Final Report Summary - BRAINVECTORS (From Brain Gene Transfer Towards Gene Therapy: Pharmacological Assessment of AAV, CAV and LVV)
for a formatted document: see Attachment
Project title: From Brain Gene Transfer towards Gene Therapy: Pharmacological Assessment of AAV, CAV-2 and LVV
Project acronym: BRAINVECTORS Project website: http://www.brainvectors.org
Introduction
Gene therapy for neurological disease has recently gained increasing interest because pioneering clinical trials have demonstrated the safety and tolerability of viral vectors based on adeno-associated virus (AAV) and lentivirus (LVV) in the brain. However, in contrast to classical drugs, once a viral vector has been delivered to the brain, it will be irreversibly present. Therefore, the possibility to adjust and, if necessary, switch-off expression of the therapeutic protein represents a significant safety measure.
BrainVectors (BV) aims at delivering a clinically-acceptable viral vector with regulatable transgene expression for gene delivery in the brain. Our objectives were to optimize:
- the viral vector: a “vehicle” to deliver the therapeutic gene inside the brain cells. The vector must be non-toxic, target brain cells and provoke no immune reaction. Furthermore, the vector must be produced as a pharmaceutical agent - i.e. at a sufficient purity and according to “good manufacturing practices”.
- the adjustable system: a dimmer or genetic switch to adjust the dose of the therapeutic protein to the patient’s needs. The genetic switch must be operated by a clinically-acceptable, non-toxic and non-immunogenic drug.
Project objectives
1. To obtain a variant of the doxycycline-inducible (Tet-On) genetic switch that, when administered in the brain, responds to a clinically-acceptable drug inducer.
2. To compare three vectors based on adeno-associated virus (AAV), canine adenovirus (CAV) and lentivirus (LVV) with the selected inducible system in the brain. The following parameters were characterized: distribution of transgene expression in the brain, inducer dose range, leakiness, ON/OFF kinetics.
3. To develop methods for scalable production of CAV vectors.
4. To obtain cell type-specific inducible viral expression vectors to improve targeting specificity.
5. To evaluate the distribution of transgene expression after injection in the rodent brain.
6. To evaluate virus-cell interactions in a human neural progenitors cellular model.
7. To analyze the immune response to the inducible system.
Main Results
The vectors and the dox-inducible switch
Partner 2 generated Tet-On systems that are more active and doxycycline (dox)-sensitive than the original Tet-On system. These optimized variants, which are protected by a patent, are particularly useful for in vivo applications that require a more sensitive Tet-On system, such as the brain due to low dox penetration. Partner 2 has also previously set up a bioassay to measure dox levels in serum, which was used to define clinically-acceptable dox doses by comparison with data from patients treated with dox for different indications. Using a dox-sensitive variant, Partners 10 and 12 constructed dox-inducible AAV and CAV vectors that, when injected in the rat brain, respond to clinically-acceptable dox doses. Partner 7 developed brain cell type-specific (neurons, astrocytes) dox-inducible systems by splitting elements of the genetic switch between two LVV vectors, while Partner 3 obtained single cell type-specific AAV vectors but the inducibility of the vectors was very poor.
Vectors construction, production and purification
Several partners (1,3,7,8,10) benefitted from the biopharmaceutical know-how of Partner 5 and 6, in particular concerning vector production strategies. Whereas AAV and LVV are already in use in clinical trials in the brain, CAV vectors have never been developed for clinical applications. Partner 5, 6, 10 and 11 collectively improved the methods of vector construction and upscaling of helper-dependent (HD) CAV vector production. The hope is that HD-CAV vectors will constitute a new tool with specific advantages (large cloning capacity, lack of immunogenicity and retrograde axonal transport) in the arsenal of viral vectors for research and clinical applications.
In vitro model for vector evaluation
Partner 6 established a 3D cellular model consisting of human neural precursor cells (neurospheres) that can be differentiated into dopaminergic neurons. In addition to the analysis of cell-vector interactions (with Partner 4, 9, 10 and 11), these cells constitute a powerful in vitro model to complement animal experiments in the evaluation of therapeutic factors for Parkinson’s disease (PD).
In vivo vector evaluation
As a therapeutic model, we have chosen to use GDNF, a neurotrophic factor envisaged as a disease-modifying treatment for PD. The first clinical trials have raised questions regarding the site of GDNF administration. Partners 1,3,7,8,9 and 12 have characterized the distribution of transgene expression after injection in the rat striatum. The vectors that have been obtained during the BrainVectors project constitute of set of tools allowing one to test strategies of GDNF delivery to different types of neurons in the brain regions affected by PD.
Partner 1, 2, 7 and 12 demonstrated that the improved dox-sensitive AAV vectors can elicit therapeutically-relevant biological activities of GDNF in rats treated with clinically-acceptable dox doses. HD-CAV inducible vectors responded to similar dox doses. In contrast, LVV did not show a sufficient efficiency at the same dose.
Finally, the dox-inducible system did not elicit any cellular immune response in the rat brain since transgene expression was maintained for several months and no activation or infiltration of immune cells was detected in the brain of treated animals. Tools for analysis and monitoring of the humoral immune response against the inducible system and the GDNF transgene product were established by Partner 4 during secondments from Partner 1, 3, 9 and 12.
Conclusions
Gene therapy forms a promising novel strategy with enormous potential for the treatment of neurodegenerative disorders. However, the pharmacological aspects of this new type of medicine have thus far been addressed only marginally. The first BrainVectors output is a method to adjust the dose and period of administration of the therapeutic protein.
AAV and LVV are well-established vectors that have shown safety and tolerability in recent clinical trials. However, there are limitations to their use, i.e. prominent AAV seroprevalence in the population, risk of insertional mutagenesis by the genome-integrating LVV, limited size of the transgene. HD-CAV-2 vectors could provide a powerful tool to overcome several of these limitations. The second BrainVectors output is an improved set of methodologies and protocols for CAV-2 vector construction and scalable production, purification and characterization.
Traditional primary cultures of rodent brain cells and animal models differ significantly from the human physiology and frequently do not accurately predict the outcome of clinical trials. Partner 6 developed 3D cultures of human neural progenitor cells (hNPC ) that differentiate into complex tissue-like structures containing functional dopaminergic neurons. The third BrainVectors output is the evaluation of CAV and LVV vector-cell interactions in hNPC.
Finally, in view of the limitations faced in AAV-mediated gene therapy in peripheral organs, immunological tools to analyze and monitor immune responses to viral vectors and transgenes are becoming increasingly important for the future of gene therapy. Partner 4 shared knowledge, tools and protocols with academic researchers involved in pre-clinical gene therapy research to evaluate immune responses by a combination of in silico, in vitro and in vivo methods.
Project title: From Brain Gene Transfer towards Gene Therapy: Pharmacological Assessment of AAV, CAV-2 and LVV
Project acronym: BRAINVECTORS Project website: http://www.brainvectors.org
Introduction
Gene therapy for neurological disease has recently gained increasing interest because pioneering clinical trials have demonstrated the safety and tolerability of viral vectors based on adeno-associated virus (AAV) and lentivirus (LVV) in the brain. However, in contrast to classical drugs, once a viral vector has been delivered to the brain, it will be irreversibly present. Therefore, the possibility to adjust and, if necessary, switch-off expression of the therapeutic protein represents a significant safety measure.
BrainVectors (BV) aims at delivering a clinically-acceptable viral vector with regulatable transgene expression for gene delivery in the brain. Our objectives were to optimize:
- the viral vector: a “vehicle” to deliver the therapeutic gene inside the brain cells. The vector must be non-toxic, target brain cells and provoke no immune reaction. Furthermore, the vector must be produced as a pharmaceutical agent - i.e. at a sufficient purity and according to “good manufacturing practices”.
- the adjustable system: a dimmer or genetic switch to adjust the dose of the therapeutic protein to the patient’s needs. The genetic switch must be operated by a clinically-acceptable, non-toxic and non-immunogenic drug.
Project objectives
1. To obtain a variant of the doxycycline-inducible (Tet-On) genetic switch that, when administered in the brain, responds to a clinically-acceptable drug inducer.
2. To compare three vectors based on adeno-associated virus (AAV), canine adenovirus (CAV) and lentivirus (LVV) with the selected inducible system in the brain. The following parameters were characterized: distribution of transgene expression in the brain, inducer dose range, leakiness, ON/OFF kinetics.
3. To develop methods for scalable production of CAV vectors.
4. To obtain cell type-specific inducible viral expression vectors to improve targeting specificity.
5. To evaluate the distribution of transgene expression after injection in the rodent brain.
6. To evaluate virus-cell interactions in a human neural progenitors cellular model.
7. To analyze the immune response to the inducible system.
Main Results
The vectors and the dox-inducible switch
Partner 2 generated Tet-On systems that are more active and doxycycline (dox)-sensitive than the original Tet-On system. These optimized variants, which are protected by a patent, are particularly useful for in vivo applications that require a more sensitive Tet-On system, such as the brain due to low dox penetration. Partner 2 has also previously set up a bioassay to measure dox levels in serum, which was used to define clinically-acceptable dox doses by comparison with data from patients treated with dox for different indications. Using a dox-sensitive variant, Partners 10 and 12 constructed dox-inducible AAV and CAV vectors that, when injected in the rat brain, respond to clinically-acceptable dox doses. Partner 7 developed brain cell type-specific (neurons, astrocytes) dox-inducible systems by splitting elements of the genetic switch between two LVV vectors, while Partner 3 obtained single cell type-specific AAV vectors but the inducibility of the vectors was very poor.
Vectors construction, production and purification
Several partners (1,3,7,8,10) benefitted from the biopharmaceutical know-how of Partner 5 and 6, in particular concerning vector production strategies. Whereas AAV and LVV are already in use in clinical trials in the brain, CAV vectors have never been developed for clinical applications. Partner 5, 6, 10 and 11 collectively improved the methods of vector construction and upscaling of helper-dependent (HD) CAV vector production. The hope is that HD-CAV vectors will constitute a new tool with specific advantages (large cloning capacity, lack of immunogenicity and retrograde axonal transport) in the arsenal of viral vectors for research and clinical applications.
In vitro model for vector evaluation
Partner 6 established a 3D cellular model consisting of human neural precursor cells (neurospheres) that can be differentiated into dopaminergic neurons. In addition to the analysis of cell-vector interactions (with Partner 4, 9, 10 and 11), these cells constitute a powerful in vitro model to complement animal experiments in the evaluation of therapeutic factors for Parkinson’s disease (PD).
In vivo vector evaluation
As a therapeutic model, we have chosen to use GDNF, a neurotrophic factor envisaged as a disease-modifying treatment for PD. The first clinical trials have raised questions regarding the site of GDNF administration. Partners 1,3,7,8,9 and 12 have characterized the distribution of transgene expression after injection in the rat striatum. The vectors that have been obtained during the BrainVectors project constitute of set of tools allowing one to test strategies of GDNF delivery to different types of neurons in the brain regions affected by PD.
Partner 1, 2, 7 and 12 demonstrated that the improved dox-sensitive AAV vectors can elicit therapeutically-relevant biological activities of GDNF in rats treated with clinically-acceptable dox doses. HD-CAV inducible vectors responded to similar dox doses. In contrast, LVV did not show a sufficient efficiency at the same dose.
Finally, the dox-inducible system did not elicit any cellular immune response in the rat brain since transgene expression was maintained for several months and no activation or infiltration of immune cells was detected in the brain of treated animals. Tools for analysis and monitoring of the humoral immune response against the inducible system and the GDNF transgene product were established by Partner 4 during secondments from Partner 1, 3, 9 and 12.
Conclusions
Gene therapy forms a promising novel strategy with enormous potential for the treatment of neurodegenerative disorders. However, the pharmacological aspects of this new type of medicine have thus far been addressed only marginally. The first BrainVectors output is a method to adjust the dose and period of administration of the therapeutic protein.
AAV and LVV are well-established vectors that have shown safety and tolerability in recent clinical trials. However, there are limitations to their use, i.e. prominent AAV seroprevalence in the population, risk of insertional mutagenesis by the genome-integrating LVV, limited size of the transgene. HD-CAV-2 vectors could provide a powerful tool to overcome several of these limitations. The second BrainVectors output is an improved set of methodologies and protocols for CAV-2 vector construction and scalable production, purification and characterization.
Traditional primary cultures of rodent brain cells and animal models differ significantly from the human physiology and frequently do not accurately predict the outcome of clinical trials. Partner 6 developed 3D cultures of human neural progenitor cells (hNPC ) that differentiate into complex tissue-like structures containing functional dopaminergic neurons. The third BrainVectors output is the evaluation of CAV and LVV vector-cell interactions in hNPC.
Finally, in view of the limitations faced in AAV-mediated gene therapy in peripheral organs, immunological tools to analyze and monitor immune responses to viral vectors and transgenes are becoming increasingly important for the future of gene therapy. Partner 4 shared knowledge, tools and protocols with academic researchers involved in pre-clinical gene therapy research to evaluate immune responses by a combination of in silico, in vitro and in vivo methods.