Engineered arylsulfatase A with increased activity, stability and brain delivery for therapy of metachromatic leukodystrophy.

A deficiency of human arylsulfatase A (hASA) causes metachromatic leukodystrophy (MLD), a lysosomal storage disease characterized by sulfatide accumulation and central nervous system (CNS) demyelination. Efficacy of enzyme replacement therapy (ERT) is increased by genetic engineering of hASA to elevate its activity and transfer across the blood-brain barrier (BBB), respectively. To further improve the enzyme's bioavailability in the CNS, we mutated a cathepsin cleavage hot spot and obtained hASAs with substantially increased half-lives. We then combined the superstabilizing exchange E424A with the activity-promoting triple substitution M202V/T286L/R291N and the ApoEII-tag for BBB transfer in a trimodal modified neoenzyme called SuPerTurbo-ASA. Compared with wild-type hASA, half-life, activity, and M6P-independent uptake were increased more than 7-fold, about 3-fold, and more than 100-fold, respectively. ERT of an MLD-mouse model with immune tolerance to wild-type hASA did not induce antibody formation, indicating absence of novel epitopes. Compared with wild-type hASA, SuPerTurbo-ASA was 8- and 12-fold more efficient in diminishing sulfatide storage of brain and spinal cord. In both tissues, storage was reduced by ∼60%, roughly doubling clearance achieved with a 65-fold higher cumulative dose of wild-type hASA previously. Due to its enhanced therapeutic potential, SuPerTurbo-ASA might be a decisive advancement for ERT and gene therapy of MLD.

Brain cell type-specific endocytosis of arylsulfatase A identifies limitations of enzyme-based therapies for metachromatic leukodystrophy.

Enzyme replacement therapies, allogeneic bone marrow transplantation and gene therapies are treatment options for lysosomal storage diseases caused by inherited deficiencies of soluble lysosomal enzymes. Independent from the approach, the enzyme must be delivered to lysosomes of deficient patient cells. Little is known about the dissemination of enzyme within a tissue where cells compete for uptake via different receptor systems, binding affinities and endocytic rates. To evaluate dissemination and lysosomal targeting of a lysosomal enzyme in the CNS, we analysed receptor-mediated endocytosis of arylsulfatase A (ASA) by different types of brain-derived cell lines and primary murine brain cells. For ASA expressed by chinese hamster ovary cells for enzyme replacement therapy of metachromatic leukodystrophy, endocytic rates decline from microglia to neurons and astrocytes and to oligodendrocytes. Only immature oligodendrocytes endocytose significant amounts of enzyme. Uptake by non-microglial cells is due to mannose 6-phosphate receptors, whereas several receptor systems participate in endocytosis by microglial cells. Interestingly, ASA expressed by microglial cells cannot be taken up in a mannose 6-phosphate dependent manner. The resulting failure to correct non-microglial cells corroborates in vivo data and indicates that therapeutic effects of allogeneic bone marrow transplantation and hematopoietic stem cell gene therapy on metachromatic leukodystrophy are independent of metabolic cross-correction of neurons, astrocytes and oligodendrocytes by receptor-mediated endocytosis.

Deletion of fatty acid amide hydrolase reduces lyso-sulfatide levels but exacerbates metachromatic leukodystrophy in mice.

An inherited deficiency of arylsulfatase A (ASA) causes the lysosomal storage disease metachromatic leukodystrophy (MLD) characterized by massive intralysosomal storage of the acidic glycosphingolipid sulfatide and progressive demyelination. Lyso-sulfatide, which differs from sulfatide by the lack of the N-linked fatty acid, also accumulates in MLD and is considered a key driver of pathology although its concentrations are far below sulfatide levels. However, the metabolic origin of lyso-sulfatide is unknown. We show here that ASA-deficient murine macrophages and microglial cells express an endo-N-deacylase that cleaves the N-linked fatty acid from sulfatide. An ASA-deficient astrocytoma cell line devoid of this activity was used to identify the enzyme by overexpressing 13 deacylases with potentially matching substrate specificities. Hydrolysis of sulfatide was detected only in cells overexpressing the enzyme fatty acid amide hydrolase (FAAH). A cell-free assay with recombinant FAAH confirmed the novel role of this enzyme in sulfatide hydrolysis. Consistent with the in vitro data, deletion of FAAH lowered lyso-sulfatide levels in a mouse model of MLD. Regardless of the established cytotoxicity of lyso-sulfatide and the anti-inflammatory effects of FAAH inhibition seen in mouse models of several neurological diseases, genetic inactivation of FAAH did not mitigate, but rather exacerbated the disease phenotype of MLD mice. This unexpected finding was reflected by worsening of rotarod performance, increase of anxiety-related exploratory activity, aggravation of peripheral neuropathy, and reduced life expectancy. Thus, we conclude that FAAH has a protective function in MLD and may represent a novel therapeutic target for treatment of this fatal condition.

Evolutionary redesign of the lysosomal enzyme arylsulfatase A increases efficacy of enzyme replacement therapy for metachromatic leukodystrophy.

Protein engineering is a means to optimize protein therapeutics developed for the treatment of so far incurable diseases including cancers and genetic disorders. Here we report on an engineering approach in which we successfully increased the catalytic rate constant of an enzyme that is presently evaluated in enzyme replacement therapies (ERT) of a lysosomal storage disease (LSD). Although ERT is a treatment option for many LSDs, outcomes are lagging far behind expectations for most of them. This has been ascribed to insufficient enzyme activities accumulating in tissues difficult to target such as brain and peripheral nerves. We show for human arylsulfatase A (hARSA) that the activity of a therapeutic enzyme can be substantially increased by reversing activity-diminishing and by inserting activity-promoting amino acid substitutions that had occurred in the evolution of hominids and non-human mammals, respectively. The potential of this approach, here designated as evolutionary redesign, was highlighted by the observation that murinization of only 1 or 3 amino acid positions increased the hARSA activity 3- and 5-fold, with little impact on stability, respectively. The two kinetically optimized hARSA variants showed no immunogenic potential in ERT of a humanized ARSA knockout mouse model of metachromatic leukodystrophy (MLD) and reduced lysosomal storage of kidney, peripheral and central nervous system up to 3-fold more efficiently than wild-type hARSA. Due to their safety profile and higher therapeutic potential the engineered hARSA variants might represent major advances for future enzyme-based therapies of MLD and stimulate analogous approaches for other enzyme therapeutics.

Potential of surfactant-coated nanoparticles to improve brain delivery of arylsulfatase A.

The lysosomal storage disorder (LSD) metachromatic leukodystrophy (MLD) is caused by a deficiency of the soluble, lysosomal hydrolase arylsulfatase A (ASA). The disease is characterized by accumulation of 3-O-sulfogalactosylceramide (sulfatide), progressive demyelination of the nervous system and premature death. Enzyme replacement therapy (ERT), based on regular intravenous injections of recombinant functional enzyme, is in clinical use for several LSDs. For MLD and other LSDs with central nervous system (CNS) involvement, however, ERT is limited by the blood-brain barrier (BBB) restricting transport of therapeutic enzymes from the blood to the brain. In the present study, the potential of different types of surfactant-coated biodegradable nanoparticles to increase brain delivery of ASA was evaluated. Three different strategies to bind ASA to nanoparticle surfaces were compared: (1) adsorption, (2) high-affinity binding via the streptavidin-biotin system, and (3) covalent binding. Adsorption allowed binding of high amounts of active ASA. However, in presence of phosphate-buffered saline or serum rapid and complete desorption occurred, rendering this strategy ineffective for in vivo applications. In contrast, stable immobilization with negligible dissociation was achieved by high-affinity and covalent binding. Consequently, we analyzed the brain targeting of two stably nanoparticle-bound ASA formulations in ASA-/- mice, an animal model of MLD. Compared to free ASA, injected as a control, the biodistribution of nanoparticle-bound ASA was altered in peripheral organs, but no increase of brain levels was detectable. The failure to improve brain delivery suggests that the ASA glycoprotein interferes with processes required to target surfactant-coated nanoparticles to brain capillary endothelial cells.