Methods for delivery of polynucleotides by adeno-associated virus for lysosomal storage disorders

Abstract

The present invention relates to methods and materials useful for systemically delivering polynucleotides across the blood brain barrier using adeno-associated virus as a vector. For example, the present invention relates to methods and materials useful for systemically delivering α-N-acetylglucosamidinase polynucleotides to the central and peripheral nervous systems, as well as the somatic system. Use of these methods and materials is indicated, for example, for treatment of the lysosomal storage disorder mucopolysaccharidosis IIIB. As another example, the present invention relates to methods and materials useful for systemically delivering N-sulphoglucosamine sulfphohydrolase polynucleotides to the central and peripheral nervous systems, as well as the somatic system. Use of this second type of methods and materials is indicated, for example, for treatment of the lysosomal storage disorder mucopolysaccharidosis IIIA.

Claims

1. A method of treating mucopolysaccharidosis IIIB (MPS IIIB) in a patient in need thereof comprising: (i) intravenously administering to the patient about 1×10.sup.13 to about 1×10.sup.16 vg/kg of a recombinant adeno-associated virus 9 (rAAV9) comprising a single-stranded genome expressing an α-N-acetylglucosaminidase (NAGLU) polynucleotide, wherein the rAAV9-NAGLU genome consists essentially of the polynucleotide sequence of SEQ ID NO: 5; and (ii) expressing the encoded NAGLU polypeptide in neurons, glia cells, and endothelial cells of the central nervous system of the patient; wherein the administration and expression are effective to: (a) alleviate lysosomal storage lesions in the central nervous system, peripheral nervous system and other somatic tissues, (b) alleviate neuropathology, astrocytosis and/or neurodegeneration in the central nervous system and peripheral nervous system, and (c) restore α-N-acetylglucosaminidase (NAGLU) activity in somatic tissues; wherein mannitol is not administered to the patient prior to administering the rAAV9.

2. The method of claim 1, further comprising testing the patient for neuroinflammation after the rAAV9 has been administered.

3. The method of claim 1, further comprising testing for systemic expression of the polynucleotide in the peripheral central nervous system of the patient.

4. The method of claim 1, comprising administering about 1×10.sup.13 vg/kg of the rAAV9.

5. The method of claim 1, comprising administering about 1×10.sup.14 vg/kg of the rAAV9.

6. The method of claim 1, comprising administering about 1×10.sup.15 vg/kg of the rAAV9.

7. The method of claim 1, comprising administering about 1×10.sup.16 vg/kg of the rAAV9.

Description

BRIEF DESCRIPTION OF THE DRAWING

(1) FIG. 1 is a map of the rAAV-CMV-hNAGLU vector genome.

(2) FIG. 2 shows improved behavior and extended survival in MPS IIIB mice after systemic gene transfer by rAAV-CMV-hNAGLU. FIG. 2a. Hidden task in water maze (n.sup.=11/group). Day 1: test trial. FIG. 2b. Latency to fall from a rotarod (n.sup.=11/group). FIG. 2c. Survival (i 5/group, P<0.001). +/+: wt; −/−: MPS IIIB; AAV9-L, AAV9-H: MPS IIIB mice treated with 5×10.sup.12 or 1.5×10.sup.13 vg/kg rAAV9-hNAGLU vector, respectively. *: P<0.05 (vs. +/+); #: P<0.05 (vs. AAV9-L); {circumflex over ( )}: P<0.05 (vs. AAV9-H); &: P>0.05 (vs. −/−). Repeated measures ANOVA analyses: @: day effect P<0.01; $: group (treatment effect) P<0.01; %: day-group interaction P41.020 (rotarod).

(3) FIG. 3 shows rAAV9-mediated expression of functional rNAGLU in tissues. Tissues from MPS IIIB mice treated with rAAV9-hNAGLU were assayed for NAGLU activity (6 and 9 mo pi)(n=5-6/group). FIG. 3a. Dose-response. +/+: wt; AAV9-11, AAV9-L: MPS IIIB mice treated with 1.5×10.sup.13(AAV9-H) or 5×10.sup.12 vg/kg (AAV9-L) vector; FIG. 3b. Impact of mannitol pretreatment. M+/M−: MPS IIIB mice treated with 2×10.sup.13 vg/kg vector with (M+) or without (M−) mannitol pretreatment. FIG. 3c. Plasma NAGLU activity (n=3-4). +/−: heterozygotes. No significant difference in tissue NAGLU activity was detected at 6 and 9 months pi. Data shown are means±SD of combined data on tissues from mice at 6 and 9 mo pi. *: P<0.01 vs. +/+; #: P<0.05 vs. AAV9-H or M+: P>0.05 vs. +/+. @: P<0.05 vs. +/−.

(4) FIG. 4 shows the significant reduction of GAG content in the CNS and somatic tissues. Tissues from MPS IIIB mice treated with rAAV9-hNAGLU were assayed to quantify GAG content (6 and 9 mo pi). FIG. 4a. Dose response. FIG. 4b. Impact of mannitol pretreatment. +/Ai wt; −/−: MPS IIIB; AAV9-H, AAV9-L: MPS MB mice treated with 1.5×1013 vg or 5×1012 vg/kg vector; M+, M−: MPS IIIB mice treated with rAAV9 vector (2×1013 vg/kg) with or without mannitol pretreatment. Data shown are means±SD (n=5-6), combining data from tissues collected at 6 and 9 mo pi. *: P<0.01 vs. +/+; #: P<0.05 vs. AAV9-H or M+; {circumflex over ( )}: P<0.05 vs. AAV9-L or M−; +: P>0.05 vs. +/+.

(5) FIG. 5 shows rAAV9-mediated correction of astrocytosis and neurodegeneration in MPS IIIB mice. Brain sections of MPS IIIB mice treated with rAAV9-CMV-hNAGLU vector (6 mo pi) were assayed for GFAP by immunofluorescence and stained with toluidine blue for histopathology. FIG. 5a. Number of astrocytes: Data are means±SD of GFAP+ cells per 330×433 pm on 6-8 IF-GFAP-staining sections/mouse, from 3 mice/group. FIG. 5b. Number of purkinje cells: Data are means±SD of purkinje cells/200 p.m (in length) in ansiform lobules in cerebellum on 6 toluidine blue stained sections/mouse, from 3 mice/group. NT: non-treated MPS IIIB mouse; AAV9: MPS IIIB mouse treated with rAAV9. CTX: cerebral cortex; ST: Striatum; TH: thalamus; BS: Brain stem. *: P<0.01 vs. non-treated.

(6) FIG. 6 shows rAAV9-mediated expression of functional rSGSH in tissues of treated MPSIIIA mice. For each tissue, AAV9, rh74 and untreated result bars are respectively shown from left to right.

(7) FIG. 7 shows a significant reduction of GAG content in tissues of treated MPSIIIA mice. For each tissue, AAV9, rh74 and untreated result bars are respectively shown from left to right.

(8) FIG. 8 shows an improvement in cognitive behavior assays after treatment of one-month old MPSIIIA mice with low dose scAAV9 or rh74-U1a-SGSH. In the lower graphs, for each tissue, untreated, wild type and either AAV9 or rh74 result bars are respectively shown from left to right.

(9) FIG. 9 shows a significant reduction of GAG content in tissues of MPSIIIA mice treated at 2 or 6 months of age. For each tissue, wild type, untreated, two-month and six-month result bars are respectively shown from left to right.

(10) FIG. 10 shows an improvement in cognitive behavior assays after treatment of two- or six-month old MPSIIIA mice with high dose scAAV9 or rh74-U1a-SGSH. In the lower graphs, for each tissue, untreated, wild type and either AAV9 or rh74 result bars are respectively shown from left to right.

DETAILED DESCRIPTION

(11) The present invention is illustrated by the following examples relating to delivery of human NAGLU (hNAGLU) genes to the spinal cord via intravenous delivery of rAAV9. Example 1 describes rAAV encoding hNAGLU. Example 2 describes the administration of the rAAV encoding hNAGLU to MPSIIIB mice. Examples 3 through 6 describe the beneficial results of administration of the rAAV. Example 7 discusses the significance of the results. Example 8 describes rAAV encoding SGSH. Examples 9 through 11 describe administration of various dosages of rAAV encoding SGSH to MPSIIIA mice of varying ages, as well as the beneficial effects of the administration.

Example 1

Recombinant AAV (rAAV) Viral Vectors Encoding NAGLU

(12) A rAAV vector plasmid, containing AAV2 ITRs, an immediate early CMV promoter/enhancer, an SV40 intron, a human α-N-acetylglucosaminidase coding region, a bGH polyadenylation signal sequence, and ampicillin resistance gene, was used to produce a rAAV9-CMV-hNAGLU viral vector.

(13) Recombinant AAV9 viral vectors with the hNAGLU-encoding genome were produced in 293 cells using three-plasmid co-transfection, and purified as described in Zolotukhin et al., Gene Ther., 6: 973-985 (1999). This vector is referred to as “rAAV9-CMV-hNAGLU” herein. The vector genomes contained minimal elements required for transgene expression, including AAV2 terminal repeats, a human cytomegalovirus (CMV) immediate-early promoter, SV40 splice donor/acceptor signal, a human NAGLU coding sequence (SEQ ID NO: 1), and bGH polyadenylation signal. SEQ ID NO: 5 is the DNA sequence of the vector genome. FIG. 1 is a map of the vector genome wherein the length of the various elements of the genome is indicated below the element.

(14) A control self-complementary AAV encoding green fluorescent protein, scAAV9-CMV-GFP was also produced, containing AAV2 terminal repeats, a human cytomegalovirus (CMV) immediate-early promoter, SV40 splice donor/acceptor signal, a eGFP coding sequence, and SV40 polyadenylation signal.

Example 2

Administration of Viral Vectors

(15) An MPS IIB3 knock-out mouse colony [Li et al., Proc. Natl. Acad. Sci. USA, 96: 14505-14510 (1999) was maintained on an inbred background (C57BL/6) of backcrosses of heterozygotes. All care and procedures were in accordance with the Guide for the Care and Use of Laboratory Animals [DHHS Publication No. (NIH) 85-23]. The genotypes of progeny mice were identified by PCR.

(16) To assess the therapeutic efficacy of rAAV9 gene delivery, 4-6-week-old MPS IIIB mice were treated with an IV injection of rAAV9-CMV-hNAGLU (5×10.sup.12 or 1.5×10.sup.13 vg/kg, n=11/group). Separately, other MPS IIIB mice were treated with 2×10.sup.13 vg/kg rAAV9-CMV-hNAGLU, with or without mannitol pretreatment (n=5/group), to assess the impact on CNS entry. Controls were sham-treated (phosphate-buffered saline) wild type (wt) and MPS IIIB littermates (n=11). Tissue analyses were carried out at 6 months and 9 months (n=2-4/group) post-injection (pi).

(17) Additionally, self-complementary AAV (scAAV) vector carrying a cytomegalovirus-green fluorescent protein (CMV-GFP) transgene (5×10.sup.12 vg/kg) was injected IV into 6-8-week-old wt mice (n=4/group) to determine the distribution of transgene expression (1 month pi), as a comparison to rAAV9-hNaGlu treatment.

(18) Results are presented below.

Example 3

Behavioral Tests

(19) The rAAV9-CMV-hNaGlu-treated MPS BIB mice and controls were tested for behavioral performance at approximately 5.0-5.5 months of age as follows.

(20) Hidden Task in the Morris Water Maze

(21) A water maze (diameter=122 cm) was filled with water (45 cm deep, 24-26° C.) containing 1% white TEMPERA paint, located in a room with numerous visual cues. See Warburton et al., J. Neurosci, 21: 7323-7330 (2001). Mice were tested for their ability to find a hidden escape platform (20×20 cm) 0.5 cm under the water surface. Each animal was given four trials per day, across three days, as described previously. Measures were taken of latency to fmd the platform (sec) via an automated tracking system (San Diego Instruments). Results are shown in FIG. 2a.

(22) Rotarod

(23) Mice were tested on an accelerating rotarod (Med Associate, Inc.) to assess motor coordination. See Lijam et al., Cell, 90895-905 (1997). Rotation speed was set at an initial value of 3 revolutions per minute (rpm), with a progressive increase to a maximum of 30 rpm across five minutes (the maximum trial length). For the first test session, animals were given three trials, with 45 seconds between each trial. Two additional trials were given 48 hours later. Measures were taken for latency to fall from the top of the rotating barrel. Results are shown in FIG. 2b.

(24) Statistical Analyses

(25) Means, standard deviation (SD) and unpaired student t-test were used to analyze quantitative data. Behavioral measures were taken by an observer blind to experimental treatment. Behavioral testing data were also analyzed using repeated measures ANOVA (SAS 9.1.3) to determine the significance of the variances among treatment and control groups and testing days.

(26) Results of Behavioral Tests

(27) All mice treated IV with 5×10.sup.12 or 1.5×10.sup.13 vg/kg rAAV9-NAGLU were tested for behavior at 5-5.5 mo of age to assess the neurological impacts. Both dosage groups exhibited significant decreases in latency to find a hidden platform in a water maze (FIG. 2a), and significantly longer latency to fall from an accelerating rotarod (FIG. 2b), compared with non-treated MPS IIIB mice, indicating the correction of cognitive and motor function. There were no significant differences in behavior performance between these two dose groups.

Example 4

Longevity Assessment

(28) Following the rAAV9-hNaGlu vector injection(s), mice were continuously observed for the development of endpoint symptoms, or until death occurred. The endpoint was when the symptoms of late stage clinical manifestation (urine retention, rectal prolapse, protruding penis) in MPS IIIB mice became irreversible, or when wt control mice were 24 months or older. Longevity data were analyzed using Kaplan-Meier method. The significance level was set at P<0.05. Results are shown in FIG. 2c.

(29) Ten rAAV9-treated MPS IIIB mice, five from each dose group, were observed for longevity. All ten survived >16.9 months (with one mouse of the low-dose group dying at age of 16.1 months) and the majority of them survived 18.9-27.4 months within the normal range of lifespan, while all non-treated MPS IIIB mice died at 8-12 months of age (P<0.001) (FIG. 2c). These data demonstrate that a single IV rAAV9 vector injection alone is functionally beneficial in treating the CNS disease and increasing longevity in MPS IIIB mice.

Example 5

Tissue Analyses

(30) In the therapeutic studies above, tissue analyses were carried out at 6 mo and 9 mo post injection (pi). Mice were anesthetized with 2.5% Avertin before tissue collection. Brain, spinal cord and multiple somatic tissues were collected on dry ice or embedded in OCT compound and stored at −70° C., before being processed for analyses. Tissues were also processed for paraffm sectioning.

(31) Tissue samples from scAAV9-GFP vector-treated mice were collected for analysis 4-5 weeks pi. The mice were anesthetized with 2.5% Avertin and then perfused transcardially with cold PBS (0.1M, pH7.4), followed by 4% paraformaldehyde in phosphate buffer (0.1M, pH7.4). The entire brain and spinal cord, as well as multiple somatic tissues (including liver, kidney, spleen, heart, lung, intestine and skeletal muscles), were collected and fixed in 4% paraformaldehyde overnight at 4° C. before being further processed for vibratome sectioning.

(32) NAGLU Activity Assay

(33) Tissues were analyzed at 6 mo and/or 9 mo pi by NAGLU activity assay to determine the distribution and level of rAAV9-mediated transgene expression. Tissue samples were assayed for NaGlu enzyme activity following a published procedure with modification. The assay measures 4-methylumbelliferone (4MU), a fluorescent product formed by hydrolysis of the substrate 4-methylumbellireyl-N-acetyla-D-glucosaminide. The NaGlu activity is expressed as unit/mg protein. 1 unit is equal to 1 nmol 4MU released/h at 37° C. Results are shown in FIG. 3.

(34) GAG Content Measurement

(35) GAG was extracted from tissues following published procedures [van de Lest et al., Anal. Biochem. 221: 356-361(1994)] with modification [Fu et al., Gene Ther., 14: 1065-1077 (2007). Dimethylmethylene blue (DMB) assay was used to measure GAG content [de Jong et al., Clin. Chem., 35: 1472-1477 (1989)]. The GAG samples (from 0.5-1.0 mg tissue) were mixed with H.sub.2O to 40 m1 before adding 35 nM DMB (Polysciences CEO 03610-1) in 0.2 mM sodium formate buffer (SFB, pH 3.5). The product was measured using a spectrophotometer (0D535). The GAG content was expressed as μg/mg tissue. Urine GAG content was also measured. Heparan sulfate (Sigma, H9637) was used as standard. Results are shown in FIG. 4.

(36) Immunofluorescence

(37) Immunofluorescence (IF) was performed to identify cells expressing hNAGLU, GFP or glial fibrillary acidic protein (GFAP) for astrocytes, using antibodies against hNaGlu (a kind gift from Dr. E F Neufeld, UCLA), GFP (Invitrogen) or GFAP (Chemicon), and corresponding secondary antibody conjugated with AlexaFluor.sup.568 or AlexaFluor.sup.488 (Molecular Probes). The IF staining was performed on thin cryostat sections (8 p.m) of tissue samples following procedures recommended by the manufacturers. The sections were visualized under a fluorescence microscope.

(38) Histopathology

(39) Tissues were assayed for histopathology to visualize the impact of IV rAAV9-NAGLU gene delivery on the lysosomal storage pathology in MPS IIIB mice. Histopathology was performed following standard methods. Paraffm sections (41.un) were fixed with 4% paraformaldehyde in phosphate buffer (0.1 M, pH 7.2) at 4° C. for 15 min and stained with 1% toluidine blue at 37° C. for 30 min to visualize lysosomal GAG. The sections were mounted, and imaged under a light microscope.

(40) Quantitative Real Time PCR

(41) Total DNA was isolated from tissue samples of treated and nontreated MPS IIIB mice using Qiagen DNeasy columns. Brain DNA was isolated from midbrain tissue. The DNA samples were analyzed by quatitative real-time PCR, using Absolute Blue QPCR Mix (Thermo Scientific, Waltham, Mass.) and Applied Biosystems 7000 Real-Time PCR System, following the procedures recommended by the manufacturer. Taqman primers specific for the CMV promoter were used to detect rAAV vector genomes: f: GGCAGTACATC AAGTGTATC (SEQ ID NO: 6); r: ACCAATGG TAATAG CGATGAC (SEQ ID NO: 7); probe: [6˜FAM]AATGACGGTAAAT GGCCCGC[TAMRA˜6˜FAM] (SEQ ID NO: 8). Genomic DNA was quantified in parallel samples using β-actin specific primers: f: GTCATCAC TATTG GCAACGA (SEQ ID NO: 9); r: CTCAGGAGTTTTGTCACCTT (SEQ ID NO: 10); probe: [6˜FAM]TTCCGATGCCCT GAGGCTCT[Tamra˜Q] (SEQ ID NO: 11). Genomic DNA from nontreated MPS IIIB mouse tissues was used as controls or background and absence of contamination. Global CNS and widespread somatic restoration of NAGLU.

(42) Tissue Analysis Results

(43) Tissues were analyzed at 6 months and/or 9 months pi by immunofluorescence (IF) and NAGLU activity assay to determine the distribution and level of rAAV9-mediated transgene expression. NAGLU-specific IF was detected throughout the brains of treated mice, in neurons, glia, and abundant endothelial cells in capillaries and larger blood vessels, in an apparently dose-dependent fashion. No significant differences were observed in the distribution or levels of rNaGlu signal between 6 months and 9 months pi. NAGLU-positive glial cells were not costained with anti-glial fibrillary acidic protein (GFAP) Ab, and were likely to be oligodendrocytes, based on their morphology. Importantly, while rNAGLU IF was observed in the brains of all rAAV9-treated mice, mannitol pretreatment did appear to increase the number of transduced cells in the CNS.

(44) Differential transduction levels were observed in peripheral organs. The rNAGLU protein was detected in 20-40% of hepatocytes, >95% of cardiomyocytes, and 10-30% of skeletal myocytes. The distribution of rAAV9-transduced hepatocytes was uniform throughout the liver. Transduction in abundant neurons in myenteric plexus and submucosal plexus of the intestine was observed, suggesting efficient targeting of the peripheral nervous system (PNS). The rNAGLU signals were mostly present in granules, whereas scAAV9-mediated GFP signals were uniform in the cytoplasm of transduced cells, suggesting correct lysosomal trafficking of rNAGLU. Transduction of endothelial cells was also observed in peripheral tissues of rAAV9-GFP vector-treated mice.

Example 6

Enzyme Function Assays

(45) Function of the recombinant NAGLU and resulting effects in animals were also analyzed in the therapeutic studies above.

(46) rNAGLU Enzymatic Function

(47) Transgene enzymatic activity was assayed to quantify the expression and the functionality of rAAV9-mediated rNAGLU. There were no significant differences in tissue NAGLU activity at 6 and 9 months pi, suggesting stable transduction. The rAAVmediated rNaGlu was metabolically functional and the tissue rNAGLU activity was dose-dependent, with approximately normal levels in the brains of mice receiving 5×10.sup.12 vg/kg vector, and supra-physiologic levels in the brains of mice receiving 1.5×10.sup.13 vg/kg (FIG. 3a). In both dose groups, we detected NAGLU activity at normal or subnormal levels in the liver, lung and intestine (FIG. 3a), supra-physiologic levels in the skeletal muscles (FIG. 3a) and heart (40 & 100 units/mg protein, data not shown), and low levels in the spleen, but no detectable NAGLU activity in the kidney. A low level of NAGLU activity was detected in the kidneys of the mice treated with 2×10.sup.13 vg/kg vector (FIG. 3b). Mannitol pretreatment led to an increase in NAGLU activity in the brain (though not significant due to high individual variation), liver, spleen, lung and intestine, but a decrease in the heart and skeletal muscle (FIG. 3b). No detectable NAGLU activity (<0.03 unit/mg) was observed in tissues from non-treated MPS IIIB mice.

(48) rNaGlu Secretion

(49) Plasma samples were assayed for NAGLU activity to assess the secretion of the enzyme. Activity was detected in the plasma of all rAAV9-treated MPS MB mice at or near heterozygote levels, though lower than homozygous wt levels (FIG. 3c). Mannitol pretreatment resulted in significant reduction in plasma NAGLU activity (FIG. 3c). These data indicate that the rNAGLU was secreted, though the source tissue or cell type is not clear.

(50) GAG Content Reduction

(51) Tissues were assayed for GAG content to quantify the impact of IV rAAV9-NAGLU gene delivery on the lysosomal storage pathology in MPS IIIB mice. The single IV rAAV9-NAGLU injection led to a reduction of GAG content to normal levels in the brain, liver, heart, lung, intestine and skeletal muscle in mice of all four treatment groups (FIG. 4). Doses of 5×10.sup.12 μg or 1.5×10.sup.13 vg/kg resulted in partial GAG reduction in the spleen but had no impact in kidney (FIG. 4a).

(52) Treatment with 2×10.sup.13 vg/kg led to a decrease of GAG to normal levels in the spleen, and partial GAG reduction in the kidney (FIG. 4b), consistent with the observed enzyme activity levels.

(53) Histopathology Correction

(54) Histopathology showed complete clearance or reduction of lysosomal storage lesions in the vast majority of CNS areas, including cerebral cortex, thalamus, brain stem, hippocampus, and spinal cord in all four treatment groups. There were decreases in the size, number of vacuoles, and number of cells with lysosomal storage lesions, even in the few brain areas that did not show a complete correction, such as purkinje cells and cells in the striatum and hypothalamus. Importantly, the majority of brain and spinal cord parenchymal cells exhibited a well defined normalized morphology. Immunofluorecence detection for the lysosomal marker, LAMP-1, showed that IV infusion of rAAV9-NAGLU vector at all doses also resulted in marked reduction of LAMP-1 signal, especially in neurons, throughout the brain. This further supports the conclusion that the amount of vector crossing the BBB was sufficient for efficient correction of CNS lysosomal storage pathology.

(55) In somatic tissues, complete clearance of lysosomal storage lesions in the livers of all rAAV9-hNaGlu treated mice was observed as well as attenuation of nuclear shrinkage, a marker of cell stress and damage.

(56) Correction of Gliosis and Neurodegeneration

(57) In order to determine whether the rAAV9-hNaGlu vector delivery had an impact on astrocytosis, a major secondary neuropathology of MPS IIIB, brain sections were assayed by immunofluorecence for GFAP expressing cells. Significant decreases in astrocyte numbers in gray matter throughout the brain of treated mice were observed compared to untreated at 6 mo and 9 mo pi (FIG. 5a). Histopathology also revealed significant increases in the numbers of neurons, such as Purkinje cells (FIG. 5b), in the brains of treated MPS IIIB mice. These data strongly indicate the amelioration of astrocytosis and neurodegeneration, which are hallmarks of secondary neuropathologies in MPS IIIB, in response to the rAAV9 treatment.

(58) Vector Genome Distribution

(59) Quantitative real-time PCR was performed to compare the amount of rAAV9-CMV-hNaglu vector entering the CNS versus somatic tissues. Table 1 shows the distribution of the vector genome in different tissues/organs of MPS IIIB mice treated with IV vector injection at varying doses. The highest concentrations of vector genome were detected in liver (8.20±4.73-32.09±3.93 copies/cell), followed by heart (0.07-0.22 copies/cell), and brain (0.06±0.001-0.15±0.02 copies/cell), and very low copy numbers were detected in other tissues/organs (Table 1). This differential vector distribution in rAAV9-treated MPS IIIB mice largely correlated with the distribution of rNAGLU IF and enzymatic activity. Notably, mannitol pretreatment increased the vector copy numbers in the brain, correlating with brain NAGLU activity levels. Furthermore, these data reflect persistence of vector genome distribution in treated mice at 6 months pi, supporting a stable long-term transduction. Levels of vector genome copies correlating with rNAGLU activity and distribution were not detectable, possibly due to difficulties in quantitative isolation of DNA from muscle tissue.

(60) TABLE-US-00003 TABLE 1 Estimated vector genome in the liver and brain of rAAV9-treated mice Vector genome (copy/cell) Mice n Liver Brain Heart rAAV9-L 2  8.20 ± 4.73 0.07 ± 0.07 0.07* rAAV9-H 3 10.86 ± 2.94 0.09-10.47 0.13 ± 0.07 rAAV9-M− 2 21.97 ± 6.43 0.06 ± 0.001 0.22* rAAV9-M+ 3 32.09 ± 3.93 0.15 ± 0.02 0.14* Non-treated 1 0.000 0.000 0.00  Mouse tissue samples (6 mo pi) were assayed in duplicates for vector genome copy numbers by qPCR. Data is expressed as vector copy/cell (means ± SD). rAAV9-L: IV infusion of 5 × 10.sup.12 vg/kg; rAAV9-H: IV infusion of 1.5 × 10.sup.13 vg/kg; rAAV9-M−: IV infusion of 2 × 10.sup.13 vg/kg without mannitol pretreatment; rAAV9-M+: IV infusion of 2 × 10.sup.13 vg/kg following mannitol pretreatment. *Data from 1 sample in duplicates.

Example 7

Discussion

(61) This study demonstrates the first significant therapeutic benefit for treating MPS IIIB in adult animals from systemic gene delivery to the CNS without additional treatment to disrupt the BBB. A single IV injection of hNAGLU-expressing rAAV9 vector was sufficient to significantly improve cognitive and motor functions, and greatly prolong survival in MPS IIIB mice. In the present study using rAAV9, the increased longevity exceeds the outcome of previous studies using rAAV2 vector delivered through either intracisternal injection, or systemic injection following mannitol pretreatment. The rNAGLU enzyme was clearly secreted and functional, leading to a significant bystander effect, and efficient degradation of heparan sulfate GAG in CNS tissues. Importantly, the clinically meaningful therapeutic benefits of the IV-delivered rAAV9 vector in MPS IIIB mice were achieved at a lower dose than the mannitol-facilitated, systemically delivered rAAV2 vector. The enhanced rAAV9-CNS transduction in response to mannitol pretreatment suggests further potential for reducing the vector dose, and the attendant risk and burden to patients.

(62) The IV vector injection resulted in a ubiquitously diffuse, global rAAV9-NaGlu transduction throughout the CNS, reflecting the expected distribution pattern for vascular delivery. This contrasts sharply with the focal gradient distribution typically achieved through direct brain parenchymal injection, or the periventricular transduction pattern from intracisternal and intraventricular injection. While similar to the pattern of transgene expression from IV-delivered rAAV2 after mannitol pretreatment, the extent of rAAV9 transduction was significantly higher in all areas of the brain. This correlates with the increased effects on longevity and cognitive function compared to that previously achieved using rAAV2-mannitol treatment, and the normal or above normal levels of NAGLU activity in the CNS. These findings strongly support the use of the trans-BBB neurotropic rAAV9 as a vector for CNS gene therapy and reinforce the view that efficient CNS delivery is the most critical issue for developing therapies to treat MPS IIIB.

(63) The rAAV9-transduced CNS cells include neurons, glia and endothelia. Neuronal cell transduction appears to be non-preferential, including most types of neurons throughout the brain. In contrast, the transduction of glial cells appears to be cell-type specific, targeting predominantly oligodendrocyte-like cells, though it is unclear whether this is a receptor- or promoter-specific phenomenon. In a previous report [Faust et al., supra] describing predominant transduction of astocytes after systemic injection of rAAV9 vector in adult mice, a hybrid chicken J3-actin/CMV-enhancer promoter was used, rather than the CMV enhancer-promoter used in the present study.

(64) In normal cells, 5-20% of newly synthesized lysosomal protein is secreted and available to be taken up by neighboring cells, leading to the by-stander effect. The widespread clearance/reduction of lysosomal storage pathology, and normalized tissue GAG content, strongly support an efficient by-stander effect from the rAAV9-′mediated rNAGLU. The abundant transduction of endothelial cells in the brain may be an important contributor to the effectiveness of rAAV9 gene delivery for MPS IIIB because of the close association between CNS cells and brain microvascular endothelial cells, which together constitute the neurovascular unit. While the observed high levels of rNAGLU expression stem from the transduction of a relatively small number of CNS cells, it is sufficient to correct the neuropathology leading to functional correction of the neurological disorders.

(65) The rAAV9 treatment also led to a regular morphology in CNS cells, and the correction of major secondary neuropathology, astrocytosis, and neurodegeneration. It is worth noting that this level of correction of CNS pathology was not achieved in previous studies using rAAV2-hNAGLU vector with mannitol. While neuropathology is the primary cause of mortality in MPS IIIB patients, somatic correction may provide additional therapeutic benefits, since lysosomal storage pathology inevitably manifests in virtually all organs. The IV-delivered rAAV9 exhibited broad tropism in peripheral tissues in a distinct pattern, as previously reported, reflecting extensive extravasation and cell-type specific transduction. This led to complete, longterm correction of lysosomal storage in multiple somatic tissues even at a relatively low dose. Again, relatively low levels of transduction in some tissues were associated with clearance of lysosomal storage of GAGs in the organs, consistent with a significant contribution from the by-stander effect of secreted rNAGLU enzyme. It is not clear whether the by-stander correction in peripheral tissues is mediated by enzyme secreted from neighboring cells within the same tissue, or circulating rNAGLU secreted by more extensively transduced tissues, in a manner analogous to enzyme replacement therapy. However, the observation of partial GAG reduction in the kidney only at the highest vector dose, correlating with detectable transduction in the kidney only at that dose, suggests that the by-stander effect may be primarily local in this tissue. The primary source of circulating NAGLU may be liver, muscle, or endothelium. However, the decrease in plasma levels in response to mannitol pretreatment correlated with decreased transduction in muscle rather than liver, suggesting that liver may not be the primary source.

(66) Another important observation is the efficient transduction of neurons in myenteric plexus and submucosal plexus of the intestine, potentially enabling correction of not only the CNS but also the PNS at all levels via systemic delivery. This suggests that neurotropism is a general property of the AAV9 serotype, and not dependent on the specific structure of the brain neurovascular unit. Broad neurotropism is a valuable property in gene therapy for the treatment of MPS IIIB, considering that lysosomal storage pathology manifests not only in the CNS but also in the PNS.

Example 8

Recombinant AAV (rAAV) Viral Vectors Encoding SGSH

(67) A rAAV vector plasmid was used to produce three different rAAV9-CMV-hSGSH viral vectors.

(68) The three self-complementary AAV hSGSH vector-producing plasmids were constructed using conventional plasmid cloning techniques. Each vector genome contains an SGSH coding region (SEQ ID NO: 3) and either the mouse U1a promoter, with or without an intron, or a CMV promoter without an intron, Each vector genome also contains a bGH polyadenylation signal. Each self-complementary vector plasmid construct contains one intact AA2 terminal repeat and one modified AAV2 terminal repeat missing the terminal resolution site, thereby forcing the replication of dimeric self-complementary DNA genomes. Self-complementary AAV hSGSH viral vectors were produced and packaged in AAV serotype 9 capsids. The viral vectors were tested for expression of hSGSH protein and reduction of GAG storage in human MPS IIIA fibroblasts.

(69) Self-complementary AAV hSGSH viral vectors were tested in an MPS IIIA mouse model having a missense mutation in the SGSH gene [Bhaumik et al., Glycobiology, 9(12):1389-1396 (1999)] as described in the examples below.

Example 9

(70) MPS IIIA mice were injected at 10 weeks of age with 5×10.sup.12 vgp/kg) of scAAV-U1a-hSGSH vector encapsidated in either AAV9 or AAVrh74 serotype. At 10 days post-injection, the mice were euthanized and assays were performed to determine the effects of the treatment.

(71) hSGSH transgene expression was assayed. Tissues analyzed included kidney (Kid), heart (Hrt), intestine (Int), skeletal muscle (Mus), lung, brain, Liver (Liv), spleen (Spl) and serum. FIG. 6 shows enzyme expression relative to untreated MPS IIIA mice at the same age (−/−). The scAAV-SGSH vectors reached the CNS and expressed the transgene within days of administration. FIG. 7 shows GAG content measured in the kidney (Kid), heart (Hrt), muscle (Mus), lung, brain, Liver (Liv) and spleen (Spl).

(72) Sections of CNS and somatic tissues were stained with the lysosomal marker, Lamp1, revealing clearance of lysosomal storage pathology. Histopathology additionally revealed numerous clear vacuoles present in untreated mice but corrected in treated animals.

Example 10

(73) The therapeutic effects of scAAV-SGSH treatment at a low vector dose were examined.

(74) Vector was administered by tail vein injection in MPS IIIA mice at one month of age at an approximately 25-fold lower dose than in Example 9. MPS IIIA mice were treated with 1.7×10.sup.11 vgp/kg scAAV9-U1a-hSGSH or 2.7×10.sup.11 vgp/kg scAAVrh74-U1a-hSGSH vector.

(75) At three months post-injection, expression of SGSH in the CNS was observed by immunofluroescence staining. Correction of astrocytocis, a hallmark of neuroinflamation associated with MPS IIIA pathology, was also observed.

(76) At 7-7.5 months age, the animals were tested for learning ability in the Morris water maze. As shown in FIG. 8, compared to untreated controls, treated animals were similar to wt mice in their latency to locate the hidden platform (upper charts) and spent more time in the zone (4) where the platform had been in the previous tests when the platform was removed (lower charts).

Example 11

(77) Therapeutic effects of scAAV treatment at high dose at late stage of disease were also examined.

(78) MPS IIIA mice were treated with a high dose (5×10.sup.12 vgp/kg) of scAAV9-U1a-hSGSH vector at 6 months of age, after significant neuropathology had already developed. At 7-7.5 months age, the animals were tested for learning ability in the Morris water maze. At 7.5 months of age, the mice were euthanized and tissues assayed for glycosaminoglycan (GAG) content. Tissues analyzed include liver (Liv), kidney (Kid), heart (Hrt), brain, spleen (Spl), lung, skeletal muscle, and intestine.

(79) FIG. 9 shows clearance of accumulated GAGs in different tissues, including CNS. FIG. 10 shows, compared to untreated controls, treated animals were similar to wt mice in their latency to locate the hidden platform (upper charts) and spent more time in the zone (4) where the platform had been in the previous tests when the platform was removed (lower charts).

(80) While the present invention has been described in terms of various embodiments and examples, it is understood that variations and improvements will occur to those skilled in the art. Therefore, only such limitations as appear in the claims should be placed on the invention.

(81) All documents referred to in this application are hereby incorporated by reference in their entirety.