Skip Navigation


Human Molecular Genetics Advance Access originally published online on March 16, 2005
Human Molecular Genetics 2005 14(9):1139-1152; doi:10.1093/hmg/ddi126
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
14/9/1139    most recent
ddi126v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (26)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Matzner, U.
Right arrow Articles by Gieselmann, V.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Matzner, U.
Right arrow Articles by Gieselmann, V.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2005. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oupjournals.org

Enzyme replacement improves nervous system pathology and function in a mouse model for metachromatic leukodystrophy

Ulrich Matzner1,*, Eva Herbst2, Kerstin Khalaj Hedayati2, Renate Lüllmann-Rauch2, Carsten Wessig3, Stephan Schröder1, Carl Eistrup4, Christer Möller4, Jens Fogh4 and Volkmar Gieselmann1

1Institut für Physiologische Chemie, Rheinische Friedrich-Wilhelms Universität, Nussallee 11, 53115 Bonn, Germany, 2Anatomisches Institut, Christian-Albrechts-Universität, Otto-Hahn-Platz 8, 24043 Kiel, Germany, 3Neurologische Klinik und Poliklinik, Bayerische Julius-Maximilians Universität, Josef-Schneider-Str. 11, 97080 Würzburg, Germany and 4Zymenex A/S, Roskildevej 12C, 3400 Hillerød, Denmark and Dalénum 13, 18170 Lidingö, Sweden

* To whom correspondence should be addressed. Tel: +49 228734744; Fax: +49 228732416; Email: matzner{at}institut.physiochem.uni-bonn.de

Received January 7, 2005; Accepted March 8, 2005


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
A deficiency of arylsulfatase A (ASA) causes the lysosomal storage disease metachromatic leukodystrophy, which is characterized by accumulation of the sphingolipid 3-O-sulfogalactosylceramide (sulfatide). Sphingolipid storage results in progressive demyelination and severe neurologic symptoms. The disease is lethal, and curative therapy is not available. To assess the therapeutic potential of enzyme replacement therapy (ERT), ASA knockout mice were treated by intravenous injection of recombinant human ASA. Plasma levels of ASA declined with a half-time of ~40 min, and enzyme was detectable in tissues within minutes after injection. The uptake of injected enzyme was high into liver, moderate into peripheral nervous system (PNS) and kidney and very low into brain. The apparent half-life of endocytosed enzyme was ~4 days. A single injection led to a time- and dose-dependent decline of the excess sulfatide in PNS and kidney by up to 70%, but no reduction was seen in brain. Four weekly injections with 20 mg/kg body weight not only reduced storage in peripheral tissues progressively, but also were surprisingly effective in reducing sulfatide storage in brain and spinal cord. The histopathology of kidney and central nervous system was ameliorated. Improved neuromotor coordination capabilities and normalized peripheral compound motor action potential demonstrate the benefits of ERT on the nervous system function. Enzyme replacement may therefore be a promising therapeutic option in this devastating disease.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
Metachromatic leukodystrophy (MLD) is an inborn error of the sphingolipid catabolism and results from an inherited functional deficiency of the lysosomal enzyme arylsulfatase A (ASA; E.C. 3.1.6.8 [EC] .) (1Go). ASA-deficient cells accumulate the acidic sphingolipid 3-O-sulfogalactosylceramide (sulfatide), which is hydrolyzed by ASA under normal circumstances. The accumulating sulfatide is incorporated into lysosomal storage deposits, which exhibit typical ultrastructure and metachromatic staining characteristics. Lysosomal storage is observed in a variety of tissues and cell types including epithelial cells and myelinating cells. As the disease advances, Schwann cells and oligodendrocytes are compromised, resulting in a widespread and devastating demyelination in the peripheral (PNS) and central nervous system (CNS). In the most frequent form of the disease (late infantile form), neurologic symptoms start around the age of 2, progress rapidly and cause premature death within the first decade. A considerable number of patients suffer from a juvenile or adult form of the disease. Allogeneic hematopoietic stem cell transplantation is the only therapeutic option for MLD to date. However, while deceleration of disease progression may occur in late-onset patients, transplantation usually has no effect on severely affected early-onset patients (2Go).

ASA knockout mice develop a disease which resembles MLD (3Go–5Go). Thus, they display storage deposits with a distribution and ultrastructure which is virtually identical to those in patients. The mice develop neurologic symptoms reminiscent of the human disease comprising gait disturbances, reduced motor coordination abilities and hyperactivity (3Go,6Go,7Go). The symptoms become apparent at ~1 year of age, but they do not reduce the life expectancy of the mice. The mild phenotype has been explained by the lack of widespread demyelination (3Go,8Go,9Go). The limited demyelination in mice can be attributed to the short life span, which does not allow for the development of cellular dysfunctions, causative for demyelination. The mice therefore represent an appropriate animal model, particularly to investigate therapeutic interventions in an early stage of the human disease.

ASA knockout mice have been recently treated by bone marrow stem cell gene therapy (7Go,10Go–13Go). Bone marrow cells from ASA-deficient donor mice were transduced with retroviral vectors, which mediate overexpression of the human ASA, and the genetically modified cells were subsequently transplanted into ASA knockout mice. The gene therapy studies demonstrated that the human ASA can compensate for the lack of the murine ASA in vivo. Lysosomal storage of sulfatide could be reduced, for example, in epithelial cells of kidney and liver (7Go). The metabolic correction of these cells is due to a transfer of human ASA from donor-type cells to recipient cells via a release/uptake mechanism (14Go). This cell-to-cell transfer pathway is shared by other soluble lysosomal enzymes and involves delivery of a fraction of the newly synthesized enzyme from enzyme-producing cells, the diffusion of the released enzyme to recipient cells and its receptor-mediated endocytosis and targeting to the lysosomal compartment.

As well as transplantation-based therapy approaches enzyme replacement therapy (ERT) depends on receptor-mediated endocytosis to deliver a therapeutic enzyme to the lysosomes of deficient cells (15Go,16Go). In contrast to transplantation-based regimens the enzyme is, however, recombinantly expressed by cultured cells and injected into the circulation as a purified product. The curative potential of ERT has been demonstrated for a number of lysosomal storage diseases (LSDs). ERT with glucocerebrosidase is routinely used in the non-neuronopathic type I variant of Gaucher disease. More than 10 years of clinical experience indicate that the therapy is a big success in managing this disease (17Go). More recently ERTs for Fabry disease, mucopolysaccharidosis type I and other LSDs have been developed (15Go). ERT has been shown to reduce lysosomal storage in the visceral tissues, but not in the CNS. The inaccessibility of the CNS by ERT has been considered to be a general limitation of ERT for LSDs. However, recent preclinical studies in mouse models for aspartylglucosaminuria and {alpha}-mannosidosis revealed that intravenous injection of the therapeutic enzyme can also reduce cerebral storage in certain LSDs (18Go,19Go). The reasons why the CNS is amenable to ERT in some species and/or diseases are entirely unclear. Data on ERT in MLD are so far lacking. To assess the corrective potency of ERT in MLD, we therefore analyzed the pharmacokinetics and pharmacodynamics of recombinant human ASA (rhASA) in ASA knockout mice. We provide the first proof-of-principle for the feasibility of ERT in an animal model for MLD.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
Characterization of the CHO-rhASA
SDS–PAGE and MALDI-TOF analysis of preparations of rhASA expressed in Chinese hamster ovary cells (CHO-rhASA) detected a compound of correct size and the absence of contaminants (Fig. 1A and B). Wild-type human ASA has three N-linked carbohydrates with a combined molecular mass of ~5–6 kDa (20Go). Treatment with PNGase F reduced the apparent molecular mass of CHO-rhASA, indicating glycosylation of the recombinant enzyme (Fig. 1A). After reacting the CHO-rhASA with low concentrations of PNGase F, MALDI-TOF analysis revealed four deglycosylation intermediates (Fig. 1C), possibly representing the polypeptide linked to three, two, one and no N-glycan(s). The mass difference between the products with the highest and the lowest mass is in the range of 5 kDa, suggesting full glycosylation of the enzyme. To evaluate the mannose phosphorylation of CHO-rhASA, the mannose-6-phosphate (M6P)-dependent endocytosis of the enzyme was evaluated by an in vitro feeding assay. The assay revealed efficient endocytosis of CHO-rhASA by BHK cells (Fig. 1D). Furthermore, uptake could be completely blocked by M6P, but not by glucose-6-phosphate (G6P). It can be concluded that CHO-rhASA bears M6P residues.



View larger version (38K):
[in this window]
[in a new window]
  		Figure 1. In vitro analyses of CHO-rhASA. (A) SDS–PAGE of naive (–) and deglycosylated (+) enzyme. Gels were stained with Coomassie blue. The masses of protein standards (std) are indicated. Treatment with PNGase F causes a shift of the ASA band owing to the loss of N-linked carbohydrates. (B) MALDI-TOF analysis. rhASA isolated from secretions of CHO cells exhibits a correct size of 57 kDa and the enzyme preparation lacks major contaminants. The minor peak at 29 kDa (ASA/2) represents the doubly charged molecule. (C) MALDI-TOF analysis of CHO-rhASA treated with a low concentration of PNGase F. The limited deglycosylation yields four products. They presumably represent rhASA, which bears three, two, one or no N-glycan(s). The mass pattern therefore suggests that all three potential N-glycosylation sites of the CHO-rhASA are glycosylated. (D) M6P-dependent endocytosis of CHO-rhASA by BHK cells analyzed by ELISA. BHK cells were incubated with 1 µg CHO-rhASA per ml medium for 20 h in the presence or absence of 10 mM M6P. Other dishes were supplemented with 10 mM G6P as a control. The complete block of endocytosis by M6P indicates that BHK cells internalize CHO-rhASA via M6P receptors. This indicates the presence of M6P residues on CHO-rhASA. The data are expressed as mean ±SD, n=3.

 
Pharmacokinetics and biodistribution of CHO-rhASA after single dosing
ASA knockout mice were first treated by a single injection of CHO-rhASA into the tail vein.

To determine the rate of rhASA clearance from the circulation, plasma levels of enzyme were analyzed at different times after infusion of 20 or 40 mg enzyme per kg body weight (Fig. 2A). For both doses, the plasma levels reached a maximum in the first minutes after injection and declined from then on. Irrespective of the administered dose, rhASA was cleared from plasma with a half-time of ~40 min. To evaluate the kinetics of tissue uptake, mice were perfused at different times after infusion and several organs were analyzed for rhASA concentrations. Immunoreactivity for rhASA could be detected 10 min after treatment with 40 mg/kg in all tissues (Fig. 2B and C). In liver which acquired the highest enzyme concentration (discussed subsequently), the enzyme levels increased ~4-fold within the next 5 h and dropped thereafter until day 14 to ~4% of the maximum level (Fig. 2B). The kinetics was similar for kidney, sciatic nerve and brain (Fig. 2C). Independent of the differences between the tissue-specific uptake rates, the enzyme was eliminated from all tissues within a comparable time course. Thus, the half-life of immunologically detectable rhASA in liver, kidney, sciatic nerve and brain ranged ~4 days (Fig. 2B and C). It is striking that the maximum concentration of rhASA differed by more than three orders of magnitude between liver and brain 5 h after infusion (Fig. 2B and C). To analyze the biodistribution of rhASA in more detail, ASA knockout mice were infused with increasing doses of rhASA and tissues were analyzed 8 days later by ELISA. A roughly linear, dose-dependent increase of the enzyme concentration could be detected in kidney and peripheral nerves (Fig. 2D). However, the majority of the infused enzyme was found in liver (shown for 20 mg/kg in Fig. 2E). When compared with liver, the rhASA concentrations were ~7% in kidney, <0.05% in brain and 12–15% in sciatic nerve and brachial plexus. Taking the different masses of these tissues into consideration, it can be calculated that ~97% of the retrievable enzyme was found in liver, ~3% in kidneys and <0.1% in the CNS and peripheral nerves.



View larger version (23K):
[in this window]
[in a new window]
  		Figure 2. Pharmacokinetics of CHO-rhASA after a single injection into the tail vein of ASA-deficient mice. All data are expressed as mean ±SD. (A) rhASA levels in plasma in the first hour after injection of 20 mg/kg (closed circles, n=9) or 40 mg/kg (open circles, n=11). The concentration was determined by ELISA. (B) rhASA levels in liver at various times after administration of 40 mg/kg. The concentration was determined by ELISA (n=3). (C) Tissue kinetics of rhASA after administration of 40 mg/kg in kidney (open bars), sciatic nerve (hatched bars) and brain (closed bars) as determined by ELISA (n=3) (nd, not determined). (D) Tissue levels of rhASA 8 days after injection of different enzyme doses (n=5). The concentration was determined by ELISA for kidney (open bars), brachial plexus (closed bars) and sciatic nerve (hatched bars). Tissues from untreated wild-type mice and mock-treated ASA knockout mice were analyzed as negative controls. Control homogenates from the sciatic nerve show some unspecific background signal because the incubation times had to be prolonged to quantify the specific immunoreactivity in the very small nerve samples from treated mice. (E) Relative tissue distribution of rhASA 8 days after injection of 20 mg/kg. The rhASA concentration was measured by ELISA and normalized on the level in liver (n=3).

 
Reduction of sulfatide levels after single dosing of CHO-rhASA
To evaluate the therapeutic potential of CHO-rhASA treatment, ASA knockout mice were intravenously infused with a single dose of 40 mg CHO-rhASA per kg body weight and lipids were extracted from kidney 8 days later. Thin-layer chromatography (TLC) of the lipid extracts revealed a prominent decline of sulfatide levels in enzyme-treated ASA knockout mice when compared with mock-treated controls (Fig. 3A).



View larger version (41K):
[in this window]
[in a new window]
  		Figure 3. Sulfatide clearance from tissues after single treatment of ASA-deficient mice with CHO-rhASA. The levels of sulfatide (closed bars), cholesterol (open bars) and sphingomyelin (hatched bars) were determined by TLC and are expressed as mean of arbitrary units ±SD. Asterisks indicate a statistically significant difference to mock-treated controls (Student's t-test, P<0.05). (A) Analysis of kidney lipids by TLC. Lipids were extracted from kidneys of different experimental groups and incubated under alkaline conditions to hydrolyze phosphoglycerolipids and cholesterylester. The reaction products were separated by TLC and visualized and analyzed by densitometry. The loading volumes were normalized on the protein concentration of the crude tissue homogenate used for lipid extraction. Increasing amounts of cholesterol (chol), sulfatide (sulf) and sphingomyelin (sm) were loaded as standards. (B) Lipid levels in kidney at different times after injection of 40 mg/kg (n=3). (C) Lipid levels in kidney 8 days after injection of different enzyme doses (n=5). (D) Lipid levels in brain at different times after injection of 40 mg/kg (n=3). (E) Lipid levels in the sciatic nerve 8 days after injection of different enzyme doses (n=5). (F) Lipid levels in the brachial plexus 8 days after injection of different enzyme doses (n=5).

 
The time dependence of sulfatide reduction was investigated in a second experiment. For that purpose, sulfatide levels were determined in kidneys at different times after injection of 40 mg/kg (Fig. 3B). Significant clearance of sulfatide could be detected as soon as 5 h after infusion and the extent of reduction increased until day 8. At that time, around two-thirds of the excess sulfatide were cleared from kidney. Six days later sulfatide re-appeared and the residual mean sulfatide level rose by ~22%. To determine the dose-dependence of sulfatide reduction, mice were treated with different doses of CHO-rhASA and analyzed 8 days later. Ten mg/kg resulted in a significant decline of sulfatide storage in kidney (Fig. 3C). The extent of sulfatide clearance increased with increasing doses, and a roughly linear relation between dose and loss of sulfatide was detectable.

To evaluate effects of enzyme replacement on the lipid catabolism of the nervous system, total brain, sciatic nerve and brachial plexus from the differently treated animals were also analyzed. When compared with kidney, where sulfatide levels increased ~10-fold in aged ASA knockout mice (Fig. 3B and C), the nervous system showed only a roughly 2-fold elevation of sulfatide levels (Figs 3D–F and 4B–D). Single dosing with 40 mg/kg had no effect on sulfatide storage in brain (Fig. 3D). However, a significant decline was detectable in the sciatic nerve (Fig. 3E) and the brachial plexus (Fig. 3F). Enzyme doses of 10 and 20 mg/kg reduced the mean sulfatide storage in peripheral nerves, to a lower extent and the difference to control tissues was statistically not significant (Fig. 3E and F).



View larger version (32K):
[in this window]
[in a new window]
  		Figure 4. Lipid levels in (A) kidney (B) brain, (C) sciatic nerve and (D) brachial plexus after repeated dosing of 20 mg CHO-rhASA/kg once a week. The knockout mice were treated with up to four infusions of CHO-rhASA. Controls were mock-treated with four injections of buffer. The levels of sulfatide (closed bars), cholesterol (open bars) and sphingomyelin (hatched bars) were analyzed 8 days after the last dosing and are expressed as mean ±SD (n=3). Asterisks indicate a significant difference in the sulfatide level in enzyme-treated knockout mice when compared with mock-treated controls (Student's t-test, P<0.05).

 
Reduction of sulfatide storage after repeated dosing of CHO-rhASA
The unexpected high efficacy of single enzyme doses in reducing sulfatide levels in peripheral tissues provided the rationale to evaluate the therapeutic potential of repeated injections. We chose a treatment schedule based on up to four injections of 20 mg CHO-rhASA/kg once a week. Sulfatide levels were analyzed 8 days after the last injection in kidney, peripheral nerves and brain of mice treated by one, two, three or four injections (Fig. 4).

TLC revealed that sulfatide declined progressively with an increasing number of infusions in all peripheral tissues. After the fourth treatment, ~65% of excess sulfatide was cleared from kidney and brachial plexus (Fig. 4A and D) and ~82% from the sciatic nerve (Fig. 4C). In brain, no decline of sulfatide was detectable after the first, second and third treatments (Fig. 4B). However, after the fourth treatment, the sulfatide level was significantly reduced by 13% on average. This represents a clearance of 30% of excess sulfatide from brain.

To verify the sulfatide reduction in the CNS of mice treated by four injections, brain and spinal cord were histologically analyzed. In the CNS white and gray matter of mock-treated knockout mice, the distribution of sulfatide storage as visualized in 100 µm sections by selective staining with alcian blue was identical to that previously described for ASA knockout mice (9Go). According to size, two types of storage patterns could be distinguished: (a) large (>20 µm) alcianophilic profiles corresponding to phagocytes and macrocellular neurons stuffed with sulfatide-storing lysosomes and (b) small (<5 µm) alcianophilic profiles corresponding to oligodendrocytic processes filled with numerous storage lysosomes and possibly other cell types with low amounts of storage material (Fig. 5). The large alcianophilic profiles could be ascribed to phagocytes and neurons, on the basis of their appearance in the thick sections (Fig. 5, inset). In neuronal perikarya, the nucleus is clearly visible as an unstained structure surrounded by alcianophilic material, whereas phagocytes appear as bizarre-shaped spots, in which the nucleus usually cannot be seen because it is superimposed by numerous large storage lysosomes. The ultrastructural equivalents of these light microscopic features have been described (9Go). The visual inspection of 100 µm slices from brain and spinal cord of rhASA-treated mice revealed that the large sulfatide-storing phagocytes could no longer be identified neither in the white nor in the gray matter, whereas the large alcianophilic neurons in the gray matter were still observed (as shown for the brain stem in Fig. 5D). Small alcianophilic profiles were abundant in rhASA-treated mice, both in the white and gray matter.



View larger version (149K):
[in this window]
[in a new window]
  		Figure 5. Sulfatide storage in the CNS as histochemically demonstrated by incubation with alcian blue (Materials and Methods). Coronal thick sections (100 µm) through the brain stem of a mock-treated ASA knockout mouse (A, C, E and G) and an enzyme-treated ASA knockout mouse (B, D, F and H). (A and B) Overview to outline the regions shown at higher magnification in the photomicrographs subsequently. 7n, root of facial nerve; CbN, cerebellar nucleus; icp, inferior cerebellar pedunculus; PnC, pontine reticular nucleus; VC, ventral cochlear nucleus; Ve, vestibular nucleus. (C and D) Abducens nucleus (6 N) and adjacent regions (7g, genu of facial nerve). In the mock-treated mouse, alcianophilic material (sulfatide) is seen in many cells, of which phagocytes and neurons can be identified (marked by triangles and circles, respectively, in the inset). In the rhASA-treated mouse, alcianophilic material is seen mainly in neurons. (E and G) and (F and H) Inferior cerebellar pedunculus (icp) as an example of a white matter tract. In the mock-treated mouse, numerous large alcianophilic phagocytes are seen, only a few of which are in the optic focus in (G) (some are marked by circles). The small alcianophilic granules may be associated with oligodendrocytes, which cannot be identified at this magnification, however. In the rhASA-treated mouse, the icp shows only small alcianophilic structures suggesting that the sulfatide-storing phagocytes are decreased in size and/or number. The overall staining of the cerebellar and vestibular nuclei in (F) is reduced because alcianophilia is restricted mainly to neurons and has largely disappeared from phagocytes. Bars: 500 µm in (A and B); 100 µm in (C and D); 200 µm in (E and F); 10 µm in (G, H and insets of C and D).

 
Storage was further investigated by automated quantification of storage deposits in spinal cord using the image-process software Image J. For a separate analysis of the large storage deposits typical for phagocytes and neurons and the small deposits of oligodendrocytes, two size categories of alcian blue-positive particles were differentiated. A diameter of 8 µm (resembling an area of 20 pixels under the chosen settings) was taken as threshold for the allocation into the two categories. The statistical analysis of the raw data revealed a significant decline of the frequency of particles ≥8 µm by 86 and 68% in the white and gray matter of enzyme-treated mice, respectively (Table 1). In contrast, the frequency of particles <8 µm was unchanged in the two regions.


View this table:
[in this window]
[in a new window]
  		Table 1. Quantification of sulfatide storage deposits in gray and white matter of spinal cord
 
In the sciatic nerves (data not shown), storage was seen in Schwann cells and macrophages, the latter of which occurred, however, rather infrequent. Mock-treated and rhASA-treated animals did not show clear differences.

Apart from the nervous system, the kidney was also histologically analyzed. Kidneys of mock-treated mice displayed the same sulfatide storage patterns as previously described (21Go). Storage was intense in thin limbs and thick ascending limbs of Henle's loop and moderate in distal convoluted tubules and collecting ducts (Fig. 6). After four enzyme injections, the storage material was almost completely cleared from the distal convoluted tubules (Fig. 6J) and storage was clearly reduced in the upper portions of the thick ascending limbs (Fig. 6G). However, the storage in thin limbs and collecting ducts persisted and resembled that of mock-treated mice (Fig. 6B–E).



View larger version (150K):
[in this window]
[in a new window]
  		Figure 6. Sulfatide storage in the kidney as histochemically demonstrated by incubation of 100 µm slices with alcian blue (Materials and Methods). (A) Wild-type mouse: weak staining is seen in the inner stripe of the outer medulla (iS-oM), whereas the outer stripe of outer medulla (oS-oM) and the cortex (C) are unstained. (B and C) Mock-treated ASA knockout mouse: severe sulfatide storage (alcianophilic material) is seen in the tubules of the inner stripe of outer medulla, whereas several profiles show sulfatide storage in the outer stripe and cortex. (D and E) ASA knockout mouse treated with four doses of 20 mg CHO-rhASA/kg. The cortex is devoid of alcianophilic material. In the outer stripe, staining is reduced. In the inner stripe, staining appears unchanged when compared with the mock-treated animal. Bars represent 500 µm in (A, B and D) and 200 µm in (C and E). (F–J) Semithin sections stained with toluidine blue. (F and H) Mock-treated ASA knockout mouse, thick ascending limb (TAL) in the outer stripe and distal convoluted tubules (DCT) in the cortex. The intensely stained cytoplasmic inclusions correspond to lysosomes filled with sulfatide as shown previously (21Go). (G and J) ASA knockout mouse treated with four doses of 20 mg CHO-rhASA/kg, representative segments of the nephron corresponding to those in (F and H). The storage material is reduced in the TAL and absent from the DCT profiles. G, glomerulus; PT, proximal tubule. Bar represents 20 µm in (F–J).

 
The analysis of the kidney also revealed a significant 1.4-fold increase of the kidney wet weight in 9-month-old ASA knockout mice when compared with wild-type controls (data not shown). Interestingly, enzyme replacement reduced and partially normalized the increased kidney size. The extent of reduction was statistically significant after the third and fourth treatment (Student's t-test, P<0.05), and the kidney weight declined to 1.2-fold of normal after four injections (data not shown). In a second independent experiment, the mean kidney weight of 12-month-old knockout mice was 1.5-fold increased when compared with wild-type mice (data not shown). It declined significantly (Student's t-test, P<0.05) to 1.1-fold of normal after four injections of enzyme (data not shown). Liver and brain were weighed as controls and no significant differences were detectable between the experimental groups for these organs (data not shown).

Enzyme replacement improves the nervous system function
ASA knockout mice develop nerve conduction impairments and a number of neurologic symptoms. To measure putative therapeutic effects on neurologic parameters, the compound motor action potential (CMAP) and the rotarod performance were examined.

Neurophysiological studies of the sciatic nerve were done 6 days after the fourth treatment of 12-month-old mice. After distal stimulation, age-matched wild-type control animals showed a normal CMAP with an amplitude of 19.0±1.7 mV (mean±SD, n=8), a latency of 0.84±0.11 ms and a duration of 3.3±0.36 ms (Fig. 7A). Mock-treated ASA-deficient animals showed a less compact motor response with significantly reduced mean amplitude (15.6±3.9 mV, P<0.05) and increased duration (4.1±0.31 ms, P<0.01). The mean latency and the nerve conduction velocity were 0.81±0.09 ms and 45.4±8.8 m/s, respectively, and not significantly different from that of wild-type mice (Fig. 7A). Treatment with CHO-rhASA resulted in an increase of the amplitude to normal values (20.4±5.9 mV, P<0.05) and a significant decrease of the duration (3.8±0.35 ms, P<0.05). Thus, the impaired conduction of the sciatic nerve was virtually normalized after treatment with rhASA. Similar data were obtained after proximal stimulation, and the data were reproduced in an independent experiment using mice of another treatment series (data not shown).



View larger version (13K):
[in this window]
[in a new window]
  		Figure 7. Functional effects of repeated dosing with 20 mg CHO-rhASA/kg once weekly. (A) Neurophysiological parameters determined in wild-type mice (closed bars), mock-treated ASA knockout mice (open bars) and ASA knockout mice 6 days after the fourth injection of CHO-rhASA (hatched bars). The indicated parameters were measured in the intrinsic foot musculature after distal stimulation of the sciatic nerve. Data are expressed as mean ±SD (n=7–8). Asterisks indicate a significant difference (Student's t-test, P<0.05). (B) Rotarod performance of mice at a mean age around 9 months (9.2±1.1 months). ASA knockout mice (closed triangles, n=7) were analyzed at days 2 and 3 after the third treatment with 20 mg CHO-rhASA/kg. Age-matched wild-type mice (circles, n=10) and mock-treated ASA knockout mice (open triangles, n=10) were analyzed as controls in parallel. The percentages of mice which were able to balance on a slowly rotating rod for at least 4 min were determined in four consecutive trials. (C) Rotarod performance of mice at a mean age of 12 months (11.8±1.1 months). The treated mice were tested at days 2 and 3 after the third injection of 20 mg CHO-rhASA/kg. Legend and group sizes are as in (B), except n=14 for rhASA-treated knockout mice.

 
Previous behavioral tests revealed progressive deficits of ASA knockout mice in balancing on a slowly rotating rod (6Go,7Go). To determine effects of treatment on motor coordination, mice were tested before the first and after the third infusion of CHO-rhASA by rotarod experiments. In a first study, mice at a mean age of ~9 months were analyzed. In the test before treatment, wild-type mice were successful in 32 of 40 trials (80%), whereas the two groups (yet untreated) of ASA knockout mice were successful in 22 of 40 (55%) and 25 of 40 (63%) trials (data not shown). Thus, the data confirmed that behavioral deficits of ASA knockout mice are already detectable, but still comparably mild at 9 months of age (6Go). After treatment of one group of knockout mice with three weekly doses of 20 mg CHO-rhASA/kg, the same three groups were re-analyzed ~4 weeks later. When compared with the first test, the mean success of rhASA-treated mice was improved by 27% and reached 82%. In contrast to this group, the mean performance of wild-type and mock-treated controls was only improved by 10 and 5%, respectively (Fig. 7B). Thus, by a combination of treatment and training, 9-month-old ASA knockout mice acquired the ability to stay on the rod with a higher frequency than untrained wild-type mice of the same age.

To investigate effects on more advanced motor coordination disabilities, 12-month-old mice (3 months older than the mice in the first study) were analyzed in another experiment. Now the percentages of successful mice before treatment were 43% (wild-type controls), 18% (ASA knockouts destined for mock treatment) and 10% (ASA knockouts destined for treatment with rhASA) on average (data not shown). After treatment, 65% of wild-type controls and 13% of mock-treated ASA knockout mice were successful (Fig. 7C). However, the mean percentage of rhASA-treated knockout mice increased to 31%. Thus, also in elder mice with progressed coordination impairments, motor coordination could be substantially improved by three treatments with CHO-rhASA.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
MATERIALS AND METHODS
 REFERENCES
 
The corrective potency of ERT depends on the internalization and correct lysosomal targeting of the therapeutic enzyme after systemic administration. Several receptor systems, including the M6P/insulin-like growth factor II receptor, are involved in the uptake and intracellular sorting of soluble lysosomal enzymes (15Go,22Go). The histological analysis of mice, which have been recently treated by gene therapy, revealed that human ASA can diminish sulfatide storage in hepatic and renal epithelia after expression from genetically modified cells of the hematopoietic system in vivo (7Go). These results suggested that also the human ASA can be internalized via cell surface receptors and delivered to the lysosomal compartment. ERT might therefore also be a therapeutic option for MLD. A recently established protocol, which allows purification of rhASA on a large scale, enabled us to assess the potential of ERT in a mouse model for MLD.

Consistent with the concept of endocytosis via cell surface receptors, intravenously injected CHO-rhASA was efficiently cleared from the systemic circulation and started to accumulate in tissues within minutes (Fig. 2A–C). Cell culture experiments revealed that CHO-rhASA bears M6P residues and uses the M6P receptor-dependent pathway for cell entry (Fig. 1D). The BHK cells which were utilized in this in vitro assay do not express other receptors for lysosomal enzymes. Therefore, it cannot be decided to which extent other receptors, such as the mannose receptor or the asialoglycoprotein receptor, compete for the binding and endocytosis of substituted CHO-rhASA under in vivo conditions. Approximately 30% of the total amount of injected rhASA could be retrieved from dissected mouse organs 5 h after intravenous injection of 40 mg/kg (data not shown). Among the retrievable fraction, >90% was localized to liver, whereas kidney and peripheral nerves shared the vast majority of the remaining enzyme (Fig. 2E). A comparison with previous data about ASA activities in wild-type mice (10Go) suggests that enzyme levels after rhASA treatment were ~95-fold (liver), ~1.2-fold (kidney), ~0.6-fold (peripheral nerves) and ~0.001-fold (brain) of normal on average.

One intravenous injection of rhASA led to a pronounced time- and dose-dependent decline of sulfatide storage in kidney and peripheral nerves (Fig. 3). This is the first proof that ERT using ASA is effective in reducing the sulfatide storage in vivo. Notably, as soon as 5 h after injection of 40 mg/kg, a significant decline of sulfatide storage in kidney was detectable (Fig. 3B). Eight days after treatment storage was diminished to a minimum and up to 70% of the excess sulfatide had vanished from kidney (Fig. 3B) and peripheral nerves (Fig. 3E and F).

When compared with previous gene therapy experiments, in which TLC did not reveal a significant decline of the mean concentration of sulfatide in total kidney (7Go), both the velocity and the extent of storage reduction were surprising. The difference between the two studies is striking because the steady state level of human ASA in kidney, which was achieved by transplantation, was 1.3-fold of normal on average and thus virtually the same as the maximum level reached by a single injection of 40 mg CHO-rhASA/kg (discussed earlier). Thus, CHO-rhASA which was present in kidney only for a couple of days eliminated more than two-thirds of excess sulfatide, whereas a similar amount of enzyme did not reduce the mean storage when it was stably expressed from cells of the hematopoietic system for almost 1 year. The dependence of the therapeutic efficacy on the cell type which expresses the enzyme points to cell type-specific differences in the biosynthesis of human ASA. Cell culture studies suggested that human ASA is inefficiently phosphorylated by cells of the hematopoietic system, but efficiently phosphorylated by CHO and BHK cells (23Go) (Fig. 1D). Phosphorylation is, however, important for therapeutic efficacy in ASA knockout mice and animal models for other LSDs (11Go,24Go). A low phosphorylation of human ASA by hematopoietic cells and a high phosphorylation by CHO cells may thus explain the partial failure of bone marrow stem cell gene therapy and the success of ERT in ASA knockout mice.

The single-dose experiments indicated that clearance of storage is only transient and sulfatide re-accumulated in the second week after treatment (Fig. 3B). The partial re-accumulation of storage can be explained by the limited half-life of the internalized rhASA. It was in the range of 4 days (Fig. 2B and C) and thus very similar to the tissue half-lives of other lysosomal enzymes in preclinical ERT studies (18Go,25Go,26Go). The re-accumulation of sulfatide necessitated a regimen based on repeated enzyme injections to maintain or even enhance storage reduction over the long range. Repeated treatment with 20 mg CHO-rhASA per kg resulted in a step-wise decline of sulfatide storage in peripheral tissues (Fig. 4A, C and D). Up to 65 and 82% of excess sulfatide could be eliminated from kidney and peripheral nerves by four injections, respectively. The histological analysis of kidney revealed the most prominent decline of sulfatide storage in the cortex (Fig. 6). Here, storage was abolished or greatly reduced in distal convoluted tubules and in the upper portion of the thick ascending limbs of Henle's loop, respectively. At present, it is unclear why these segments of the nephron respond more clearly to ERT than other segments. Possibly, the region specificity is determined by the expression pattern of the receptor(s), which endocytose rhASA.

To our surprise, repeated dosing did reduce storage not only in peripheral tissues, but also in the CNS. This was first evidenced by TLC of brain lipids, which showed a decline of sulfatide by 13% after the fourth treatment (Fig. 4B). However, reduction of CNS storage was more clearly seen in the morphological analysis. Histology of brain and spinal cord revealed a strikingly reduced frequency of the enlarged sulfatide-storing phagocytes throughout the CNS (shown for brain stem in Fig. 5). Thus, the reduction of sulfatide levels in the CNS appears to be mainly due to clearance of lipid from these phagocytes representing activated microglial cells (3Go) rather than from neurons or oligodendroglia. This observation was confirmed by software-assisted analysis of sulfatide storage deposits in spinal cord. The recorded alcian blue-positive deposits were divided into two size categories (<8 and ≥8 µm) to differentiate between the small (<5 µm) profiles of oligodendrocytes and the large (>20 µm) profiles typical of phagocytes and neurons. The statistical analysis of the frequencies of the two categories revealed a significant decline of the large deposits from the white and gray matter by roughly 90 and 70%, respectively (Table 1). Because neuronal storage need not be considered in the white matter, the reduction by 90% seen in this region can be ascribed to the loss of sulfatide storage from phagocytes. The less dramatic clearance of large deposits from the gray matter may be explained by the persistence of storage in neuronal perikarya. In contrast to large deposits, the frequency of small deposits did not decline. At present, we cannot decide whether and to which extent other cells but oligodendroglia may have, in enzyme-treated mice, contributed to the small alcianophilic profiles, such as phagocytes with greatly reduced amounts of storage material.

It has been shown recently in animal models for Krabbe and Sandhoff diseases that microglial activation plays a major role in the pathogenesis of these related sphingolipid storage diseases (27Go,28Go). Microglial activation is also prominent in ASA knockout mice in the second year of life (3Go). Reduction of sulfatide storage in microglial cells can therefore be expected to be beneficial even in the absence of detectable clearance of sulfatide from other glial cells and neurons.

The sulfatide clearance from the CNS is difficult to explain. Because the blood–brain barrier prevents direct transfer of rhASA from the circulation to the CNS, the brain did not acquire enzyme levels exceeding 0.1% of wild-type levels during the treatment period of 4 weeks (Fig. 2C and E). Such a low enzyme level is typical for the most severe form of MLD (1Go), and it is therefore unlikely that it can be effective in correcting the catabolic defect of phagocytes in the mouse CNS. Alternatively, the clearance might be explained by a displacement of the storing phagocytes through blood-derived phagocytes, which have endocytosed rhASA prior to immigration into the CNS. However, although macrophage progenitors migrate continuously from the periphery into the adult brain and differentiate into subpopulations of the microglial compartment (29Go–33Go), this process is very slow and does not allow for the replacement of a substantial fraction of microglia within 4 weeks. Thus, it was calculated that immigration adds less than one blood-derived microglial progenitor to 1000 resident microglial cells per day in wild-type mice (30Go). Therefore, 4 weeks should permit replacement of only 2–3% of the microglial cell population. This estimate has been experimentally verified in wild-type mice (32Go), and a comparably low rate of cerebral cell immigration has been measured in ASA knockout mice by quantitative PCR following bone marrow transplantation (10Go). It is therefore not probable that the phagocytes with reduced amounts of sulfatide as seen in the enzyme-treated mice were mainly blood-derived. A third, although highly speculative, possibility is that the sulfatide is not hydrolyzed in the CNS, but exported from the phagocytes to peripheral cells by a still unrecognized transport mechanism. The driving force for this export might be an increasing imbalance of the equilibrium between sulfatide storage in the CNS and peripheral tissues owing to the ASA-catalyzed hydrolysis of sulfatide in the periphery. This hypothesis postulates the communication of cellular sulfatide pools across the blood–brain barrier and the existence of a transport machinery for sulfatide. Further experiments are required to test these hypotheses and to determine the mechanism by which sulfatide storage in the CNS is reduced.

The functional consequences of the reduced sulfatide concentration in the nervous system were investigated by electrophysiological and behavioral studies. Recordings of the CMAP in the sciatic nerve of untreated ASA knockout mice suggested an impaired conduction of a subset of axonal fibers. This was indicated by a significantly lower and broader amplitude in the presence of a normal nerve conduction velocity (Fig. 7A). Treatment decreased the duration and increased the height of the flattened amplitude demonstrating the abrogation of inhibitory effects. This corrective effect might be associated with the critical role of sulfatide in the organization of paranodal axoglial junctions and the correct clustering of voltage-gated Na+ and K+ channels along the axolemma (34Go–36Go). The possibility to reverse changes of the CMAP by a comparably short exposure to recombinant enzyme might have great implications for the treatment of MLD. Because PNS symptoms prevail before the end stage of MLD, ERT might substantially retard the disease progression and improve the quality of life. This notion is supported by the rotarod data in mice, indicating improvement of the motor coordination both at an early and at a more advanced stage of the disease (Fig. 7B and C).

Notably, the therapeutic effects described in this study were achieved in a treatment period of only 1 month. Long-term treatment over several months might be even more effective and is required to determine the full therapeutic potential of ERT. Nowadays, long-term studies are not realizable, because ASA knockout mice develop an immunological response to the repeatedly infused CHO-rhASA. Anti-ASA antibodies were detectable for the first time after the third infusion and their serum titer increased after the fourth injection (data not shown). Anaphylactic reactions prevented the treatment of the mice for more than 4 weeks. The available ASA knockout mouse model is completely deficient for ASA (3Go). In contrast, most of the human MLD patients express a mutated variant of the ASA polypeptide (1Go). The constitutive expression of such a mutated variant is likely to induce a partial or complete immunological tolerance to the correctly folded wild-type enzyme. ASA knockout mice, which express an enzymatically inactive mutant of the human ASA from a stably integrated transgene, would therefore represent a better model for most MLD patients than the conventional ASA knockout mouse and possibly allow long-term enzyme replacement studies without severe immunological complications.

In summary, the present study provides the first proof-of-principle for the feasibility of ERT in an animal model for MLD. Quantitative biochemical assays and histological techniques demonstrated that intravenous injection of rhASA partially restores the normal sulfatide catabolism in peripheral tissues of ASA-deficient mice. Furthermore, ERT also has a clear sulfatide clearance effect in the PNS and in the CNS. Repeated treatment also results in the improvement of the increased kidney size, the CNS histopathology and behavioral parameters. The extent of metabolic correction substantially exceeds the therapeutic efficacy of previous BMT experiments including bone marrow stem cell gene therapy.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
REFERENCES
 
Production, purification and analysis of CHO-rhASA
CHO-rhASA was purified from secretions of chinese hamster ovary (CHO) cells overexpressing the human ASA from the expression plasmid pASAExp1 (Zymenex A/S, Hillerød, Denmark). The purification protocol will be outlined elsewhere. The specific activity of the enzyme preparation was >60 U/mg. The CHO-rhASA was rebuffered in 1x TBS, pH 7.4, to a concentration of 2.5–4.3 mg/ml and was analyzed by SDS–PAGE and MALDI-TOF spectroscopy. MALDI mass spectra were collected using a Voyager-DE STR BioSpectrometry workstation (Perspective Biosystems, Inc., Framingham, MA, USA) equipped with a 337 nm nitrogen laser. Measurements were taken manually in linear, positive ion mode at a 20–24 kV acceleration voltage, 90% grid voltage and 200 ns delayed ion extraction. Each mass spectrum obtained was the sum of 300 unselected laser profiles on one sample preparation. Sinapinic acid was used as matrix. For partial or complete deglycosylation of CHO-rhASA, 1 µg enzyme was reacted with 1 or 500 mU PNGase F (Roche Diagnostics, Mannheim, Germany) for 20 h at 37°C. The endocytosis assay was done with 1 µg CHO-rhASA per ml medium for 20 h, as described earlier (37Go). ASA was measured by an indirect sandwich ELISA and an activity assay (10Go,38Go).

Treatment of the mice
ASA knockout mice and wild-type controls with the mixed genetic background C57Bl/6Jx129ola (3Go) were kept under standard housing conditions in accordance with the current German law on the protection of animals. All experiments were approved by the local committee for animal welfare (Bezirksregierung Köln, reference number 50.203.2-BN 24, 18/04). Experiments were done with 8 to 12-month-old animals. Depending on the animal weight and the concentration of the CHO-rhASA stock, 200–300 µl enzyme solution (CHO-rhASA in 1x TBS pH 7.4) was administered by an intravenous bolus injection into the tail vein. Control animals were injected with 250 µl 1x TBS pH 7.4.

Analysis of mice
During the treatment period, blood was taken from the tail vein. For the final analysis, mice were deeply anesthetized using an intraperitoneal injection of tribromoethanol and transcardially perfused. For histological investigations, mice were first perfused with phosphate-buffered saline (PBS) and then with 6% glutaraldehyde in 100 mM phosphate buffer pH 7.4. Tissues were then dissected and processed as described subsequently. For biochemical analyses, mice were perfused with PBS alone. Kidneys, liver, brain, brachial plexus and sciatic nerves were dissected, weighed and frozen. Tissue samples were homogenized in 1x TBS pH 7.4. Aliquots of the homogenates were used for lipid extraction (discussed subsequently), protein determination (BioRad Dc assay, BioRad, Hercules, USA) and measurements of ASA by ELISA (10Go).

Lipid analysis
Aliquots of tissue homogenates (discussed earlier) were centrifuged at 100 000g for 1 h, and the pellet was first extracted with 5 ml chloroform/methanol (C/M) 2:1 (v/v) and then with 5 ml C/M 1:1 at 60°C for 4 h in each case. Following evaporation of the solvent, the dry lipids were redissolved in 5 ml MeOH. Alkaline methanolysis was started with 125 µl 4 N NaOH at 37°C and stopped after 2 h with 20 µl 100% acetic acid. Lipids were dried and dissolved in 1 ml MeOH. For desalting by reverse-phase chromatography, Lichroprep RP-18 columns (Merck, Darmstadt, Germany) with a bed volume of 1 ml were equilibrated with C/M/0.1 M KCl 6:96:94. After adding one volume of 0.3 M ammonium acetate to the lipid solution, the mixture was loaded onto the column. After washing with 6 ml H2O, lipids were eluted with 1 ml MeOH and then with 6 ml C/M 1:1. Aliquots of the lipid extracts were sprayed onto silica gel 60 plates (Merck) using the Automatic TLC Sampler 4 from CAMAG (Muttenz, Switzerland). Loading volumes were normalized on the protein concentration of the crude homogenates used for lipid extraction. Different amounts (0.5–8 µg) of lipid standards (cholesterol, sphingomyelin and sulfatide, all standards from Sigma) were loaded on separate lanes. After TLC with C/M/H2O 70:30:4 as a solvent system, lipids were visualized according to Yao and Rastetter (39Go). The plates were scanned with a flat bed scanner (PowerLook III from UMAX Data Systems, Hsinchu, Taiwan), and the intensities of lipid bands were determined with the analysis software Aida 2.11 (Raytest, Straubenhardt, Germany). The amount of cholesterol, sphingomyelin and sulfatide are expressed as arbitrary units, representing the intensities of the respective TLC band after background correction. Statistical analysis was performed using Student's t-test.

Histology
Kidneys, spinal cord and brain were dissected from perfusion-fixed mice. For the detection of sulfatides, tissue slices (100 µm thick) were prepared with a vibratome and incubated with alcian blue (Alcec Blue, Sigma-Aldrich, Taufkirchen, Germany) as described earlier (9Go). The histochemical conditions (pH 5.7, 300 mM MgCl2) were such as to warrant specific staining of sulfatides (40Go). Paraffin sections from kidney blocks were prepared after pre-embedding incubation with alcian blue. Sciatic nerves and kidney samples were embedded in araldite according to routine methods for preparing semithin sections, either with or without pre-embedding incubation in alcian blue. Histological examinations were performed in a blinded manner, without knowledge of the treatment status. Storage in spinal cord was quantitatively analyzed using the image-process software Image J (http://rsb.info.nih.gov/ij/). For this purpose, a 100x100 µm2 grid was applied to black-and-white images of alcian blue-stained cross-sections (100 µm thick, 100-fold magnification) and the numbers and sizes of the alcian blue-positive profiles were determined in eight squares located in the gray matter and eight squares located in the white matter. Profiles with a diameter below and above 8 µm (20 pixels under the chosen settings) were separately quantitated to differentiate between (a) the small (<5 µm) sulfatide-storing profiles of oligodendrocyte processes and other cells with low amounts of storage material and (b) the large (>20 µm) deposits typical of phagocytes and neurons severely congested with sulfatide. Statistical analysis was performed using Student's t-test.

Behavior
Rotarod tests were performed as described previously (7Go). Briefly, the ability of mice to balance on a slowly rotating rod for >4 min was determined. This was done four times on two consecutive days at a constant speed of six rotations per minute.

Electrophysiological investigations
Nerve conduction of sciatic nerves was studied under anesthesia by established electrophysiological methods (41Go). In brief, the CMAP was recorded with two needle electrodes in the foot muscles after distal stimulation of the tibial nerve at the ankle and proximal stimulation of the sciatic nerve at the sciatic notch. Statistical analysis was performed using Student's t-test.


   ACKNOWLEDGEMENTS
 
We thank Dagmar Niemeier and Pia Hydén for excellent technical assistance and Dr Joachim Kappler for his help in the histological evaluation.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 

  1. von Figura, K., Gieselmann, V. and Jaeken, J. (2001) Metachromatic leukodystrophy. In Scriver, C.R., Beaudet, A.L., Sly, W.S., Valle, D., Childs, B., Kinzler, K.W. and Vogelstein, B. (eds), The Metabolic and Molecular Bases of Inherited Disease. Mc Graw-Hill, New York, USA, pp. 3695–3724.

  2. Krivit, W. (2004) Allogeneic stem cell transplantation for the treatment of lysosomal and peroxisomal metabolic diseases. Springer Semin. Immunopathol., 26, 119–132.[CrossRef][ISI][Medline]

  3. Hess, B., Saftig, P., Hartmann, D., Coenen, R., Lullmann-Rauch, R., Goebel, H.H., Evers, M., von Figura, K., D'Hooge, R., Nagels et al. (1996) Phenotype of arylsulfatase A-deficient mice: relationship to human metachromatic leukodystrophy. Proc. Natl Acad. Sci. USA, 93, 14821–14826.[Abstract/Free Full Text]

  4. Gieselmann, V., Matzner, U., Hess, B., Lullmann-Rauch, R., Coenen, R., Hartmann, D., D'Hooge, R., DeDeyn, P. and Nagels, G. (1989) Metachromatic leukodystrophy: molecular genetics and an animal model. J. Inherit. Metab. Dis., 21, 564–574.

  5. Gieselmann, V., Franken, S., Klein, D., Ma°nsson, J.E., Sandhoff, R., Lullmann-Rauch, R., Hartmann, D., Saravanan, V.P., de Deyn, P.P., D'Hooge, R. et al. (2003) Metachromatic leukodystrophy: consequences of sulphatide accumulation. Acta Paediatr. Suppl., 92, 74–79.[CrossRef][Medline]

  6. D'Hooge, R., van Dam, D., Franck, F., Gieselmann, V. and de Deyn, P.P. (2001) Hyperactivity, neuromotor defects, and impaired learning and memory in a mouse model for metachromatic leukodystrophy. Brain Res., 907, 35–43.[CrossRef][ISI][Medline]

  7. Matzner, U., Hartmann, D., Lullmann-Rauch, R., Coenen, R., Rothert, F., Ma°nsson, J.E., Fredman, P., D'Hooge, R., de Deyn, P.P. and Gieselmann, V. (2002) Bone marrow stem cell-based gene transfer in a mouse model for metachromatic leukodystrophy: effects on visceral and nervous system disease manifestations. Gene Ther., 9, 53–63.[CrossRef][ISI][Medline]

  8. Coenen, R., Gieselmann, V. and Lullmann-Rauch, R. (2001) Morphological alterations in the inner ear of the arylsulfatase A-deficient mouse. Acta Neuropathol. (Berl.), 101, 491–498.[Medline]

  9. Wittke, D., Hartmann, D., Gieselmann, V. and Lullmann-Rauch, R. (2004) Lysosomal sulfatide storage in the brain of arylsulfatase A-deficient mice: cellular alterations and topographic distribution. Acta Neuropathol. (Berl.), 108, 261–271.[CrossRef][Medline]

  10. Matzner, U., Harzer, K., Learish, R.D., Barranger, J.A. and Gieselmann, V. (2000) Long-term expression and transfer of arylsulfatase A into brain of arylsulfatase A-deficient mice transplanted with bone marrow expressing the arylsulfatase A cDNA from a retroviral vector. Gene Ther., 7, 1250–1257.[CrossRef][ISI][Medline]

  11. Matzner, U., Schestag, F., Hartmann, D., Lullmann-Rauch, R., D'Hooge, R., de Deyn, P.P. and Gieselmann, V. (2001) Bone marrow stem cell gene therapy of arylsulfatase A-deficient mice, using an arylsulfatase A mutant that is hypersecreted from retrovirally transduced donor-type cells. Hum. Gene Ther., 12, 1021–1033.[CrossRef][ISI][Medline]

  12. Biffi, A., de Palma, M., Quattrini, A., Del Carro, U., Amadio, S., Visigalli, I., Sessa, M., Fasano, S., Brambilla, R., Marchesini, S. et al. (2004) Correction of metachromatic leukodystrophy in the mouse model by transplantation of genetically modified hematopoietic stem cells. J. Clin. Invest., 113, 1118–1129.[CrossRef][ISI][Medline]

  13. Matzner, U. and Gieselmann, V. (2005) Gene therapy for metachromatic leukodystrophy. Expert Opin. Biol. Ther., 5, 55–65.[CrossRef][ISI][Medline]

  14. Kornfeld S. (1992) Structure and function of the mannose 6-phosphate/insulin-like growth factor II receptors. Annu. Rev. Biochem., 61, 307–330.[CrossRef][ISI][Medline]

  15. Neufeld, E.F. (2004) Enzyme replacement therapy. In Platt, F.M. and Walkley, S.V. (eds), Lysosomal Disorders of the Brain. Oxford University Press, Oxford, UK, pp. 327–338.

  16. Matzner, U. (2005) Therapy of lysosomal storage diseases. In Saftig, P. (ed.), Lysosomes. Landes Bioscience, Georgetown, USA, in press.

  17. Barton, N.W., Brady, R.O., Dambrosia, J.M., Di Bisceglie, A.M., Doppelt, S.H., Hill, S.C., Mankin, H.J., Murray, G.J., Parker, R.I. and Argoff, C.E. (1991) Replacement therapy for inherited enzyme deficiency—macrophage-targeted glucocerebrosidase for Gaucher's disease. N. Engl. J. Med., 324, 1464–1470.[Abstract]

  18. Dunder, U., Kaartinen, V., Valtonen, P., Vaananen, E., Kosma, V.M., Heisterkamp, N., Groffen, J. and Mononen, I. (2000) Enzyme replacement therapy in a mouse model of aspartylglycosaminuria. FASEB J., 14, 361–367.[Abstract/Free Full Text]

  19. Roces, D.P., Lullmann-Rauch, R., Peng, J., Balducci, C., Andersson, C., Tollersrud, O., Fogh, J., Orlacchio, A., Beccari, T., Saftig, P. et al. (2004) Efficacy of enzyme replacement therapy in alpha-mannosidosis mice: a preclinical animal study. Hum. Mol. Genet., 13, 1979–1988.[Abstract/Free Full Text]

  20. Gieselmann, V., Schmidt, B. and von Figura K. (1992) In vitro mutagenesis of potential N-glycosylation sites of arylsulfatase A. Effects on glycosylation, phosphorylation, and intracellular sorting. J. Biol. Chem., 267, 13262–13266.[Abstract/Free Full Text]

  21. Lullmann-Rauch, R., Matzner, U., Franken, S., Hartmann, D. and Gieselmann, V. (2001) Lysosomal sulfoglycolipid storage in the kidneys of mice deficient for arylsulfatase A (ASA) and of double-knockout mice deficient for ASA and galactosylceramide synthase. Histochem. Cell Biol., 116, 161–169.[ISI][Medline]

  22. Ghosh, P., Dahms, N.M. and Kornfeld, S. (2003) Mannose 6-phosphate receptors: new twists in the tale. Nat. Rev. Mol. Cell Biol., 4, 202–212.[CrossRef][ISI][Medline]

  23. Muschol, N., Matzner, U., Tiede, S., Gieselmann, V., Ullrich, K. and Braulke, T. (2002) Secretion of phosphomannosyl-deficient arylsulphatase A and cathepsin D from isolated human macrophages. Biochem. J., 368, 845–853.[CrossRef][ISI][Medline]

  24. Sands, M.S., Vogler, C.A., Ohlemiller, K.K., Roberts, M.S., Grubb, J.H., Levy, B. and Sly, W.S. (2001) Biodistribution, kinetics, and efficacy of highly phosphorylated and non-phosphorylated beta-glucuronidase in the murine model of mucopolysaccharidosis VII. J. Biol. Chem., 276, 43160–43165.[Abstract/Free Full Text]

  25. Vogler, C., Sands, M., Higgins, A., Levy, B., Grubb, J., Birkenmeier, E.H. and Sly, W.S. (1993) Enzyme replacement with recombinant beta-glucuronidase in the newborn mucopolysaccharidosis type VII mouse. Pediatr. Res., 34, 837–840.[ISI][Medline]

  26. Crawley, A.C., Brooks, D.A., Muller, V.J., Petersen, B.A., Isaac, E.L., Bielicki, J., King, B.M., Boulter, C.D., Moore, A.J., Fazzalari, N.L. et al. (1996) Enzyme replacement therapy in a feline model of Maroteaux–Lamy syndrome. J. Clin. Invest., 97, 1864–1873.[ISI][Medline]

  27. Matsushima, G.K., Taniike, M., Glimcher, L.H., Grusby, M.J., Frelinger, J.A., Suzuki, K. and Ting, J.P. (1994) Absence of MHC class II molecules reduces CNS demyelination, microglial/macrophage infiltration, and twitching in murine globoid cell leukodystrophy. Cell, 78, 645–656.[CrossRef][ISI][Medline]

  28. Wada, R., Tifft, C.J. and Proia, R.L. (2000) Microglial activation precedes acute neurodegeneration in Sandhoff disease and is suppressed by bone marrow transplantation. Proc. Natl Acad. Sci. USA, 97, 10954–10959.[Abstract/Free Full Text]

  29. Hickey, W.F. and Kimura, H. (1988) Perivascular microglial cells of the CNS are bone marrow-derived and present antigen in vivo. Science, 239, 290–292.[Abstract/Free Full Text]

  30. Lawson, L.J., Perry, V.H. and Gordon, S. (1992) Turnover of resident microglia in the normal adult mouse brain. Neuroscience, 48, 405–415.[CrossRef][ISI][Medline]

  31. Unger, E.R., Sung, J.H., Manivel, J.C., Chenggis, M.L., Blazar, B.R. and Krivit, W. (1993) Male donor-derived cells in the brains of female sex-mismatched bone marrow transplant recipients: a Y-chromosome specific in situ hybridization study. J. Neuropathol. Exp. Neurol., 52, 460–470.[ISI][Medline]

  32. Kennedy, D.W. and Abkowitz, J.L. (1997) Kinetics of central nervous system microglial and macrophage engraftment: analysis using a transgenic bone marrow transplantation model. Blood, 90, 986–993.[Abstract/Free Full Text]

  33. Vallieres, L. and Sawchenko, P.E. (2003) Bone marrow-derived cells that populate the adult mouse brain preserve their hematopoietic identity. J. Neurosci., 23, 5197–5207.[Abstract/Free Full Text]

  34. Hansson, C.G., Karlsson, K.A. and Samuelsson, B.E. (1978) The identification of sulphatides in human erythrocyte membrane and their relation to sodium–potassium dependent adenosine triphosphatase. J. Biochem. (Tokyo), 83, 813–819.