doi:10.1006/mthe.2002.0650, available online at http://www.idealibrary.

com on IDEAL
ARTICLE

Demonstration of Feasibility of In Vivo Gene Therapy

for Gaucher Disease Using a Chemically Induced
Mouse Model
John Marshall,1,* Kerry Anne McEachern,1 Julie A. Cavanagh Kyros,1 Jennifer B. Nietupski,1
Tracey L. Budzinski,1 Robin J. Ziegler,1 Nelson S. Yew,1 Jennifer Sullivan,1 Abraham Scaria,1
Nico van Rooijen,2 John A. Barranger,3 and Seng H. Cheng1
1
Genzyme Corporation, Framingham, Massachusetts 01701-9322, USA
2
Vrije Universiteit, 1081 BT Amsterdam, The Netherlands
3
University of Pittsburgh, Pennsylvania 15261, USA

*
To whom correspondence and reprint requests should be addressed. Fax: (508) 872-4091. E-mail: john.marshall@genzyme.com.

Progress towards developing gene therapy for Gaucher disease has been hindered by the lack
of an animal model. Here we describe a mouse model of Gaucher disease which has a chemically
induced deficiency of glucocerebrosidase and that accumulates elevated levels of glucosylce-
ramide (GL-1) in the lysosomes of Kupffer cells. Administration of mannose-terminated gluco-
cerebrosidase (Cerezyme) resulted in the reduction of GL-1 levels in the livers of these animals.
Gene transduction of hepatocytes with a plasmid DNA vector encoding human glucocerebrosi-
dase (pGZB-GC) generated high-level expression and secretion of the enzyme into systemic cir-
culation with consequent normalization of Kupffer cell GL-1 levels. This suggested that the de
novo synthesized and unmodified enzyme produced by hepatocyte transduction was also capa-
ble of being delivered to the cells that are primarily affected in Gaucher disease.
Immunolocalization studies also revealed that preferential transduction and expression of human
glucocerebrosidase in the Kupffer cells with subsequent reduction in the GL-1 levels could be
attained with a low dose of a recombinant adenoviral vector encoding the human enzyme
(Ad2/CMV-GC). This observation raises the possibility of gene therapy for Gaucher disease that
involves directly transducing the affected histiocytes using recombinant adenoviral vectors.
Together, these data demonstrate the potential for use of in vivo gene therapy vectors for treat-
ing Gaucher disease.

Key Words: Gaucher disease, glucocerebrosidase, glucosylceramide, gene therapy,

recombinant adenoviral vectors, plasmid vectors

INTRODUCTION characterized by neurological disease and by an earlier

Gaucher disease is the most prevalent of the more than
40 lysosomal storage disorders that have been described
thus far. It is caused by a deficiency of the lysosomal
hydrolase, glucocerebrosidase, resulting in the accumu-
lation of the glycolipid glucosylceramide (GL-1) in tissue
macrophages [1–3]. Delineation of the three types of
Gaucher disease is based on the severity and age of
disease onset. Type 1, which is the most prevalent, is the
nonneuropathic form of the disease and is characterized
by the accumulation of GL-1 primarily in Kupffer cells,
osteoclasts, and splenic macrophages. The abnormal
storage of GL-1 leads to hepatosplenomegaly, pancy-
topenia, lesions in the bone and, in some cases, pul-
monary involvement. Types 2 (acute) and 3 (chronic) are
age of onset and significantly more severe systemic clin-
ical symptoms [1].
The involvement of macrophages and the presence of
a mannose receptor on the surface of these cells provided
the basis for enzyme replacement therapy for Gaucher dis-
ease using glucocerebrosidase containing ␣-mannosyl-ter-
minated oligosaccharides [4]. Replacement therapy using
such a modified enzyme (Cerezyme) is effective for Type
1 patients but less so for Type 3 and not Type 2 patients
because of the inability of the recombinant enzyme to tra-
verse the blood-brain barrier. Clinical therapy is highly
effective but response is heterogenous, with some organs
being less responsive to treatment. Consequently, alter-
native and adjunctive therapies are also being evaluated,

MOLECULAR THERAPY Vol. 6, No. 2, August 2002 179

Copyright © The American Society of Gene Therapy
1525-0016/02 $35.00
 ARTICLE doi:10.1006/mthe.2002.0650, available online at http://www.idealibrary.com on IDEAL
A P Ι 0.0001
B D
C cessful [17–19]. The introduced mutations were all lethal
such as using bisphosphonates to treat osteopenic bone
disease [5], glucosylceramide synthase inhibitors to abate
the accumulation of GL-1 [6,7], and cell and gene [8–13]
therapies to augment the levels of glucocerebrosidase. Here
we report our efforts to evaluate the feasibility of gene
therapy for Gaucher disease.
The characteristics of the pathobiology of Gaucher
disease make it a logical candidate for gene therapy
strategies that are directed at correcting the enzyme defi-
ciency in macrophages. Indeed, the majority of the early
gene therapy efforts for Gaucher disease were focused on
retroviral-mediated transduction of bone marrow stem
cells or peripheral blood monocytes [10–13]. Preclinical
studies in mice showed that high-efficiency gene trans-
fer and stable engraftment of gene-modified bone mar-
row stem cells could be achieved [10,14]. However, sub-
sequent clinical studies using autologous stem cells from
Gaucher patients showed that the efficiency of retrovi-
ral transduction of these cells was markedly lower, and
that engraftment of the modified stem cells in the
absence of ablation was transient and occurred at a low
180
FIG. 1. Localization of GL-1 in the induced Gaucher mouse liver (A). BALB/c
mice were treated with CBE (delivered intraperitoneally) and GL-1-containing
liposomes (delivered intravenously) and 24 hours later, administered with
either clodronate liposomes or PBS. After an additional 24 hours, the livers were
harvested, and the levels of GL-1 in the liver were determined. Treatment with
clodronate liposomes results in depletion of the Kupffer cells. The data shown
are for individual tissue GL-1 values (E) and the mean (▲) ± SEM (n = 6).
Unpaired t-test P values are indicated for the corresponding groups. (B, C, D)
Localization of a fluorescent analog of GL-1. BALB/c mice were injected intra-
venously with GL-1-containing liposomes containing 1% of the fluorescent ana-
log, GL-1B (BODIPY FL C12-glucocerebroside). After 24 hours the livers were
harvested and processed for histological staining. (B) The 5 ␮m sections were
incubated with a rat anti-mouse F4/80 (macrophage marker) antibody and
detected with an AlexaFluor488 goat anti-rat IgG secondary antibody. (C) GL-
1B fluorescence was visualized using a rhodamine filter. The image in (D) is
an overlay of images in (B) and (C), showing co-localization of GL-1B and
macrophages. Arrows depict the co-localization of GL-1B in two separate
macrophages.
rate [15]. These observations suggested that further
improvements in vector design and in the conditions for
transducing human stem cells were necessary.
Evaluation of new therapeutic approaches for Gaucher
disease has been hampered by the lack of a suitable ani-
mal model. Previous attempts to generate either transgenic
knockout mice [16] or mice with targeted mutations akin
to those that cause disease in humans have been unsuc-
within hours of birth because of epidermal abnormalities.
These have been characterized as defective development
of the stratum corneum, which results in a compromised
permeability barrier that leads to death by dehydration
[20]. An alternative is an artificially induced model in
which glucocerebrosidase is specifically and irreversibly
inhibited by conduritol B epoxide (CBE), and the glycol-
ipid levels in the Kupffer cells increased by administering
GL-1 [21–27]. Mice treated with CBE and GL-1-containing
liposomes lack glucocerebrosidase activity [24] and display
elevated levels of the glycolipid, primarily in Kupffer cells
[25]. As such, these animals display one of the major
pathological lesions observed in Gaucher disease.
To evaluate the efficacy of alternative gene therapy
approaches for Gaucher disease, we validated the utility
of the artificially induced murine model. Using this
model, we asked if direct gene transduction of the liver
could act as a “depot” of recombinant glucocerebrosi-
dase and if the secreted enzyme cleared GL-1 from
Kupffer cells. Unlike Cerezyme, gene therapy-derived glu-
cocerebrosidase will not be further modified to expose
the mannose residues on the oligosaccharide side chains.
It was, therefore, important to determine whether or not
the gene therapy-derived enzyme was delivered to the
macrophages. Finally, since a proportion of recombinant
adenoviral vectors delivered systemically is internalized
by macrophages [28], we investigated whether or not
clearance of GL-1 levels could be realized using a low
dose of the viral vector.
MOLECULAR THERAPY Vol. 6, No. 2, August 2002
Copyright © The American Society of Gene Therapy
doi:10.1006/mthe.2002.0650, available online at http://www.idealibrary.com on IDEAL
P = 0.05
Cerezyme
RESULTS
Characterization of a Chemically Induced Murine
Model of Gaucher Disease
An evaluation of the feasibility of gene therapy for
Gaucher disease would benefit from the availability of an
animal model. Because mice whose gene for glucocere-
brosidase has been disrupted by insertional mutagenesis
are nonviable [16], we elected to adapt an induced
Gaucher mouse model that was first described by Kanfer
et al. [21]. In this model, the animals are treated with CBE
to deplete endogenous enzyme activity and liposomes con-
taining GL-1 to augment the accumulation of the gly-
cosphingolipid in the lysosomes of Kupffer cells. Following
treatment with CBE and GL-1-containing liposomes, the
levels of glucocerebrosidase in the livers were reduced by
greater than 90% and the levels of GL-1 were increased
5- to 10-fold over those observed in normal mice (~ 5 ␮g
GL/g tissue) (Fig. 1A). These elevated levels of GL-1 in the
induced mouse liver were sustained for up to 5 days post-
treatment (data not shown).
As Gaucher disease manifests in the liver as a result of
accumulation of GL-1 primarily in the Kupffer cells, we
asked whether or not the elevated levels of GL-1 in the arti-
ficially-induced model of Gaucher disease were due to
accumulation of the glycosphingolipid in this cell type. To
test this, mice that had been administered CBE and GL-1-
containing liposomes were treated 24 hours later with clo-
dronate-encapsulated liposomes to transiently deplete the
liver of macrophages [29,30]. Treatment with clodronate
liposomes reduced the elevated levels of GL-1 to basal lev-
els (Fig. 1A), indicating that the majority of the liver accu-
mulation of GL-1 in the induced Gaucher mouse was in
the Kupffer cells. This result was confirmed using a fluo-
rescent analog of GL-1 (BODIPYFL-tagged C12-glucocere-
broside). Liposomes containing this fluorescently labeled
MOLECULAR THERAPY Vol. 6, No. 2, August 2002
Copyright © The American Society of Gene Therapy
ARTICLE
FIG. 2. Effect of adding exogenous recombinant glucocere-
brosidase on GL-1 levels in the induced Gaucher mouse liver.
BALB/c mice were treated with CBE (delivered intraperi-
toneally) and GL-1-containing liposomes (delivered intra-
venously) and 24 hours later administered either 40 U/kg
Cerezyme or PBS. After an additional 24 hours the livers were
harvested and the levels of GL-1 determined. The data shown
are for individual tissue GL-1 values (O) and the mean (▲) ±
SEM (n = 5). Unpaired t-test P values are indicated for the cor-
responding groups.
GL-1 (GL-1B) were injected intravenously into
CBE-treated mice and the livers were harvested
24 hours later. Histological sections (5 ␮m) of
the livers were probed with two rat anti-mouse
macrophage antibodies (MoMa-1 and F4/80)
followed by a goat anti-rat AlexaFluor350 sec-
ondary antibody. Sections were viewed by flu-
orescence microscopy using Ex503/Em512 for
the GL-1B and Ex346/Em442 for the
AlexaFluor350. All of the GL-1B fluorescence
coincided with the AlexaFluor350 fluorescence, indicat-
ing that the GL-1B was localized in the Kupffer cells (Fig.
1B). Together, these results demonstrate that treatment
with CBE and GL-1-containing liposomes leads to elevated
levels of GL-1 in the Kupffer cells of mice for up to 5 days.
Histiocyte storage of GL-1 is one of the hallmarks of
Gaucher disease, and these animals represent a working
model to evaluate the feasibility of gene therapy for treat-
ing this disease.
To confirm that the GL-1 delivered to the Kupffer cells
of the induced Gaucher mouse model was located within
the lysosomal compartment, the mice were injected intra-
venously with 40 U/kg Cerezyme. Cerezyme is a recombi-
nant form of glucocerebrosidase in which the oligosac-
charide side chains have been enzymatically modified to
expose terminal mannose residues. Modification of the
enzyme in this manner facilitates preferential targeting of
the enzyme to the mannose receptors on the macrophages
[4]. Since glucocerebrosidase is only active at a pH less
than 6, a reduction in GL-1 levels in the liver of the
induced Gaucher mouse would indicate that the glycol-
ipids were located in the lysosomes. Treatment with
Cerezyme resulted in a reduction of the GL-1 levels in the
livers of the induced Gaucher mice compared to those of
naïve animals (Fig. 2). For the GL-1 to be hydrolyzed by
Cerezyme, most of the glycolipid delivered to the Kupffer
cells of the induced Gaucher mouse must consequently
have been localized within the acidic lysosomal compart-
ment.
Cellular Localization of Modified Cerezyme and
Unmodified Recombinant Human Glucocerebrosidase
in the Liver
Enzyme replacement therapy using Cerezyme has been
shown to be efficacious in the treatment of Gaucher Type
1 patients [31]. This suggests that gene transduction of an
181
 ARTICLE doi:10.1006/mthe.2002.0650, available online at http://www.idealibrary.com on IDEAL
A B
D E
G H
appropriate depot organ with subsequent secretion of glu-
cocerebrosidase could also be effective. However, as indi-
cated earlier, Cerezyme is a modified form of glucocere-
brosidase in which the oligosaccharides have been altered
to improve delivery to the Kupffer cells. Since glucocere-
brosidase produced following gene therapy is unlikely to
be similarly processed, it was important to determine
whether unmodified enzyme also localized to the Kupffer
cells.
To address this question, Cerezyme and unmodified
recombinant glucocerebrosidase (purified from CHO cells)
were localized in rat livers following intravenous infusion.
Animals were administered 80 U/kg Cerezyme and the liv-
ers were harvested 1 hour later and processed for histo-
logical analysis. The liver sections were probed with anti-
bodies specific for human glucocerebrosidase,
macrophages, and endothelial cells. Figures 3A and 3D are
representative liver sections stained with antibodies spe-
cific for human glucocerebrosidase showing abundant
staining of the enzyme in these tissue sections. Based on
182
FIG. 3. Localization of systemically deliv-
C ered recombinant human glucocerebrosi-
dase in the liver. Fischer 344 rats were
administered intravenously with 80 U/kg
Cerezyme (A–F) or with 80 U/kg unmodi-
fied glucocerebrosidase (G–I). The livers
were harvested 1 hour later and processed
for histological staining. Liver sections were
incubated with two anti-human gluco-
cerebrosidase monoclonal antibodies (1B5
and 8E4) and detected with an
AlexaFluor568 goat anti-mouse IgG1 sec-
ondary antibody (A, D, G). The same sec-
tions were also incubated with either a
mouse anti-rat CD11b (macrophage
F marker) antibody (B, H), or a sheep anti-rat
von Willebrand factor (endothelial cell
marker) antibody (E). These were detected
using an Alexafluor488 goat anti-mouse
IgG2a and an AlexaFluor488 donkey anti-
sheep secondary antibody, respectively.
Images in (C), (F), and (I) are overlays of the
corresponding images from (A), (D), and
(G) with (B), (E) and (H), respectively.
Arrows indicate examples of cells that show
human glucocerebrosidase localized in
either the macrophages (C, I) or endothe-
lial cells (F).
I
cell morphology, very little of the enzyme appeared to be
localized in the hepatocytes. The same liver sections were
costained with either CD11b, a macrophage-specific anti-
body (Fig. 3B), or with an anti-rat von Willebrand factor
antibody, an endothelial cell marker (Fig. 3E). Figure 3C
is an overlay of the images from Figs. 3A and 3B showing
co-localization of Cerezyme in a number of the
macrophages (shown as yellow cells), and Figure 3F is an
overlay of Figs. 3D and 3E showing that the majority of
the endothelial cells had internalized Cerezyme. These
data indicate that, following intravenous infusion of
Cerezyme, most of the enzyme is internalized by endothe-
lial cells, some by macrophages, and very little by hepa-
tocytes. This biodistribution of Cerezyme in the different
liver cell types is concordant with published data [32].
This experiment was repeated using unmodified recom-
binant human glucocerebrosidase that had been purified
from CHO cells. To determine if unmodified enzyme could
be internalized by macrophages, liver sections of rats treated
with 80 U/kg unmodified recombinant glucocerebrosidase
MOLECULAR THERAPY Vol. 6, No. 2, August 2002
Copyright © The American Society of Gene Therapy
doi:10.1006/mthe.2002.0650, available online at http://www.idealibrary.com on IDEAL
ARTICLE

Ad2/CMV-GC
and supports the feasibility of gene therapy for treating
Gaucher disease.

Efficacy of Adenoviral-Mediated Gene Therapy of

FIG. 4. Tissue expression levels of glucocerebrosidase following systemic deliv-
ery of the recombinant adenoviral vector Ad2/CMV-GC. BALB/c mice received
an intravenous injection of 5 ⫻ 1012 particles/kg of Ad2/CMV-GC. Serum,
liver, spleen, and lungs were analyzed for glucocerebrosidase activity 2 days
post-injection. Levels of the enzyme in naïve BALB/c mice tissues were also
assayed. The dashed line indicates the steady state concentration of Cerezyme
attained in the serum of Gaucher patients during a therapeutic infusion. Data
shown are expressed as mean ± SEM (n = 8).
and processed 1 hour later were stained for the enzyme (Fig.
3G) and for macrophages (Fig. 3H). The staining pattern for
the unmodified glucocerebrosidase (Fig. 3G) was similar to
that for Cerezyme (Figs. 3A and 3D), except that some of
the hepatocytes also stained positive. Figure 3I is an over-
lay of 3G and 3H showing the localization of unmodified
glucocerebrosidase in some of the macrophages (shown as
yellow cells). This result indicates that unmodified gluco-
cerebrosidase retains the ability to localize to Kupffer cells
Gaucher Disease
To extend these observations to gene therapy-derived
human glucocerebrosidase, a recombinant adenoviral vec-
tor encoding human glucocerebrosidase (Ad2/CMV-GC)
was generated. Intravenous administration of 5 ⫻ 1012 par-
ticles/kg of Ad2/CMV-GC into BALB/c mice resulted in
high-level expression of glucocerebrosidase in the liver
(100-fold endogenous levels), spleen (10-fold endogenous
levels), and lungs (10-fold endogenous levels) as well as
secretion of significant levels into the serum (Fig. 4). The
level of glucocerebrosidase attained in the serum was deter-
mined to be approximately 10-fold higher than the serum
steady state levels of Cerezyme in Gaucher patients dur-
ing a therapeutic infusion (Fig. 4). This suggests that gene
transduction using a high dose of Ad2/CMV-GC can result
in the production of therapeutic levels of glucocerebrosi-
dase in the serum as well as supraphysiological levels of
enzyme in various Gaucher-affected tissues.
The distribution of human glucocerebrosidase among
the different liver cell types following treatment with
Ad2/CMV-GC was also examined. Rats were administered
either 8 ⫻ 1012 (Figs. 5A–5C) or 2 ⫻ 1012 (Figs. 5D–5F) par-
ticles/kg of Ad2/CMV-GC, and the livers were examined 2
days later. Immunofluorescence staining revealed that the
majority of the human glucocerebrosidase in the livers of
rats that received the higher dose of the recombinant ade-
novirus vector was in the hepatocytes (Fig. 5A).

A B C

FIG. 5. Localization of recombinant human

glucocerebrosidase in the liver following sys-
temic delivery of Ad2/CMV-GC. Fischer 344
rats were administered intravenously with
either 8 ⫻ 1012 particles/kg (A–C) or 2 ⫻
1012 particles/kg (D–F) Ad2/CMV-GC. The
livers were processed for histological stain-
ing2 days post-injection. Liver sections were
incubated with two anti-human glucocere-
brosidase monoclonal antibodies (1B5 and D E F
8E4) and detected with an Alexafluor568
goat anti-mouse IgG1 secondary antibody
(A, D). The same sections were also incu-
bated with a mouse anti-rat CD11b
(macrophage marker) antibody (B, E), and
detected using AlexaFluor488 goat anti-
mouse IgG2a. The image in (C) is an overlay
of the images in (A) and (B), and the image
in (F) is an overlay of the images in (D) and
(E). Arrows indicate examples of cells show-
ing both human glucocerebrosidase and
macrophage staining.

MOLECULAR THERAPY Vol. 6, No. 2, August 2002 183

Copyright © The American Society of Gene Therapy
 ARTICLE doi:10.1006/mthe.2002.0650, available online at http://www.idealibrary.com on IDEAL

in GL-1 was most likely a consequence of viral

P = 0.001
transduction of the affected Kupffer cells.
Hence, administration of a low dose of a
recombinant adenoviral vector encoding glu-
cocerebrosidase may be efficacious in reduc-
ing the lysosomal storage abnormality in the
liver.

Ability of Kupffer Cells to Internalize

Recombinant Glucocerebrosidase Secreted
from Genetically Modified Hepatocytes
Because recombinant adenoviral vectors
directly transduced the Kupffer cells, we were
unable to surmise whether glucocerebrosidase
secreted from the transduced hepatocytes was
able to cross-correct GL-1 accumulation in the
Ad2/CMV-GC Kupffer cells. To address this question, a plas-
mid DNA vector encoding human glucocere-
brosidase (pGZB-sGC) was constructed and
FIG. 6. Effect of intravenous administration of Ad2/CMV-GC on liver GL-1 levels in the delivered hydrodynamically into the livers of
induced Gaucher mouse. BALB/c mice were treated with CBE (delivered intraperitoneally) and mice. This procedure results in the transduc-
GL-1-containing liposomes (delivered intravenously) and 24 hours later administered either 5 tion of 5–10% of hepatocytes and less than
⫻ 106, 1.5 ⫻ 106, or 0.5 ⫻ 1012 particles/kg Ad2/CMV-GC. The animals were killed, and the lev-
els of GL-1 in their livers were assayed 2 days later. The data shown are for individual tissue GL-
0.1% of Kupffer cells (data not shown)
1 values (E) and the mean (F) ± SEM (n = 6). Unpaired t-test P values are indicated for the cor- [35,36]. Hydrodynamic injection of 10 ␮g of
responding groups. pGZB-sGC into the liver of the induced
Gaucher mice resulted in an approximately
fourfold higher level of glucocerebrosidase in
Superimposing the image of this section with that stained the liver and an approximately 20-fold higher level in the
for macrophages (Fig. 5B) revealed co-localization (shown serum than that observed in untreated mice (Fig. 7A).
as yellow cells in Fig. 5C) in 10–20% of the Kupffer cells. Concomitant with these levels of expression was a reduc-
Rats treated with the lower dose of Ad2/CMV-GC (2 ⫻ 1012 tion in the GL-1 content in the livers of these animals to
particles/kg) showed little evidence of glucocerebrosidase basal levels (Fig. 7B). That the reduction in GL-1 was due
in the parenchymal cells (Fig. 5D). Instead, the majority to expression of glucocerebrosidase, rather than to the pro-
of the enzyme was localized in the Kupffer cells (Fig. 5F). cedure used to transfect the liver, was demonstrated by the
These observations are consistent with the notion that a finding that mice treated with an irrelevant plasmid
proportion of adenoviral vectors administered systemi- (pGZB-s␣Gal, a plasmid DNA vector encoding another
cally are sequestered by the reticuloendothelial cells, and lysosomal enzyme, human ␣-galactosidase A) did not
that administration of doses above this threshold results affect GL-1 levels. Since hydrodynamic-mediated trans-
in the additional transduction of hepatocytes [28,33,34]. duction resulted in expression of glucocerebrosidase pri-
This pattern of transgene expression suggests that gene marily in the hepatocytes, our finding of complete clear-
therapy using a relatively low dose of Ad2/CMV-GC may ance of GL-1 strongly suggests that the secreted enzyme
suffice to reverse the accumulation of GL-1 in histiocytes was susceptible to uptake by the neighboring and possi-
that are primarily affected in Gaucher disease. bly the distant Kupffer cells. Hence, gene delivery vectors
To determine if administration of low doses of recom- that effect genetic modification of depot organs with sub-
binant adenoviral vectors could be therapeutic, a dose- sequent secretion of unmodified glucocerebrosidase
response study was performed in the induced Gaucher should be efficacious in treating Gaucher disease.
mouse model. Mice treated with CBE and GL-1-contain-
ing liposomes were infused increasing amounts of Clearance of GL-1 from Kupffer Cells Following
Ad2/CMV-GC ranging between 0.5 and 5 ⫻ 1012 parti- Intranasal Instillation of Ad2/CMV-GC
cles/kg. Analysis of the livers of these mice 2 days later To further demonstrate that glucocerebrosidase secreted
showed that a low dose of Ad2/CMV-GC (0.5 ⫻ 1012 par- from gene-transduced cells is capable of being internalized
ticles/kg) was equally effective at clearing the GL-1 from by Kupffer cells, mice were instilled intranasally with 5 ⫻
the Kupffer cells as the higher doses (Fig. 6). Since mice 1011 particles/kg of Ad2/CMV-GC. It has been shown that
treated with 0.5 ⫻ 1012 particles/kg of Ad2/CMV-GC did pulmonary instillation of a recombinant adenoviral vector
not result in secretion of detectable levels of glucocere- encoding the lysosomal enzyme ␣-galactosidase A results
brosidase into the serum (data not shown), the reduction in transduction of the lung with subsequent secretion of

184 MOLECULAR THERAPY Vol. 6, No. 2, August 2002

Copyright © The American Society of Gene Therapy
doi:10.1006/mthe.2002.0650, available online at http://www.idealibrary.com on IDEAL
A FIG. 7.
pGZB-sGC
B macrophages results in the clearance of accumulated GL-
P = 0.001
P Ι 0.001
pGZB-sGC
the enzyme into systemic circulation [37]. Since no viral
vector was detected outside of the lung following this route
of delivery, the detection of enzyme in other visceral organs
indicated that enzyme was secreted from the lung and rein-
ternalized by these other tissues. Figure 8 shows that mice
administered intranasally with Ad2/CMV-GC can effect a
reduction of GL-1 in the Kupffer cells of the chemically
induced Gaucher mouse. This suggests that glucocere-
brosidase was similarly secreted from the transduced lung
into systemic circulation and confirms that the glucocere-
brosidase produced by gene transfer can be delivered effec-
tively to Kupffer cells.
DISCUSSION
The primary pathological manifestation of Gaucher disease
is the accumulation of GL-1 in the lysosomes of tissue
macrophages resulting in enlarged, so-called “Gaucher
cells” [1–3]. The resultant perturbation causes significant
enlargement of the liver and spleen, as well as other tis-
sue defects such as pulmonary complications and bone
lesions. The nonneuropathic form of Gaucher disease can
be treated successfully with enzyme replacement therapy
[31]. Therapy using recombinant glucocerebrosidase that
MOLECULAR THERAPY Vol. 6, No. 2, August 2002
Copyright © The American Society of Gene Therapy
ARTICLE
Plasmid DNA-mediated transduction of the livers of induced
Gaucher mice. (A) Tissue expression levels of glucocerebrosidase following
hydrodynamic delivery of the plasmid pGZB-sGC. BALB/c mice were treated
with CBE (delivered intraperitoneally) and GL-1-containing liposomes (deliv-
ered intravenously) and then 24 hours later, injected hydrodynamically with
10 ␮g of pGZB-sGC. The levels of glucocerebrosidase were assayed in the
serum and liver 2 days post-injection. The levels of glucocerebrosidase in naïve
mice and in mice injected with pGZB-␣gal (encoding human ␣-galactosidase
A) were also determined. Data shown are expressed as mean ± SEM (n = 8).
(B) Effect of hydrodynamic administration of pGZB-sGC on GL-1 levels in the
induced Gaucher mouse. Mice were treated as described above. The levels of
GL-1 in the livers were assayed 2 days later. The data shown are for individ-
ual tissue GL-1 values (O) and the mean (▲) ± SEM (n = 8). Unpaired t-test P
values are indicated for the corresponding groups.
has been modified to facilitate preferential targeting to
1 from the affected organs with consequent clinical
improvement. However, because the half-life of gluco-
cerebrosidase is relatively short, sustained clinical improve-
ment requires biweekly infusions of the recombinant
enzyme for life. Because gene therapy provides the possi-
bility for prolonged expression of the enzyme, we have
investigated the feasibility of this approach as an alterna-
tive modality for treating Gaucher disease.
Much of the early gene therapy effort for Gaucher dis-
ease was directed at evaluating ex vivo approaches [10–13].
Because the clinical experience using this approach has
been unsatisfactory [15], the focus of the current studies
was centered around the use of in vivo gene delivery vec-
tors. Gene therapy for Gaucher disease using in vivo gene
transfer vectors poses several technical challenges. Firstly,
glucocerebrosidase is not normally secreted unless the cells
express very high levels of the enzyme. As such, the gene
delivery vectors chosen need to be extremely efficient in
their ability to effect expression of the enzyme. Secondly,
unlike the majority of lysosomal enzymes, glucocere-
brosidase is targeted to the lysosomes by way of a man-
nose-6-phosphate-independent pathway. The exact mech-
anism by which the enzyme is delivered to the lysosomal
compartment is still unclear. Efficient targeting of recom-
binant glucocerebrosidase to the affected macrophages is
currently realized by modifying the purified enzyme to
expose mannose residues on the oligosaccharides as is the
case with Cerezyme. However, this additional processing
is untenable with glucocerebrosidase that is produced by
gene therapy. Therefore, it was important to determine if
de novo synthesized and secreted glucocerebrosidase
obtained following in vivo gene transduction could be
internalized by tissue macrophages. Finally, progress
toward evaluating novel therapeutic approaches for
Gaucher disease has been hindered by the absence of a
viable animal model in which to test efficacy.
In this report, we validated the utility of an induced
mouse model of Gaucher disease. The generation of such
a model with accumulation of GL-1 in the liver has been
described [21,25], though there are no reports of this
185
 ARTICLE doi:10.1006/mthe.2002.0650, available online at http://www.idealibrary.com on IDEAL

P = 0.001
GL-1 could be hydrolyzed readily following a
bolus intravenous injection of Cerezyme. This
enzyme is weakly active at neutral pH and only
functions optimally in the acidic environment
of a lysosome. We surmise that Cerezyme must
have accumulated within the lysosomes of the
tissue macrophages for treatment with it to
lower the liver GL-1 levels. This demonstration
also helped validate the induced mouse as a
suitable model of Gaucher disease because it
was successfully treated with the same thera-
peutic used for Gaucher patients.
Using the chemically induced Gaucher
mouse, studies were performed to examine the
feasibility of gene transfer for treating this dis-
ease. Gene transduction using either recombi-
Ad2/CMV-GC IV nant adenoviral or plasmid DNA vectors
Ad2/CMV-GC IN resulted in expression and secretion of thera-
peutic levels of glucocerebrosidase into the
FIG. 8. Effect of intranasal delivery of Ad2/CMV-GC on liver GL-1 levels in the induced Gaucher
mouse. BALB/c mice were treated with CBE (delivered intraperitoneally) and GL-1-containing
serum. Moreover, levels of the enzyme were
liposomes (delivered intravenously) and 24 hours later administered 5 ⫻ 1011 particles/kg also detected in the liver, spleen and lung, the
Ad2/CMV-GC either intravenously (i.v.) or intranasally (i.n.). The animals were killed and the organs that are primarily affected in Gaucher
levels of GL-1 in the livers were assayed 2 days later. The data shown are for individual tissue disease. Systemic delivery of 5⫻1012 particles/kg
GL-1 values (O) and the mean (▲) ± SEM (n = 6). Unpaired t-test P values are indicated for
of Ad2/CMV-GC generated levels of glucocere-
the corresponding groups.
brosidase in the serum that are predicted to be
therapeutic with consequent clearance of GL-1
from the livers of the induced Gaucher mouse.
model being used to evaluate therapeutics. Model valida- Interestingly, low doses of the recombinant adenoviral
tion was achieved by demonstrating that the elevated lev- vector were equally effective at reducing GL-1 levels, pre-
els of the GL-1 in the liver were primarily associated with sumably because the virus transduced the macrophages
Kupffer cells, localized within the lysosomal compartment, following uptake [28]. This observation suggests that a low
and metabolically accessible. That the GL-1 was located in dose of a recombinant adenoviral vector may be effica-
the liver macrophages was demonstrated by treating the cious at reducing the storage abnormality in the liver.
animals with clodronate-encapsulated liposomes [29] and However, because this dose did not result in secretion of
by showing a concomitant loss in liver GL-1 levels. This the enzyme, it is unlikely that it would alter the GL-1
conclusion was further verified by immunolocalization accumulation in other affected histiocytes that were not
studies showing that fluorescently tagged GL-1 was pri- transduced by the adenoviral vector. Thus, use of low doses
marily located in the Kupffer cells. These observations are of adenoviral vectors may have limited utility.
congruent with those reported earlier showing that GL-1 Because the primary affected cells in Gaucher disease
delivered as an emulsion or liposomal formulation results are the tissue macrophages, a key question was whether
primarily in liver deposition, with approximately equal or not gene transfer-derived glucocerebrosidase containing
amounts going to hepatocytes and Kupffer cells [25–27]. unmodified oligosaccharide side chains could be inter-
However, the GL-1 in the hepatocytes is lost within 24 nalized by these cells. Systemic administration of recom-
hours through biliary excretion and the remaining levels binant adenoviral vectors resulted in the delivery of a pro-
of liver GL-1, which are relatively stable for several days, portion of the virus to the macrophages; thus, we were
are associated primarily with the Kupffer cells [26,38]. It unable to address this issue using this viral vector. We,
should be noted that the GL-1 used in these studies was therefore, employed pressure-mediated delivery of plas-
derived from plants. However, this material contains a mid DNA encoding glucocerebrosidase to the liver. The
C-16 ceramide, which is one of the more common forms results demonstrated that unmodified glucocerebrosidase
of GL-1 derived from the visceral organs of Gaucher generated by hydrodynamic-mediated delivery of pGZB-
patients. sGC reduced the levels of GL-1 in the Kupffer cells. Because
We also demonstrated that the GL-1 in the induced hydrodynamic-mediated delivery of plasmid DNA results
Gaucher mouse was appropriately stored within the lyso- primarily in transduction of hepatocytes [35,36], the com-
somes of the Kupffer cells and that it was accessible for plete clearance of GL-1 strongly suggested that low levels
enzymatic hydrolysis. The subcellular localization was of the enzyme secreted from the hepatocytes were taken
determined indirectly by demonstrating that the stored up by the macrophages. That glucocerebrosidase produced

186 MOLECULAR THERAPY Vol. 6, No. 2, August 2002

Copyright © The American Society of Gene Therapy
doi:10.1006/mthe.2002.0650, available online at http://www.idealibrary.com on IDEAL
by gene transfer can be targeted to Kupffer cells was veri-
fied by the observation that Ad2/CMV-GC delivered
intranasally into the lung also resulted in a reduction of
GL-1 levels in the Kupffer cells. In this instance, the
enzyme was produced in the lung and secreted into sys-
temic circulation. The notion that the unmodified gluco-
cerebrosidase generated by gene transfer could be inter-
nalized by macrophages in vivo was further supported by
studies using purified recombinant enzyme showing sim-
ilar uptake, albeit at only 10–20% the efficiency of man-
nose-terminated enzyme [39–41]. However, it should be
noted that unmodified glucocerebrosidase is reportedly
less labile in serum and more efficiently delivered to the
lysosomal compartment, and exhibits a longer half-life
than the oligosaccharide-remodeled enzyme. Thus,
although less unmodified enzyme may be delivered to
macrophages, it acts for a longer time. The observation
that gene transfer-derived glucocerebrosidase could be
taken up by macrophages is consistent with previous
reports showing that enzyme derived from gene-modified
myoblasts contain high mannose-type oligosaccharide
chains [42]. Together, these observations indicate that
gene therapy strategies that result in the transduction of
depot organs with a resultant secretion of glucocerebrosi-
dase should be efficacious in the treatment of Gaucher
disease.
We have validated the utility of a chemically induced
mouse model of Gaucher disease and showed that it is a
valuable surrogate for the evaluation of therapeutic
approaches. Moreover, gene transduction using recombi-
nant adenoviral and plasmid DNA vectors can result in the
expression and secretion of therapeutic levels of gluco-
cerebrosidase. Importantly, we demonstrate that the
secreted enzyme is in a form that is compatible with
uptake by macrophages, the cell type that is primarily
affected in Gaucher disease. We have demonstrated the
potential of in vivo gene transfer approaches for treating
this disease.
MATERIALS AND METHODS
Formulation of GL-1-containing emulsions and liposomes. To generate
an emulsion of glucosylceramide (Matreya, Inc., Pleasant Gap, PA), the
glycolipid was formulated with the detergent Myrj-52 (Sigma, St. Louis,
MO) at a ratio of 4:3 (w/w) as described [25]. Briefly, GL-1 was first evap-
orated from chloroform and a saline solution (145 mM sodium chloride
solution) of Myrj-52 was then added to achieve a final glycolipid concen-
tration of 20 mg/ml. The mixture was sonicated using a probe sonicator
(VirSonic 100, Virtis Inc., Gardiner, NY) until an even dispersion was
achieved. To generate the liposomal formulation, glucosylceramide (GL-1)
was formulated with phosphatidylcholine (PC) and phosphatidylserine
(PS) at a ratio of 1:8:1 (w/w/w). Chloroform solutions of the appropriate
ratios of PC, PS and GL-1 were first mixed and then evaporated to dryness
on a rotary evaporator. The dried lipid film was then hydrated in saline
(GL-1 final concentration 5mg/ml) using a bath sonicator until an even sus-
pension was achieved. The GL-1-containing liposomes and emulsions were
stored at 4⬚C until ready for use. Liposomal formulations containing a flu-
orescent analog of GL-1, BODIPY FL C12-glucocerebroside (GL-1B; Molecular
Probes, Eugene, OR) were prepared by the same methodology. The amount
MOLECULAR THERAPY Vol. 6, No. 2, August 2002
Copyright © The American Society of Gene Therapy
ARTICLE
of GL-1B used in the formulations was approximately 1% (by mass) of the
total GL-1.
Generating an artificially induced Gaucher mouse model. All procedures
involving animals were reviewed and approved by an Institutional Animal
Care and Use Committee (IACUC) following Association for Assessment of
Accreditation of Laboratory Animal Care (AAALAC), State and Federal
guidelines. To generate the induced Gaucher mouse model, female BALB/c
mice (6–8 weeks old; Taconic, Germantown, NY) were first injected
intraperitoneally with 100 mg/kg CBE (Sigma, St. Louis, MO) as described
[22]. The mice were challenged either intraperitoneally with an emulsion
of GL-1 (GL-1:Myrj-52 emulsion; 500 mg GL-1/kg body weight) or intra-
venously into the tail vein with a liposomal formulation of GL-1 (PC:PS:GL-
1 liposomes; 50 mg GL-1/kg body weight) 1 hour later. It has been shown
that injection of CBE results in the complete inhibition of glucocerebrosi-
dase activity in all tissues within 1 hour and that the inhibitor is cleared
rapidly (in approximately 5 hours) from the animals through hydrolysis
[21,22,24]. As such, CBE should not have any effect on the exogenously
administered glucocerebrosidase or on the enzyme produced by gene trans-
fer.
Transient depletion of Kupffer cells. Transient depletion of tissue
macrophages was achieved by intravenous administration of clodronate
(gift from Roche Diagnostics) encapsulated in liposomes (clodronate lipo-
somes) as described [29,30]. Liposomes containing clodronate (5 mg/ml)
were injected in a total volume of 200 ␮l. Liver sections taken 24 hours
after treatment with clodronate liposomes and analyzed by immunohis-
tochemical staining confirmed that greater than 90% of the macrophages
in the liver had been eliminated [29].
Recombinant glucocerebrosidase. Recombinant human glucocerebrosi-
dase was purified from the media of Chinese hamster ovary cells that had
been stably transfected with an expression vector encoding the human
enzyme. We used two forms of the enzyme: unmodified glucocerebrosidase,
and Cerezyme (Genzyme Corp, Cambridge, MA), which is glucocerebrosi-
dase modified by exoglycosidases to generate mannose-terminated oligosac-
charide side chains. This modification imparts more efficient targeting of
the enzyme to Kupffer cells and other tissue macrophages by way of man-
nose receptors on the cell surface [4]. The purified enzymes (40 U/kg in 5%
mannitol) were administered into the tail veins of mice as a 100 ␮l bolus
injections, or to rats (80 U/kg) as 500 ␮l bolus injections.
Gene delivery vectors. The recombinant adenoviral vector, Ad2/CMV-GC
encoding human glucocerebrosidase was constructed essentially as
described [43]. Briefly, the E1 region of adenovirus serotype 2 was replaced
with an expression cassette consisting of the human cytomegalovirus
(CMV) immediate early promoter and enhancer, the human glucocere-
brosidase cDNA and a simian virus 40 small t polyadenylation signal
sequence. For intravenous delivery, the recombinant adenoviral vectors
were administered as a 100 ␮l bolus injection into the tail vein in an excip-
ient of 5% sucrose in phosphate-buffered saline (PBS). For pulmonary deliv-
ery, a 100 ␮l volume of the viral vector was instilled intranasally through
the nares by inspiration.
The plasmid DNA vector, pGZB-sGC encoding a synthetic human glu-
cocerebrosidase was generated as described [44]. The vector was modified
to contain a minimum number of CpG motifs to improve its toxicity pro-
file. pGZB-sGC contains a minimal origin of replication, a kanamycin resist-
ance gene and an expression cassette composed of the CMV promoter and
enhancer, a hybrid intron, the human glucocerebrosidase cDNA (that had
been codon optimized for the most abundant mammalian tRNAs), and a
bovine growth hormone polyadenylation signal sequence. The plasmid
DNA vector, pGZB-saGal is identical to pGZB-sGC except that the gluco-
cerebrosidase cDNA was replaced with that for human ␣-galactosidase A.
Hydrodynamic delivery of the plasmid DNA vectors was performed using
the TransIT in vivo gene delivery system (Mirus Corp., Madison, WI) as out-
lined by the manufacturer and essentially as previously described [35,36].
A complex was formed between 10 ␮g of plasmid DNA and TransIT poly-
mer which was then rapidly delivered (6–8 seconds) into the tail vein of a
mouse. The volume administered (in ml) was equal to 1/10th the weight of
the mouse (in g).
187
 ARTICLE doi:10.1006/mthe.2002.0650, available online at http://www.idealibrary.com on IDEAL
Measurement of glucocerebrosidase activity. Tissues were homogenized
in 150 mM citrate-phosphate buffer, pH 5.4, containing sodium tauro-
cholate (0.25%, w/w) and Triton X-100 (0.25%, w/w) using a Tissue Tearor
(BioSpec Products Inc., Bartlesville, OK). The homogenates were clarified
by centrifugation for 15 minutes at 10000g and then assayed for gluco-
cerebrosidase activity. The enzymatic activity of glucocerebrosidase was
determined using the synthetic substrate 4-methylumbelliferyl-glucopyra-
noside (Sigma, St. Louis, MO) at 10 mM in homogenization buffer con-
taining 1% bovine serum albumin at 37⬚C. The reaction was terminated
by adding 0.5 volume of 1M glycine buffer, pH 12.5, and the cleaved 4-
methylumbelliferone measured using a fluorimeter (Ex365/Em445).
Quantitation of tissue glucosylceramide levels. Tissue GL-1 levels were
quantified using a modification of the procedures described [45–47]. Briefly,
total lipid from 100 mg portions of the test tissue was extracted by homog-
enizing in 2 ml of a 2:1 (v/v) chloroform:methanol mixture using a
TissueTearor (BioSpec Products Inc., Bartlesville, OK). Following a 15-
minute incubation at 37⬚C, tissue debris was pelleted by centrifugation for
5 minutes at 1000g. The supernatant was removed and saved (sample 1)
and the pellet was re-extracted with 2 ml of the 2:1 chloroform:methanol
mix. This was again incubated for 15 minutes at 37⬚C, after which the
sample was re-centrifuged, and the supernatant was removed and combined
with sample 1. The combined supernatants were extracted with 0.22 vol-
umes of 0.9% saline by vortexing for 1 minute. The phases were separated
by centrifugation for 5 minutes at 1000g, which resulted in a lower organic
phase (sample 2) and an upper aqueous phase. The aqueous phase was
removed and re-extracted with 4.5 volumes of the 2:1 chloroform:methanol
mixture. Following centrifugation, the aqueous phase was carefully
removed and the organic phase was combined with sample 2. The com-
bined organic phases were re-extracted with 0.22 volumes of 0.9% saline,
after which the final organic phase was dried down to a film. The film was
solubilized in 1 ml chloroform and a 200 ␮l aliquot passed over a LiChrolut
RP-18 solid-phase extraction column (Merck, Darmstadt, Germany). After
washing the column with 2 ml of chloroform, neutral glycolipids were
eluted with 1 ml of acetone:methanol (9:1, v/v). The eluted samples were
dried and solubilized in 200 ␮l of chloroform, 50 ␮l of which was then spot-
ted onto an HPTLC plate (#13728 ; Merck, Darmstadt, Germany) along with
known GL-1 standards (Matreya, Inc., Pleasant Gap, PA). The lipids were
chromatographed in a tank equilibrated with chloroform:methanol:water
(70:30:10, v/v) for approximately 25 minutes. The plates were dried,
sprayed with 3% (w/v) cupric acetate monohydrate and 15% (v/v) phos-
phoric acid, and baked for 20–30 minutes at 180⬚C to develop the sepa-
rated lipid bands. The baked TLC plates were scanned as high-resolution
TIFF images, and the bands were quantified using ScanAnalysis (Biosoft,
Ferguson, MO). Statistical significance was determined using an unpaired
t-test (Statview; SAS Inst. Inc., Cary, NC).
Histology and immunofluorescence staining. Female Fischer 344/NTac
rats (6–8 weeks old) were obtained from Taconic (Germantown, NY).
Following treatment with either the recombinant enzymes or gene delivery
vectors, the rats were anesthetized with an intraperitoneal injection of ket-
amine/xylamine. A whole body perfusion was performed by way of the left
ventricle of the heart using a butterfly needle. PBS at 4⬚C containing 10 U/ml
heparin was perfused at a rate of 82 ml/minute until the lungs were
blanched. The inferior vena cava was cut and the flow of PBS/heparin con-
tinued until the color of the liver changed from dark red to a light brown.
The rat was then perfused with approximately 200 ml/kg fixative (2% (v/v)
formaldehyde, 0.0156% (v/v) glutaraldehyde in PBS) at 4⬚C. Following per-
fusion, the liver was excised, covered with fixative solution and incubated
at room temperature for 1 hour. It was then immersed in a 30% sucrose solu-
tion at 4⬚C for 24 hours before being embedded in Tissue-Tek OCT com-
pound (Sakura Inc., Torrance, CA) and microtomed into 5 ␮m thick sections.
All antibody incubations were performed in block solution (PBS con-
taining 3% normal goat serum). To localize human glucocerebrosidase,
liver sections were first incubated for 1 hour with an equal mixture of 1B5
and 8E4 [48], two human specific anti-glucocerebrosidase monoclonal anti-
bodies. Both of these antibodies are of the IgG1␬ isotype. To visualize the
enzyme, the sections were treated with an AlexaFluor568-labeled goat anti-
mouse IgG1 secondary antibody. The liver sections were simultaneously
probed with either a mouse anti-rat CD11b (IgG2a; macrophage marker)
188
antibody (BD Biosciences, San Diego, CA) or a sheep anti-rat von
Willebrand factor (endothelial cell marker) antibody (Accurate Chemical
Co., Westbury, NY). Following three 5-minute washes with 1X Cadenza
buffer (Shandon, Pittsburgh, PA), the slides were incubated for 1 hour with
the appropriate secondary antibodies. These were either a goat anti-mouse
IgG2a AlexaFluor488, or a donkey anti-sheep IgG AlexaFluor488 (Molecular
Probes). Sections were viewed using an Olympus BX60 fluorescent micro-
scope, and images were captured using Magnafire v1.0 (Lead Technologies)
with an Optronics digital imager.
ACKNOWLEDGMENTS
We thank Ronald Scheule (Genzyme Corporation) for his insights and review of
the manuscript and the Vector Production, AC&TS, and Histology groups for
their technical assistance.
RECEIVED FOR PUBLICATION MARCH 18, 2002; ACCEPTED MAY 22, 2002.
REFERENCES
1. Beutler, E., and Grabowski, G. A. (2001). Gaucher disease. In The Metabolic and Molecular
Basis of Inherited Disease (C. R. Scriver, A. L. Beaudet, D. Valle, and W. S. Sly, Eds.), pp.
3635-3668. McGraw Hill, New York.
2. Morales, L. E. (1996). Gauchers disease: a Review. Ann. Pharmacother. 30: 381–388.
3. Grabowski, G. A., and Horowitz, M. (1997). Gauchers disease: molecular, genetic and
enzymological aspects. Baill. Clin. Hæmatol. 10: 635–656.
4. Furbish, F. S., Steer, C. J., Krett, N. L., and Barranger, J. A. (1981). Uptake and distribu-
tion of placental glucocerebrosidase in rat hepatic cells and effects of sequential degly-
cosylation. Biochim. Biophys. Acta 673: 425–434.
5. Ciana, G., Cuttini, M., and Bembi, B. (1997). Short-term effects of pamidronate in
patients with Gaucher’s disease and severe skeletal involvement. New Engl. J. Med. 337:
712.
6. Lee, L., Abe, A., and Shayman, J. A. (1999). Improved inhibitors of glucosylceramide syn-
thase. J. Biol. Chem. 274: 14662–14669.
7. Cox, T., et al. (2000). Novel oral treatment of Gauchers disease with N-butyldeoxyno-
jirimycin (OGT 918) to decrease substrate biosynthesis. Lancet 355: 1481–1485.
8. Liu, C., Dunigan, J. T., Watkins, S. C., Bahnson, A. B., and Barranger, J. A. (1998). Long-
term expression, systemic delivery, and macrophage uptake of recombinant human glu-
cocerebrosidase in mice transplanted with genetically modified primary myoblasts. Hum.
Gene Ther. 9: 2375–2384.
9. Ross, C. J. D., Bastedo, L., Maier, S. A., Sands, M. S., and Chang, P. L. (2000). Treatment
of a lysosomal storage disease, mucopolysaccharidosis VII, with microencapsulated
recombinant cells. Hum. Gene Ther. 11: 2117–2127.
10. Correll, P. H., Colilla, S., Dave, H. P. G., and Karlsson, S. (1992). High levels of human
glucocerebrosidase activity in macrophages of long-term reconstituted mice after retro-
viral infection of hematopoietic stem cells. Blood 80: 331–336.
11. Xu, L. C., et al. (1995). Long-term in vivo expression of the human glucocerebrosidase
gene in nonhuman primates after CD34+ hematopoietic cell transduction with cell-free
retroviral vector preparations. Proc. Natl. Acad. Sci. USA 92: 4372–4376.
12. Dunbar, C. E., et al. (1998). Retroviral transfer of the glucocerebrosidase gene into
CD34+ cells from patients with Gaucher disease: In vivo detection of transduced cells
without myeloablation. Hum. Gene Ther. 9: 2629–2640.
13. Havenga, M. J. E., Valerio, D., Hoogerbrugge, P., and Van Es, H. H. G. (1999). In vivo
methotrexate selection of murine hemopoietic cells transduced with a retroviral vector
for Gaucher disease. Gene Ther. 6: 1661–1669.
14. Ohashi, T., et al. (1992). Efficient transfer and sustained high expression of the human
glucocerebrosidase gene in mice and their functional macrophages following trans-
plantation of bone marrow transduced by a retroviral vector. Proc. Natl. Acad. Sci. USA
89: 11332–11326.
15. Novelli, E.M., and Barranger, J.A. (2001). Gene therapy for lysosomal storage disorders.
Expert Opin. Biol. Ther. 1: 857–867.
16. Tybulewicz, V. L. J., et al. (1992). Animal model of Gauchers disease from targeted dis-
ruption of the mouse glucocerebrosidase Gene. Nature 357: 407–410.
17. Bornstein, P., et al. (1995). Metaxin, a gene contiguous to both thrombospondin 3 and
glucocerebrosidase, is required for embryonic development in the mouse: Implications
for Gaucher disease. Proc. Natl. Acad. Sci. USA 92: 4547–4551.
18. Liu, Y., et al. (1998). Mice with type 2 and 3 Gaucher disease point mutations gener-
ated by a single insertion mutagenesis procedure (SIMP). Proc. Natl. Acad. Sci. USA 95:
2503–2508.
19. Sidransky, E., and Ginns, E. I. (1997). Gauchers disease: the best laid schemes of mice
and men. Baill. Clin. Hæmatol. 10: 725–737.
20. Holleran, W. M., et al. (1994). Consequences of ␤-glucocerebrosidase deficiency in epi-
dermis: ultrastructure and permeability barrier alteration in Gaucher disease. J. Clin.
Invest. 91: 1736–1764.
21. Kanfer, J. N., Legler, G., Sullivan, L., Raghavan, S., and Mumford, R. A. (1975). The
Gaucher mouse. Biochem. Biophys. Res. Comm. 67: 85–90.
MOLECULAR THERAPY Vol. 6, No. 2, August 2002
Copyright © The American Society of Gene Therapy
doi:10.1006/mthe.2002.0650, available online at http://www.idealibrary.com on IDEAL
22. Stephens, M. C., Bernatsky, A., Burachinsky, V., Legler, G., and Kanfer, J. N. (1978). The
Gaucher mouse: differential action of conduritol B epoxide and reversibility of its effects.
J. Neurochem. 30: 1023–1027.
23. Legler, G. (1990). Glycoside hydrolases: mechanistic information from studies with
reversible and irreversible inhibitors. Adv. Carbohydr. Chem. Biochem. 48: 319–384.
24. Hara, A., and Radin, N. S. (1979). Destruction and resynthesis of mouse ␤-glucosidases.
Biochim. Biophys. Acta 582: 412–422.
25. Datta, S. C., and Radin, N. S. (1988). Normalization of liver glucosylceramide levels in
the Gaucher mouse by phosphatidylserine injection. Biochem. Biophys. Res. Comm. 152:
155–160.
26. Pentchev, P. G., et al. (1981). Biliary excretion of glycolipid in induced or inherited glu-
cosylceramide lipidosis. Biochim. Biophys. Acta 665: 615–618.
27. Ellens, H., Morselt, H., and Scherpof, G. (1981). In vivo fate of large unilamellar sphin-
gomyelin-cholesterol liposomes after intraperitoneal and intravenous injection into rats.
Biochim. Biophys. Acta 674: 10–18.
28. Worgall, S., Wolff, G., Falck-Pedersen, E., Crystal, R.G. (1997). Innate immune mecha-
nisms dominate elimination of adenoviral vectors following in vivo administration. Hum.
Gene Ther. 8: 37–44.
29. van Rooijen, N. and Sanders, A. (1994). Liposome mediated depletion of macrophages:
mechanism of action, preparation of liposomes and applications. J. Immunol. Methods
174: 83–93.
30. van Rooijen, N., Sanders, A., and van der Berg, T. K. (1996). Apoptosis of macrophages
induced by liposome-mediated intracellular delivery of clodronate and propamidine. J.
Immunol. Methods 193: 93–99.
31. Barton, N.W., Furbish, F.S., Murray, G.J., Garfield, M., and Brady, R.O. (1990).
Therapeutic response to intravenous infusions of glucocerebrosidase in a patient with
Gaucher disease. Proc. Natl. Acad. Sci USA 87: 1913–1990.
32. Xu, L. C., et al. (1996). Turnover and distribution of intravenously administered man-
nose-terminated human acid ␤-glucosidase in murine and human tissues. Pediatr. Res.
39: 313–322.
33. Kuzmin, A. I., Finegold, M. J., and Eisensmith, R. C. (1997). Macrophage depletion
increases the safety, efficacy and persistence of adenovirus-mediated gene transfer in vivo.
Gene Ther. 4: 309–316.
34. Wolff, G., et al. (1997). Enhancement of in vivo adenovirus-mediated gene transfer and
expression by prior depletion of tissue macrophages in the target organ. J. Virol. 71:
624–629.
MOLECULAR THERAPY Vol. 6, No. 2, August 2002
Copyright © The American Society of Gene Therapy
ARTICLE
35. Zhang, G., Budker, V., and Wolff, J. A. (1999). High levels of foreign gene expression in
hepatocytes after tail vein injections of naked plasmid DNA. Hum. Gene Ther. 10:
1735–1737.
36. Liu, F., Song, Y. K., and Liu, D. (1999). Hydrodynamics-based transfection in animals by
systemic administration of plasmid DNA. Gene Ther. 6: 1258–1266.
37. Li, C., et al. (2002). Adenovirus-transduced lung as a portal for delivering ␣-galactosi-
dase A into systemic circulation for Fabry disease. Mol. Ther. 5: 745–754.
38. Tokoro, T., Gal, A. E., Galo, L. L., and Brady, R. O. (1987). Studies of the pathogenesis
of Gauchers disease: tissue distribution and biliary excretion of [14C]L-glucosylceramide
in rats. J. Lipid Res. 28: 968–972.
39. Brady, R. O., Pentchev, P. G., Gal, A. E., Hibbert, S. R., and Dekaban, A. S. (1974).
Replacement therapy for inherited enzyme deficiency. Use of purified glucocerebrosi-
dase in Gauchers disease. New Eng. J. Med. 291: 989–993.
40. Morrone, S., Pentchev, P. G., Baynes, J., and Thorpe, S. (1981). Studies in vivo of the
tissue uptake, cellular distribution and catabolic turnover of exogenous glucocerebrosi-
dase in rat. Biochem. J. 194: 733–742.
41. Doebber, T. W., et al. (1982). Enhanced macrophage uptake of synthetically glycosy-
lated human placental ␤-glucocerebrosidase. J. Biol. Chem. 257: 2193–2199.
42. Liu, C., Duniga, J. T., Watkins, S. C., Bahnson, A. B., and Barranger, J. A. (1998). Long-
term expression, systemic delivery, and macrophage uptake of recombinant human glu-
cocerebrosidase in mice transplanted with genetically modified primary myoblasts. Hum.
Gene Ther. 9: 2375–2384.
43. Armentano, D., et al. (1997). Effect of the E4 region on the persistence of transgene
expression from adenovirus vectors. J. Virol. 71: 2408–2416.
44. Yew, N. S., et al. (2000). Reduced inflammatory response to plasmid DNA vectors by
elimination and inhibition of immunostimulatory CpG motifs. Mol. Ther. 1: 255–262,
doi:10.1006/mthe.2002.0036.
45. Miller, S. P. F., Zirzow, G. C., Doppelt, S. H., Brady, R. O., and Barton, N. W. (1996).
Analysis of the lipids of normal and Gaucher bone marrow. J. Lab. Clin. Med. 127:
353–358.
46. Doering, T., et al. (1999). Sphingolipid activator proteins are required for epidermal per-
meability barrier formation. J. Biol. Chem. 274: 11038–11045.
47. Abe, A., et al. (2000). Glycosphingolipid depletion in Fabry disease lymphoblasts with
potent inhibitors of glucosylceramide synthase. Kidney Int. 57: 446–454.
48. Aerts, J. M. F. G., et al. (1990). Comparative study on glucocerebrosidase in spleens from
patients with Gaucher disease. Biochem. J. 269: 93–100.
189