Abstract
Gene delivery to the central nervous system has remained challenging, particularly for diseases such as lysosomal storage disease, in which the gene product has to reach virtually the entire central nervous system (CNS), including the potential to target other organs. In the past 5 years, gene therapy studies have focused on evaluating the transduction efficiency of different AAV serotypes in the CNS of nonhuman primates. These studies have led to the conclusion that AAV5, AAV9, and recently AAVrh.10 are likely to be the serotypes of choice for clinical gene delivery.
In a comparative study with AAV 1 and AAV8, AAV5 was shown to be the most efficient in transducing neurons in nonhuman primate striatum (Dodiya et al., 2010), which was confirmed by other preclinical studies (Colle et al., 2010).
Recent studies have focused on the ability of certain AAV serotypes to cross the blood–brain barrier, leading to the revolutionary finding that AAV9 could be injected intravenously and still efficiently infect neurons and glial cells in the brain and spinal cord in nonhuman primates (Bevan et al., 2011; Gray et al., 2011). Furthermore, recent studies have also demonstrated that AAV9 could be placed in the cerebrospinal fluid to efficiently target motor neurons (Bevan et al., 2011).
Another emerging candidate for CNS gene delivery is AAVrh.10, a rhesus macaque isolate belonging to clade E (AAV8 family) (Cearley and Wolfe, 2006), which seems to have a preferential tropism for neurons and oligodendrocytes, favoring the use of this vector for gene delivery in white matter disorders.
Metachromatic leukodystrophy (MLD) is a fatal lysosomal storage disorder characterized by accumulation of sulfatides in glial cells and neurons resulting from an inherited deficiency of arylsulfatase A (ARSA), the enzyme responsible for sulfatide catalysis. The most common form of MLD affects children between 1 and 2 years of age. A period of apparently normal early development is followed by hypotonia, weakness, and motor impairment. Later signs are spasticity, cognitive regression, and decline in fine motor skills, as well as optic atrophy and seizures. Death usually results from a respiratory infection. Although sulfatide accumulation is also observed in neurons, the cells mainly affected in MLD are the oligodendrocytes, resulting in demyelination of the central and peripheral nervous system (Batzios and Zafeiriou, 2012), the histopathological hallmark of MLD. At present no effective cure is available, but rapid delivery of ARSA enzyme to brain oligodendrocytes is likely to arrest and even reverse disease progression.
In this issue of Human Gene Therapy, Piguet and colleagues evaluate the effects of two different viral vectors, AAV5 and AAVrh.10, driving the expression of ARSA in a mouse model of the disease.
In a previous study, the same research group elegantly showed that ARSA delivery to the brain using AAV5 under the phosphoglycerate kinase (PGK) promoter has limited transduction but robust effects in alleviating disease progression in MLD mice, likely due to the secretion and uptake of the ARSA protein (Sevin et al., 2006).
Here Piguet and colleagues evaluate the advantages of using AAVrh.10 driving ARSA expression under cytomegalovirus/β-actin hybrid promoter to brain oligodendrocytes and neurons in MLD mice in comparison to AAV5 delivery with PGK promoter. The cytomegalovirus/β-actin hybrid promoter, which appears to result in a robust transduction profile, is composed of the enhancer from the cytomegalovirus immediate early gene, the promoter, the splice donor and intron from the chicken β-actin gene, and the splice acceptor from rabbit β-globin.
The authors convincingly demonstrate that after single intrastriatal injection of 2.3×109 vector genome copies (VG) of AAVrh.10cuARSA or AAV5-PKG-ARSA, AAVrh.10 resulted in 3.9-fold higher expression of ARSA in brains, interestingly in areas distant from the injection site. Also, AAVrh.10 resulted in transduction of both oligodendrocytes and neurons, as opposed to AAV5, with which only neurons were transduced. The wider expression of ARSA in the brain of MLD animals treated with AAVrh.10 resulted in a faster correction of sulfatide accumulation in oligodendrocytes. Remarkably, the same effect is obtained when mice are treated at young (8 months) or old (16 months) age, which can have important implications in the translation to the clinic.
The striking improvement in enzyme distribution is likely due to two main components. The first is the ability of AAVrh.10 to be efficiently retrotransported through synapses, increasing the number of neurons taking up the virus even in areas far from the injection site. The second is not related to the viral vector used, but to the ability of ARSA to be secreted by neurons and taken up by other cells, most importantly by oligodendrocytes.
In this context, using a vector that can efficiently transduce both neurons and oligodendrocytes is essential to obtain a wide distribution. The use of AAVrh.10 in myelinating disorders is very promising because the target cells, i.e. oligodendrocytes, can be reached by direct virus transduction in the areas close to the injection site, but also, and most important, this vector is efficiently transported along axons and through synapses, increasing the targeted area. In this way, more oligodendrocytes can be “indirectly” targeted by taking up the enzyme produced by surrounding neurons. In diseases of this kind, in which neurons are not primarily affected, they can be used as vehicles and “factories” to obtain a wider distribution of therapeutic molecules.
The distribution study carried out with AAVrh.10 GFP, however, showed that when the protein delivered is not secreted or taken up, the use of this vector results in an equally efficient neuronal transduction, but much more limited distribution in glial cells, where the virus cannot be transported. This limitation would probably also occur should the molecule of interest be under an oligodendrocyte-specific promoter. This strategy could increase the number of glial cells expressing the gene of interest in the areas surrounding the injection site but would limit the potential of the molecules that are secreted by neurons to be taken up in areas far from the injection site. However, this hypothesis still needs to be tested in the case of ARSA, and these experiments would provide important information on the advantages of using cell type–specific promoters to target glial cells. The demonstration of targeting oligodendrocytes is exciting since limited reports have shown that AAV vectors target oligodendrocytes efficiently (Lawlor et al., 2009; Piguet et al., 2012). Indeed oligodendrocytes are a cellular target that would be optimal for diseases such as multiple sclerosis and have most recently been implicated in motor neuron disease progression in amyotrophic lateral sclerosis (Lee et al., 2012).
In MLD, Piguet and colleagues demonstrated in this issue the ability to target the CNS early and at a mid-stage of disease progression. What remains to be fully determined is whether the extent of benefit is equal at these two time points (8 vs. 16 months of age) and what is the potential point of no benefit in this disease. Additionally, it would be highly relevant in view of clinical trials to determine whether targeting more cells at a symptomatic stage dramatically changes disease progression and whether it might even revert the phenotype. Nevertheless, this study reports an important finding to improve gene delivery for severe lysosomal storage disorders and thus sets the stage for clinical trial development to treat this devastating genetic disorder.
AAVRh.10 is currently in trials to treat children with late infantile neuronal ceroid lipofuscinosis, a lysosomal storage disorder caused by mutations in the CLN2 gene and a deficiency of tripeptidyl peptidase I. To date there have been no adverse events, leading to excitement and hope for lysosomal storage disorders using this delivery approach.