Introduction

GM1-gangliosidosis is a neurodegenerative disorder, which belongs to a group of hereditable human disorders called lysosomal storage diseases (LSDs). They are characterized by the accumulation of undigested macromolecules in the lysosomal compartment. It has been estimated that 60% of these diseases exhibit some degree of neurological impairment owing to global storage in the central nervous system (CNS). GM1-gangliosidosis is an autosomal recessive deficiency of acid -galactosidase (gal),¹ biochemically characterized by accumulation of GM1-ganglioside and its asialo derivative GA1 in the CNS,² and partially degraded glycoproteins, keratan sulfate, and oligosaccharides in visceral organs.³ The most severe form of this disease (infantile or type I) has a very early onset, with biochemical and histopathological abnormalities already present in utero.^{4, 5} This disease is characterized by a rapid neurological decline with death occurring usually before 2 years of age. The available knockout mouse models^{6, 7} replicate the neuro-biochemical aspects of type I GM1-gangliosidosis with gal activity <4% of normal and extensive storage of GM1-ganglioside and GA1 throughout the brain.^{6, 7} Storage can be detected as early as post-natal day 5,⁸ and it progresses rapidly to several fold normal levels by 3 months of age.^{6, 9} However, the neurological involvement is considerably milder than its human counterpart¹⁰ and mice survive to about 8–9 months of age.

The basis for treatment of lysosomal storage diseases by gene therapy is the release of lysosomal enzymes from normal cells and their subsequent transport into lysosomes in enzyme deficient cells by mannose-6-phosphate receptor mediated endocytosis.¹¹ This cross-correction mechanism underlies the rationale that not all cells in the CNS have to be genetically modified to achieve therapeutic efficacy. The validity of this concept has been amply demonstrated in experimental gene therapy studies where lysosomal enzymes are delivered to the brain by normal or genetically modified cells, or their cDNAs to endogenous cells by viral vectors. Viral vector–mediated gene delivery to the brain holds great potential for the treatment of neuronopathic lysosomal storage diseases as the genetic modification of small numbers of endogenous cells has proven to be sufficient to deliver corrective levels of enzyme to large regions of the CNS (for review see ref. ¹²). The widespread distribution of the lysosomal enzyme in the brain results from its continued release from vector-transduced cells over long periods of time, diffusion in the brain parenchyma,¹³ axonal transport within neurons from the site of production,¹⁴ and through cerebrospinal fluid flow in the perivascular space of Virchow-Robin.¹⁵ Infusion of adeno-associated virus (AAV) vectors into the brain lateral ventricles of neonatal mice has proven to be an effective approach to obtain widespread distribution of vector-transduced cells^{16, 17, 18} and lysosomal enzymes^{16, 17, 19} throughout the brain. Injection of an AAV1 vector encoding human -glucuronidase in a mouse model of mucopolysaccharidosis type VII resulted in complete correction of lysosomal storage throughout the brain.¹⁷

As GM1-ganglioside storage is already evident at post-natal day 5 in GM1-gangliosidosis mice, in this study we evaluated the therapeutic potential of neonatal (P0) intracerebroventricular (i.c.v.) infusion of an AAV2/1 vector encoding mouse lysosomalgal.

Results

In this study, we evaluated the therapeutic potential of neonatal i.c.v. infusion of an AAV2/1 vector encoding mouse lysosomal gal in GM1-gangliosidosis mice (gal^-/-). To this end, we injected 2 l of AAV2/1-CBA-gal vector (dose=1.65 10¹¹ genome copies (g.c.)) into each cerebral lateral ventricle of P0 gal^-/- mice (N=7), and assessed gal activity and GM1-ganglioside storage in the brain at 3 months of age. Bilateral i.c.v. injections of AAV vectors in neonatal mice have been shown to generate widespread symmetrical patterns of transduction of the brain hemispheres.^{16, 17, 18} Therefore we used the left hemisphere for histological assessment of enzyme distribution and GM1-ganglioside storage, and the right hemisphere was used for quantitative biochemical assessment of the same two parameters.

Distribution of gal in the brain

Distribution of gal throughout the brain in AAV-treated gal^-/- mice (N=7), untreated gal^-/- (N=3), and C57BL/6 wild-type mice (N=3) was evaluated by staining histological sections with X-gal solution at pH 5.0 (Figure 1). In AAV-treated gal^-/- mice we observed robust staining throughout the entire brain (Figure 1, treated gal^-/-). The most intense staining was associated with the olfactory bulb, neocortex, hippocampal formation, thalamic, and sub-thalamic nuclei (Figure 1, regions 1 and 4 in treated gal^-/-). Also, in the midbrain, the superior colliculus was strongly stained (Figure 1, arrow in region 5 in treated gal^-/-). In rostral parts of the brain (Figure 1, regions 2 and 3 in treated gal^-/-), the strongest staining was observed in layers II, V, and VI in prefrontal cortex (Figure 1, region 2 in treated gal^-/-), and layer V in sensorimotor cortical areas (Figure 1, region 3 in treated gal^-/-). The brains of untreated gal^-/- mice were devoid of any staining (Figure 1, untreated gal^-/-), whereas those of wild-type mice showed faint staining throughout most of the brain (Figure 1, untreated WT). The strongest staining in wild-type controls was observed in the hippocampus (CA1-CA3), choroid plexus (Figure 1, arrows in region 4 in untreated WT), piriform cortex (Figure 1, arrows in regions 2 and 3 in untreated WT), and mitral cell layer in the olfactory cortex (Figure 1, arrow in region 1 in untreated WT).

Figure 1.

Distribution of gal in the brains of GM1-gangliosidosis mice injected at P0 with AAV vector. Mice were killed at 3 months of age and the left hemisphere was used for histological analysis of gal expression by X-gal staining at pH 5.0. Representative sections from regions 1–5 in (a) are shown in (b). Black arrows in untreated wild-type controls indicate the nuclei with the highest intensity of staining in regions 1 (mitral cell layer in the olfactory bulb), 2–3 (piriform cortex), 4 (hippocampus and choroid plexus). The black arrow in region 5 of AAV-treated gal^-/- brains indicates the superior colliculus. Picture in a is a sagital section of the brain counterstained with nuclear fast red. Bar=1 mm.

Full figure and legend (172K)

Next we measured gal activity in 5 brain regions (same as above) in the right hemisphere and cerebellum (Figure 2). gal activity (nmol/h/mg protein) in AAV-treated gal^-/- mice was 7- to 65-fold higher than in wild-type controls (Figure 2a). Conversely, gal activity in untreated gal^-/- mice was only 1–4% of wild-type levels (Figure 2a). In contrast to the results in the brain, gal activity in the liver of AAV-treated and untreated gal^-/- mice was about 3% of wild-type levels (Figure 2b).

Figure 2.

gal activity in brain and liver at 3 months after intraventricular injection of AAV vector. (a) Enzymatic activity was measured in five regions (same as in Figure 1a) of the right hemisphere and cerebellum. In all analyzed brain regions of treated gal^-/- mice, the gal activity was more than 7-fold higher than found in normal brain regions and 400- to 1000-fold higher than found in the brain regions of untreated gal^-/- mice. (b) In the livers of both AAV1-treated and untreated gal^-/- mice, the enzyme activity was 100-fold lower than in wild-type mice. Shown are mean1SEM.

Full figure and legend (62K)

Biochemical quantification of GSLs in the brain

The effect of AAV2/1-CBA-gal vector on lysosomal storage in gal^-/- mice was quantified by measuring the concentration of different glycosphingolipids (GSLs) in region 4 of the brain by high-performance thin layer chromatography (HPTLC) (Figure 3 and Table 1). The GM1-ganglioside content in untreated gal^-/- mice (475.842.0 g/100 mg dry weight; mean1 SEM) was 9-fold higher than in wild-type C57BL/6 controls (53.21.1 g/100 mg dry weight) (P<0.01 with analysis of variance (ANOVA) for gal^-/- versus wild-type). In contrast, the GM1-ganglioside content in AAV-treated gal^-/- mice (46.92.1 g/100 mg dry weight) was similar to that in the wild-type (Figure 3a and Table 1) (P<0.01 with ANOVA for AAV-treated gal^-/- versus gal^-/-). Of the other gangliosides analyzed, only GD1b was significantly elevated in untreated gal^-/- mice compared to wild-type C57BL/6 controls (P<0.01 with ANOVA for gal^-/- versus wild-type), and was normalized in AAV-treated gal^-/- mice (Table 1; P<0.01 with ANOVA for AAV-treated gal^-/- versus gal^-/-). The GSL GA1 was present at an exceptionally high level in untreated gal^-/- mice (1.640.5 mg/100 mg dry weight), but it was not detectable in AAV-treated gal^-/- mice or wild-type controls (Figure 3b and Table 1). Also, AAV treatment normalized levels of GM1, GT1b, and GQ1b gangliosides in the cerebellum (P<0.01 with ANOVA for AAV-treated gal^-/- versus gal^-/-), whereas GA1 was not detectable in 5 out of 6 samples analyzed (Supplementary Figure S1 and Table S1).

Figure 3.

HPTLC of brain glycosphingolipids in 3-month-old gal^-/- mice. The distribution of (a) gangliosides and (b) GA1 was measured in region 4 of the brains of wild-type control mice (no. 24.2 and 24.1), untreated gal^-/- (no. 186 and 188), and AAV1-treated gal^-/- mice (no. 139, 157, and 173). The amount of gangliosides and GA1 spotted per lane was equivalent to approximately 1.5 g sialic acid and 0.2 mg brain dry weight, respectively. The individual gangliosides were labeled according to the nomenclature system of Svennerholm⁴³ (right side of the chromatograms).

Full figure and legend (72K)

Table 1 - Effect of AAV vector on the concentration of lipids in the brain of adult gal mice.

Full table

Histological evaluation of GM1-ganglioside storage throughout the brain

Recombinant Alexa 555-conjugated cholera toxin subunit B (CTX-B) was used to stain tissue sections from AAV-treated gal^-/- and control mice to assess GM1-ganglioside storage throughout the brain (Figure 4). In untreated gal^-/- mice, CTX-B staining revealed large numbers of cells with strong staining restricted to the perikarya throughout the cortex, hippocampus, thalamus, and other subcortical nuclei (arrowheads in Figure 4a–d), which were absent in wild-type C57BL/6 mice (Figure 4e–h). In these animals CTX-B staining was uniformly distributed throughout the tissue. Interestingly, fiber tracts in untreated gal^-/- mice appeared to be weakly stained compared to the same tracts in wild-type mice. Examples of this were the cortical-fugal axonal bundles in the striatum (arrows in Figure 4b and f), fiber tracts in the lateral geniculate (arrows in Figure 4d and h). In AAV-treated gal^-/- mice, the CTX-B staining pattern was identical to that observed in wild-type controls with complete absence of cells with staining restricted to the perikarya anywhere in the brain, and restoration of fiber tract staining patterns (Figure 4i–l).

Figure 4.

Histological assessment of GM1-ganglioside storage throughout the brain. (a–d) Sections from untreated gal^-/-, (e–h) wild-type, and (i–l) AAV1-treated gal^-/- brains were stained with Alexa 555-conjugated CTX-B. Strong perikaryal staining was observed in untreated gal^-/- mice throughout the brain (arrowheads in a, c, and d). White matter in these mice (arrows in b and d) appeared to be less intensely stained than in wild-type mice (arrows in f and h). AAV-treated gal^-/- mice were indistinguishable from wild-type mice with uniformly diffuse staining throughout the neuropil and white matter tracts that could be easily differentiated from the rest of the brain. Black holes in pictures correspond to cell nuclei. Abbreviations: CA1—CA1 field of the hippocampus; DLG—dorsal lateral geniculate nucleus; VLG—ventral lateral geniculate nucleus; VPM—vental posteromedial thalamic nucleus. Bars: a, b, d–f, h–j, and l=200 m; Insets and c, g, and k=50 m.

Full figure and legend (466K)

Histological evaluation of cholesterol storage

Lysosomal storage in GSL storage diseases has been shown to be accompanied by cholesterol storage/trapping in the endosomal/lysosomal compartment.^{20, 21} Therefore we analyzed cholesterol distribution in the brains of AAV-treated gal^-/- and control mice by Filipin staining (blue) of histological sections (Figure 5). In untreated gal^-/- mice there was strong Filipin labeling of cells throughout the brain (Figure 5a–f), in a pattern consistent with cholesterol storage/trapping in the endosomal/lysosomal compartment (white arrows in Figure 5a, c, and f). In contrast, in wild-type controls (Figure 5g–i) and AAV-treated gal^-/- mice (Figure 5j–o) there was only diffuse low-intensity staining that could not be associated with any particular cell.

Figure 5.

Histological assessment of cholesterol storage throughout the brain. (a–f) Sections from untreated gal^-/-, (g–i) wild-type, and (j–o) AAV1-treated gal^-/- brains were stained for unesterified cholesterol with Filipin (blue), and counterstained for nuclei with TO-PRO3 (red). Untreated gal^-/- mice showed strong perikaryal staining (white arrows) throughout the brain, consistent with endosomal/lysosomal accumulation of cholesterol. In wild-type and AAV1-treated gal^-/- mice there was only weak diffuse Filipin staining. Abbreviation: CA1—CA1 field of the hippocampus. Bar=50 m.

Full figure and legend (490K)

Biochemical analysis of non-polar neutral and acidic lipid levels in the brain

Given our histological results showing abnormal staining of white matter with CTX-B and cholesterol storage throughout the brain, we analyzed the levels of non-polar neutral and acidic lipids in region 4 of the brain in AAV-treated gal^-/-, untreated gal^-/-, and wild-type control mice (Figure 6). Most lipids appeared to be present at comparable levels in all mice, including cholesterol (Figure 6a and Table 1). However, cerebrosides (arrow in Figure 6a) and sulfatides (arrow in Figure 6b) were present at significantly lower levels in untreated gal^-/- mice than in wild-type controls (Table 1; P<0.05 with ANOVA for gal^-/- versus wild-type). In AAV-treated gal^-/- mice both lipids were present at levels similar to those measured in wild-type controls (Table 1). The cholesterol level in the cerebellum was similar in all three groups of mice, whereas cerebrosides and sulfatides appeared to be reduced in untreated gal^-/- mice compared to wild-type and AAV-treated gal^-/- mice (Supplementary Table S1).

Figure 6.

HPTLC of brain neutral and acidic lipids in 3-month-old gal^-/- mice. (a) The levels of non-polar neutral and (b) acidic lipids in region 4 of the brain were assessed in wild-type control mice (no. 24.2), untreated gal^-/- (no. 186), and AAV1-treated gal^-/- mice (no. 139). The amount of neutral and acidic lipids spotted per lane was equivalent to approximately 0.07 and 0.2 mg tissue dry weight, respectively. Arrowheads show the position of a cerebrosides and b sulfatides on the chromatograms. Abbreviation: CE, cholesterol esters; TG, triglycerides; IS, internal standard; C, cholesterol; Cer, Ceramide; CB, cerebrosides (doublet); PE, phosphatidylethanolamine; PC, phosphatidylcholine; SM, sphingomyelin; FA, fatty acids; CL, cardiolipin; PA, phosphatidic acid; Sulf, sulfatides (doublet); PS, phosphatidylserine; PI, phosphatidylinositol.

Full figure and legend (59K)

Discussion

The etiology of GM1-gangliosidosis, like most other lysosomal storage diseases with neurological involvement, has been known for a long time, but effective therapies have yet to be developed. A number of therapeutic strategies for GM1-gangliosidosis are being explored, including substrate reduction therapy,^{8, 22} chemical chaperone therapy,²³ viral vector–mediated in vivo gene therapy²⁴ and transplantation of genetically modified hematopoetic stem cells.²⁵

In the present study, we evaluated the therapeutic potential of an AAV2/1 vector encoding mouse lysosomal gal injected into the cerebral lateral ventricles of neonatal (P0) GM1-gangliosidosis mice (gal^-/-). This is a highly effective experimental strategy to achieve widespread gene delivery to the brain,^{16, 17, 18, 19} and it has been successfully used to prevent or reverse lysosomal storage throughout the brain in two other mouse models of lysosomal storage diseases.^{16, 17, 19} At 3 months of age we found that enzymatic activities in AAV-treated gal^-/- brains were 7- to 65-fold higher than in wild-type mice, and at least 1000-fold higher than in non-treated gal^-/- mice (Figure 2). The pattern of gal distribution in the brain (Figures 1 and 2) is consistent with that reported for other lysosomal enzymes after neonatal i.c.v. delivery of AAV2/1 vectors.^{16, 17, 19} Interestingly, the regions with highest gal staining intensity in this study are the same regions that were transduced at high efficiency in wild-type animals after neonatal i.c.v. injection of an AAV2/1 vector expressing green fluorescent protein.¹⁸ However, as we have not determined the distribution of AAV-transduced cells in the brain, we can only speculate that the highest intensities of gal staining (dark blue in Figure 1) are observed in brain regions that were transduced at high efficiency.

gal expression at higher-than-normal levels in the brains of AAV-treated gal^-/- mice did not appear to cause any obvious neuroanatomical or behavioral abnormalities. It has been reported that massive overexpression of human beta-glucuronidase in transgenic mice leads to its storage in lysosomes and moderate secondary elevation of other lysosomal enzymes. However overexpression of that enzyme was considered "generally benign" as no deleterious effects were observed.²⁶ Presently we are unable to exclude the possibility that gal overexpression may have caused lysosomal distension in some regions of the brain. Further studies will be necessary to evaluate the long-term safety of gal overexpression in the brain.

Accumulation of GM1-ganglioside in the brain of the gal^-/- mouse model used in the present study can be detected as early as post-natal day 5,⁸ and it has been reported to reach levels 5-fold higher than normal by 3.5 months of age.⁶ Also the GSL GA1 is present in the brains of gal^-/- mice at considerably higher levels than in human patients.⁶ Here we performed a detailed biochemical analysis of GSL levels in region 4 of the cerebrum and cerebellum of AAV-treated gal^-/- mice and controls. Our results indicate that AAV treatment completely eliminated GM1-ganglioside storage and normalized secondary GSL alterations (e.g., GA1, cerebrosides, and sulfatides) in both structures analyzed. We also analyzed GM1-ganglioside storage in other regions of the brain by staining tissue sections with fluorescently labeled CTX-B, which has been used to detect GM1-ganglioside in cultured cells.^{21, 23} CTX-B staining results showed that GM1-ganglioside storage was resolved throughout the brain in AAV-treated gal^-/- mice.

Lysosomal storage in sphingolipid storage diseases is accompanied by cholesterol storage/trapping in the endosomal/lysosomal compartment,²⁰ which can be assessed histologically by staining with Filipin. Our analyses of cholesterol content in the untreated GM1-gangliosidosis brain are apparently contradictory, with Filipin staining showing a pattern consistent with endosomal/lysosomal storage of cholesterol, while the HPTLC quantification clearly shows that cholesterol levels are normal in gal^-/- brains (Figure 6 and Table 1). Our quantification results are consistent with previous studies in these gal^-/- mice.^{8, 22} Similar results have been reported in a mouse model of Niemann–Pick type C, another sphingolipid storage disease, where Filipin staining suggests extensive cholesterol storage,²⁷ but biochemical quantification shows unchanged cholesterol levels in the brain, despite considerable elevation in the liver and spleen.²⁸ Therefore, it seems reasonable to conclude that in the brain, the trapping effect of sphingolipid storage shifts cholesterol localization from the plasma membrane to the endosomal/lysosomal compartment while maintaining cholesterol homeostasis. The mechanism by which this takes place is not clear but it is interesting to consider that astrocytes appear to be the major source of cholesterol in the brain²⁹ whereas sphingolipid storage occurs primarily in neurons.²⁷ It is noteworthy that in the present study AAV treatment also corrected this secondary alteration in cholesterol localization, an aspect that has been previously reported to occur after AAV-mediated correction of the primary enzymatic defect in the brain of Niemann–Pick A mice.³⁰

An interesting observation made during histological analysis of GM1-ganglioside storage was that CTX-B staining of fiber tracts in untreated gal^-/- mice was considerably lower than in the wild-type controls or the AAV-treated gal^-/- mice. Biochemical analysis of neutral and acidic lipids in the brain showed that cerebrosides and sulfatides, which are myelin-enriched lipids,^{31, 32} were significantly reduced in untreated gal^-/- mice compared to wild-type controls (Figure 6 and Table 1). AAV treatment restored cerebrosides and sulfatides in the gal^-/- mice to wild-type levels (Table 1). A neuropathological feature of type I (or infantile form) GM1-gangliosidosis in humans^{33, 34, 35, 36} and dogs³³ is the abnormal or delayed myelin formation in the CNS. A significant deficiency in myelin formation as is suggested by the biochemical quantification of cerebrosides and sulfatides, would be expected to result in overt neurological symptoms such as tremor and spasticity. However, neurological symptoms in these mice only become apparent by 6 months of age and the onset seems to coincide with CNS inflammation.³⁷ Therefore it is possible that myelination in gal^-/- mice proceeds in a relatively normal way but myelin composition and possibly its function are altered. Further experiments will be necessary to clarify these issues.

It appears that gal^-/- mice reproduce several neurochemical and developmental features of type I GM1-gangliosidosis and that neonatal i.c.v. AAV-mediated gene delivery can correct the enzymatic deficiency throughout the brain and restore normal brain neurochemistry. Additional experiments will be necessary to determine the long-term impact of this therapeutic approach on behavior and lifespan of gal^-/- mice.

To our knowledge this is the first report showing complete correction of enzymatic deficiency and lysosomal storage throughout the GM1-gangliosidosis mouse brain. In conclusion, AAV vectors encoding lysosomal acid gal should be further evaluated in larger animal models of GM1-gangliosidosis to devise gene delivery approaches that can be translated into human clinical trials.

Materials and Methods

AAV vector design and preparation. The AAV vector used in this study, AAV2/1-chicken beta-actin promoter-gal, carries AAV2 inverted terminal repeats and the mouse lysosomal acid gal cDNA³⁸ under transcriptional control of a hybrid CMV enhancer/chicken beta-actin promoter (CBA). This AAV vector was derived from the plasmid pTR-UF12.1³⁹ (from Richard Snyder, University of Florida, FL) by removing the internal ribosome entry site-green fluorescent protein cassette and replacing it with a woodchuck hepatitis virus post-transcriptional regulatory element amplified from the plasmid CSCGW.⁴⁰ The AAV2 vector pseudotyped with an AAV1 capsid (hence AAV2/1) was prepared with a titer of 4.12 10¹³g.c./ml as described previously.¹⁸

Animal procedures. The GM1-gangliosidosis mouse model⁶ was obtained from Dr Kunihiko Suzuki (Neuroscience Center, University of North Carolina, Chapel Hill, NC). Newborn mice for these experiments were generated by breeding gal^+/- females to gal^-/- males. Mouse pups (P0) were cryoanesthetized and injected with 2 l of viral vector into each cerebral lateral ventricle with a glass micropipette (70–100 m diameter at the tip), using a Narishige IM300 microinjector (Narishige, Japan). Pups were then placed on a warming pad and returned to the dam after regaining normal color and full activity typical of newborn mice. The genotype of injected mice was determined at post-natal day 25 by PCR. The experimental protocol was approved by the Institutional Animal Care and Use Committee at The Massachusetts General Hospital and followed guidelines set forth in the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Tissue preparation. At 3 months of age AAV-treated gal^-/- mice (n=7), age-matched non-treated gal^-/- mice (n=3), and age-matched wild-type C57BL/6 mice (n=3) were killed by CO₂ asphyxiation and the livers and brains harvested. The left hemisphere of the brain was embedded in tissue freezing medium (Triangle Biomedical Sciences, Durham, NC) and rapidly frozen in a 2-methyl-butane/dry-ice bath. Consecutive 20-m thick coronal cryosections were prepared and stored at -20°C. The right hemisphere and liver were rapidly frozen in the vapor phase of liquid nitrogen and stored at -80°C.

Histological procedures. One series of frozen sections representing the entire left hemisphere from each AAV-treated gal^-/- mouse and control mice was fixed for 10 min in 0.25% glutaraldehyde in phosphate-buffered saline (PBS) at room temperature followed by two washes in 1 PBS. Sections were incubated overnight at 37°C in X-gal solution (5 mM K₄Fe(CN)₆, 5 mM K₃Fe(CN)₆, 2 mM MgCl₂, 1 mg/ml 5-bromo-4-chloro-3-indolyl-D-galactosidase (X-gal) in PBS, pH 5.0).⁴¹ X-gal reacted sections were washed in PBS, and counterstained with Nuclear Fast Red (Vector Laboratories, Burlingame, CA). Images of X-gal reacted hemispheres were acquired in a Nikon Supercoolscan 9000 slide scanner at 3,000 dpi.

For analysis of GM1-ganglioside storage, one series of sections representing the entire left hemisphere from each mouse was fixed for 10 min in 4% paraformaldehyde in PBS, pH 7.4, at room temperature. After 3 5 min washes in PBS, sections were blocked with 1.5% normal goat serum (NGS, Vector Laboratories, Burlingame, CA) in PBS for 1 h at room temperature. After this, sections were incubated for 1 h at room temperature with Alexa 555-conjugated cholera toxin B in PBS, (1:200; Molecular Probes). After 3 10 min washes in PBS they were counterstained with 4',6-diamidino-2-phenylindole, and coverslipped with fluorescent mounting media (DakoCytomation, Carpinteria, CA).

Cholesterol levels in the brain were evaluated by Filipin staining of one series of sections representing the entire left hemisphere from each mouse as follows. Sections were fixed for 10 min in 4% paraformaldehyde in PBS at room temperature, followed by 3 5 min washes in PBS and 10 min incubation in 1.5% glycine in PBS. After 3 5 min washes in PBS, sections were incubated for 1 h at room temperature with Filipin (0.05 mg/ml; Sigma, St Louis, MO) and TO-PRO-3 (1:1,000; Molecular Probes, Eugene, OR) in PBS. After this, sections were washed and coverslipped with fluorescent mounting media (DakoCytomation).

Pictures of Cholera toxin B-stained sections were acquired with a Laser Scanning Microscope LSM 5 Pascal (Zeiss, Germany) and corresponding Software (Release 3.2). Filipin staining of experimental and control brains were analyzed in a Nikon TE2000U fluorescent microscope equipped with UV (Filipin) and Cy5 (TO-PRO-3) filters. Images were acquired with a Retiga EXi camera (QImaging, Burnaby, BC, Canada) and MetaVue software version 6.2r4 (Molecular Devices, Downingtown, PA).

Enzymatic assays. The right hemisphere was cut in 50 m sections, which were separated so that one section was used for enzymatic assay, and the next section was used for lipid quantification. Measurements were carried out in pooled sections sampling 2 mm of tissue throughout the cerebrum (five regions). The cerebellum and brain stem were analyzed as a single unit. For gal enzymatic assays, tissues were homogenized in four volumes of water, followed by three cycles of freeze-thawing and centrifugation at 1,000 g for 5 min at 4°C. Supernatants were assayed for gal activity and protein concentration using 4-methylumberlliferyl--D-galactopyranoside in a micro-plate fluorimetric assay,⁹ and a Bio-Rad Dc protein assay (Bio-Rad) with bovine serum albumin as standard, respectively, using a VICTOR³ multilabel counter (Perkin-Elmer, Wellesley, MA). gal activity was represented as nmol/hr/mg protein.

Purification and analysis of brain total lipids. The lipid composition of region 4 (see Figure 1) and the cerebellum from each brain was determined as described previously.⁸ Briefly, total lipids were isolated and purified from freeze-dried brain samples using chloroform: methanol (1:1, by vol). Neutral and acidic lipids were separated using a DEAE-Sephadex column. Total gangliosides were separated from other acidic lipids and quantified using the resorcinol assay as described.⁸ Individual lipids were quantified by HPTLC as described.⁸ Briefly, lipids were spotted on Silica gel 60 HPTLC plates (E Merck, Darmstadt, Germany) using a Camag Linomat V auto-TLC spotter (Camag Scientific Inc., Wilmington, NC). Purified lipid standards were either purchased from Matreya Inc. (Pleasant Gap, PA) or were a gift from Dr. Robert Yu (Medical College of Georgia, Augusta, GA). For gangliosides and GA1, the HPTLC plates were developed by a single ascending run with choloroform:methanol:water (55:45:10, by vol for gangliosides and 65:35:8, by vol for GA1) containing 0.02% CaCl₂ 2H₂O. Plates were sprayed with either resorcinol-HCl or orcinol-H₂SO₄ and heated on an aluminum block heater at 95°C for approximately 30 min to visualize gangliosides or GA1, respectively. The neutral and acidic HPTLC plates were developed to a height of 4.5 and 6 cm, respectively, with chloroform:methanol:acetic acid:formic acid:water (35:15:6:2:1 by vol), and then developed to the top with hexanes:diisopropyl ether:acetic acid (65:35:2 by vol). The lipids were visualized by charring with 3% cupric acetate in 8% phosphoric acid solution, followed by heating the plate at 165°C. The percentage distribution of individual bands was determined by scanning the plate on a ScanMaker 4800 with ScanWizard5 V7.00 software (Microtek, Carson, CA) for GA1 or on a Personal Densitometer SI with ImageQuant software (Molecular Dynamics, Sunnyvale, CA) for gangliosides and other lipids. The total brain ganglioside content was normalized to 100% and the percentage distribution values were used to calculate sialic acid concentration of individual gangliosides as described.⁴² Gangliosides were identified according to the nomenclature system of Svernnerholm.⁴³ The density values for the other lipids were fit to a standard curve of the respective lipid and used to calculate individual concentrations. One-way ANOVA was used to evaluate the significance of differences between all the groups for all lipids using Statview 5.0 (SAS, Cary, NC).

References

1. Okada, S and O'Brien, JS (1968). Generalized gangliosidosis: beta-galactosidase deficiency. Science 160: 1002–1004. | Article | PubMed | ChemPort |
2. O'Brien, JS, Stern, MB, Landing, BH, O'Brien, JK and Donnell, GN (1965). Generalized gangliosidosis: another inborn error of ganglioside metabolism. Amer J Dis Child 109: 338–346. | PubMed | ChemPort |
3. Suzuki, K (1968). Cerebral GM1-gangliosidosis: chemical pathology of visceral organs. Science 159: 1471–1472. | Article | PubMed | ChemPort |
4. Kudoh, T et al. (1978). Prenatal diagnosis of GM1-gangliosidosis: biochemical manifestations in fetal tissues. Hum Genet 44: 287–293. | Article | PubMed | ChemPort |
5. Bieber, FR, Mortimer, G, Kolodny, EH and Driscoll, SG (1986). Pathologic findings in fetal GM1 gangliosidosis. Arch Neurol 43: 736–738. | PubMed | ChemPort |
6. Hahn, CN et al. (1997). Generalized CNS disease and massive GM1-ganglioside accumulation in mice defective in lysosomal acid beta-galactosidase. Hum Mol Genet 6: 205–211. | Article | PubMed | ISI | ChemPort |
7. Matsuda, J et al. (1997). Beta-galactosidase-deficient mouse as an animal model for GM1-gangliosidosis. Glycoconj J 14: 729–736. | Article | PubMed | ChemPort |
8. Kasperzyk, JL et al. (2004). N-butyldeoxygalactonojirimycin reduces neonatal brain ganglioside content in a mouse model of GM1 gangliosidosis. J Neurochem 89: 645–653. | Article | PubMed | ISI | ChemPort |
9. Hauser, EC, Kasperzyk, JL, d'Azzo, A and Seyfried, TN (2004). Inheritance of lysosomal acid beta-galactosidase activity and gangliosides in crosses of DBA/2J and knockout mice. Biochem Genet 42: 241–257. | Article | PubMed | ChemPort |
10. Matsuda, J et al. (1997). Neurological manifestations of knockout mice with beta-galactosidase deficiency. Brain Dev 19: 19–20. | Article | PubMed | ChemPort |
11. Neufeld, EF and Fratantoni, JC (1970). Inborn errors of mucopolysaccharide metabolism. Science 169: 141–146. | Article | PubMed | ChemPort |
12. Sands, MS and Davidson, BL (2006). Gene therapy for lysosomal storage diseases. Mol Ther 13: 839–849. | Article | PubMed | ChemPort |
13. Taylor, RM and Wolfe, JH (1997). Decreased lysosomal storage in the adult MPS VII mouse brain in the vicinity of grafts of retroviral vector-corrected fibroblasts secreting high levels of beta-glucuronidase. Nat Med 3: 771–774. | Article | PubMed | ISI | ChemPort |
14. Passini, MA, Lee, EB, Heuer, GG and Wolfe, JH (2002). Distribution of a lysosomal enzyme in the adult brain by axonal transport and by cells of the rostral migratory stream. J Neurosci 22: 6437–6446. | PubMed | ISI | ChemPort |
15. Liu, G, Martins, I, Wemmie, JA, Chiorini, JA and Davidson, BL (2005). Functional correction of CNS phenotypes in a lysosomal storage disease model using adeno-associated virus type 4 vectors. J Neurosci 25: 9321–9327. | Article | PubMed | ChemPort |
16. Passini, MA and Wolfe, JH (2001). Widespread gene delivery and structure-specific patterns of expression in the brain after intraventricular injections of neonatal mice with an adeno-associated virus vector. J Virol 75: 12382–12392. | Article | PubMed | ISI | ChemPort |
17. Passini, MA et al. (2003). Intraventricular brain injection of adeno-associated virus type 1 (AAV1) in neonatal mice results in complementary patterns of neuronal transduction to AAV2 and total long-term correction of storage lesions in the brains of beta-glucuronidase-deficient mice. J Virol 77: 7034–7040. | Article | PubMed | ISI | ChemPort |
18. Broekman, ML, Comer, LA, Hyman, BT and Sena-Esteves, M (2006). Adeno-associated virus vectors serotyped with AAV8 capsid are more efficient than AAV-1 or -2 serotypes for widespread gene delivery to the neonatal mouse brain. Neuroscience 138: 501–510. | Article | PubMed | ChemPort |
19. Rafi, MA et al. (2005). AAV-mediated expression of galactocerebrosidase in brain results in attenuated symptoms and extended life span in murine models of globoid cell leukodystrophy. Mol Ther 11: 734–744. | Article | PubMed | ChemPort |
20. Puri, V et al. (1999). Cholesterol modulates membrane traffic along the endocytic pathway in sphingolipid-storage diseases. Nat Cell Biol 1: 386–388. | Article | PubMed | ISI | ChemPort |
21. Puri, V et al. (2003). Sphingolipid storage induces accumulation of intracellular cholesterol by stimulating SREBP-1 cleavage. J Biol Chem 278: 20961–20970. | Article | PubMed | ISI | ChemPort |
22. Kasperzyk, JL, d'Azzo, A, Platt, FM, Alroy, J and Seyfried, TN (2005). Substrate reduction reduces gangliosides in postnatal cerebrum-brainstem and cerebellum in GM1 gangliosidosis mice. J Lipid Res 46: 744–751. | PubMed | ISI | ChemPort |
23. Matsuda, J et al. (2003). Chemical chaperone therapy for brain pathology in G(M1)-gangliosidosis. Proc Natl Acad Sci USA 100: 15912–15917. | Article | PubMed | ChemPort |
24. Takaura, N et al. (2003). Attenuation of ganglioside GM1 accumulation in the brain of GM1 gangliosidosis mice by neonatal intravenous gene transfer. Gene Ther 10: 1487–1493. | Article | PubMed | ChemPort |
25. Sano, R, Tessitore, A, Ingrassia, A and d'Azzo, A (2005). Chemokine-induced recruitment of genetically modified bone marrow cells into the CNS of GM1-gangliosidosis mice corrects neuronal pathology. Blood 106: 2259–2268. | Article | PubMed | ISI | ChemPort |
26. Vogler, C et al. (2003). Transgene produces massive overexpression of human beta -glucuronidase in mice, lysosomal storage of enzyme, and strain-dependent tumors. Proc Natl Acad Sci USA 100: 2669–2673. | Article | PubMed | ChemPort |
27. Zervas, M, Dobrenis, K and Walkley, SU (2001). Neurons in Niemann–Pick disease type C accumulate gangliosides as well as unesterified cholesterol and undergo dendritic and axonal alterations. J Neuropathol Exp Neurol 60: 49–64. | PubMed | ISI | ChemPort |
28. te Vruchte, D et al. (2004). Accumulation of glycosphingolipids in Niemann–Pick C disease disrupts endosomal transport. J Biol Chem 279: 26167–26175. | Article | PubMed | ChemPort |
29. Pfrieger, FW (2003). Outsourcing in the brain: do neurons depend on cholesterol delivery by astrocytes? Bioessays 25: 72–78. | Article | PubMed | ChemPort |
30. Passini, MA et al. (2005). AAV vector-mediated correction of brain pathology in a mouse model of Niemann–Pick A disease. Mol Ther 11: 754–762. | Article | PubMed | ChemPort |
31. Matthieu, JM, Widmer, S and Herschkowitz, N (1973). Biochemical changes in mouse brain composition during myelination. Brain Res 55: 391–402. | Article | PubMed | ChemPort |
32. Norton, WT and Poduslo, SE (1973). Myelination in rat brain: changes in myelin composition during brain maturation. J Neurochem 21: 759–773. | Article | PubMed | ChemPort |
33. Kaye, EM et al. (1992). Dysmyelinogenesis in animal model of GM1 gangliosidosis. Pediatr Neurol 8: 255–261. | Article | PubMed | ISI | ChemPort |
34. Shen, WC, Tsai, FJ and Tsai, CH (1998). Myelination arrest demonstrated using magnetic resonance imaging in a child with type I GM1 gangliosidosis. J Formos Med Assoc 97: 296–299. | PubMed | ChemPort |
35. Folkerth, RD, Alroy, J, Bhan, I and Kaye, EM (2000). Infantile G(M1) gangliosidosis: complete morphology and histochemistry of two autopsy cases, with particular reference to delayed central nervous system myelination. Pediatr Dev Pathol 3: 73–86. | Article | PubMed | ChemPort |
36. Gururaj, A et al. (2005). Magnetic resonance imaging findings and novel mutations in GM1 gangliosidosis. J Child Neurol 20: 57–60. | PubMed |
37. Jeyakumar, M et al. (2003). Central nervous system inflammation is a hallmark of pathogenesis in mouse models of GM1 and GM2 gangliosidosis. Brain 126: 974–987. | Article | PubMed | ISI | ChemPort |
38. Nanba, E and Suzuki, K (1990). Molecular cloning of mouse acid beta-galactosidase cDNA: sequence, expression of catalytic activity and comparison with the human enzyme. Biochem Biophys Res Commun 173: 141–148. | Article | PubMed | ChemPort |
39. Matalon, R et al. (2003). Adeno-associated virus-mediated aspartoacylase gene transfer to the brain of knockout mouse for canavan disease. Mol Ther 7: 580–587. | Article | PubMed | ChemPort |
40. Sena-Esteves, M, Tebbets, JC, Steffens, S, Crombleholme, T and Flake, AW (2004). Optimized large-scale production of high titer lentivirus vector pseudotypes. J Virol Methods 122: 131–139. | PubMed | ISI | ChemPort |
41. Sena-Esteves, M, Camp, SM, Alroy, J, Breakefield, XO and Kaye, EM (2000). Correction of acid beta-galactosidase deficiency in GM1 gangliosidosis human fibroblasts by retrovirus vector-mediated gene transfer: higher efficiency of release and cross-correction by the murine enzyme. Hum Gene Ther 11: 715–727. | Article | PubMed | ChemPort |
42. Seyfried, TN, Yu, RK and Miyazawa, N (1982). Differential cellular enrichment of gangliosides in the mouse cerebellum: analysis using neurological mutants. J Neurochem 38: 551–559. | Article | PubMed | ChemPort |
43. Svennerholm, L (1963). Chromatographic separation of human brain gangliosides. J Neurochem 10: 613–623. | Article | PubMed | ISI | ChemPort |

Acknowledgements

MLDB was supported by a Fulbright Scholarship. TNS supported by a grant from NIH (HD39722) and the NTSAD Association.

SUPPLEMENTARY MATERIAL

Figure S1.

Table S1. Effect of AAV vector on the concentration of lipids in the cerebellum of adult -gal mice.