In all experiments animals were counterbalanced per genotype and gender. The research was performed in accordance with 86/609/EEC regarding the care and use of animals for experimental procedures. Animals were weighed before each test.
MPSVII mice show poorest physical condition as compared to wt animals, which was improved by HDCAV2GUSB-treatment. Poorest daily life activities were shown in the nesting behavior and seemed to be also improved in HDCAV2GUSB-treated animals. The assessment of animals in an abbreviated version of SHIRPA, wood and wire rod tests and hanger tests evidenced impairment of sensorimotor functions. MPSVII animals showed lower reactivity to handling as well as poorest prehensility in the use of elements of support and strength. Visual and forelimb reflexes, motor coordination and equilibrium were normal.
Summary: MPSVII animals showed impairment in all the domains studied with main affectation on cognition and HDCAV2GUSB was able to successfully improve and/or reverse the major deficits including total learning and memory capacities.
D3.5: HD-GUSB effect based on histological analysis in the dog MPS VII brain
The brains of seventeen dogs were injected with HD CAV-2 vectors. The initial cohort was injected with HD-GFP to determine efficacy, level of retrograde transport, immunogenicity, neuropathology, and optimize doses and coordinates in the healthy and MPS VII dog brains. Initially, stereotactic injection in dogs, and particularly in MPS VII dogs, was challenging because of the variation in the size of pups, the lack of a detailed atlas for 2 to 3-month-old mixed breed dogs, and MPS-associated bone deformation that did not allow a trustworthy location of the bregma or brain structures. Because MRI facilities were not available to determine stereotactic coordinates, injections were targeted to large structures and, when relevant, multiple coordinates along the needle track. The above constraints certainly led to vector delivery into ventricles in some MPS VII dogs, which in this case would lead to loss of extra-chromosomal vector genomes in dividing ependymal cells. At the site of striatal injections, cells with quintessential neuron morphology were predominant. We detected GFP+ cells throughout each hemisphere in particular on the striatum; thalamus; hippocampus; frontal, temporal, occipital cortex and nuclei of Meynert. In dogs injected unilaterally, we detected GFP+ cells throughout the ipsilateral hemispheres and in the contralateral cortex, which has efferent projections to the injected striatum. To distinguish between neurons and glia we stained sections with anti-NeuN and anti-GFAP antibodies. In afferent regions all GFP+ cells were neurons, while at the site of injection the majority of GFP+ cells also expressed NeuN. Combine, these data demonstrated that CAV-2 vectors preferentially transduced neurons in the large dog brain and are capable of efficient axonal transport, which is consistent with the tropism of CAV-2 in rodent brains and in organotypic cultures of human brain biopsies.
Conclusion: we assayed the efficacy of a HD CAV-2 vector containing a human GUSB expression cassette in the canine young MPS VII brain. As little as approximately 2 x 10 8 infectious particles of HD-RIGIE led to histological improvement throughout the MPS VII dog brain. A striking result was the comparable efficacy of HD-RIGIE: the dose injected into the MPS VII dogs was 50 to 200-fold lower than that used for AAV vectors in the brains of Sanfilippo and Hurler syndrome dogs. To put this dose into a more global gene transfer perspective, it would not be surprising to see 8 x 109 physical particles of a HD vector used in the rodent brain. Equally clinically relevant, and in contrast to studies using AAV vectors for MPS therapy, only transient immunosuppression appeared necessary and we found no obvious reduction in efficacy in the MPS VII dogs that were sacrificed 4 versus 1 month postinjection.
D3.6: Behavior modifications following HD-GUSB injection in the dog MPS VII brain
Due to the severity of the joint and cardiac problem associated with MPS VII in these dogs we were unable to assay behaviour modifications associated with brain gene transfer. Eventually this could have been overcome with glu enzyme replacement therapy, but the costs, limited supply of enzyme and minimum time postinjection did not allow this.
WP4: Developing novel genetic Parkinson's disease models and treating neuron death
Objectives:
i) characterize the biological functions of LRRK2;
ii) establish a high-content cell-based array for drug development;
iii) develop models for Parkinson's disease using ectopic expression of wild type and mutant LRRK2; and
iv) test the efficacy of HD CAV-2-mediated delivery of GDNF for neuronal protection against acute and genetically-induced parkinsonism.
D4.1 Report describing HD-LRRK2 vector efficacy in hmNSCs
In the CNS of rodents, dogs and primates, CAV-2 vectors preferentially transduce neurons when injected into the parenchyma. Whether CAV-2 vectors also transduce precursors of human brain cells that maintain the potential to differentiate into neurons or glia was unknown. To address this, we assayed human neural stem cells (hNSCs), which can be derived from specific brain regions and give rise to specific neurons such as dopaminergic neurons, when tissue is derived from fetal midbrain (hmNSCs).
Summary: While HD-LRRK2 transduced MDCK cells efficiently, hmNSCs were transduced with lower efficacy.
D4.2 Report describing the effect of HD-LRRK2 on proliferation and differentiation in hmNSCs
Vector dose of HD-CAV-2 expressing the mutated LRRK2 G2019S gene (HD-CAV-LRRK2) was studied in undifferentiated and differentiated hmNSCs cultures, in comparison to the mock vector HD-GFP. For initial screen of transduction parameters, HD-GFP was used. Different transduction settings were assessed in undifferentiated and differentiated cultures, by combining MOIs and transduction time.
Summary: Overall LRRK2G2019S did not induce significant specific alterations in proliferation apoptotic and neuronal markers in our cellular model.
D4.3: Report describing the cell based HCS system using LRRK2 expression in hmNSCs
Partner 8 did not perform these assays.
D4.4 Report describing HD-LRRK2∗ efficacy in the rat brain
The striatum of 16 rats were injected unilaterally at 2 coordinates with HD-LRRK2 and HD-LRRK2 G2019S. The controls were sham and HD-GFP injected (n = 8).
After the baseline of each individual animal was measured 3 days pre-injection by the amphetamine-induced rotation test, the rats were retested 1-month post-injection vector. We found no significant difference in the rotational behavior depending on the injected virus. At similar times, motor learning was tested using the cylinder test. Again there was no significant difference between groups. Rats were sacrificed 1-month postinjection for histological analyses. Tyrosine hydroxylase (TH)-immunoreactivity in the striatum as a measurement of dopaminergic terminal integrity was not reduced in virus injected animals. The number of TH-positive cells in the SN was also not affected. In addition, there was little GFP fluorescence in injected striata and none in the SN. Thus, we concluded that the amount of injected particles was not sufficient to efficiently transduce dopaminergic neurons in the SN and also that there was no behavioral or histological effect of such injections.
Summary: 16 rats were injected with LRRK2 containing vectors plus controls. We found no changes in the brain. Proposed corrective actions (intranigral injections, inject more particles and/or transiently immunosuppress) were not performed by partner 8.
D4.5: Report describing HD-LRRK2∗ vector efficacy in microcebes
The task was separate into different experiments
1st year: we developed behavioural tests after the injection of 6-OHDA:
1) hourglass: The animal is placed in a cylinder that is turned upside down. We measure the time taken by the microcebe to return to the upright position. Usually, unimpaired animals do it in less than 3 seconds;
2) tower with 7 levels of rungs where the microcebes have to jump from the ground to the 7th level, and where the difficulty increases progressively
3) reward searching in one of 6 aligned tubes; it consisted in reward searching in one of 6 aligned tubes. The animals have to jump on a small platform to reach the reward. We measure the time to catch the reward during 30 trials and the hand used.
4) hill and valley staircase task.
2nd year: we performed pilot studies with HD-LRRK2∗ (G2019S) vectors injected in the caudate putamen of two microcebes (4.5 x 109pp, 3µl injected, coordinates A:5 mm, L: 2mm, P: 8mm, 5 mm, 4mm). The microcebes were sacrificed 15 days after injection. We then investigated the loss of dopaminergic neurons in the SN. We observed a loss of dopaminergic neurons of 24% and 42%. The behavioural study showed that we need to train the animals before injection and follow up them during all the incubation time.
3rd - 4th year: We injected unilaterally at 2 coordinates with 2 x 109 physical particles of HD-LRRK2 and HD-LRRK2∗ G2019S (controls: 6 lemurs HD-GFP, 4 lemurs HD-LRRK2, 8 lemurs HD-LRRK2∗ injected). Prior to vector application, all the animals were trained to perform motor tasks. Then the mouse lemurs were followed up with behavioural tests during 6 months post-injection at the end of the follow up, the animals were sacrificed and the brains were processed for IHC studies.
Behavioral studies
Hourglass test:
G2019S LRRK2 lemurs are considerably impaired from 1 week to 3 weeks after injection in the hourglass test compared to performance before injection and this deficit is observed during the 6 month follow up. The G2019 lemurs did never reach again the baseline (before injection). For wild type LRRK2, the impairment occurs 3 weeks after injection, the animals recover good performances compared to the baseline. Control animals keep baseline performance during the follow up.
Tower test
Control animals show constant performance along the follow up meaning that animals did not encounter any difficulty to jump. G2019S lemurs (particularly 3/8 mouse lemurs) present an important impairment 3 weeks after the injection then recover progressively during 6 weeks and show again motor deficits along the follow up. For wild type animals, important motor deficits are observed 4 weeks after injection until 18 weeks and then the animals are less impaired but never reach again the baseline.
Post-mortem studies
Histological studies were done at the end of the 6 month-follow-up. It is important to note that the behavioural impairment was not maximal at the end of protocol.
Detection of protein expression with immunohistochemistry
We have performed different labelling with various antibodies: anti-LRRK2, anti-Flag, anti-GFP to detect HD-CAV2 transduce neurones, anti-TH to detect dopaminergic neurones, anti-GFAP to detect astrocytic inflammation.
Dosage of dopamine into the striatum of 4 control HD-GFP and 4 HD-LRRK2∗ injected animals
Lemurs brains were rapidly removed and tissue dosages of monoamines (DA, 5-HT) and their metabolites (DOPAC, HVA, 5-HIAA) were performed.
The injection into the striatum of 2 x 109 physical particles of HD-LRRK2 and HD-GFP have clearly an impact of the dopamine level of the striatum whatever the vector injected. The dopamine content is not significantly changed depending on the virus injected into the striatum. However the injection into the ipsilateral side in the G2019S lemurs involved a decrease of the dopamine level in the contralateral side. This decrease is not observed in the control animal suggesting that it is an effect of the LRRK2 expression.
Level of dopamine into the SN of control, WT, and LRRK2∗ lemurs
We have done mapping of TH-immunoreactive neurons into the SN of 4 controls, 4 Wild type and 8 LRRK2∗ (G2019S) and counted the number with the software Mercator (Explora Nova).
We observed a weak loss of TH neurons in the SN in the ipsilateral side whatever the virus injected. But we observed also a decrease in the contralateral side in LRRK2WT and LRRK2∗ compared to control suggesting a role of the overexpression of the LRRK2 gene. This is in agreement with the results observed into the striatum. We need to investigate the number of neurons into the SN in non-injected lemurs to confirm this result.
Morphology of the dopaminergic neurons in the SN
We detected a loss of dopaminergic fibers in G2019S animals compared to control and wild type animals. We observed dystrophic neurites and swollen axons, characteristic of neurodegeneration.
Summary: The injection of HD-CAV-LRRK2∗ into the striatum of mouse lemurs leads to transient motor deficits in prehension, postural stability and rigidity of the body. The loss of dopaminergic neurons is weak but we observe a clear decrease of the number of fibers and apparition of swollen axons and dystrophic neurons, characteristic of neurodegeneration. We observe also a glial activation on the ipsilateral SN. 6 months postinjection into the striatum, this suggested that the HD-LRRK2∗ lemurs present early signs of PD.
D4.6: Report describing HD-LRRK2∗ vector efficacy in Macaca fascicularsi
Injection of HD-LRRK2∗
We performed a pilot study in two Macaques to determine the distribution and axonal transport of CAV-2 into the monkey brain.
Stereotaxic delivery of CAVGFP and CAVCherry was performed into the caudate nucleus, putamen and thalamus of 2 intact animals. Convection enhanced delivery (CED) was used for putamen and thalamus injection while the injection into the caudate nucleus was undertaken by conventional method. GFP and Cherry expression were found throughout the brain. We observed no obvious advantage for CED versus conventional delivery.
Summary:
-CAV-2 was efficiently transported in a retrograde manner in the monkey brain and it showed a neuronal selectivity
-CAV-2 delivery into the brain induced the overexpression of GFP and LRRK2∗ proteins in brain areas projecting to striatum
-HD-LRRK2∗ injection into the putamen induced a loss of striatal 11C-DTBZ uptake and dopaminergic cell loss (VMAT-ir cells) in the homolateral SN
-HD-LRRK2∗ injections into the putamen induced overexpression of LRRK2∗ into the SN cells and provoked damage of the axons and neurites but no parkinsonism
-HD-LRRK2∗ injection into the striatum induced increased levels of phosphorylated Tau protein in several brain structures of the homolateral hemisphere.
-Striatal injections HD-LRRK2∗ induced gliosis and microglial activation in striatum and SN even 6 months after vector injection
-HD-LRRK2 and HD-GFP injection into the SN does not induce any behavioral manifestation 12 months after the surgery
-The decreased density of dopaminergic nigral cells observed after HD-LRRK2 and CAV-2 GFP delivery into the SN may be related to the vector injection itself
WP 5: Production and purification of HD CAV-2 vectors
D5.1 MDCK-E1-Cre recombinase cell line
An MDCK-E1-Cre recombinase cell line for the amplification of CAV-2 derived vectors was established from MDCK from ECACCC (No 84121903, lot 05G029). The cells were maintained and genetically modified in MEM 10% (v/v) FBS, full traceability records of the animal origin reagents used were retained.
Starting with MDCK cells, MDCK-E1-Cre cells were created in a 2-step procedure to constitutively express the CAV-2 E1 region and Cre recombinase. In step 1, MDCK cells were transfected with a plasmid called pCI-neoE1K9 provided by Partner 1-CNRS. pCI-neoE1K9 contains a CAV-2 E1 region expression cassette and a neomycin resistance gene. These cells (MDCK-E1) were selected with G418 and cloned by limiting dilution. E1 expression was functionally screened by the clone's ability to amplify E1-deleted CAV-2 vectors. A high E1-expressing MDCK-E1 clone was then transfected, in step 2, with a plasmid called pZeoCre (provided by Partner 1-CNRS), which expresses an nls-Cre recombinase and Zeocin resistance genes.
MDCK-E1-Cre cells are available and will be tested for production of HD CAV-2 vectors (currently under development by consortium partners CNRS and UAB).
D5.2: Report describing set-up of differential packaging system
The goal of D5.2 was to generate a CAV-2 vector that contained one of the phiC31 recombinase recognition sequences (attB) near the packaging signal and determine if this sequence, as in the case of human adenovirus serotype 5, delayed the production of mature CAV-2 particles.
CNRS designed a 1 kb sequence that contained the inverted terminal repeat (ITR), lox61, attB, a mutated packaging signal (based on CAV4.19 Soudais et al Mol Ther 2001) and lox67. The lox61 and lox67 sequences were chosen to avoid reinsertion of the excised floxed sequence into CAVattBCherry during the final step of HD (HD) vector production This sequence was synthesized, and sub-cloned upstream to a Cherry expression cassette in pTCAV-12VK. The resulting construct, pTCAVattBCherry was recombined with pTG5412, which contains the entire CAV-2 genome, to create a plasmid (pCAVattBCherry) that harboured the E1-deleted vector genome of the CAVattBCherry. CAVattBCherry was removed from the plasmid and transfected into DKZeo cells top produce the vector.
During the amplification it became obvious that CAVattBCherry had a delayed production of mature particles: the amplification cycles were optimal after collecting the cells 4-5 days post-infection instead of the normal 36 hours cycle for CAV-2. CAVattBCherry amplification was not only slower, but also less efficient, likely due to the incorporation of the mutated packaging signal. These characteristics are difficult for production, but optimal for CAVattBCherry use as a helper vector for HD vector production. Post-vector purification, CNRS and UAB tested the packaging delay using CsCl purified vectors via FACS and limiting dilution assays. Both assays showed that CAVattBCherry has a delayed packaging. Restriction digestion of vector DNA purified from infected DKZeo and DKCre cells showed that the genome is also sensitive to Cre recombinase removal of the packaging domain.
Conclusion: D5.2 completed successfully and BRAINCAV has generated a CAV-2 helper vector with characteristics that are optimal for HD vector production.
D5.3 Optimized bioreactor protocol
This technical deliverable described the optimized procedure for the cultivation of MDCK E1 cells adapted to serum-free medium and the production of CAV-2 vectors using a stirred tank bioreactor. The protocol also supports the procedure for production of HD (HD) CAV-2 vectors.
D5.5 Helper and HD master virus bank and GMP master cell bank
Helper and HD virus banks were manufactured in April 2011 and May 2012, respectively. ?E1 helper vector, denominated as JB?5, contains loxP sites flanking packaging domains and a RSV-lacZ expression cassette. For JB?5 master virus bank, DKZeo cells maintained in DMEM supplemented with 10 % (v/v) fetal bovine serum and 1% (v/v) of non-essential amino acids were infected with MOI 5. Viral vectors were harvested 48 hpi and purified by CsCl2 gradients. Thirty vials, 500 μl each, of Helper vector JB?5 in phosphate buffered saline with 10% (v/v) glycerol were produced and stored at -85ºC in the premises of partner 4. HD vector denominated as HD-GFP, contains an GFP expression cassette. HD-GFP bank was prepared after infecting DKCre cells in same culture medium using an MOI ratio between HD-GFP and JB?5 of 5 to 1.
GMP master MDCKE1Cre cell bank:
A GMP-certified Master Cell Bank (MCB) for MDCK-E1 Cre was manufactured in February 2012 using the procedures described in the Batch Production Record PF-012. The manufactured MCB batch number is 00412.
The QC tests used for the approval and release of the GMP-certified MDCK-E1 Cre Master Cell Bank were defined according to the European Pharmacopoeia. 100 vials with, 1 mL each, of the GMP-certified MDCK-E1 Cre Master Cell Bank were produced. They are stored in the vapor phase of liquid nitrogen. Process details are described in the Batch production record (PF-012).
WP 0: Management
Consortium management tasks and achievements
The Project Manager assumed the follow-up of the Grant Agreement signature between the EC and the coordinator CNRS, providing all necessary information/documents to both parties. Initially scheduled on September 2008, the GA was finally signed December 18, 2008. (With the authorization of the EC Project Officer, the project had started October 1st, 2008.) The manager also did a follow up of the GA accession forms to be established by the beneficiaries of the project. The EC funds were transferred to the coordinator end January 2009, and the Project Management team (Coordinator + Manager) processed with the administrative and financial follow-up of the distribution of the EC grant to the beneficiaries. All funds were distributed by beginning February 2009.
Report preparation
PMT has been heavily involved in the preparation of the 3 periodic and the final reports. The Manager provided the templates and sending reminders to the partners. Together with the Coordinator, all elements put on the online reporting tools set up by the European Commission (SESAM and FORCE) were controlled. This has lead so far to a limited number of comments from the European Commission because the major issues were dealt with before submission of the report. In a second step, the PMT ensured that the comments and remarks of the BRAINCAV Scientific Officer were addressed in the best delays.
Grant Agreement amendment
BRAINCAV had to request 4 GA amendments to take in consideration the changes in the consortium. This included the following changes:
-As the coordinator CNRS actually employed personnel from Inserm, it was suggested to adjust the previous budget by diminishing the Inserm budget and increasing the CNRS budget. The manager prepared an amendment that was agreed by the EC on April 2009. Beneficiary 2 (CRUK) asked to receive the whole EC contribution at once, as its work was concentrated on the first period. After formal agreement by the EC project Officer, it was agreed to do so.
-Some beneficiaries to the project disagreed to provide the Project Manager with all information regarding the IP rules internal to their institutions as the deliverable was set as 'PU' (public) when it was seen as strategic. The consortium then agreed to mark the deliverable as 'RE' (restricted access) or 'CO' (confidential).
-Name of Legal Entity Authorized Representative of CNRS has changed from M. Michel Portefaix to Mrs Ghislaine Gibello, since January 1st, 2009.
-To address technically difficult experiments, budget changes were agreed upon between the partners during the ExB meetings and validated by the partners:
-P5, who was responsible for mid-size HD vector production, has not been able to produce HD vectors. To date, P1 is the only partner that can make HD vectors, but did not budget resources to produce vectors for 4 years. After discussing possible options, P5 proposed to reallocate some of its EC contribution to the P1 so that the P1 can continue the production of the vectors. P5 sent an evaluation of the amount that could be reallocate to the P1. As a consequence, P5 EC contribution will be reduced by 17,680 Euros in favour of the P1; the initial EC request will not be otherwise modified. The P1 notes that this sum will not cover any related functioning costs and that it has already devoted extra resources to HD vector production during the first 30 months. At the M36 ExB meeting, P6 also offered to redistribute 20,000 Euros to P1 to ensure the production of the vectors. For similar reasons, the P8 will evaluate if a redistribution of its funding would be appropriate to support this production. The ExB members validated the proposed budget redistributions.
-Significant delays took place in the animal studies as explained in the RTD WP3 and 4 progress reports. This created an insurmountable problem into the ability to deliver on time. In spite of the alternatives explored to maximize the possibilities to stay within the limits of the initial work plan, it was not possible to deliver within 48 months. This was also foreseen as a major worry in the initial proposal. The animal studies were completed thanks to the 6-month extension of the GA. This lead to a 6-month extension, postponing the end of the project to month 54
-Departure of P8-ULEI from the Consortium
-Change in the Coordinator's contact
-Addition of the ICREA third party to P5-UAB
-Several limited changes to the Annexe 1 including the budget redistribution to take into account the departure of P8 and the redistribution of some tasks this partner was taking care of.
All these elements were changed in 4 GA amendments which were prepared for the coordinator by the manager and submitted by the coordinator.
Deviations from the original work plan and proposed corrective actions:
The main difficulties that occurred during the course of the project were solved or addressed by appropriate corrective action. The main issues were linked to the difficulties of working with MPS VII dogs and non-human primates. In addition, we had to deal with the departure of P8, who was WP4 leader and responsible for several tasks in WP4. P8 left the Consortium during the 3rd period and had to be replaced in order to ensure that the work could be finalized in the best conditions by the other BRAINCAV participants. Tasks and budget had to be redistributed to ensure that P8 work could be finished in the best conditions and full respect of the initial workplan.
Potential Impact:
WP6 (WP leader EJ Kremer CNRS) addressed the need to inform scientists, health care professionals, the general public, patient associations, and families about the initiative of the European Union via BRAINCAV.
Health care professionals:
Part-1 CNRS recently presented BRAINCAV'S goals to the University Hospital in Montpellier on period 1.
General public:
Part-1 participated in British Society of Gene Therapy public engagement day (March 29th 2010) to talk about BRAINCAV'S goals in for brain therapies. Part-4-IBET participated in a public engagement day at ITQB.
Patient associations: Part-1 is in regular contact with the French lysosomal storage disorders society (VML) to communicate BRAINCAV'S goal to its members. Part-5 is in regular contact with the Spanish MPS society concerning three MPS VII children in Spain.
Expected results, potential impact and use
Due PD's prevalence, the greatest socio-economic impact BRAINCAV could have is for the understanding and treatment of Parkinson's disease. Over 1 million Europeans suffer from Parkinson's disease with 40% of the victims under the age of 60. Parkinson's disease costs the EU over 50 billion Euros annually in health care, disability, loss of productivity, and long-term care - a cost that will only increase as the population ages. At present approximately 1% of the population greater than 50 years old has Parkinson's disease. This number will increase in concert with the number of older people. Since the submission of BRAINCAV, the use of CAV-2 vectors in the context of PD (fundamental and clinical applications) has attracted significant interest from the scientific community. Numerous colleagues have contacted BRAINCAV members. Recently, there has been a North American led initiative to invest in better PD animal models. BRAINCAV'S WP4 is therefore primed to make a significant contribution in this respect.
List of Websites:
http://www.braincav.eu