CNS Penetration of Intrathecal-Lumbar Idursulfase in the
Monkey, Dog and Mouse: Implications for Neurological
Outcomes of Lysosomal Storage Disorder
Pericles Calias1*., Mikhail Papisov2,3,4., Jing Pan1, Nancy Savioli1, Vasily Belov2,3,4, Yan Huang1, Jason
Lotterhand1, Mary Alessandrini1, Nan Liu1, Alan J. Fischman3,4, Jan L. Powell1, Michael W. Heartlein1
1Shire Human Genetic Therapies, Inc., Lexington, Massachusetts, United States of America, 2Department of Radiology, Massachusetts General Hospital, Boston,
Massachusetts, United States of America, 3Department of Radiology, Harvard Medical School, Cambridge, Massachusetts, United States of America, 4Department of
Nuclear Medicine, Shriners Burns Hospital, Boston, Massachusetts, United States of America
A major challenge for the treatment of many central nervous system (CNS) disorders is the lack of convenient and effective
methods for delivering biological agents to the brain. Mucopolysaccharidosis II (Hunter syndrome) is a rare inherited
lysosomal storage disorder resulting from a deficiency of iduronate-2-sulfatase (I2S). I2S is a large, highly glycosylated
enzyme. Intravenous administration is not likely to be an effective therapy for disease-related neurological outcomes that
require enzyme access to the brain cells, in particular neurons and oligodendrocytes. We demonstrate that
intracerebroventricular and lumbar intrathecal administration of recombinant I2S in dogs and nonhuman primates
resulted in widespread enzyme distribution in the brain parenchyma, including remarkable deposition in the lysosomes of
both neurons and oligodendrocytes. Lumbar intrathecal administration also resulted in enzyme delivery to the spinal cord,
whereas little enzyme was detected there after intraventricular administration. Mucopolysaccharidosis II model is available
in mice. Lumbar administration of recombinant I2S to enzyme deficient animals reduced the storage of glycosaminoglycans
in both superficial and deep brain tissues, with concurrent morphological improvements. The observed patterns of enzyme
transport from cerebrospinal fluid to the CNS tissues and the resultant biological activity (a) warrant further investigation of
intrathecal delivery of I2S via lumbar catheter as an experimental treatment for the neurological symptoms of Hunter
syndrome and (b) may have broader implications for CNS treatment with biopharmaceuticals.
Citation: Calias P, Papisov M, Pan J, Savioli N, Belov V, et al. (2012) CNS Penetration of Intrathecal-Lumbar Idursulfase in the Monkey, Dog and Mouse:
Implications for Neurological Outcomes of Lysosomal Storage Disorder. PLoS ONE 7(1): e30341. doi:10.1371/journal.pone.0030341
Editor: Maria A. Deli, Biological Research Center of the Hungarian Academy of Sciences, Hungary
Received August 3, 2011; Accepted December 14, 2011; Published January 18, 2012
Copyright: ? 2012 Calias et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This study was sponsored by Shire Human Genetic Therapies, Inc. The funders participated in the study design and conduct, data collection and
analysis, decision to publish and preparation of the manuscript.
Competing Interests: The authors have read the journal’s policy and have the following conflicts: PC, JP, YH, MA, NL, NS, JL, JLP and MWH are full-time
employees of Shire Human Genetic Therapies, Inc. MP, VB and AJF have no competing interests to declare. This does not alter the authors9 adherence to all the
PLoS ONE policies on sharing data and materials.
* E-mail: firstname.lastname@example.org
. These authors contributed equally to this work.
The brain is protected by the blood-brain barrier (BBB) , the
blood-cerebrospinal fluid (CSF) barrier  and the avascular
arachnoid epithelium . Together, these barriers provide
physical, transport and metabolic regulation by restricting the
entry of macromolecules and polar solutes from the blood to the
brain and spinal cord . Most pharmacological agents are unable
to penetrate the brain in sufficient amounts to have therapeutic
benefits, with only highly lipophilic, small molecules (,500 Da)
usually able to cross the BBB . Thus, a major challenge for
treatment of central nervous system (CNS) disorders, many of
which are debilitating and life threatening, is the lack of
convenient and effective methods for delivery of therapeutic
agents to widespread regions of the brain.
Various noninvasive brain-targeting methods using endogenous
molecular transport mechanismshavebeen explored as drug delivery
strategies. These have included fusion proteins that target delivery by
transcytosis using insulin , , or transferrin , , receptors.
Encapsulation technologies, such as pegylated immunoliposomes,
have been used to deliver plasmids to the brain through the BBB
utilizing monoclonal antibody ligands as the targeting agent , .
More recent efforts have explored the use of nanoparticles (polymers,
emulsions or suspensions) to traverse the BBB through various
endocytotic pathways . While promising from a mechanistic
point of view, few of these innovative strategies have progressed
beyond preclinical evaluation, leaving direct administration into the
CSF or braintissue as the only clinically viable method for delivery of
therapeutics to the brain and spinal cord.
The two main routes of direct delivery to the CNS are intrathecal
(IT) drug administration and direct intracerebroventricular (ICV)
injection. IT drug administration is an established route for
treatment of disorders such as chronic pain due to cancer or other
conditions , , , and spasticity , . Several IT drug
delivery devices are marketed for these applications, with the
benefits and inherent complications generally well understood .
ICV administration has also been used therapeutically, most
notably for treatment of Parkinson’s disease , , delivery of
PLoS ONE | www.plosone.org1 January 2012 | Volume 7 | Issue 1 | e30341
opioids for pain , and chemotherapy in children .
However, this administration route has not achieved the same
level of use in clinical practice compared to IT drug delivery
devices. While the disadvantages of ICV administration (invasive-
ness, need for specialized neurosurgical skills) are readily apparent,
it is not clear whether benefits are to be expected when
administering a drug, especially a large protein, into the ICV
space versus the IT space at the midthoracic region. One possible
advantage is that drugs delivered in the midthoracic region might
be absorbed into the bloodstream or degraded locally before the
CSF flow delivers them to the brain tissues whereas ICV-delivered
drugs might have a better penetration rate into the brain
parenchyma. Thus, it becomes a clinical imperative to compare
the drug delivery and distribution patterns after IT versus ICV
administration, in order to offer patients with devastating CNS
diseases requiring protein therapy at the level of the brain a
treatment modality with an optimal risk/benefit ratio.
As early as the 1960s, investigators reported that large
molecular weight molecules may be distributed throughout the
brain via the CSF following IT-lumbar delivery. Rieselbach et al
 imaged the movement of radioactive colloidal gold (Au198)
delivered into the lumbar sac of both monkeys and humans,
showing widespread cerebral subarachnoid and ventricular system
distribution of the particles primarily from CSF bulk flow.
However, the distribution and tissue deposition of large molecules
are complex, with variable CSF flow in different brain regions
influencing the diffusion of molecules into surrounding tissues .
More recent investigations suggest that the mechanisms by which
large molecules may be transferred from the CSF to brain
parenchyma involve active transport mechanisms, including
receptor mediated uptake, axonal transport and translocation
along the Virchow-Robin channels , , .
Mucopolysaccharidosis (MPS) II (Hunter syndrome) is a rare
inherited lysosomal storage disorder (LSD) primarily affecting
children. The disease is caused by a deficiency of iduronate-2-
sulfatase (I2S), with an estimated incidence of 1 per 162,000 live
births . Recombinant human I2S (idursulfase, ElapraseH; Shire
Human Genetic Therapies, Inc., Cambridge, MA) is approved for
treatment of certain somatic symptoms of Hunter syndrome but
there is no pharmacological therapy for treatment of the
neurological manifestations, which can include delayed develop-
glycosylated enzyme (approximate molecular weight: 76 kDa) ,
does not traverse the BBB following intravenous (IV) administration
. However, upon direct administration to the brain parenchyma
of animal models, the enzyme is readily absorbed by target cells via
mannose-6-phosphate (M6P) receptors .
A clinical debate remains as to whether large molecular weight
proteins can be delivered as effectively by the IT-lumbar route
as by ICV administration. While studies in rodents suggest
equivalency in brain tissue biodistribution of large molecules
delivered via IT or ICV routes into the CSF , to our
knowledge, this comparison has not been performed in large
animal models or in humans. Consequently, ICV or intracisternal-
IT routes both remain preferred methods of administration for
delivery of recombinant proteins to deep brain tissues of animal
models. Studies in the MPS I canine model, where recombinant
iduronidase was administered by intracisternal injection, provide
supportive evidence that enzyme delivered to the CSF may
penetrate into deeper regions of the brain and modify disease
Here, we are the first to report results from a series of animal
studies comparing the IT-lumbar and ICV delivery routes for a
large molecule biologic, I2S. Direct comparisons were made using
positron emission tomography (PET) in nonhuman primates and
enzyme biodistribution in dogs. We observed similarly widespread
distribution and cellular localization of I2S in the brain after both
IT-lumbar and ICV delivery. In contrast, notably more I2S was
detected in the spinal cord following IT delivery. Chronic CNS
delivery of a specific I2S-IT formulation was effected via a SC port
connected to an IT-lumbar catheter. This study revealed in greater
detail the extent of enzyme biodistribution and cellular localization
in the brain and spinal cord of nonhuman primates. Moreover, in a
mouse model of MPS II, IT-delivered I2S produced morphological
improvements in all areas of the brain evaluated.
Thus, contrary to the prevailing viewpoint that flow dynamics of
the parenchyma interstitium  and CSF  would prevent
penetration of IT-lumbar–administered proteins to the white
matter of the brain, we clearly demonstrate that IT delivery of a
lysosomal enzyme results in protein penetration to all brain tissues
and deposition in the lysosomal compartment of target cells, the
site of pathological glycosaminoglycan (GAG) accumulation. The
less invasive nature of IT-lumbar delivery and the successful
precedent of its usage in chronic pain and spasticity, make this
route an attractive, clinically relevant means of delivering
biological therapeutics to the brain, particularly in children.
1. PET Imaging of I2S Translocation in Cynomolgus
labeled I2S at three doses, 3 (n=4), 10 (n=1), and 20 (n=1) mg by
IT-lumbar injection and 3 mg (n=1) by ICV injection;124I-labeled
I2Swas alsoadministeredIV at1 (n=4) and 0.1(n=4)mg/kg.PET
imaging data showed that124I -labeled I2S administered through
the IT-lumbar catheter spread immediately and uniformly in the
CSF over 15–20 cm of the length of the spine, 10–15 cm cranially
and 3–5 cm dorsally from the catheter tip (Figure 1A-D). Within 30
minutes after injection, I2S was distributed over the entire
leptomeningeal space. While distribution of protein in the
parenchyma was comparable for IT and ICV administrations, the
ICV route resulted in notably less deposition within the spinal
column (Figure 1A). By 1 hour, 55620% of the dose was in the
cranial region following IT-lumbar administration. Sequential PET
imaging of the brain after the IT-lumbar injection (Figure 1B)
demonstrated that I2S had moved from the CSF into the superficial
(20–100 mg/ml as estimated from PET data) and then into deeper
brain tissues (3–20 mg/ml). The inflow continued for 461 hours, as
exemplified by the differential (0.5 to 5 hour) PET image in
Figure 1C. In the cranial segments, the clearance was faster
(Figure 1C), which is consistent with CSF drainage to the system
predominantly in the arachnoid granulations of the superior
longitudinal sinus . No residual I2S deposition was detected
near the catheter opening.
I2S transfer from CSF to the systemic circulation started
immediately after the injection, suggesting significant uptake of
drug from the lumbar CSF into the blood and distribution to the
periphery. As no I2S deposition was observed in lymph nodes
along the spine, the data are consistent with direct drainage of
CSF into venous blood in the arachnoid as the major route of I2S
clearance from the CNS. Approximately 40610% of the dose was
transferred to the systemic circulation within 5 hours, at which
time I2S concentrations in the liver and heart were approximately
50% of the respective concentrations in the same organs resulting
from IV injection at the same dose (Table 1). The kinetics of I2S
accumulation outside the CNS were consistent with absorption
from the CSF, with Tmaxof 461 hours and Cmaxof 0.160.5%
Intrathecal Delivery of Lysosomal Enzymes
PLoS ONE | www.plosone.org2 January 2012 | Volume 7 | Issue 1 | e30341
injected dose/g for lung, heart and kidneys and 0.360.1% injected
dose/g in the liver. I2S concentrations in the blood (measured by
radioactivity at 5 hours postinjection) were similar to residual I2S
concentrations in blood at the same time point after IV
administration (1265% of the initial concentration after IV
administration). Therefore, these data suggest that the CSF serves
as an intermediate reservoir for I2S, from which approximately
half of the dose is gradually transferred into the systemic
circulation over the first 4 hours after injection.
Overall, data from PET analyses suggests that IT-lumbar
administration is an efficacious delivery route for I2S to
the CNS and, in parallel, to other organs. The results of
enzyme delivery to the systemic circulation and other organs
following IT-lumbar administration will be reported in detail
2. Cellular Localization of I2S in the CNS
Healthy cynomolgus monkeys received six consecutive monthly
doses of I2S-IT (3, 30 and 100 mg/dose administered via an
implanted IT drug delivery device with a SC port) as well as
weekly IV I2S doses of 0.5 mg/kg. Device and vehicle controls
(n=6, all groups) were treated similarly with respect to dosing
regimen. Both the IT device placement and the dosing regimen at
the highest dose tested were well tolerated .
Assessment of vehicle treated control animals revealed that
the immunohistochemistry (IHC) procedure did not detect
Figure 1. In vivo distribution of124I -labeled I2S (3 mg/animal) in cynomolgus monkeys by PET. (A) Distribution of I2S administered
through the lumbar (left) and ICV (right) catheters 30 minutes after the administration as demonstrated by a projection PET image (sum of all slices).
Relative linear color scale. (B) The distribution of I2S in the brain at 0.5, 2.5, 5 and 24 hours after lumbar administration; PET image, 1.2 mm slice
through the corpus callosum region in the plane parallel to the occipital bone. The color scale is calibrated in mg/ml of I2S. (C) Changes in the cerebral
I2S distribution between 0.5 and 5 hours after lumbar administration shown in monochrome linear color scale. The image was obtained by
subtraction of the quantitative data matrix obtained at 5 hours from the one obtained at 0.5 hours. Neutral orange color represents no change.
Clearance of I2S from the CSF is seen as black, and accumulation in the parenchyma and arachnoid as white color. (D) An example of single-animal
dynamics of I2S clearance from the leptomeningeal compartment and CNS.
Intrathecal Delivery of Lysosomal Enzymes
PLoS ONE | www.plosone.org3 January 2012 | Volume 7 | Issue 1 | e30341
endogenous monkey iduronate-2-sulfatase and was specific for
recombinant human I2S (Figure 2A). In contrast, there was
widespread cellular deposition of I2S throughout the CNS in
treated animals. In the gray matter, I2S was detected in neurons
of the cerebrum, cerebellum, brainstem, and spinal cord of all
groups in a dose-dependent manner. In the surface gray matter of
the higher dose groups, large numbers of cerebral neurons were
positive for I2S staining in the surface cortex (Figure 2B). I2S was
also detected in neurons in the thalamus (Figure 2C), hippocam-
pus (Figure 2D), caudate nucleus (Figure 2E) and spinal cord
(Figure 2F). Meningeal and perivascular cells also stained positive
for I2S (Figure 2G). These data are consistent with the results
from our PET studies described above. While the I2S staining
density in the white matter was generally lower than that in the
(Figure 3A). As shown in Figure 3B, I2S was located within the
lysosomes of oligodendrocytes and detected specifically within the
lysosomes of neurons. I2S was also detected within axons
In order to discern whether the delivered I2S retained biological
activity, levels of I2S in the brain were measured utilizing a specific
activity assay. The enzyme activity in the brain of the 3 mg IT
dose group 24 hours after the last dose was not apparently
different from the basal levels in the device- and vehicle-control
animals. Enzyme activity in the brain of 30 mg and 100 mg IT-
dosed animals was above baseline at necropsy (24 hours postdose).
Results of the quantitative analysis of I2S in brain tissues from this
study are described in detail elsewhere .
The observed I2S distribution pattern was analogous to
healthy beagle dogs given a single IT or ICV dose. No
endogenous dog iduronate-2-sulfatase was detected in vehicle
treated control animals by IHC confirming the specificity of
the immunostaining procedure for recombinant human I2S
(Figure 4A). I2S was widely distributed throughout the gray
matter of both IT and ICV groups as determined by IHC. In the
cerebral cortex, neurons were positive for I2S in all six neuronal
layers, from the surface molecular layer to the deep internal layer
in both IT and ICV groups (Figure 4B and 4C). In the cerebellar
cortex of the IT and ICV groups, I2S was detected in neurons,
including Purkinje cells (Figure 4D and 4E). In both IT and ICV
groups, a large population of neurons in the hippocampus was
positive for I2S (Figure 4F and 4G). I2S-positive neurons were
also found in the thalamus and caudate nucleus in both groups
(Figure 4H and 4I).
3. IT-Lumbar–Administered I2S Improves Brain Pathology
in an MPS II Model
Currently, the only animal model of MPS II disease is an I2S
gene knockout generated in mice . I2S was administered to 8-
12 week old male mice via direct lumbar injection (260 mg; two
injections at study days 1 and 8 or three injections at study days 1,
8 and 15). Mice were sacrificed 1 hour after the final injection
followed by tissue preparation for IHC and histopathological
examinations. Following the third injection, there was widespread
reduction of cellular vacuolation in the surface cerebral cortex,
caudate nucleus, thalamus, cerebellum, and the white matter in
I2S-treated mice compared to control (untreated) mice (Figure 5A,
a-j). The IT-treated mice also had marked reductions in lysosomal-
associated membrane protein-1 (LAMP-1) immunoreactivity, a
lysosomal protein marker used for the detection of lysosomal
storage disorders and an indicator of disease state, in the surface
cerebral cortex, caudate nucleus, thalamus, cerebellum and white
matter (Figure 5B, a-j). Morphometrical analysis of LAMP-1
immunostaining of various brain regions confirmed that there
were significant reductions in the LAMP-1 positive immunostain-
ing in all areas of the brain evaluated (cortex, caudate nucleus,
thalamus, cerebellum and white matter; Figure 5C). As demon-
strated by IHC, I2S was not detected in uninjected MPS II control
animals (Figure 6A), but was detected in the neurons of the
cerebral and cerebellar cortex and the meningeal cells (Figure 6B
and 6C) of treated animals. Additionally, electron microscopy
showed there was a reduction in the presence of storage inclusions
in neurons in the gray matter and vacuolation in oligodendrocytes
in the white matter (Figure 7A-D).
While much progress has been made in treating the somatic
symptoms of MPS disorders , , development of a therapy
for the devastating neurological involvement is greatly needed. In
severe Hunter syndrome, histological changes in the brains of
affected patients include atrophy, cortical neuronal swelling,
cerebral white matter reduction, dilated perivascular spaces and
Purkinje cell dendrite swelling . Magnetic resonance imaging/
spectroscopy studies have shown that severe diffuse lesions
involving the white matter, brain atrophy and hydrocephalus
were more common in patients with cognitive impairment than in
those without impairment . However, even patients without
cognitive impairment or developmental delays were shown to have
Table 1. Distribution of 124I-labeled I2S following positron emission tomography in cynomolgus monkeys at 5 hours after
Injection Route Number of AnimalsTotal Dose Gray matter (%ID/g)Liver (%ID/g) Heart (%ID/g)
IL4 3 mg0.7760.430.1460.02 0.0960.02
IL1 10 mg 0.2420.240.11
IL1 20 mg0.465 0.150.08
IL1 30 mg 0.4790.14 0.10
IV4 1 mg/kg**0.02560.004 0.3760.100.1360.07
IV4 0.1 mg/kg**0.05860.0530.4460.140.1460.02
ICV1 3 mg 0.6350.097 0.04
*Data normalized to 2 kg animal body weight.
**The total injected doses in the IV groups were 3.561 mg and 0.3460.1 mg, respectively.
ICV, intracerebroventricular; ID, injected dose; IL, intralumbar; IV, intravenous.
Intrathecal Delivery of Lysosomal Enzymes
PLoS ONE | www.plosone.org4 January 2012 | Volume 7 | Issue 1 | e30341
brain abnormalities that included atrophy, ventriculomegaly and
enlarged perivascular spaces .
Several strategies have been evaluated to correct the neurolog-
ical aspects of LSDs but these have either been unsuccessful or are
still in preclinical development. Bone marrow transplants (BMT)
or hematopoietic stem cell transplants (HSCT) appear to have
limited or no success in treating the neurological and cognitive
impairment associated with Hunter syndrome. Guffon et al 
found improvements in somatic symptoms in children with Hunter
syndrome who were treated with BMT, with two patients with the
attenuated phenotype having normal social and scholastic
development; four patients with the severe syndrome had postgraft
declines in intelligence/developmental quotients and three lost the
ability to walk in their early teens. Limitations to BMT/HSCT
may be that only a small number of transplanted cells migrate into
the CNS and differentiate into microglia, and the time required for
the number of cells to be transplanted may be too long to prevent
disease progression .
Viral vectors, primarily adeno-associated virus (AAV) , 
and lentiviruses , ,  expressing recombinant enzymes
have also been evaluated, generally in mouse models, for virus-
mediated enzyme replacement therapy in the brain , ,
, . Cross-correction (in which corrected cells release
enzyme into extracellular space that is taken up by receptor-
mediated endocytosis on neighboring cells) and enzyme transport
along neural projections are reported to be the primary
mechanisms responsible for widespread brain distribution .
Secretion of enzyme from transduced cells into the CSF may also
contribute to the distribution of enzyme replacement therapy .
However, one limitation to the use of viral vectors is that direct
injection into the brain parenchyma is required . It is
estimated that as many as 368 injections would be needed to
distribute AAV vectors to the infant brain . Another limitation
is that lentivirus vectors can integrate into the genome of dividing
and nondividing cells [51–52]. Since lentivirus vectors strongly
prefer transcriptional units for integration , , there is the
potential for inactivation of essential genes or activation of proto-
oncogenes. Although postmitotic cells such as neurons are unlikely
to be affected, neural stem cells in the brain may be susceptible
. AAV vectors also integrate into active genes, with frequent
chromosomal deletions of up to 2 kB at integration sites .
The current studies are the first to demonstrate that direct CNS
administration of a recombinant lysosomal protein results in the
delivery of a significant fraction of the administered protein to the
brain and widespread deposition in neurons of the brain and
spinal cord in both cynomolgus monkeys and dogs. The
similarities observed in brain distribution patterns achieved after
IT-lumbar and ICV administration of I2S are suggestive of bulk
flow and active remixing of the CSF. Thus in a clinical setting,
both administration routes might be effective in delivering drug to
target cells in the CNS. Spinal injection ports have two benefits
over ICV injection: (1) they are perceived as less invasive and (2)
they have been used successfully for many years in the treatment of
chronic pain and spasticity, the latter indication being especially
relevant for the pediatric population. Additionally, the observed
deposition of I2S in the spinal cord following IT administration
indicates that the IT-lumbar route may provide a third advantage
over ICV administration in addressing spinal components of
complex CNS disorders.
Evidence from perivascular cell staining and protein transloca-
tion dynamics observed by PET imaging demonstrates that the
enzyme moves within the perivascular space, presumably by
pulsation-assisted convective mechanisms. An additional mecha-
nism of transport is suggested by the observed association of I2S
with neurofilaments, indicative of active axonal transport. The
latter presumably begins with protein interaction with neuronal
M6P receptors , which are widely expressed on cells of the
spinal cord and brain . Axonal transport of lysosomal enzymes
has been implied by indirect methods in vivo  and by imaging
in vitro . The current study provides the first direct evidence of
axonal transport of nonvirally mediated enzyme replacement
delivered via the CSF. Thus, protein delivery from the CSF to the
brain surface and deeper into the brain seems to depend on active
transfer processes, none of which have been previously described
for protein delivery to the brain. Further elucidation of these
transport mechanisms and their dependence on protein structure
will establish the general applicability of this approach for
treatment of other CNS disorders.
We have demonstrated that IT-lumbar administration of I2S
results in cellular deposition of the enzyme in the brain and spinal
Figure 2. Cellular uptake of I2S in the neurons and vascular
cells of the brain of monkeys. (A) I2S immunohistochemical
staining of the cerebral cortex of vehicle control monkeys was negative.
Representative images showing I2S following 6 monthly IT injections of
100 mg/dose detected in the neurons by immunohistochemistry in the
cerebral cortex (B), thalamus (C), hippocampus (D), caudate nucleus (E)
and spinal cord (F). (G) Meningeal cells (arrow) covering the surface of
the cerebral cortex and perivascular cells surrounding the blood vessel
(V) were also positive for I2S. Green, I2S; blue, 49-6-diamidino-2-
phenylindole (DAPI)-stained nuclei. Scale bar: 25 mm.
Intrathecal Delivery of Lysosomal Enzymes
PLoS ONE | www.plosone.org5 January 2012 | Volume 7 | Issue 1 | e30341
cord of cynomolgus monkeys and dogs, and associates with the
lysosomes. Moreover, in a knockout mouse model of MPS II, I2S
treatment by IT-lumbar administration led to a decrease in cellular
vacuolation, a hallmark of the pathology associated with this
disease. Overall, these data demonstrate the physiological transport
of a therapeutic protein from CSF to the brain parenchyma after
IT-lumbar administration, thus suggesting a clinically feasible route
for delivery of enzyme replacement therapeutics to the brain.
Studies in humans utilizing this approach are currently under way.
Materials and Methods
The studies involving animals were performed at Association
for Assessment and Accreditation of Laboratory Animal Care
(AAALAC)–accredited facilities and were performed in accordance
with the Guide for the Care and Use of Laboratory Animals (7th
Edition, 1996, National Research Council, U.S.). The following
detailsregardingthe welfare of non human primatesand steps taken
to ameliorate suffering are provided in accordance with the
recommendation of the Weatherall report, ‘‘The use of non-human
primates in research’’. Non human primates were housed under
temperature, humidity, and lighting controlled conditions in
accordance with recommendations in the Guide for the Care and
Use of Laboratory Animals. For example, rooms were set to
maintain 2262uC, 50% 630% relative humidity, 12 hour light:
dark cycle, and with 10 room air changes per hour. Appropriate
food, water, treats and vitamin supplements were provided and
animals were given access to environmental enrichment such as
approved toys, swings, perches, mirrors, television, or music to
Figure 3. I2S is detected within the lysosomes of oligodendrocytes and in the axons of white matter. (A) Representative images
showing I2S uptake in oligodendrocytes in the white matter of the 100 mg/dose IT-injected monkeys as demonstrated by colocalization of I2S (a,
green) with glutathione-S-transferase-pi, an oligodendrocyte marker (b, red) and the overlay image (c). (B) I2S was located within lysosomes of the
oligodendrocytes as demonstrated by colocalization of I2S (a, green) with lysosomal associated membrane protein-1 LAMP-1 (b, red) and the overlay
image (c). I2S is located within the lysosomes of neurons, as demonstrated by colocalization of I2S (d, green) with LAMP-1 (e, red) and the overlay
image (f). (C) I2S was also detected in some axons in the white matter as demonstrated by colocalization of I2S (a, green) with neurofilament, an
axonal marker (b, red) and the overlay image (c). Scale bars: A, 25 mm; B, a–c, 10 mm, d–f, 5 mm; C, 30 mm.
Intrathecal Delivery of Lysosomal Enzymes
PLoS ONE | www.plosone.org6 January 2012 | Volume 7 | Issue 1 | e30341
promote psychological well-being. Every effort was made to
minimize pain, discomfort and suffering through the use of
appropriate methods and agents for analgesia, anesthesia and
euthanasia as described below. All animals were under the care and
supervision of a veterinarian.
The 6-month cynomolgus monkey and Beagle dog studies were
performed at Northern Biomedical Research, Inc. (Muskegon, MI)
in accordance with the Guide for the Care and Use of Laboratory
Animals, United States Department of Health and Human
Services, No. 86-23, and the Animal Welfare Act (9 CFR Part
3); USDA No. 34-R-0025. The studies were approved by the
Northern Biomedical Research, Inc. Institutional Animal Care
and Use Committee (IACUC) (Cynomolgus monkey study
protocol number 047-004; Beagle study protocol number 047-
001). Non-invasive PET imaging in cynomolgus monkeys was
performed at Massachusetts General Hospital (Boston, MA) in
accordance with the Guide for the Care and Use of Laboratory
Animals and the Animal Welfare Act; USDA No. 14-R-0014. The
study was approved by the Massachusetts General Hospital
IACUC (Protocol number 2003N00128). The studies in mice
were performed at Shire Human Genetic Therapies (Lexington,
MA) and were approved by the Shire Human Genetic Therapies
IACUC (Protocol number ACUP 47).
1. Iduronate-2-sulfatase (I2S)
Recombinant human I2S was expressed and purified from a
human-derived cell line  and was provided in IT (idursulfase-
IT) , and IV (commercially available idursulfase) formulations
for the non-human primate and Beagle studies. For the mouse
studies, I2S was concentrated and suspended in phosphate
buffered saline (PBS).
2. Positron Emission Tomography Study in Cynomolgus
Male and female cynomolgus monkeys (Macaca fascicularis)
between 1.8–5.6 kg were administered124I-labeled I2S at three
doses, 3 mg (6 mg/ml; n=4), 10 mg (20 mg/ml; n=1) and 20 mg
(40 mg/ml; n=1), by IT-lumbar injection and 3 mg (6 mg/ml;
n=1) by ICV injection. In addition,124I-labeled I2S was
administered IV at two doses (1 mg/kg, n=4; 0.1 mg/kg, n=4;
see Table 1). Over the entire duration of the study, the monkeys
were segregated from other nonhuman primates and housed in a
separate room at the Massachusetts General Hospital primate
facility. Animals were fasted for 24 hours before each experiment.
At the housing site, the animals were sedated with ketamine
hydrochloride IM combined with xylazine (IM; Rompun, Bayer
AG; Leverkusen, Germany) and then transported to the imaging
site. The animals were intubated and given continuous Isoflurane
(Halocarbon Products Corp., River Edge, NJ)/oxygen anesthesia.
Heart rate, breathing rate and carbon dioxide content in the
exhaled air were monitored continuously; isoflurane flow was
adjusted as needed. Animals were given nonradioactive iodine
solution (0.2 ml, 15 mM sodium iodide) SC immediately before
the study to suppress124I uptake in the thyroid. The radioiodin-
ated proteins were administered IV or IT.
The study used a microPET P4 scanner (Siemens/CTI
Concorde Microsystems, Knoxville, TN) designed for small
primates (opening 22 cm, axial field of view [FOV] 7.89 cm,
transaxial FOV 19 cm). The energy window of the detector for the
entire study was set to 350–650 keV and coincidence timing
window was set to 6 nanoseconds. We have determined previously
that the imaging data under these settings are fully quantitative
and linear at
supplied as sodium124I solution in 0.02 M sodium hydroxide, 0.3–
2.7 mL/MBq (IBA Molecular, Richmond, VA). The nominal
radiochemical purity was 95% (,5% of iodate and diiodate by
high-performance liquid chromatography [HPLC]) and the
nominal radionuclide purity was .99% at calibration (,0.5% of
spectroscopy). The chemical purity was Te,1 mg/ml by ultravi-
olet (UV)-visible spectroscopy.
I2S was labeled with124I, up to 185 MBq/mg, using Iodogen
(Pierce Biotechnology, Inc., Rockford, IL) precoated iodination
tubes (pH, 7.4; 25uC) for 20 minutes, with subsequent metabisul-
fite treatment desalting and purification of the iodinated protein by
size exclusion HPLC (SEC HPLC). Typically, the reaction
mixture consisted of 0.5 M sodium phosphate buffer solution,
0.05 ml; protein (6–50 mg/ml), 0.05 ml; and [124I] sodium iodide,
0.02–0.07 ml (1.5–3 mCi). After iodination, the reaction mixture
124I doses of up to 1 mCi per animal.
125I, none detected by high-purity germanium gamma
Figure 4. Widespread distribution in the brains of dogs
following either intracerebroventricular (ICV) or intrathecal
(IT) administration. (A) I2S immunohistochemical staining of the
cerebral cortex of vehicle control dogs was negative. Representative
fluorescent immunohistochemical images showed uptake of I2S in the
neurons of both ICV-dosed (C,D,F,H) and IT-dosed (C,E,G,I) dogs. I2S
was detected in neurons in the deep internal layer of the cerebral
cortex (B,C), Purkinje cells of the cerebellum (D,E) and neurons in the
hippocampus (F,G) and thalamus (H,I). Green, I2S; blue, DAPI-stained
nuclei. Scale bar: 25 mm.
Intrathecal Delivery of Lysosomal Enzymes
PLoS ONE | www.plosone.org7 January 2012 | Volume 7 | Issue 1 | e30341
Figure 5. Reversal of pathology in I2S knockout (mucopolysaccharidosis II) mice after three IT-lumbar injections of I2S. (A)
Hematoxylin and eosin-stained brain tissues of uninjected (left panels) and injected (right panels) mice showed numerous cellular storage vacuoles
(arrows) in the uninjected brain that were markedly reduced in injected mice in the cerebral cortex (a,b), caudate nucleus (c,d), thalamus (e,f), white
Intrathecal Delivery of Lysosomal Enzymes
PLoS ONE | www.plosone.org8 January 2012 | Volume 7 | Issue 1 | e30341
was transferred to a polypropylene tube containing 1.060.2 mg of
dry sodium metabisulfite to deactivate the oxidants present in the
aqueous phase. The solution was then centrifuged at 5 rpm for
1 minute and desalted on Sephadex G-25 (GE Healthcare,
Milwaukee, WI) equilibrated with unbuffered isotonic saline.
The iodinated protein was analyzed by SEC HPLC using a Bio-Sil
SEC 125 column (BioRad Laboratories, Inc., Hercules, CA),
8067.8 mm, with UV (wavelength 280 nm) and gamma radioac-
tivity detection. The elution buffer was 20 mM sodium citrate,
pH=6.5, in 0.9% sodium chloride solution. The amount of
protein was determined on the basis of the UV absorption at
280 nm; yield by protein was found to be nearly 100%.
Radiochemical yield (by124I) was 8567% and the radiochemical
purity was .98%. In a separate experiment, we have determined
that iodination under these conditions did not affect the rate of
protein uptake by cells or the subsequent intracellular activity as
determined in vitro in cell-based assays (data not shown).
The radiolabeled preparation (containing 0.5–0.7 mCi of124I
per animal) was mixed with a calculated amount of unlabeled I2S
to obtain the desirable protein dose. The solution was distributed
into 0.5 or 1 ml syringes. The total administered volume was 0.4–
1 ml, depending on the protein dose.
For IV administration, a catheter equipped with a T connector
was inserted in the saphenous vein nonsurgically. The animals
were set on a microPET bed and a whole body transmission image
was acquired before the injection. The animals then were
positioned for dynamic imaging of the lower thoracic (heart and
liver) area. The protein solution was administered through the T
connector cap and flushed with 1 ml saline simultaneously with
the start of the dynamic imaging data acquisition. The latter was
carried out for 20 minutes. Static whole body images were
acquired (section by section) continuously for 5 hours and then at
24 and 48 hours. Blood samples were taken with each imaging
session and were studied by HPLC with gamma detection.
For CNS administration, I2S was administered through SC
injection ports equipped with catheters surgically implanted into
the upper lumbar segment of the leptomeningeal space or the left
ventricle for IT-lumbar and ICV delivery, respectively. A needle
equipped with a T connector was inserted into the port. The
animals were set on a microPET bed and a whole body
transmission image was acquired before the injection. The animals
were positioned for dynamic imaging of the area of port opening
into the cerebrospinal fluid. The I2S solution was administered
through the T connector cap and flushed with saline (1 ml +
0.5 ml per kg of body weight) simultaneously with the start of the
dynamic imaging data acquisition. The dynamic acquisition was
carried out for 20 minutes. Whole body images were acquired
(section by section) continuously for 5 hours and then at 24 and
48 hours. Blood samples were taken at 40 minutes and 2 hours
and studied by HPLC with gamma detection.
Data acquisition, prereconstruction and reconstruction were
carried out on a Dell Precision PWS690 Workstation (Dell, Inc.,
Round Lake, TX; 3 GB RAM and 8 Xeon 3.20 GHz processors
running under a 64-bit Windows XP [Microsoft Corp., Redmond,
WA]) using Siemens MicroPET Firmware/Software, release 2.4
(Siemens Medical Solutions, Inc., Malvern, PA). All subsequent
image processing and analyses were performed on nonhost
workstations using the ASIProVM software (Siemens/CTI Con-
corde Microsystems, Knoxville, TN) under Windows XP and
MacOS 10.5 (Apple Computer, Inc., Cupertino, CA) and AMIDE
 running under MacOS 10.5. The data were reconstructed into
the image matrix with the pixel size of 0.95 mm and fixed slice
thickness of 1.2 mm using a 3-dimensional (3-D) ordered-subset
expectation maximization/maximum a posteriori (OSEM3D/MAP)
protocol ,  with the hyperparameter b value of 1.516. The
data were also reconstructed with Fourier rebinning 2-D filtered
backprojection (FORE-2DFBP) to ensure that the numerical data
derived from OSEM3D/MAP and FORE-2DFBP reconstructed
images were identical and thus excluded possible reconstruction
artifacts (none were identified). FORE-2DFBP was performed with
a ramp filter cutoff at the Nyquist spatial sampling frequency
(0.5 mm21) . Whole body images were composed of the
acquired section images with a 12 mm overlap.
The images were processed to obtain numerical data from
hand-drawn 3-D regions covering all organs and tissues of interest.
Typically, each region of interest was 0.2–5 ml in volume. The
data (expressed in nCi per ml) were converted into protein
concentration and percent injected dose/ml, and tabulated. For
the spine, the data were also obtained for each 1.2 mm slice of the
spinal column and expressed as protein amount per spinal column
length (mg/mm). Data were processed to evaluate the statistics
(mean values 6 standard deviations). Corrections typical for PET
Figure 6. Cellular uptake of I2S in IT-injected (three injections) I2S knockout mice. (A) I2S immunohistochemical staining of the cerebral
cortex of untreated control mice was negative. In IT-injected mice, I2S positive staining was found in neurons of the cerebral cortex (B) and Purkinje
cells of the cerebellum (C). Meningeal cells (arrows) were also I2S positive in IT-injected animals. Scale bar: 25 mm.
matter (g,h), and cerebellum (i,j). Scale bar: 25 mm. (B) As demonstrated by immunohistochemical staining of lysosomal-associated membrane-1
(LAMP-1), there was a marked reduction of LAMP-1 immunoreactivity in the brains after three IT injections of I2S (right panels) compared with
uninjected mice (left panels). There was a decrease in the number of LAMP-1 positive cells and lighter staining intensity in the cerebral cortex (a,b),
caudate nucleus (c,d), thalamus (e,f), white matter (g,h), and cerebellum (i,j). Scale bar: 25 mm. (C) A comparison of the mean LAMP-1 positive area
between uninjected and I2S (two or three IT injections) injected wild-type (WT) mice in the cerebral cortex (cortex), caudate nucleus (CP), thalamus
(TH), white matter (WM) and cerebellum (CBL). Data are represented as the mean 6 s.d. # P,0.05; * P,0.01; ** P,0.001.
Intrathecal Delivery of Lysosomal Enzymes
PLoS ONE | www.plosone.org9 January 2012 | Volume 7 | Issue 1 | e30341
data evaluation (radionuclide decay, scatter, randoms, etc.) were
made as in our previous studies .
3. Six-month Cynomolgus Monkey Study
Forty-eight male cynomolgus monkeys were randomized into
one of five groups. For the experimental design of the study, see
Table 2. All had catheters implanted in the subarachnoid space at
the lumbar spine that terminated in a SC titanium access port.
Prednisolone sodium succinate (intravenous [IV], 30 mg/kg) and
flunixin meglumine (intramuscular [IM], 2 mg/kg) were admin-
istered prior to the surgical procedure and monkeys were
pretreated with SC atropine sulfate (0.04 mg/kg), sedated with
ketamine hydrochloride (IM, 8 mg/kg; Ketalar, Pfizer, Inc. New
York, NY), intubated and maintained on approximately1 L/
minute of oxygen and 2% isoflurane. An incision was made over
the dorsal processes of the lumbar spine (L4, L5or L6), and a
hemilaminectomy was made for the insertion of a tapered
polyurethane catheter (58.4 cm in length with six side holes of
0.33 mm diameter) at L3, L4or L5. The catheter was inserted
through a small dural incision and was advanced approximately
10 cm anterograde to the area of the thoracolumbar junction. A
titanium SC port was attached to the IT catheter and implanted in
the SC tissue. Proper catheter placement was confirmed by
myelogram using Isovue-300 (0.8 ml; Bracco Diagnostics, Inc.,
Princeton, NJ). After recovering from surgery, animals received
butorphanol tartrate (IM, 0.05 mg/kg) and ceftiofur sodium (IM,
5 mg/kg twice daily [BID] for 2 d). IV infusions were initiated at
least 6 days after device implantation, with IT doses administered
2 days before every fourth weekly IV dose.
The I2S-IT formulation was supplied in a vehicle of 154 mM
sodium chloride and 0.005% polysorbate 20 (pH 6.0) at
concentrations of 3, 30 and 100 mg/ml. Six monthly doses of
I2S-IT were administered as a 1 ml bolus followed by a flush of
0.5 ml phosphate buffered saline (PBS). Commercially available
I2S was used for IV administration (weekly doses of 0.5 mg/ml).
Monkeys in the vehicle-control group received I2S-IT vehicle and
I2S-IV vehicle. Device control animals received PBS (pH 7.2) for
IT administration and 0.9% sodium chloride for IV administra-
tion. Clinical signs were monitored. Monkeys were sacrificed
24 hours after the final dose or following a 1-month recovery
period. For a detailed description regarding the harvesting of
tissues for histopathology, IHC, and quantitative I2S analysis,
please see Felice et al .
3.1. Determination of I2S activity in tissue extracts.
activity was determined with a 2-step fluorometric assay 
using the substrate 4-methylumbelliferyl-a-iduronate-2-sulfate
(Moscerdam Substrates, Rotterdam, the Netherlands). Tissue
extracts were diluted in 0.2% bovine serum albumin (pH and
heat-treated to inactivate lysosomal enzymes) supplemented with
0.004% sodium azide. Substrate was desulfated by duplicate
incubations of 10 ml of diluted sample with 20 ml of 1.25 mM
substrate for 4 hours at 37uC. After I2S inhibition by the addition
of 20 ml of phosphate/citrate buffer (0.2 M Na2HPO4/0.1 M
citric acid, 0.02% sodium azide, pH 4.7), samples were incubated
Figure 7. Electron micrographs of brain cells in uninjected and IT-injected (3 doses) I2S knockout mice. Pathological improvements
occurred at the ultrastructural level. (A) Neurons of uninjected mice had lamellated inclusions, zebra body-like structures and vacuoles containing
granular storage material (insert), which was reduced in I2S injected mice (B). Oligodendrocytes of uninjected mice showed large electron-lucent
storage vacuoles (arrow; C) while oligodendrocytes of I2S-injected mice had minimal vacuolation (D).
Intrathecal Delivery of Lysosomal Enzymes
PLoS ONE | www.plosone.org10 January 2012 | Volume 7 | Issue 1 | e30341
with 10 ml of lysosomal enzymes purified from bovine testis for
24 hours at 37uC to liberate 4-methylumbelliferyl from substrate
desulfated in the first reaction. This second reaction was stopped
by adding 200 ml of stop buffer (0.5 M NaHCO3/Na2CO3,
pH 10.7 with 0.025% Triton X-100) to each well. Fluorescence
was measured in 96-well fluorometry plates using a SpectraMax
M2 fluorescent plate reader (Molecular Devices, Sunnyvale, CA).
Biomedicals) as a standard and were normalized to total protein
in tissue extracts as determined by bicinchoninic acid assay. I2S
activity was expressed as nmol of substrate hydrolyzed per hour
per mg of total protein (nmol/h/mg protein).
4. Beagle Dog Study
Male beagle dogs were randomized using computer-generated
numbers into two groups (group 1 [ICV], n=3; group 2 [IT];
n=4). All had catheters implanted in the subarachnoid space at
the lumbar spine or in the left lateral cerebral ventricle (for dosing)
and in the cisterna magna (for sampling). All catheters terminated
in a SC titanium access port. Dogs were pretreated with atropine
sulfate (SC, 0.04 mg/kg) followed by sodium thiopental (IV,
16 mg/kg) and were masked to a surgical plane of anesthesia,
intubated and maintained on 2% halothane or isoflurane.
Prednisolone sodium succinate (IV, 30 mg/kg) and flunixin
meglumine (IM, 2 mg/kg), were administered prior to surgery.
To place the catheter in the cisterna magna, a longitudinal
incision was made from approximately C4to the occipital crest
and the musculature was reflected. A hemilaminectomy (approx-
imately 5 mm) was performed in the posterior portion of C1. An
incision was made in the dura, and the catheter (0.9 mm outside
diameter [OD] x 0.5 mm inside diameter [ID] stepped polyure-
thane catheter with polished stainless steel tip) was directed toward
the cisterna and secured in place using dental acrylic. To place the
catheter in the IT lumbar space, an incision was made over the
dorsal process of the lumbar spine at approximately L4, L5or L6.
The muscle was dissected, and a hemilaminectomy was performed
for the insertion of a 0.9 mm OD x 0.5 mm ID stepped
polyurethane, fenestrated catheter. The catheter was advanced
to the area of the thoracolumbar junction. Proper catheter
placement was confirmed by myelogram with Isovue-300. For
group 1 dogs, the insertion was at L5with the catheter tip located
at L1. For group 2 dogs, the insertion was at L3, L4or L5with the
catheter tip located at L1or T12. The skin was closed with sutures
and tissue adhesive.
Magnetic resonance imaging (MRI) was performed to deter-
mine the coordinates for ICV catheter placement in the group 1
dogs. A dorsal sagittal incision was made over the calvarium, and a
0.9 mm OD x 0.5 mm ID polyurethane, fenestrated catheter was
inserted in the left lateral cerebral ventricle through a craniotomy
using stereotaxic techniques and anchored with dental acrylic.
Dogs received postsurgical MRIs while under anesthesia to verify
catheter placement; because one dog died during the postsurgical
scan due to prolonged anesthesia, subsequent scans were not
Upon recovery from anesthesia, the dogs received butorphanol
tartrate (IM, 0.05 mg/kg), for analgesia and ceftiofur sodium (IM,
5 mg/kg BID; one injection during or prior to surgery followed by
three injections). The iduronate-2-sulfatase (I2S) infusions were
started a minimum of 2 days after implantation of the delivery
devices. An additional dog was used as an undosed surgical
control. A single bolus 1 ml injection of I2S (30 mg/ml in 20 mM
sodium phosphate, pH 6.0; 137 mM sodium chloride; 0.02%
polysorbate-20) was administered IT or ICV followed by a 0.3 ml
flush with PBS, pH 7.2. Clinical signs were monitored. Sacrifice
occurred 24 hours postdose.
4.1. Tissue processing for IHC in the dog study.
3 mm formalin-fixed brain slices (numbers 1, 4, 7, 10, 13 and 16)
from each dog were numbered 1–6 (rostral to caudal). Brain slices
1–4 were from the levels of the forebrain, caudate putamen and
thalamus/hypothalamus of the cerebrum, respectively. The
posterior two slices contained the midbrain, cerebellum and
brain stem (medulla oblongata) tissues. Cervical, thoracic, lumbar
spinal cord and liver samples were also collected. At the time of
tissue trimming, brain slice 1 was separated into left and right
hemispheres and processed in two separate cassettes. Brain slices
2–6 were cut into quadrants due to their larger size. Each section
was separated into quadrants (left and right hemispheres with
upper and lower sections) and processed in separate cassettes.
Transverse sections of each spinal cord sample were processed
5. Mouse MPS II Model
The I2S knockout mouse model of MPS II was developed using
a targeted disruption of the I2S locus that results in an
accumulation of GAG in tissues and organs . Six groups of
male I2S knockout mice, 8–12 weeks old, were either treated with
I2S (10 ml; 26 mg/ml) or left untreated. Groups A and B (n=3)
were administered three doses (at study days 1, 8 and 15) and two
doses (at study days 1 and 8) of I2S, respectively. Group C and E
(n=3) were untreated control groups and group F (n=3) was an
untreated wild-type control. Group D was also treated with three
doses at study days 1, 8 and 15. Animals were sacrificed 1 hour
after the final dose of I2S. For the study, a concentrated I2S
solution was prepared by dialyzing I2S against four changes of 2 L
Table 2. Design of I2S biodistribution study in cynomolgus monkeys.
GroupNumber of animals IV Dose (mg/kg)*
IT Dose (mg)*
Last Day on Study (number of animals)
6 Months Recovery
16 DC (NS)DC (PBS)6-
2 12 0 (vehicle) 0 (IT vehicle)66
3 12 0.5366
5 12 0.510066
*I2S unless otherwise specified.
DC, device control; IT, intrathecal; IV, intravenous; NS, normal saline; PBS, phosphate-buffered saline, pH 7.2.
Intrathecal Delivery of Lysosomal Enzymes
PLoS ONE | www.plosone.org11 January 2012 | Volume 7 | Issue 1 | e30341
phosphate-buffered saline (PBS) followed by centrifugation using a
Vivaspin column (Sigma-Aldrich, St. Louis, MO). The solution
was resuspended in 1 ml PBS for a concentration of 51 mg/ml,
which was subsequently diluted to the final dosing concentration.
For the injection procedure, mice were anesthetized with 1.25%
2,2,2 tribromoethanol (intraperitoneally, 240–350 mg/kg) and the
skin was prepped with povidone iodine followed by isopropyl
alcohol. A small (1–2 cm) midline incision was made over the
lumbosacral spine, and I2S (10 ml) was injected using a 32-gauge
needle with a GASTIGHTH 10–20 ml glass Hamilton syringe
(Hamilton Medical, Inc., Reno, NV) at a rate of 2 ml/20 seconds.
The skin was closed with wound clips. Sacrifice was performed 1
hour after the final injection. Before tissue collection, mice were
perfused with saline followed by 10% neutral buffered saline
(groups A, B and C) or 4% paraformaldehyde (groups D, E and F).
Histology, IHC and electron microscopy were performed on brain
tissues. The area of LAMP-1–positive cells was analyzed with
Image-Pro Plus software (Media Cybernetics, Inc., Bethesda, MD),
and comparative statistics were performed using the Student’s t -
6. Histological and Immunohistochemical Staining
Tissues were embedded in paraffin and serial sections were cut
and stained with hematoxylin and eosin for histological evaluation.
For IHC analysis in the dog and monkey studies, deparaffinized
slides were incubated overnight with mouse monoclonal antibody
2C4-2B2 (Maine Biotechnology Services, Portland, ME) as the
primary antibody to detect injected I2S (or mouse IgG as a control
antibody; Vector Laboratories, Burlingame, CA). Following an
overnight incubation at 2-8uC, a secondary goat anti-mouse IgG
conjugated with horseradish peroxidase was added (Promega).
After additional 30 minutes of incubation at 37uC, Tyramide-
Alexa Fluor 488 labeling solution (Invitrogen Corp., Carlsbad,
CA) was added for an additional 10 minutes. Slides were cover
slipped using an antifading mounting medium (VectaShield;
Vector Laboratories) containing 1.5 mg/ml 49-6-diamidino-2-
phenylindole as a nuclear counterstain and observed with a
multiple channel Nikon fluorescent microscope. Representative
digital images were captured for documentation.
Additional primary antibodies used for IHC included rabbit
anti-glutathione-S-transferase-pi (Assay Designs, Inc., Ann Arbor,
MI) for the detection of oligodendrocytes, rabbit anti-lysosomal-
associated membrane protein-1 (anti-LAMP-1; Santa Cruz
Biotechnology, Inc., Santa Cruz, CA) for the detection of
LAMP-1 and rabbit anti-neurofilament 200 (Sigma-Aldrich) for
the detection of axons. Rabbit IgG was used as a control primary
antibody (Vector Laboratories). The secondary antibody was goat
anti-rabbit Alexa Fluor 568 (Invitrogen).
For the mouse studies, 5 mm paraffin sections were prepared for
hematoxylin and eosin I2S IHC staining. The fluorescence
procedure was similar to that used above for the dog and monkey
studies. For the LAMP-1 staining procedure, deparaffinized slides
were incubated overnight with rat anti-LAMP-1 IgG (Santa Cruz
Biotechnology) as the primary antibody or rat IgG2a as a control
antibody (AbD Serotec, Raleigh, NC). Following overnight of
incubation at 2–8uC, biotinylated rabbit anti-rat IgG (H&L)
mouse adsorbed (Vector Laboratories) was added. Following 30
minutes of incubation at 37uC, samples were washed and then
treated with avidin-biotin-peroxidase complex (Vector Laborato-
ries) for 30 minutes. Labeled protein was localized by incubation
We thank J. Titus and M. Gagne for their expert technical assistance and
E. Belova for imaging data processing. R. Marini is gratefully
acknowledged for expert surgical training and veterinary oversight. We
also thank P. Dickson, D. Begley and A. Barbier for their critical review of
the manuscript and insightful comments. The authors are grateful to B.
Felice and T. McCauley for their logistical and editorial support. N. Barton
and E. Crombez are thanked for scientifically challenging us and for their
vigorous debate. Editorial support, including contributing to the first draft
of the paper, revising the paper based on author comments and styling the
paper for journal submission, was provided by L. Whetter of Zola
Conceived and designed the experiments: PC MP. Performed the
experiments: JP YH MA NL NS JL MP VB AJF. Analyzed the data: PC
MP JLP JP. Wrote the paper: PC MP JLP JP. Provided intellectual support
and conceptual advice: MWH. Critically reviewed and provided revisions
to the manuscript content: NS VB YH JL MA NL AJF MWH.
1. Abbott NJ, Patabendige AA, Dolman DE, Yusof SR, Begley DJ (2010) Structure
and function of the blood-brain barrier. Neurobiol Dis 37: 13–25.
2. Brown PD, Davies SL, Speake T, Millar ID (2004) Molecular mechanisms of
cerebrospinal fluid production. Neuroscience 129: 957–970.
3. Laterra J, Goldstein GW (2000) Ventricular organization of the cerebrospinal
fluid: blood-brain barrier, brain edema, and hydrocephalus In: Kandel E,
Schwartz JH, Jessel TM, eds. Principles of neural science. New York: McGraw-
Hill. pp 1288–1301.
4. Pardridge WM (2003) Blood-brain barrier drug targeting: the future of brain
drug development. Mol Interv 3: 90–105, 151.
5. Boado RJ, Pardridge WM (2010) Genetic engineering of IgG-glucuronidase
fusion proteins. J. Drug Target 18: 205–211.
6. Boado RJ, Hui EK, Lu JZ, Pardridge WM (2009) AGT-181: expression in CHO
cells and pharmacokinetics, safety, and plasma iduronidase enzyme activity in
Rhesus monkeys. J Biotechnol 144: 135–141.
7. Boado RJ, Zhang Y, Wang Y, Pardridge WM (2009) Engineering and
expression of a chimeric transferrin receptor monoclonal antibody for blood-
brain barrier delivery in the mouse. Biotechnol Bioeng 102: 1251–1258.
8. Boado RJ, Zhou QH, Lu JZ, Hui EK, Pardridge WM (2010) Pharmacokinetics
and brain uptake of a genetically engineered bifunctional fusion antibody
targeting the mouse transferrin receptor. Mol Pharm 7: 237–244.
9. Zhang Y, Jeong Lee H, Boado RJ, Pardridge WM (2002) Receptor-mediated
delivery of an antisense gene to human brain cancer cells. J Gene Med 4:
10. Coloma MJ, Lee HJ, Kurihara A, Landaw EM, Boado RJ, et al. (2000)
Transport across the primate blood-brain barrier of a genetically engineered
chimeric monoclonal antibody to the human insulin receptor. Pharm Res 17:
11. Barbu E, Molnar E, Tsibouklis J, Gorecki DC (2009) The potential for
nanoparticle-based drug delivery to the brain: overcoming the blood-brain
barrier. Expert Opin Drug Deliv 6: 553–565.
12. Ghafoor VL, Epshteyn M, Carlson GH, Terhaar DM, Charry O, et al. (2008)
Intrathecal drug therapy for long-term pain management. Am J Health Syst
Pharm 64: 2447–2461.
13. Belverud S, Mogilner A, Schulder M (2008) Intrathecal pumps. Neurother-
apeutics 5: 114–122.
14. Soderquist RG, Mahoney MJ (2010) Central nervous system delivery of large
molecules: challenges and new frontiers for intrathecally administered
therapeutics. Expert Opin Drug Deliv 7: 285–293.
15. Hsieh JC, Penn RD (2006) Intrathecal baclofen in the treatment of adult
spasticity. Neurosurg Focus 21: 1–6.
16. Gooch JL, Oberg WA, Grams B, Ward LA, Walker ML (2003) Complications of
intrathecal baclofen pumps in children. Pediatr Neurosurg 39: 1–6.
17. Nutt JG, Burchiel KJ, Comella CL, Jankovic J, Lang AE, et al. (2003)
Randomized, double-blind trial of glial cell line-derived neurotrophic factor
(GDNF) in PD. Neurology 60: 69–73.
18. Patel NK, Gill SS (2007) GDNF delivery for Parkinson’s disease. Acta Neurochir
Intrathecal Delivery of Lysosomal Enzymes
PLoS ONE | www.plosone.org12 January 2012 | Volume 7 | Issue 1 | e30341
19. Ballantyne JC, Carwood CM (2005) Comparative efficacy of epidural, Download full-text
subarachnoid, and intracerebroventricular opioids in patients with pain due to
cancer. Cochrane Database Syst Rev CD005178.
20. Kerr JZ, Berg S, Blaney SM (2001) Intrathecal chemotherapy. Crit. Rev Oncol
Hematol 37: 227–236.
21. Rieselbach RE, Di Chiro G, Freireich EJ, Rall DP (1962) Subarachnoid
distribution of drugs after lumbar injection. N Engl J Med 267: 1273–1278.
22. Ghersi-Egea JF, Finnegan W, Chen JL, Fenstermacher JD (1996) Rapid
distribution of intraventricularly administered sucrose into cerebrospinal fluid
cisterns via subarachnoid velae in rat. Neuroscience 75: 1271–1288.
23. Proescholdt MG, Hutto B, Brady LS, Herkenham M (2000) Studies of
cerebrospinal fluid flow and penetration into brain following lateral ventricle and
cisterna magna injections of the tracer [14C]inulin in rat. Neuroscience 95:
24. Rennels ML, Blaumanis OR, Grady PA (1990) Rapid solute transport
throughout the brain via paravascular fluid pathways. Adv Neurol 52: 431–439.
25. Rennels ML, Gregory TF, Blaumanis OR, Fujimoto K, Grady PA (1985)
Evidence for a ‘paravascular’ fluid circulation in the mammalian central nervous
system, provided by the rapid distribution of tracer protein throughout the brain
from the subarachnoid space. Brain Res 326: 47–63.
26. Meikle PJ, Hopwood JJ, Clague AE, Carey WF (1999) Prevalence of lysosomal
storage disorders. J Am Med Assoc 281: 249–254.
27. Martin R, Beck M, Eng C, Giugliani R, Harmatz P, et al. (2008) Recognition
and diagnosis of mucopolysaccharidosis II (Hunter syndrome). Pediatrics 121:
28. Elaprase [package insert] (2010). Cambridge, MA: Shire Human Genetic
29. Begley DJ, Pontikis CC, Scarpa M (2008) Lysosomal storage diseases and the
blood-brain barrier. Curr Pharm Des 14: 1566–1580.
30. Tsai SY, Markus TM, Andrews EM, Cheatwood JL, Emerick AJ, et al. (2007)
Intrathecal treatment with anti-Nogo-A antibody improves functional recovery
in adult rats after stroke. Exp Brain Res182: 261–266.
31. Kakkis E, McEntee M, Vogler C, Le S, Levy B, et al. (2004) Intrathecal enzyme
replacement therapy reduces lysosomal storage in the brain and meninges of the
canine model of MPS I. Mol Genet Metab 83: 163–174.
Sci 531: 29–39.
33. Chiro GD, Hammock MK, Bleyer WA (1976) Spinal descent of cerebrospinal
fluid in man. Neurology 26: 1–8.
34. Segal MB (2005) Fluid compartments of the central nervous system. In:
Zheng W, Chodobski A, eds. The blood-cerebrospinal fluid barrier Boca Raton,
Chapman & Hall/CRC. pp 83–99.
35. Felice BR, Wright TL, Boyd RB, Butt MT, Pfeifer RW, et al. (2011) Safety
evaluation of chronic intrathecal administration of idursulfase-IT in cynomolgus
monkeys. Toxicol Pathol 39: 879–892.
36. Garcia AR, Pan J, Lamsa JC, Muenzer J (2007) The characterization of a
murine model of mucopolysaccharidosis II (Hunter syndrome). J Inherit Metab
Dis 30: 924–934.
37. Kakkis ED (2002) Enzyme replacement therapy for the mucopolysaccharide
storage disorders. Expert Opin Investig Drugs 11: 675–685.
38. Hemsley KM, Hopwood JJ (2009) Delivery of recombinant proteins via the
cerebrospinal fluid as a therapy option for neurodegenerative lysosomal storage
diseases. Int J Clin Pharmacol. Ther 47 (Suppl 1): S118–123.
39. Hamano K, Hayashi M, Shioda K, Fukatsu R, Mizutani S (2008) Mechanisms
of neurodegeneration in mucopolysaccharidoses II and IIIB: analysis of human
brain tissue. Acta Neuropathol 115: 547–559.
40. Vedolin L, Schwartz IVD, Komlos M, Schuch A, Puga AC, et al. (2007)
Evaluation of chronic intrathecal administration of idursulfase-IT in cynomolgus
monkeys Am J Neuroradiol 28: 1029–1033.
41. Matheus MG, Castillo M, Smith JK, Armao D, Towle D, et al. (2004) Brain
MRI findings in patients with mucopolysaccharidosis types I and II and mild
clinical presentation. Neuroradiology 46: 666–672.
42. Guffon N, Bertrand Y, Forest I, Fouilhoux A, Froissart R (2009) Bone marrow
transplantation in children with Hunter syndrome: outcome after 7 to 17 years.
J Pediatr 154: 733–737.
43. Shihabuddin LS, Aubert I (2010) Stem cell transplantation for neurometabolic
and neurodegenerative diseases. Neuropharmacology 58: 845–854.
44. Passini MA, Lee EB, Heuer GG, Wolfe JH (2002) Distribution of a lysosomal
enzyme in the adult brain by axonal transport and by cells of the rostral
migratory stream. J Neurosci 22: 6437–6446.
45. Cearley CN, Wolfe JH (2007) A single injection of an adeno-associated virus
vector into nuclei with divergent connections results in widespread vector
distribution in the brain and global correction of a neurogenetic disease.
J Neurosci 27: 9928–9940.
46. Luca T, Givogri MI, Perani L, Galbiati F, Follenzi A, et al. (2005) Axons
mediate the distribution of arylsulfatase A within the mouse hippocampus upon
gene delivery. Mol Ther 12: 669–679.
47. Consiglio A, Quattrini A, Martino S, Bensdadoun JC, Dolcetta D, et al. (2001)
In vivo gene therapy of metachromatic leukodystrophy by lentiviral vectors:
correction of neuropathology and protection against learning impairments in
affected mice. Nat Med 7: 310–316.
48. Lattanzi A, Neri M, Maderna C, di Girolamo I, Martino S, et al. (2010)
Widespread enzymatic correction of CNS tissues by a single intracerebral
injection of therapeutic lentiviral vector in leukodystrophy mouse models. Hum
Mol Genet 19: 2208–2227.
49. Davidson BL, Breakefield XO (2003) Viral vectors for gene delivery to the
nervous system. Nat Rev Neurosci 4: 353–364.
50. Vite CH, Passini MA, Haskins ME, Wolfe JH (2003) Adeno-associated virus
vector-mediated transduction in the cat brain. Gene Ther 10: 1874–1881.
51. Wu X, Burgess SM (2004) Integration target site selection for retroviruses and
transposable elements. Cell Mol Life Sci 61: 2588–2596.
52. Woods NB, Muessig A, Schmidt M, Flygare J, Olsson K, et al. (2003) Lentiviral
vector transduction of NOD/SCID repopulating cells results in multiple vector
integrations per transduced cell: risk of insertional mutagenesis. Blood 101:
53. Hematti P, Hong BK, Ferguson C, Adler R, Hanawa H, et al. (2004) Distinct
genomic integration of MLV and SIV vectors in primate hematopoietic stem
and progenitor cells. PLoS Biol 2: 2183–2190.
54. Nakai H, Montini E, Fuess S, Storm TA, Grompe M, et al. (2003) AAV serotype
2 vectors preferentially integrate into active genes in mice. Nat Genet 34:
55. Kar S, Poirier J, Guevara J, Dea D, Hawkes C, et al. (2006) Cellular distribution
of insulin-like growth factor-II/mannose-6-phosphate receptor in normal human
brain and its alteration in Alzheimer’s disease pathology. Neurobiol Aging 27:
56. Hawkes C, Kar S (2002) Insulin-like growth factor-II/Mannose-6-phosphate
receptor in the spinal cord and dorsal root ganglia of the adult rat. Eur J Neurosci
57. Chen F, Vitry S, Hocquemiller M, Desmaris N, Ausseil J, et al. (2006) Alpha-L-
Iduronidase transport in neurites. Mol Genet Metab 87: 349–358.
58. Muenzer J, Wraith JE, Beck M, Guigliani R, Harmatz P, et al. (2006) A phase
II/III clinical study of enzyme replacement therapy with idursulfase in
mucopolysaccharidosis II (Hunter Syndrome). Genet Med 8: 465–473.
59. Loening AM, Gambhir SS (2003) AMIDE: a free software tool for multimodality
medical image analysis. Mol Imaging 2: 131–137.
60. Hudson HM, Larkin RS (1994) Accelerated image reconstruction using ordered
subsets of projection data. IEEE Trans Med Imaging 13: 601–609.
61. Qi J, Leahy RM, Cherry SR, Chatziioannou A, Farquhar TH (1998) High-
resolution 3D Bayesian image reconstruction using the microPET small-animal
scanner. Phys Med Biol 43: 1001–1013.
62. Defrise M, Kinahan PE, Townsend DW, Michel C, Sibomana M, et al. (1997)
Exact and approximate rebinning algorithms for 3-D PET data. IEEE Trans
Med Imaging 16: 145–158.
63. Papisov M, Belov V, Fischman A, Bonab A, Titus J, et al. (2010) PET imaging of
protein pharmacokinetics in the CSF of rats and monkeys with I-124. J Nucl
Med 51 (Suppl 2): 1751.
64. Voznyi YV, Keulemans JL, van Diggelen OP (2001) A fluorimetric enzyme
assay for the diagnosis of MPS II (Hunter disease). J Inherit Metab Dis 24:
65. Garcia AR, Pan J, Lamsa JC, Muenzer J (2007) The characterization of a
murine model of mucopolysaccharidosis II (Hunter syndrome). J Inherit Metab
Dis 30: 924–934.
Intrathecal Delivery of Lysosomal Enzymes
PLoS ONE | www.plosone.org13 January 2012 | Volume 7 | Issue 1 | e30341