Introduction

Globoid-cell leukodystrophy (GLD, Krabbe disease) is an autosomal-recessive neurodegenerative disease caused by a deficiency of the lysosomal enzyme galactosylceramidase (GALC).^{1, 2} This enzyme is responsible for the degradation of galactocerebroside, a major glycolipid of central and peripheral myelin. The disease is characterized by global demyelination in the central (CNS) and peripheral (PNS) nervous systems, rapidly progressing neurodegeneration, and early death. In the infantile form of GLD, symptoms begin between 3 and 6 months of age and include irritability, dysphagia, spasticity, cognitive and sensory deterioration, seizures, and eventual death usually by 2 years of age.²

The Twitcher mouse is deficient in GALC activity and shares many of the neuropathologic findings of human infantile GLD.³ Histologically, there is infiltration of periodic acid-Schiff (PAS)-positive cells in the CNS and PNS, widespread demyelination, loss of oligodendrocytes, and the presence of cytoplasmic inclusions in myelin-forming cells and macrophages.⁴ Neurologic symptoms occur between 15 and 20 days after birth and include tremors, progressive hind limbs paralysis, and loss of coordination.

We and others showed that intracranial injections of recombinant adenovirus or adeno-associated viral vectors in Twitcher mice resulted in relatively high (as much as 25-fold higher than normal) levels of GALC activity in the brain.^{5, 6, 7} However, the supraphysiological levels of GALC activity in the brains of Twitcher mice only slowed the progression of the disease and resulted in minimal improvements in myelination, behavior, and lifespan. These data suggest that simply supplying high levels of GALC to select regions of the CNS is not sufficient to correct this disease.

Bone marrow transplantation (BMT) in young Twitcher mice following high-dose (800–900 rad) radiation can extend the lifespan of these animals to an average of approximately 80–90 days.^{8, 9, 10} Allogeneic BMT has also been shown to improve the clinical course of disease in children with GLD if initiated early in life.^{11, 12} However, BMT provides only partial correction of the disease and there can be severe side effects associated with this procedure.

The CNS-directed gene therapy and BMT data suggest that alternate approaches will be required to treat this inherited neurodegenerative disease effectively. We hypothesized that the high levels of GALC activity in the brain following CNS-directed gene therapy would complement the therapeutic effects of BMT and provide more complete correction of the disease in the Twitcher mouse. Here we show that a combination of CNS-directed AAV2/5-mediated gene therapy and BMT performed using myeloreductive rather than myeloablative conditioning is more efficacious than either therapy alone. In fact, the effects of the combined therapy appear to be synergistic rather than simply additive. A combination of therapies such as this may form the basis of more effective therapies for GLD.

Results

Bone marrow engraftment following myeloreductive conditioning

Syngeneic GFP-positive bone marrow was used as a source of donor cells in order to determine the level of engraftment in the recipient animals that received sublethal doses of conditioning radiation. Donor hematopoietic chimerism in the bone marrow was evaluated in transplanted mice at 40 days of age by fluorescence-activated cell sorting. The mean donor engraftment in transplanted normal, Twitcher, and combination-treated Twitcher mice was 24.63.8, 32.711.7, and 37.111.6%, respectively (Figure 1). All of the transplanted animals had appropriate levels of myeloid engraftment with 1.50.6 and 27.55.1% of the GFP-positive cells being Mac-1 and GR-1 positive, respectively (data not shown). No wild-type (n=4) or Twitcher (n=4) mice that received 400 rad of radiation at birth and no donor cells died within 40 days of age (data not shown).

Figure 1.

Donor hematopoietic chimerism. Partial donor chimerism (25–35% GFP⁺ cells) was observed in the bone marrow of normal (WT), BMT-treated (BMT), or combination-treated (CmTx) Twitcher mice 40 days following myeloreductive conditioning (400 rad) and injection of 106 unfractionated GFP⁺ bone marrow cells. The means and 1 SD are shown.

Full figure and legend (5K)

GALC level and distribution

Brain GALC activity was measured in homogenates of one entire sagittal half of the brain from experimental and control animals (Figure 2). The GALC activity in the brains of untreated Twitcher mice was approximately 2.2 (1.9)% normal. GALC activities in the brains of untreated and BMT-treated Twitcher mice were not significantly different. In contrast, GALC activities in the brains of AAV2/5-treated and combination-treated Twitcher mice were 21766 and 38080% normal, respectively. Liver and kidney GALC activities in untreated Twitcher mice were 0 and 0.91.1% normal, respectively. However, Twitcher mice that received BMT alone had GALC activity levels of 6.55 and 0.60.9% in the liver and kidney, respectively. Twitcher mice that received intracranial injections of AAV2/5 alone had levels of GALC activity in the liver that were approximately 4.85% normal. Interestingly, Twitcher mice that received the combination treatment had GALC levels of 2915 and 2.62.4% in the liver and kidney, respectively.

Figure 2.

GALC activity. GALC activity, expressed as a percent normal (% normal), was measured in the brain, liver, and kidney of Twitcher mice following individual or combined therapies. The normal GALC levels were determined by the mean activity in each tissue measured in three normal animals at 38 days of age. Untreated Twitcher mice (Twi, 38 days of age, n=3) had 2.2, 0, and 0.9% GALC activity in the brain, liver, and kidney, respectively. Twitcher animals receiving BMT alone (BMT, 34–50 days of age, n=5) had GALC levels in the brain that were not significantly different from untreated Twitcher animals. However, BMT-treated animals had 6.5 and 0.6% normal levels in the liver and kidney, respectively. Intracranial injections of AAV2/5 alone (AAV2/5, 34 days of age, n=3) resulted in 217, 4.8, and 1.0% normal GALC levels in the brain, liver, and kidney, respectively. The combination-treated twitcher mice (AAV/BMT, 34 days of age, n=3) had 380, 29, and 2.6% normal GALC levels in the brain, liver, and kidney, respectively. The bars represent the mean1SEM.

Full figure and legend (25K)

Quantitative GALC measurements in tissue homogenates do not indicate the distribution of GALC activity. Therefore, cryosections of brain from treated and control mice were stained histochemically for GALC activity with a previously described method¹³ using X-gal as a substrate and inhibiting endogenous -galactosidase activity. Twitcher mice receiving intracranial injections of the AAV2/5 vector, either alone or in combination with BMT, showed intense staining in relatively large numbers of cells throughout the cerebral cortex, hippocampus, and thalamus (Figure 3). The most intense staining was observed in the ependyma and choroids plexus. In contrast, BMT-treated Twitcher mice were essentially indistinguishable from untreated Twitcher mice and showed only a small number of GALC-positive cells throughout the brain.

Figure 3.

GALC localization. GALC-positive (blue) cells were observed in the cerebral cortex, hippocampus, lateral ventricle, and thalamus of untreated normal mice. Numerous intensely staining GALC-positive cells were detected in the same structures from combination-treated (CmTx) and AAV2/5-treated (AAV) Twitcher mice. However, very few GALC-positive cells were observed in similar regions of either BMT-treated (BMT) or untreated (Twi) Twitcher brains.

Full figure and legend (240K)

Reduced inflammatory markers in the brain

We examined the level of inflammation in the brains of untreated and treated Twitcher mice by immunostaining for CD11b and glial fibrillary acidic protein (GFAP). Infiltration of CD11b-positive macrophages was prominent throughout the white matter tracts of the striatum and cerebellum in untreated Twitcher mice (Figure 4a). BMT alone appeared to reduce the intensity of CD11b staining throughout the neuroaxis as observed in the striatum and cerebellum. Injections of AAV2/5 dramatically reduced CD11b staining near the injection sites (e.g., cortex, corpus callosum, and striatum). However, AAV2/5 had little or no effect on CD11b staining in the cerebellum, which was not injected with vector. The overall intensity of CD11b staining in the brains of Twitcher mice receiving the combination treatment was less than either therapy alone. Widespread astrocytosis, as assessed by immunostaining for GFAP, was observed throughout the cerebrum and cerebellum, most notably in the gray matter, of untreated Twitcher mice when compared to untreated normal mice (Figure 4b). Similar to the results obtained with CD11b staining, treatment with either BMT alone or AAV2/5 alone resulted in a mild-to-moderate reduction of GFAP staining. GFAP staining in animals that received combination treatment was less intense than either treatment alone. However, Twitcher animals that received combination treatment still had more intense CD11b and GFAP staining than that observed in untreated normal animals.

Figure 4.

Astrocytosis and microglial activation. Widespread macrophage/microglial activation was observed by (a) CD11b staining in the striatum and cerebellum of untreated Twitcher (Twi) mice. Similarly, astrocytosis in the cortex and striatum of untreated Twitcher mice was detected by intense immunostaining for GFAP (b). The BMT-treated (BMT) Twitcher mice had a slight decrease in CD11b and GFAP staining in the striatum and cerebellum. The AAV2/5-treated (AAV) Twitcher mice have decreased GFAP and CD11b staining in the cortex and striatum near the injection sites. However, there is no appreciable decrease in CD11b staining in the cerebellum of AAV2/5-treated animals (a, cerebellum, AAV). The decrease in both CD11b and GFAP staining in all the brain areas examined from the combination-treated (CmTx) Twitcher mice was more dramatic than either AAV2/5 or BMT alone. There was little or no CD11b or GFAP staining in the brains of normal control animals (WT).

Full figure and legend (318K)

CNS histolopathology

Untreated Twitcher mice had decreased myelin staining and numerous PAS-positive macrophage-like cells in and around the corpus callosum compared to normal mice (Figure 5). In contrast, three Twitcher mice chosen at random that received combination treatment and were examined histologically had a marked reduction in PAS-positive globoid cells in the cerebral neocortex, corpus callosum, and internal capsule. Two of the three combination-treated Twitcher animals were indistinguishable from normal mice by light microscopy in this level of the brain. However, the brain stem structures from these combination-treated mice had persistent, albeit fewer globoid cells than those seen in untreated Twitcher mice (data not shown). Twitcher mice treated with AAV2/5 alone had similar improvements to those treated with a combination of AAV2/5 and BMT in the cerebral neocortex, corpus callosum, and internal capsule. However, the brain stem and cerebellum from these mice were indistinguishable from the untreated Twitcher mice. The three Twitcher mice treated with BMT alone had persistent pathology that was indistinguishable from untreated Twitcher mice in the cerebrum. Interestingly, the number of PAS-positive staining globoid cells was even greater than untreated Twitcher mice in the brain stem of Twitcher animals receiving BMT alone. Most of the normal and Twitcher mice that received 400 rad of conditioning radiation before BMT had a slight loss of granular cells in the cerebellum, a change that likely relates to irradiation of the immature brain.¹⁴

Figure 5.

Myelination and globoid cells. Increased numbers of PAS-positive cells (arrows, 11 cells in this panel) are present in and around the corpus callosum in untreated (Twi) Twitcher mice compared to normal (WT) animals. Comparable numbers of PAS-positive cells were observed in the corpus callosum of Twitcher animals receiving BMT alone (data not shown). There was a dramatic decrease in the number of PAS-positive cells in both the AAV2/5-treated (AAV, two positive cells) and the combination-treated (CmTx, 0 positive cells) animals. Comparable levels of myelin (luxol fast blue staining) were observed in the corpus callosum of normal and combination-treated animals. Decreased myelin staining was observed in both the untreated Twitcher and the AAV2/5-treated animals compared to normal animals.

Full figure and legend (103K)

Longevity and weight gains

Lifespan was used as another objective measure to determine the therapeutic efficacy of the various regimens (Figure 6). The average lifespan of BMT-, AAV2/5-, and combination-treated Twitcher mice was 441 days (range 34–50 days, n=15), 491 days (range 42–56 days, n=15), and 1047 days (range 38–151 days, n=17), respectively. The combination-treated Twitcher mice had a significantly (P<0.001) increased lifespan compared to any of the other treated Twitcher mice (50% of the combination-treated mice lived beyond 111 days). All the treated Twitcher groups had significantly (P<0.01) increased lifespans compared to untreated Twitcher mice (391 days; range 32–44 days, n=24). The AAV2/5-treated Twitcher mice had significantly (P<0.001) longer lifespans than BMT-treated Twitcher mice. All the Twitcher mice, untreated or treated, had significantly shorter lifespans than the normal control animals. There were no significant differences between the untreated (n=12), BMT-treated (n=14), or combination-treated (n=13) normal control groups within the time frame of this study.

Figure 6.

Life span. The mean lifespans of untreated (Twi, bold solid/dotted line, n=14), BMT-treated (BMT, bold dotted line, n=15), AAV2/5-treated (AAV2/5, bold dashed line, n=15), and combination-treated (CmTx, bold solid line, n=17) Twitcher mice are 39, 44, 49, and 104 days, respectively. Fifty percent of the combination-treated animals were still alive at 111 days of age. No untreated normal (WT, fine solid line, n=12) or normal mice that received radiation alone (data not shown, n=4) died during the study.

Full figure and legend (12K)

Body weight was used as an objective measure of disease progression. The weight gain for all animals, treated or untreated, in the study was indistinguishable up to 21 days of age. Thereafter, the weight gain in untreated and treated Twitcher mice was slower than that for the normal mice (Figure 7). At 35 days of age, both the AAV2/5-treated and combination-treated Twitcher mice had a significantly (P<0.001) greater weight than the untreated and BMT-treated Twitcher mice. By day 42, the body weights of the combination-treated Twitcher mice reached a plateau that was significantly (P<0.001) greater than any other Twitcher group and maintained that weight until the moribund stage. In contrast, both the BMT-treated and AAV2/5-treated groups started losing weight at 42 days of age. Untreated Twitcher mice were indistinguishable from BMT-treated Twitcher mice at every time point. There were no significant differences between treated and untreated normal control mice.

Figure 7.

Body weight. The body weights of all the groups of animals were indistinguishable up to 21 days of age. Beyond 28 days of age the body weights of untreated (filled triangles) and BMT-treated (open circles) Twitcher mice were identical and decreased precipitously. Beyond 35 days of age the mean body weight of the AAV2/5-treated group (open triangles) decreased rapidly. The mean body weight of the combination-treated Twitcher mice (filled diamonds) remained relatively constant for the duration of the study, albeit less than either the untreated (filled circles) or BMT-treated (open squares) normal mice.

Full figure and legend (15K)

Behavioral performance

We and others have shown previously that the rotarod assay is an effective means of evaluating the clinical effects of various therapeutic regimens in the Twitcher mouse.^{7, 15} There was no significant difference between any of the groups in the constant speed rotarod up to 30 days of age (Figure 8). All the treated groups had nearly normal performance for the first 35 days of age whereas the untreated Twitcher mice showed significantly decreased performance starting at 35 days of age. The AAV2/5-treated Twitcher mice were indistinguishable from the BMT-treated group and their performance rapidly declined after 40 days. At time points beyond 40 days of age, combination-treated Twitcher mice performed significantly (P<0.01) better than untreated, BMT-treated, and AAV2/5-treated Twitcher mice in both the constant speed and accelerating rotarod tests. The AAV2/5-treated, BMT-treated, and combination-treated Twitcher mice demonstrated significantly decreased performance compared to normal mice starting at 40, 45, and 50 days of age, respectively.

Figure 8.

Behavioral analysis. The combination-treated Twitcher mice (filled diamonds, n=17) performed significantly better than untreated (filled triangles, n=14), BMT-treated (open circles, n=15), or AAV2/5-treated (open triangles, n=15) Twitcher mice in the constant speed and accelerating rotarod paradigms. Both the BMT-treated and AAV2/5-treated Twitcher groups performed significantly better than the untreated Twitcher mice in both rotarod assays. There was no significant improvement in any of the treated Twitcher mice in the wire hang test when compared to untreated Twitcher mice. All the treated Twitcher mice had significant deficits compared to untreated normal mice (filled circles, n=12) in both rotarod paradigms and in the wire hang assay.

Full figure and legend (42K)

In the accelerating rotarod paradigm, the untreated, BMT-, and combination-treated Twitcher mice had significantly (P<0.02) impaired performance compared to normal mice starting on day 25. The AAV2/5-treated group showed significantly (P<0.01) impaired performance starting on day 35. Both the AAV2/5-treated and combination-treated groups performed significantly (P<0.05) better than untreated Twitcher mice starting on day 25, whereas the BMT-treated group showed significantly (P<0.03) better performance than untreated Twitcher mice starting on day 30. The combination-treated group had significantly (P<0.02) improved performance compared to both the AAV2/5-treated and BMT-treated group starting on day 40.

We suggested previously that the wire hang test might be a meaningful assay to measure the effect of various therapies on the PNS disease.⁷ Although the treated Twitcher mice showed improvements in other behavioral paradigms and there was a decrease in edema in the sciatic nerve, there was no significant improvement in any of the treated Twitcher groups in the wire hang test compared to untreated Twitcher mice (Figure 8). There were no significant differences between any of the treated normal animals and the untreated normal animals in any of the behavioral tests (data not shown).

Discussion

GLD is a rapidly progressing neurodegenerative disease caused by a deficiency of the lysosomal enzyme GALC. Preclinical experiments in the murine model of GLD have shown that CNS-directed gene therapy can provide high levels of GALC activity in the brain and modest therapeutic effects.^{5, 6, 7} Likewise, BMT provides some therapeutic benefit in animal models and humans and much of the clinical benefit may come from the anti-inflammatory effects of this approach. Although effective, myeloablative BMT is not without risks. Fully myeloablative conditioning regimens can have deleterious effects in young animals¹⁴ and children, and graft-versus-host disease is a common clinical consequence of allogeneic BMT.

Based on the benefits and limitations of each therapy, we hypothesized that the high level of GALC activity in the brains of AAV2/5-treated Twitcher mice might act in concert with BMT by targeting different aspects of the disease to provide additional efficacy. We also hypothesized that a mild conditioning regimen would allow therapeutic levels of engraftment while minimizing the toxicity associated with complete myeloablation. Interestingly, the effects of AAV2/5-mediated CNS-directed gene therapy combined with myeloreductive BMT appeared to be synergistic rather than simply additive. Although myeloreductive BMT and AAV2/5 alone increase lifespan by a mere 5 and 10 days, respectively, the combination of these two therapies extended the lifespan by 65 days. The mean lifespan of the combination-treated animals in this study (104 days) was greater than that observed in previous studies where Twitcher mice received myeloablative conditioning before BMT (80 days).^{8, 16}

Combinations of various therapies have been attempted previously in other murine models of lysosomal storage diseases. We previously showed in the murine model of mucopolysaccharidosis type VII that a combination of enzyme replacement therapy at birth followed by myeloreductive (600 rad) conditioning and BMT at 6 weeks of age was more effective than either therapy alone.¹⁷ Unlike the current study, however, the combined effects of that treatment regimen appeared to be additive rather than synergistic. Moderate synergy (synergy quotient=1.16, see Materials and Methods section) was observed for the lifespan in the murine model of Sandhoff disease when substrate deprivation therapy was combined with BMT.¹⁸ The synergy quotient for the lifespan in the current study with a combination of AAV2/5 and myeloreductive BMT was 4.33 (the larger the number the greater the synergy). In addition, both previous studies utilized forms of therapy that are transient in nature (enzyme replacement therapy in mucopolysaccharidosis type VII and substrate deprivation in Sandhoff disease) whereas both therapies used in the current study have the potential to provide a long-term source of the deficient enzyme.

In addition to the possible therapeutic benefits of combining intracranial injections of AAV2/5 with BMT, we also reasoned that performing BMT using myeloreductive (400 rad) rather than myeloablative (800–900 rad) conditioning regimens would decrease the radiation-induced damage during the newborn period. Although complete donor chimerism is sacrificed with myeloreductive conditioning, there was only minor cerebellar dysplasia observed in the animals receiving 400 rad as compared to the profound dysplasia observed in a previous study of newborn mice receiving 600–800 rad.¹⁴ The minor loss of granule cells in the cerebellum in the current study did not affect the behavioral performance of normal control animals receiving 400 rad and BMT.

The simplest explanation for the synergy observed with combination-treatment in the current study is that the sum of GALC activity provided by each therapy is sufficient to produce the dramatic improvements. However, the additional amount of GALC activity in the brains of Twitcher mice that received BMT was negligible compared to the levels of activity expressed from the AAV2/5 vector. Although it is possible, perhaps even likely, that the small amount of activity contributed by BMT may be localized to critical regions of the brain not accessible to the AAV2/5 vector, we believe that other factors also contribute to the synergy. It is possible that the combination of high GALC activity in the brain following intracranial injections of AAV2/5 synergized with the anti-inflammatory effects of BMT. There is convincing evidence suggesting that inflammation contributes to the CNS disease in several models of lysosomal storage disease,^{19, 20, 21} including GLD.^{22, 23, 24} It has been demonstrated that BMT can reduce the expression of proinflammatory molecules in the brains of Twitcher mice.²⁵ Each form of therapy individually reduces GFAP (astrocytosis) and CD11b (activated macrophages and microglia) staining in the brain; however, animals receiving both forms of therapy have a more complete reduction of these proinflammatory markers. It is possible that the anti-inflammatory effects of the myeloreductive BMT, with only 25–35% donor chimerism, are mediated by suppressor T cells which comprise less than 3% of all T cells.²⁶ It will be interesting to determine if GLD leads to a decrease in the number or function of suppressor T cells and if higher levels of engraftment (and presumably higher numbers of donor suppressor T cells) will synergize with CNS-directed gene therapy to a greater degree.

Although Twitcher mice receiving a combination of CNS-directed gene therapy and BMT had a better clinical outcome than mice receiving either therapy alone, they still had significant motor deficits. It should be noted that the Twitcher animals did not receive AAV2/5 injections in the cerebellum and this was an area of the brain that still had relatively high levels of proinflammatory markers. Injections of AAV2/5 into this region of the brain might increase efficacy. Alternatively, the limited clinical response could be due to the disease in the PNS. It has been shown previously that the PNS of Twitcher mice is relatively refractory to therapy. Psychosine levels in the sciatic nerve, the level of endoneurial edema, and the number of storage inclusions in Schwann cells of older (>100 days old) Twitcher animals following myeloablative conditioning and BMT were similar to that observed in 40-day-old untreated Twitcher mice.^{27, 28} In addition, it took 4–9 months for remyelination of the sciatic nerve from Twitcher mice to occur following transplantation into congenic GALC-positive recipients.²⁹ Although there was decreased edema in the sciatic nerves of the combination-treated animals in the current study, these animals were indistinguishable from untreated Twitcher mice with respect to both the number of infiltrating macrophages and the decrease in myelinated axons. The persistent abnormal histology of the peripheral nerves is consistent with the poor performance of all the treatment groups on the wire hang test. This test is a direct measure of limb strength, which can be an indicator of muscle pathology or peripheral neuropathy.

Interestingly, there was increased enzyme activity in the brain, liver, and kidney of animals receiving the combination treatment compared to either therapy alone. Several phenomena likely contributed to this finding. First, some of the increased activity was probably due to the influx of bone marrow-derived GALC-positive donor cells into these tissues. However, the increased activity in the combination-treated animals was greater than the sum of the activity from the individual therapies. Second, a portion of the increased activity in the liver and kidney was probably due to leakage of the vector into the vasculature during the intracranial injection procedure. Finally, it is possible that the conditioning radiation or the injection procedure itself contributed to the increased activity. Tissue damage caused by the radiation or injection could increase the number of infiltrating GALC-positive donor cells. In addition, it is known that a variety of genotoxic agents, including ionizing radiation, can effectively increase expression in vivo following administration of a recombinant AAV vector.^{30, 31} We are currently investigating the effects of sublethal radiation on AAV-mediated expression in vivo.

We have shown that a combination of CNS-directed AAV2/5-mediated gene therapy and BMT performed using a myeloreductive-conditioning regimen is more efficacious in the murine model of GLD than either therapy alone. It seems likely that carefully directed placement of gene transfer vectors in the CNS and PNS, high levels of expression in autologous hematopoietic cells and early initiation of therapy will provide increased efficacy for the human disease.

Materials and Methods

Animal procedures. Heterozygous (twi/+) Twitcher mice are on a congenic C57Bl/6 background and maintained by MSS at Washington University School of Medicine. Homozygous (twi/twi) Twitcher mice and normal (+/+) controls were obtained by strict brother–sister matings of heterozygous animals and the genotyping of newborn mice was determined by polymerase chain reaction specific for the Twitcher mutation.³² Mice surviving up to age 21 days were enrolled in this study. All animal procedures were carried out in accordance with the regulations established by the IACUC at Washington University School of Medicine. This study consisted of three experimental treatment groups: (1) animals receiving intracranial injections of AAV2/5 alone (AAV2/5-treated), (2) animals receiving myeloreductive BMT alone (BMT-treated), and (3) animals receiving a combination of AAV2/5 and BMT (combination-treated).

On postnatal day 3, the AAV2/5-treatment group received three injections per hemisphere, six injections total, of AAV2/5-expressing murine GALC. The injection sites are as follow: from the bregma, the anterior site was +1 mm ant/post, 2 mm medial/lateral and -1.5 mm dorsal/ventral; the middle site was +2 mm ant/post, 2 mm medial/lateral and -2.5 mm dorsal/ventral; the posterior injection site was +3.5 mm ant/post, 2 mm medial/lateral and -1.5 mm dorsal/ventral. The animals received 2.4 10⁹ particles of AAV2/5 at each injection site in a volume of 2 l. The combination-treated animals received the intracranial injections of AAV2/5 in the morning before irradiation and BMT later the same day.

In the BMT-treated and combination-treated groups, 3-day-old Twitcher mice were pretreated with a single dose of 400 rad of conditioning irradiation from a ¹³⁷Cs source. After 3–6 h, the irradiated animals received a single intravenous injection of 100 l of unfractionated bone marrow suspension containing 1 10⁶ nucleated cells via the superficial temporal vein.³³ Same-sex donors were used for the transplants. The donor animals were congenic C57Bl/6 animals that constitutively expressed GFP in most cell lineages from the CAGGS promoter. Control groups of Twitcher and normal animals received 400 rad of radiation to assess the effects of radiation alone. A group of normal mice received 400 rad of radiation and 1 10⁶ donor cells to determine the effects of radiation and BMT on performance. Finally, untreated Twitcher and normal control mice were included as controls in all of the studies.

Longevity and body weight. To determine the lifespan, each mouse was maintained humanely in accordance with the norms of ethical laboratory animal care with free access to food and water. The lifespan was measured by noting the date of death or killing. The synergistic effects of combining therapies was estimated using the following relationship: synergy quotient=(combination-treated-untreated)/(BMT-treated-untreated)+(AAV2/5-treated-untreated), where "BMT-treated" is the mean lifespan of Twitcher mice treated with BMT alone, "AAV2/5-treated" is the mean lifespan of Twitcher mice treated with AAV2/5 alone, "combination-treated" is the mean lifespan of Twitcher mice treated with both BMT and AAV2/5, and "untreated" is the lifespan of untreated Twitcher mice. A number greater than one indicates synergy, one indicates simple additive effects, and a number less than one indicates a negative effect of combining the therapies.

Body weight was determined every 7 days starting at birth until the animals died or were in a moribund state.

Fluorescence-activated cell sorting analysis. GFP-positive hematopoietic donor cells were used in these studies so that engraftment levels could accurately be determined. At 40 days of age, bone marrow cells were pelleted and resuspended in a Tris-ammonium chloride buffer to lyse the red blood cells. Samples were labeled with phycoerythrin-conjugated antibodies against Mac-1 and Gr-1 (Pharmingen, San Jose, CA) to ensure multilineage engraftment.

Recombinant AAV2/5 vector. The AAV2/5 vector used in this study was from the same stock as described previously⁷ and consisted of the cytomegalovirus enhancer, chicken--actin promoter, the murine GALC complementary DNA and the rabbit--globin polyadenylation signal. The AAV2/5 vector stocks were produced by the Vector Core Facility, University of Nantes, France.

GALC activity. Tissues from treated and control mice were removed immediately after CO₂ euthanasia and homogenized in deionized water according to previously published methods.⁷ The protein concentration was measured with a commercially available coomasie dye-binding assay (Bio-Rad, Hercules, CA) using bovine serum albumin as standard. The activity of GALC was determined with [³H]galactosylceramide substrate according to previously published methods.³⁴ Specific activity is reported as nanomole substrate cleaved/h/mg protein.

Histochemical localization of GALC activity. The histochemical staining method for in situ localization of GALC activity was modified according to a previously described method.¹³ Briefly, 16-m-thick brain cryosections were equilibrated for 15 min in citrate/phosphate buffer (pH. 4.2, citrate/phosphate buffer), then with taurodeoxycholic/oleic acids in citrate/phosphate buffer for another 15 min. Samples were incubated for 2 h in X-Gal staining solution containing taurodeoxycholic/oleic acids, and 5 mM potassium ferrocyanide/ferricyanide, in citrate/phosphate buffer at 37°C. The stained sections were rinsed in phosphate-buffered saline and counter stained with 0.1% nuclear fast red.

Immunohistochemistry. Adjacent sagittal sections of brains from control and experimental animals were stained for GFAP (Immunostar, Hudson, WI) and CD11b (BD Pharmigen, San Diego, CA). Briefly, 16-m-thick cryosections were washed in Tris-buffered saline (TBS), incubated in 1% H₂O₂, and rinsed thoroughly in TBS. The tissue was blocked for 1 h in 10% normal goat serum (Sigma, St Louis, MO), and 0.25% Triton X-100 in TBS. Sections were incubated overnight at 4°C in primary rabbit anti-mouse GFAP (1:200) antisera or rat anti-mouse CD11b (1:200) in 5% normal goat serum, 0.2% Triton X-100 in TBS. The following day the sections were rinsed in TBS and incubated with biotinylated goat anti-rabbit (1:200, Vector Laboratories, Burlingame, CA) and mouse adsorbed biotinylated goat anti-rat (1:200, Vector Laboratories) antibodies in 10% normal goat serum, 0.1% Triton X-100 in TBS for 75 min. The tissue was rinsed in TBS and then incubated in peroxidase-conjugated avidin–biotin complex (1:200, Vectastain Elite ABC, Vector Laboratories) for 1 h at room temperature. Antibody immunoreactivity was visualized with 3'-3' diaminobenzidine (Sigma) and H₂O₂ in TBS. Sections were counterstained with nuclear fast red, dehydrated, and coverslipped.

Neuropathology. Brains from treated and control mice were removed and bisected along the mid-sagittal sinus. The right hemispheres were immersion fixed in 10% neutral-buffered formalin and processed for paraffin bedding. Briefly, 10-m-thick coronal sections of the cerebrum, brainstem, and cerebellum were stained with hematoxylin and eosin as well as Luxol fast blue/PAS using standard procedures and examined under light microscopy.

Behavioral phenotyping. Experimental and control mice were challenged repeatedly in several neurobehavioral tests to evaluate muscle strength, motor, balance, coordination, and locomotor functions. Animals were randomly labeled and tested every 5 days starting at 25 days of age. The tester was blind to both genotype and treatment. The experimental groups including 15 BMT-treated, 15 AAV2/5-treated, and 17 combination-treatment Twitcher mice were analyzed in these neurobehavioral tests from 25 to 120 days of age. The same tests were applied to the control groups including 12 untreated normal, 13 BMT-treated normal, 11 combination-treated normal, and 14 untreated control Twitcher mice.

Rotarod tests were performed essentially as described previously.³⁵ In the constant speed rotarod test, the maximum latency of the mouse to fall from a constant rotating rod (2.5 r.p.m.) in three trials was recorded (maximum time=60 s). In the accelerating rotarod test, the mouse was placed on a rotating rod that accelerated from 2.5 to 8.5 r.p.m., over a 1 min period. The maximum latency of three trials was recorded (maximum time=60 s). The mice were acclimated to these tests for four consecutive days from age 21 to 24 days before actual testing. The mice were trained on the constant and accelerating rotarod for a maximum of 60 s and immediately replaced on the rotarod for five times/trial/day. The wire hang test was performed as described previously.³⁵ The ability of the mice to hang on an inverted cage lid, 50 cm above a soft bedding surface, was recorded (maximum time=60 s).

Statistical analyses. All data are expressed as mean values1 SEM Comparison of lifespan between groups was performed using Kaplan–Meier analysis. For behavioral comparisons, between group differences at each age were determined by one-way analysis of variance with post hoc Bonferroni method. Differences were considered significant at P<0.05.

References

1. Suzuki, K and Suzuki, Y (1970). Globoid cell leucodystrophy (Krabbe's disease): deficiency of galactocerebroside beta-galactosidase. Proc Natl Acad Sci USA 66: 302–309. | Article | PubMed | ChemPort |
2. Wenger, DA, Rafi, MA and Luzi, P (1997). Molecular genetics of Krabbe disease (globoid cell leukodystrophy): diagnostic and clinical implications. Hum Mutat 10: 268–279. | Article | PubMed | ChemPort |
3. Kobayashi, T, Yamanaka, T, Jacobs, JM, Teixeira, F and Suzuki, K (1980). The Twitcher mouse: an enzymatically authentic model of human globoid cell leukodystrophy (Krabbe disease). Brain Res 202: 479–483. | Article | PubMed | ISI | ChemPort |
4. Suzuki, K and Taniike, M (1995). Murine model of genetic demyelinating disease: the twitcher mouse. Microsc Res Tech 32: 204–214. | Article | PubMed | ChemPort |
5. Shen, JS, Watabe, K, Ohashi, T and Eto, Y (2001). Intraventricular administration of recombinant adenovirus to neonatal twitcher mouse leads to clinicopathological improvements. Gene Ther 8: 1081–1087. | Article | PubMed | ChemPort |
6. 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 |
7. Lin, D et al. (2005). AAV2/5 vector expressing galactocerebrosidase ameliorates CNS disease in the murine model of globoid-cell leukodystrophy more efficiently than AAV2. Mol Ther 12: 422–430. | Article | PubMed | ChemPort |
8. Yeager, AM, Brennan, S, Tiffany, C, Moser, HW and Santos, GW (1984). Prolonged survival and remyelination after hematopoietic cell transplantation in the twitcher mouse. Science 225: 1052–1054. | Article | PubMed | ISI | ChemPort |
9. Hoogerbrugge, PM, Poorthuis, BJ, Romme, AE, van de Kamp, JJ, Wagemaker, G and van Bekkum, DW (1988). Effect of bone marrow transplantation on enzyme levels and clinical course in the neurologically affected twitcher mouse. J Clin Invest 81: 1790–1794. | PubMed | ChemPort |
10. Suzuki, K, Hoogerbrugge, PM, Poorthuis, BJ, Bekkum, DW and Suzuki, K (1988). The twitcher mouse. Central nervous system pathology after bone marrow transplantation. Lab Invest 58: 302–309. | PubMed | ChemPort |
11. Krivit, W et al. (1998). Hematopoietic stem-cell transplantation in globoid-cell leukodystrophy. N Engl J Med 338: 1119–1126. | Article | PubMed | ChemPort |
12. Escolar, ML et al. (2005). Transplantation of umbilical-cord blood in babies with infantile Krabbe's disease. N Engl J Med 352: 2069–2081. | Article | PubMed | ISI | ChemPort |
13. Dolcetta, D et al. (2004). Analysis of galactocerebrosidase activity in the mouse brain by a new histological staining method. J Neurosci Res 77: 462–464. | Article | PubMed | ChemPort |
14. Sands, MS et al. (1993). Treatment of murine mucopolysaccharidosis type VII by syngeneic bone marrow transplantation in neonates. Lab Invest 68: 676–686. | PubMed | ChemPort |
15. Olmstead, CE (1987). Neurological and neurobehavioral development of the mutant "twitcher" mouse. Behav Brain Res 25: 143–153. | Article | PubMed | ChemPort |
16. Suzuki, K and Suzuki, K (1985). Genetic galactosylceramidase deficiency (globoid cell leukodystrophy, Krabbe disease) in different mammalian species. Neurochem Pathol 3: 53–68. | PubMed | ChemPort |
17. Sands, MS et al. (1997). Murine mucopolysaccharidosis type VII: long term therapeutic effects of enzyme replacement and enzyme replacement followed by bone marrow transplantation. J Clin Invest 99: 1596–1605. | PubMed | ISI | ChemPort |
18. Jeyakumar, M et al. (2001). Enhanced survival in Sandhoff disease mice receiving a combination of substrate deprivation therapy and bone marrow transplantation. Blood 97: 327–329. | Article | PubMed | ISI | ChemPort |
19. Wada, R, Tifft, CJ and Proia, RL (2000). Microglial activation precedes acute neurodegeneration in Sandhoff disease and is suppressed by bone marrow transplantation. Proc Natl Acad Sci USA 97: 10954–10959. | Article | PubMed | ChemPort |
20. Ohmi, K, Greenberg, DS, Rajavel, KS, Ryazantsev, S, Li, HH and Neufeld, EF (2003). Activated microglia in cortex of mouse models of mucopolysaccharidoses I and IIIB. Proc Natl Acad Sci USA 100: 1902–1907. | Article | PubMed | ChemPort |
21. Tsuji, D et al. (2005). Specific induction of macrophage inflammatory protein 1-alpha in glial cells of Sandhoff disease model mice associated with accumulation of N-acetylhexosaminyl glycoconjugates. J Neurochem 92: 1497–1507. | Article | PubMed | ChemPort |
22. Higashi, Y, Komiyama, A and Suzuki, K (1992). The twitcher mouse: immunocytochemical study of Ia expression in macrophages. J Neuropathol Exp Neurol 51: 47–57. | PubMed | ChemPort |
23. Ohno, M, Komiyama, A, Martin, PM and Suzuki, K (1993). Proliferation of microglia/macrophages in the demyelinating CNS and PNS of twitcher mouse. Brain Res 602: 268–274. | Article | PubMed | ChemPort |
24. Matsushima, GK et al. (1994). Absence of MHC class II molecules reduces CNS demyelination, microglial/macrophage infiltration, and twitching in murine globoid cell leukodystrophy. Cell 78: 645–656. | Article | PubMed | ChemPort |
25. Wu, YP, McMahon, EJ, Matsuda, J, Suzuki, K, Matsushima, GK and Suzuki, K (2001). Expression of immune-related molecules is downregulated in twitcher mice following bone marrow transplantation. J Neuropathol Exp Neurol 60: 1062–1074. | PubMed | ChemPort |
26. von Boehmer, H (2005). Mechanisms of suppression by suppressor T cells. Nat Immunol 6: 338–344. | Article | PubMed | ISI | ChemPort |
27. Ichioka, T, Kishimoto, Y, Brennan, S, Santos, GW and Yeager, AM (1987). Hematopoietic cell transplantation in murine globoid cell leukodystrophy (the twitcher mouse): effects on levels of galactosylceramidase, psychosine, and galactocerebrosides. Proc Natl Acad Sci USA 84: 4259–4263. | Article | PubMed | ChemPort |
28. Kondo, A, Hoogerbrugge, PM, Suzuki, K, Poorthuis, BJ, Van Bekkum, DW and Suzuki, K (1988). Pathology of the peripheral nerve in the twitcher mouse following bone marrow transplantation. Brain Res 460: 178–183. | Article | PubMed | ChemPort |
29. Scaravilli, F and Jacobs, JM (1981). Peripheral nerve grafts in hereditary leukodystrophic mutant mice (twitcher). Nature 290: 56–58. | Article | PubMed | ISI | ChemPort |
30. Yalkinoglu, AO, Heilbronn, R, Burkle, A, Schlehofer, JR and zur Hausen, H (1988). DNA amplification of adeno-associated virus as a response to cellular genotoxic stress. Cancer Res 48: 3123–3129. | PubMed | ISI | ChemPort |
31. Alexander, IE, Russell, DW and Miller, AD (1994). DNA-damaging agents greatly increase the transduction of nondividing cells by adeno-associated virus vectors. J Virol 68: 8282–8287. | PubMed | ISI | ChemPort |
32. Sakai, N et al. (1996). Molecular cloning and expression of cDNA for murine galactocerebrosidase and mutation analysis of the twitcher mouse, a model of Krabbe's disease. J Neurochem 66: 1118–1124. | PubMed | ChemPort |
33. Sands, MS and Barker, JE (1999). Percutaneous intravenous injection in neonatal mice. Lab Anim Sci 49: 328–330. | PubMed | ChemPort |
34. Wenger, D (1991). Screening for lysosomal disorders In: Hommes FA (ed). Techniques in Diagnostics Human Biochemical Genetics. Wiley-Liss: New York, pp 587–617.
35. Rogers, DC, Fisher, EM, Brown, SD, Peters, J, Hunter, AJ and Martin, JE (1997). Behavioral and functional analysis of mouse phenotype: SHIRPA a proposed protocol for comprehensive phenotype assessment. Mamm Genome 8: 711–713. | Article | PubMed | ISI | ChemPort |

Acknowledgements

This work was supported in part by National Institutes of Health Grants DK 57586(MSS) and NS43205 (MSS) and Grants MMH 9471, 9514, and NSC B195013 (DSL).