Krystof Bankiewicz - US grants

One of the characteristics of vector administration into the CNS is the ability of the transgene products and/or viral particles to travel along neural pathways. In our preliminary experiments in monkeys where the AAV-2 vector was administered into the striatum, transgene product was detected in afferent and efferent regions including the internal and external portion of the globus pallidus, subthalamic nucleus, substantia nigra reticulata and compacta, thalamus and cortex at 3 or 8 weeks after gene delivery. This strongly indicates that both retrograde and anterograde transport of vector and transgene product must be carefully evaluated as it may impact therapeutic applications of gene transfer. As the PDGTSG has identified rAAV and lentviral vectors as the initial two most promising vectors (see Principal Projects I and II) these will be the focus of this Pipeline Project. To address this transport issue this project will systematically analyze the axonal transport characteristics of AAV-2, the alternative serotype AAV-5 and lentiviral vectors in monkeys following intrastriatal administration. Each vector type will be delivered unilaterally into the striatum of normal monkeys and resulting gene expression and presence of the gene product will be detected at the site of gene delivery and projection sites. Fuorophore-labelled viral particles will be delivered in the contralateral striatum and detected at 48 hours later at the injection site and the projection regions. Results of this project will lead to an understanding of patterns of distribution of each of the viral vectors and will provide critical information regarding final design of Principal Projects.

Widespread and controllable distribution of viral vectors is essential to achieve human gene therapy for Parkinson's disease. Given the volume of the human striatum methods development is required for adequate vector coverage. Convection-enhanced delivery (CED), a recently developed approach for delivery of small and large molecules to targeted sites in solid tissues, utilizes a bulk flow mechanism to deliver and distribute macromolecules to clinically significant volumes of tissue. The technique is based on a pressure infusion system that "pushes" and replaces extracellular fluid in surrounding cells of the parenchyma. Delivery of fluid carrying the AAV-2 particles results in significant spread of the vector within CNS. This delivery system offers significant advantages by improving the volume of distribution, providing uniform distribution of drug concentration within the targeted tissue, and complete delivery of the therapeutic factor to the target site. AAV-2 binds to heparin-sulfate proteoglycans on the cell surface and enhances the interaction of AAV-2 with cells. As for other heparin-binding proteins the binding affinity of AAV-2 to heparin is low and a high-affinity receptor is required for cell transduction. Preliminary data indicate that heparin addition to AAV-2 particles allows for robust convection based delivery of AAV-2 within anatomical structures. This project will systematically examine the impact of combined CED along with heparin to distribute AAV throughout the monkey striatum. The same methodologies will be tested will also be applied to VSV-g pseudotyped lentivirus vectors.

Parkinson's disease (PD) is a debilitating neurodegenerative disorder of unknown etiology that results in the loss of dopaminergic neurons within the nigrostriatal pathway and clinically manifests as bradykinesia, resting tremor, and rigidity. Individuals suffering from PD receive variable symptomatic relief from early pharmacologic therapy in the form of the dopamine precursor, levadopa (L-DOPA). Upon entry into the CNS, L-DOPA is decarboxylated into dopamine by the critical enzyme aromatic-amino-dopa-decarboxylase (AADC). Long-term utilization of L-DOPA and progressive disease leads to side effects including dyskinesias and a narrowing of the therapeutic window. Gene transfer technologies may provide modalities by which the therapeutic window of L-DOPA therapy can be widened and side effects minimized. Adeno-associated virus (rAAV)- and lentivirus (rHIV)-based vectors have shown promise in the delivery and long-term expression of therapeutic genes within the mammalian CNS. Our hypothesis is as follows: Delivery of the AADC gene to the striata of parkinsonian monkeys via rAAV or rHIV vectors will result in extended L-DOPA efficacy and reduced occurrence of drug-induced dyskinesias. To test this hypothesis, the following experiments are to be conducted in a standardized parkinsonian non-human primate model: (i) evaluate the clinical efficacy of rAAV- and rHIV-based AADC gene transfer modalifies in combination with L-DOPA administration; (ii) evaluate these approaches in their potential to alleviate L-DOPA-induced dyskinesias; and (iii) correlate clinical efficacy with increased conversion of L-DOPA to dopamine using biochemical assessments. These experiments will provide a unique opportunity for evaluation of a combined gene therapeutic/pharmacologic approach using two promising gene transfer vectors to alleviate the potentially disabling and frequently dose-limiting complications of L-DOPA therapy.

DESCRIPTION (provided by applicant): Although L-dopa remains the mainstay of Parkinson's disease therapy, two-thirds of all patients develop motor complications within six years of initiation of therapy, the most prominent of which are so-called dopa dyskinesias. In an effort to restore the benefits of L-dopa therapy to these patients and to minimize or eliminate the side effects of chronic dopa therapy, we performed several experiments in monkey and rodent models of PD. We demonstrated that, by using an approach which combines the aromatic L-amino acid decarboxylase (AADC) gene and the pro-drug (L-dopa), the neurotransmitter involved in PD (dopamine) can be synthesized and regulated. Striatal neurons, transduced with the AADC gene by an adeno-associated viral vector (AAV), can convert peripheral L-dopa to dopamine and stimulate DA receptors. Positron emission tomography (PET) and the AADC tracer, 6-["F]fluoro-L-m-tyrosine (FMT) demonstrated complete and long-term restoration of AADC enzyme in PD monkeys. This approach to treating Parkinson's disease may reduce the need for L-Dopa/Carbidopa, thus providing a better clinical response with fewer side effects. In addition, the imbalance in dopamine production between the nigrostriatal and mesolimbic dopaminergic systems can be corrected by using AADC gene delivery to the striatum. In an effort to demonstrate clinical application of this approach, we propose the following translational experiments in the MPTP-treated Parkinsonian primate model: (i) evaluate clinical efficacy of AADC gene transfer combined with levodopa administration; (ii) evaluate the efficacy of AAV/AADC gene transfer aimed at alleviating levodopa-induced dyskinesias. More specifically, we will assess the effects of intracerebral transfer of AADC gene using AAV vector on dopa-dyskinesias using squirrel monkeys rendered Parkinsonian by systemic injection of the selective dopaminergic neurotoxin 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP). These animals develop dopa-dyskinesias, which are indistinguishable from those observed in patients with Parkinson's disease. Our therapeutic approach has the potential to restore dopamine production, even late in the disease process, at levels that can be maintained during continued nigrostriatal degeneration and may restore benefits of L-dopa treatment in patients with Parkinson's disease.

The long-term goal of this project is to apply neurotrophic factor gene therapy to patients with Parkinson's disease (PD). Extensive basic research in rodent and monkey models of PD indicates that the neurotrophic factor, glial cell ine-derived neurotrophic factor (GDNF) can slow the degeneration of dopamine (DA) neurons in the substantial nigra (SN); neurons that die in PD. Moreover, local delivery of these factors into the brain (using vector mediated gene transport) has shown protection of neurochemical, cellular and behavioral indices of DA function in 6-OHDA or MPTP lesioned animal models. Gene delivery was also shown to increase growth factor synthesis in a chronic manner within the vicinity of DA neurons; an advantage for treating slowly progressive CNS disorders such as PD. Thus, basic research strongly supports GDNF gene therapy as a novel treatment for PD. However, several issues require addressing prior to designing a clinical trial. First, identifying optimal vector system for gene delivery; this project will compare adeno-associated (AAV) and lentiviral vector systems which appear to be two most promising vector systems for transducing brain tissue. Secondly, it is not known whether GDNF gene delivery will ameliorate the parkinsonian condition in primates that have a stable, chronic lesion of the DA system. This project will use a primate model that closely reflects the mild clinical state of the disease. Thirdly, regulating gene expression will play an important role following therapeutic intervention to offset side-effects that may occur due to growth factor overexpression. Therefore, this revised project will evaluate adeno-associated viral (AAV) vectors in which the transgene can be regulated by peripheral doxycycline (Dox) administration. In summary, Dox regulated AAV and lentivral vectors will be produced and characterized in the rodent brain prior to initiating primate studies. High titer, helper-free clinical grade vectors prepared in a GLP facility will be used for primate studies. Vectors will be injected into the primate striatum using convection-enhanced delivery methods. Transgene expression and biological effects will be evaluated by a variety of behavioral, neurochemical, morphologic and molecular assays including, but not limited to in vivo FMT PET imaging, MRI, clinical rating and clinical responses to L-DOPA administration, HPLC, ELISA, in situ hybridization, immunocytochemistry, stereology and toxicology. These studies will involve collaborations within PDGTSG and will provide data leading to a clinical trial for Parkinson's disease. Results from these studies may also be applied for treating other neurodegenerative diseases and injuries related to the CNS.

DESCRIPTION (provided by applicant): The broad aim of this proposal is to test the hypothesis that L-dopa-induced dyskinesias (LID) in Parkinson's disease (PD) arise at least in part from non-uniform dopaminergic denervation of the striatum, whereby islands of dopaminergic activity (hotspots) are created within the most severely affected part of the striatum, the post-commissural putamen. Transduced with an AAV containing the cDNA for aromatic L- amino acid decarboxylase (AADC), striatal neurons gain the ability to produce dopamine (DA) from exogenous L-Dopa, a DA precursor. During testing in a non-human primate (NHP) model of PD, we observed severe LID when AAV-AADC was infused into the striatum in a way that generated focal regions of high AADC activity. More recently, we have found that the generation of a single AADC hotspot in the post-commissural putamen of a hemi-parkinsonian monkey generated LID, remarkable because this model is almost completely refractory to LID. We plan to use an inducible AADC expression vector, recently shown to be effective in a parkinsonian rodent model. With this tool, we should be able to produce hotspots in monkey brain reversibly. This in turn should allow us to ask whether turning AADC hotspots on and off correlates with induction and abatement of LID. We will also be able to ask what kinds of metabolic and molecular changes occur with onset of LID, and whether such changes are reversed upon elimination of the hotspot. This research program is important from three perspectives. First, it promises to establish an in vivo model of LID in which neural correlates of LID can be investigated in a controlled setting. Second, it seeks to test important hypotheses regarding the mechanistic and anatomical origins of L-Dopa-dependent dyskinesias. Finally, it investigates two likely causes of post-engraftment dyskinesias, which are perhaps the most important current impediment to progress in transplantation-based therapies for PD. This work will provide insight on the mechanisms behind L-dopa induced dyskinesias and will provide a basis for using gene therapy approaches to treat patients with Parkinson's disease.

DESCRIPTION (provided by applicant): The broad aim of this project is to develop an efficient means to deliver to the human brain a genetic therapy to ameliorate the neurological deficits encountered in Type A Niemann-Pick disease (NPD). Armed with mouse efficacy data, we are confident that an AAV vector that encodes human acidic sphingomyelinase (hASM) is likely to be effective in treating the disease in humans. A major challenge, however, is that widespread expression of hASM will probably be required in order to achieve significant clinical improvement in humans. Efficacy data in knockout mice, although encouraging, does not really address the technical issues that we face in the very much larger human brain. Clinical efficacy will rely considerably upon the development of techniques to deliver gene therapy vectors to such sensitive and highly problematic regions as brainstem. Recently, we have developed a method of visualizing placement of infusion cannulas on MRI, and can actually follow infusion of liposomes tagged with Gadolinium in real-time, termed Real-time Convective Delivery (RCD). In preliminary experiments, we found that these liposomes distribute very like AAV1. We hypothesize that a mixture of AAV1 containing the hASM cDNA and GDL will permit real-time tracking of AAV-mediated gene therapy. We plan to use MRI-guided delivery of AAV1-hASM in the development of a therapy for Niemann-Pick disease. We propose experiments in this application that we believe will form the basis of a major improvement in brain gene therapy in general, and more specifically in the treatment of neurological aspects of LSD's.

DESCRIPTION (provided by applicant): To advance neurological gene therapy applications into the clinic there is an urgent need to improve our understanding of gene delivery vectors, with particular emphasis on distribution and safety after direct brain delivery. While AAV2 currently dominates clinical development, alternative AAV serotypes may be better suited for some applications. The recent game-changing development of image-guided delivery techniques provides opportunity to accurately compare different vectors after direct brain delivery. Accordingly we aim to investigate the distribution and immune responses of 5 AAV serotypes (AAV1, 2, 5, 8 and 9) after confined delivery to grey or white matter structures in the non-human primate (NHP) brain. This in-depth assessment of AAV vectors will provide investigators with scientific rational for selecting a specific vector for a particular neurological application.

The goal of this proposal is to further evaluate the safety and feasibility of gene transfer to provide aromatic L- amino acid decarboxylase (AADC) enzyme into the midbrain of patients with AADC deficiency and continue Biologics License Application (BLA)-enabling studies as per FDA recommendations. AADC deficiency is a devastating genetic neurometabolic disorder which causes hypotonia, dystonia, intense and long-lasting oculogyric crises (OGC), developmental delay and chronic and severe neurological dysfunction. A gene therapy based on delivering of a recombinant adeno-associated virus carrying the DDC human gene (AAV2-AADC) to the brain structures that physiologically AADC enzyme (midbrain) could be a most needed disease-modifying treatment for AADC deficiency. Eight (8) AADC deficient patients have been treated (160 µL) in our initial NIH- funded trial in the US under BB-IND-16127 and an additional 15 subjects under an ethics committee-approved compassionate use program (CUP) in Poland. The latter received a larger infusion volume (?300 ?L) and a shorter surgical procedure. Both approaches were safe and well-tolerated regardless of dose or volume of infusion. OGCs stopped a few weeks after the surgery and subjects? sleep, mood and irritability improved. Most subjects are gaining head control and muscular tone, developing purposeful movements and some are even sitting up and starting to walk without support, regardless of their age. Encouraged by the safety and positive biomarker and clinical outcomes observed in those groups, we propose an extension of the BB-IND-16127 study to (i) determine the long-term (up to 5 years) safety and tolerability of the surgical infusion of already treated subjects (n=8) (ii) determine the safety and tolerability of a larger volume of Cohort 2 vector concentration into the SN/VTA administered via a surgical procedure optimized to increase safety by reducing surgical and anesthesia times (single-cannula insertion per hemisphere) in AADC deficient patients >4 years, and (iii) demonstrate effective restoration of AADC function by measuring CSF neurotransmitter metabolites and changes in brain FDOPA uptake on PET imaging. This will be a multi-center study with subjects to be treated at The Ohio State University and at the University of California San Francisco. As per our discussions with FDA, the study design includes a 12-month lead-in period that will serve as a natural history control group to explore potential efficacy of this novel treatment. Cohorts 3 (4-13 years, n?12) and 4 (>13 years, n?12) will then receive a larger infusion volume of AAV2-AADC at the same titer Cohort 2 received (2.6 x 1012 vg/mL; up to 300 ?L/hemisphere). Renewal of funding for this trial will enable assessment of the safety and tolerability of an optimized dose and delivery procedure to enhance distribution of AADC expression within the midbrain, which we hypothesize may lead to further clinical improvement. Completion of this exploratory clinical trial will pave the way to registration of this disease-modifying AAV2-hAADC gene therapy for AADC deficiency and future gene therapies for other neurological disorders.

CRISPR-based gene editing of the brain has the potential to revolutionize the treatment of neurological diseases. A large number of incurable brain diseases, such as Huntington's, Alzheimer's and Parkinson's disease, are caused by the over-expression of pathogenic proteins and could be treated with CRISPR based therapeutics. However, despite its potential, developing CRISPR based therapeutics for the brain has been challenging because of delivery problems. In particular, two key challenges need to be solved before gene editing in the brains of large animals and in humans is possible. First, strategies for efficiently and safely delivering Cas9 and gRNA into neurons, after an intracranial injection, need to be developed. Second, strategies that can enable a large volume of brain tissue (> 1 cm) to be transfected after an intracranial injection of CRISPR reagents also need to be developed. The central objective of this proposal is to develop a delivery strategy for gene editing the brains of large animals after an intracranial injection, termed convection-enhanced CRISPR (C-CRISPR). C-CRISPR is based on using convection-enhanced delivery (CED) to deliver an engineered Cas9 RNP, which has been fused to multiple nuclear localization signals (NLS), and has been encapsulated in PEGylated block copolymers. C-CRISPR addresses the key translational bottlenecks that have prevented CRISPR from having a translational impact in the brain. In particular, because it delivers the Cas9 RNP directly, it avoids the toxicity problems of viruses and the manufacturing challenges of using mRNA, and consequently has great translational potential. In addition, C-CRISPR uses CED to distribute the Cas9 RNP across centimeters of brain tissue, and therefore has the potential to edit the brains of large animals. C-CRISPR is based on our preliminary data demonstrating that the Cas9 RNP fused to multiple NLS signals can edit genes in murine brains after an intracranial injection, and that Cas9 RNP complexed to PEG-block copolymers can be delivered to centimeters of brain tissue, in the striatum, after delivery via CED. CED of engineered Cas9 RNP complexed to PEG block copolymers, therefore, has the potential to edit genes in human patients. We propose therefore the following aims/milestones: UG3 Specific Aim 1. Develop C-CRISPR formulations that distribute throughout the striatum of rats UG3 Specific Aim 2. Develop C-CRISPR formulations that edit centimeters of brain tissue UH3 Specific Aim 1. Develop C-CRISPR formulations that edit centimeters of tissue in pig brains The experiments in this proposal are significant because, if successful, C-CRISPR will be the first example of a non-viral delivery strategy that can edit genes in the brains of large animals. The experiments in this proposal are innovative because C-CRISPR is the first example of a delivery strategy that effectively integrates 3 complementary technologies, (1) engineered Cas9 RNPs (2) PEGylation and (3) convective enhanced diffusion, and will provide a roadmap for developing strategies for gene editing in higher animals.

CRISPR-based gene editing of the brain has the potential to revolutionize the treatment of neurological diseases. A large number of incurable brain diseases, such as Huntington's, Alzheimer's and Parkinson's disease, are caused by the over-expression of pathogenic proteins and could be treated with CRISPR based therapeutics. However, despite its potential, developing CRISPR based therapeutics for the brain has been challenging because of delivery problems. In particular, two key challenges need to be solved before gene editing in the brains of large animals and in humans is possible. First, strategies for efficiently and safely delivering Cas9 and gRNA into neurons, after an intracranial injection, need to be developed. Second, strategies that can enable a large volume of brain tissue (> 1 cm) to be transfected after an intracranial injection of CRISPR reagents also need to be developed. The central objective of this proposal is to develop a delivery strategy for gene editing the brains of large animals after an intracranial injection, termed convection-enhanced CRISPR (C-CRISPR). C-CRISPR is based on using convection-enhanced delivery (CED) to deliver an engineered Cas9 RNP, which has been fused to multiple nuclear localization signals (NLS), and has been encapsulated in PEGylated block copolymers. C-CRISPR addresses the key translational bottlenecks that have prevented CRISPR from having a translational impact in the brain. In particular, because it delivers the Cas9 RNP directly, it avoids the toxicity problems of viruses and the manufacturing challenges of using mRNA, and consequently has great translational potential. In addition, C-CRISPR uses CED to distribute the Cas9 RNP across centimeters of brain tissue, and therefore has the potential to edit the brains of large animals. C-CRISPR is based on our preliminary data demonstrating that the Cas9 RNP fused to multiple NLS signals can edit genes in murine brains after an intracranial injection, and that Cas9 RNP complexed to PEG-block copolymers can be delivered to centimeters of brain tissue, in the striatum, after delivery via CED. CED of engineered Cas9 RNP complexed to PEG block copolymers, therefore, has the potential to edit genes in human patients. We propose therefore the following aims/milestones: UG3 Specific Aim 1. Develop C-CRISPR formulations that distribute throughout the striatum of rats UG3 Specific Aim 2. Develop C-CRISPR formulations that edit centimeters of brain tissue UH3 Specific Aim 1. Develop C-CRISPR formulations that edit centimeters of tissue in pig brains The experiments in this proposal are significant because, if successful, C-CRISPR will be the first example of a non-viral delivery strategy that can edit genes in the brains of large animals. The experiments in this proposal are innovative because C-CRISPR is the first example of a delivery strategy that effectively integrates 3 complementary technologies, (1) engineered Cas9 RNPs (2) PEGylation and (3) convective enhanced diffusion, and will provide a roadmap for developing strategies for gene editing in higher animals.