www.bba-direct.com

Biochimica et Biophysica A

Review

Consequences of NPC1 and NPC2 loss of function in mammalian neurons

Steven U. Walkleya,*, Kinuko Suzukib

aSidney Weisner Laboratory of Genetic Neurological Disease Department of Neuroscience, Rose F. Kennedy Center for Research in Mental Retardation and

Human Development, Albert Einstein College of Medicine, Bronx, NY 10461, United StatesbDepartment of Pathology and Laboratory Medicine, The University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, United States

Available online 8 September 2004

Abstract

Genetic deficiency of NPC1 or NPC2 results in a devastating cholesterol-glycosphingolipidosis of brain and other organs known as

Niemann–Pick type C (NPC) disease. While NPC1 is a transmembrane protein believed involved in retroendocytic shuttling of substrate(s) to

the Golgi and possibly elsewhere in cells as part of an essential recycling/homeostatic control mechanism, NPC2 is a soluble lysosomal

protein known to bind cholesterol. The precise role(s) of NPC1 and NPC2 in endosomal–lysosomal function remain unclear, nor is it known

whether the two proteins directly interact as part of this function. The pathologic features of NPC disease, however, are well documented.

Brain cells undergo massive intracellular accumulation of glycosphingolipids (lactosylceramide, glucosylceramide, GM2 and GM3

gangliosides) and cholesterol and concomitant distortion of neuron shape (meganeurite formation). In neurons from humans with NPC

disease the metabolic defects and storage often lead to extensive growth of new, ectopic dendrites (possibly linked to ganglioside

sequestration) as well as formation of neurofibrillary tangles (NFTs) (possibly linked to dysregulation of cholesterol metabolism). Other

features of cellular pathology in NPC disease include fragmentation of the Golgi apparatus and neuroaxonal dystrophy, though reasons for

these changes remain largely unknown. As the disease progresses, neurodegeneration is also apparent for neurons in some brain regions,

particularly Purkinje cells of the cerebellum, but the basis of this selective neuronal vulnerability is unknown. The NPC1 protein is

evolutionarily conserved with homologues reported in yeast to humans; NPC2 is reported in C. elegans to humans. While neurons in

mammalian models of NPC1 and NPC2 diseases exhibit many changes that are remarkably similar to those in humans (e.g., endosomal/

lysosomal storage, Golgi fragmentation, neuroaxonal dystrophy, neurodegeneration), a reduced degree of ectopic dendritogenesis and an

absence of NFTs in these species suggest important differences in the way lower mammalian neurons respond to NPC1/NPC2 loss of

function.

D 2004 Elsevier B.V. All rights reserved.

Keywords: Niemann–Pick type C; Endosome; Lysosome; Axonal spheroid; Dendritogenesis; Neurodegeneration; Neurofibrillary tangle; Golgi fragmentation

1. Introduction

Niemann–Pick type C (NPC) disease is a neuro-

visceral lipid storage disorder originally classified along

with Niemann–Pick types A, B and D on the basis of

clinical symptoms and biochemical features [1,2]. Nie-

mann–Pick disorders were recognized as disease entities

in the early part of the 20th century but shared

numerous similarities with Gaucher disease and other

1388-1981/$ - see front matter D 2004 Elsevier B.V. All rights reserved.

doi:10.1016/j.bbalip.2004.08.011

* Corresponding author. Tel.: +1 7184304025; fax: +1 7184308821.

E-mail addresses: walkley@aecom.yu.edu (S.U. Walkley)8

kis@med.unc.edu (K. Suzuki).

so-called storage disorders and nosological confusion

was common. The insightful definition of storage

disorders as conditions caused by inherited deficiencies

of specific lysosomal hydrolases by Hers in 1965 [3] led

to the subsequent discovery of an acid sphingomyelinase

defect in Niemann–Pick types A and B [4]. Types C and

D, however, were left in limbo for an additional three

decades as a result of the failure to find a similar

lysosomal hydrolase deficiency that could account for

storage. The discovery by Pentchev and colleagues in

the early 1980s that cholesterol esterification was

defective in cultured fibroblasts from NPC patients

provided an important diagnostic tool for this disease.

This discovery also led to the view that NPC disease

cta 1685 (2004) 48–62

S.U. Walkley, K. Suzuki / Biochimica et Biophysica Acta 1685 (2004) 48–62 49

was primarily a cholesterol lipidosis [5], further splitting

type C disease from types A and B in which

sphingomyelin was the primary storage material. The

enigma of NPC disease was resolved for the majority of

cases in 1997 with the cloning of the defective gene and

the discovery of the NPC1 protein [6]. NPC1 was found

to be an integral membrane protein consisting of 1278

amino acids and 13–16 membrane-spanning domains.

NPC1 showed high homology with Patched, a critical

element in the sonic hedgehog signaling pathway, as

well as with proteins involved in cholesterol homeo-

stasis, including 3-hydroxy-3-methylglutaryl coenzyme A

(HMG-CoA) reductase, and with SCAP, the sterol

regulatory element-binding protein (SREBP)-cleavage-

activating protein (SCAP). The numerous links between

NPC1 and possible functions involving cholesterol

reinforced the view that the disorder was a primary

cholesterol lipidosis. This idea, however, could not be

substantiated in recent studies implicating NPC1 as a

eukaryotic version of a prokaryotic bpermeaseQ as in this

experimental system the NPC1 protein did not appear

capable of transporting cholesterol [7]. Further, recent

studies of the NPC1 homologue in yeast (NCR-1) have

suggested that the primordial role of this protein is

associated with subcellular sphingolipid distribution [8].

Interestingly, NPC disease has long been known to

exhibit massive storage of gangliosides and other

glycolipids in brain and, recently, it was shown that

the accumulation of unesterified cholesterol in neurons

appears largely dependent on the types of complex

gangliosides expressed [9], once again raising significant

questions as to the exact function of the NPC1 protein

as well as the identity of the primary storage material in

NPC1 disease.

While type D Niemann–Pick disease was found to be

allelic with type C1 [6], a small percentage of cases

(c5%) with remarkable similarity to NPC1 disease was

found to nonetheless have normal expression of NPC1

protein. This led to the discovery of NPC2, a second

defective protein and one with identity to HE1, a

cholesterol binding protein [10]. Interestingly, the sug-

gested linkage between NPC2 and cholesterol (see also

Ref. [11]) once again puts issues of cholesterol trafficking

and metabolism at center stage in terms of understanding

NPC disease. While it is not presently known whether

NPC1 and NPC2 proteins directly interact in cells, recent

studies on a murine NPC1–NPC2 double mutant are

consistent with direct functional co-operativity between

the two proteins [12].

In addition to human NPC1 disease, feline and murine

models of NPC disease are also known, both of which were

discovered after occurring spontaneously in normal cat and

mouse populations [13–15]. Research colonies for both

species have been established. Basic pathological features of

these animal models closely resemble those observed in

human cases of NPC disease and each model also exhibits

clinically evident neurological dysfunction. Mutations of the

NPC1 gene have also been produced in Drosophila [K.

Bhat, personal communication] and in C. elegans [16], but

detailed structural changes in neurons and the CNS of these

simpler organisms are not yet well elucidated. This review

therefore is focused on the consequences of NPC1 and

NPC2 loss of function as they occur in human, cat and

mouse neurons.

2. NPC1 and NPC2 loss of function in humans

NPC disease is clinically heterogeneous in humans,

and various phenotypes have been documented in infants

to adults [2,17,18]. Recent studies of the molecular

genetics of NPC1 have revealed functional significance

of the sterol-sensing domain in NPC1 protein and

showed that the absence of NPC1 protein can be

correlated with the most severe neurological form of

the disease which shows infantile onset and demonstra-

ble (severe) cholesterol metabolic defects [19]. Many of

the NPC1 disease clinical variants, in contrast, were

found associated with missense mutations in the

cysteine-rich luminal loop. Not surprisingly, the hetero-

geneity of NPC1 disease in terms of clinical features

and mutational characteristics is further reflected in CNS

and visceral pathology. Hepatosplenomegaly with an

accumulation of cholesterol, sphingomyelin and some

glycolipids is prominent in infantile cases, but in the

majority of juvenile or adult cases, hepatomegaly is not

a usual pathological feature. In these later onset cases,

neurodegeneration with formation of neurofibrillary

tangles, in addition to lipid storage, is the usual

pathological feature. For NPC2 cases, the number of

families with affected individuals has been too few for

detailed genotype–phenotype studies but overall the

same clinical heterogeneity characterizing NPC1 disease

also applies [18]. One recent report described adult

onset-NPC2 disease in which neuronal storage and

cognitive impairment were identified but storage in

viscera and bone marrow were absent [20]. Despite this

heterogeneity, there are numerous morphological alter-

ations that appear in common across the full spectrum

of NPC disease [21]. These include neuronal storage,

meganeurite formation and ectopic dendritogenesis, axo-

nal alterations in the form of spheroids, neurofibrillary

tangles (NFTs) and other cytoskeletal alterations, frag-

mentation of Golgi apparatus and neurodegeneration (loss

of neurons).

2.1. Neuronal storage

Neuronal storage is consistently observed in NPC disease

in humans, although enlarged, swollen neurons are more

conspicuous in early-onset than late-onset cases. On routine

paraffin sections stained with hematoxylin and eosin (H and

S.U. Walkley, K. Suzuki / Biochimica et Biophysica Acta 1685 (2004) 48–6250

E), the perikarya of affected neurons are packed with fine

vacuoles. On frozen section preparation, the neuronal

storage material reacts strongly with Periodic Acid Schiff

(PAS) and to some extent with filipin, consistent with the

presence of glycolipids and cholesterol, respectively, in

neuronal perikarya. Biochemical analysis of NPC disease-

affected brains has revealed increases of glycolipids

(glucosylceramide, lactosylceramide, gangliosides GM3

and GM2) but similar increases of total cholesterol content

have not been reported [22]. In situ analysis of gangliosides

(GM2, GM3) and unesterified cholesterol using immuno-

Fig. 1. Mammalian neurons deficient in NPC1 accumulate GM2 and GM3 gangl

normal brain. (A) Normal adult mouse brain stained with an antibody to GM2 gan

nearby glial cell. (B–C) Mouse and cat cortical neurons in NPC1 disease, respec

(arrows). (D–E) Histochemical (filipin) staining of cerebral cortex from normal m

accumulation in neuronal perikarya (arrows) and in enlarged axon hillocks/megan

the presence of GM3 ganglioside in perikarya (arrows) and in meganeurites (arro

cytochemistry and histochemical (filipin) staining, respec-

tively, has documented that these materials accumulate

directly within neuronal perikarya [23] (Fig. 1). These

storage materials are identified at the ultrastructural level in

neurons as unique inclusions consisting of concentrically

arranged lamellae resembling membranous cytoplasmic

bodies of gangliosidosis [24], although smaller and less

compact, and in some cases showing multivesiculated

profiles, hence the term polymorphous cytoplasmic bodies

[2]. Similar inclusions are found in glial cells of brain as

well as in hepatocytes of liver.

iosides and free cholesterol, none of which occurs at conspicuous levels in

glioside showing its minimal presence. Inset in A: higher magnification of

tively, showing the presence of GM2 ganglioside accumulation in neurons

ouse (D) and NPC1�/� mouse (E), with only the latter showing cholesterol

eurites (arrowheads). (F) Cortical neurons in human NPC1 disease showing

wheads). Calibration bar in F equals 25 Am for all.

S.U. Walkley, K. Suzuki / Biochimica et Biophysica Acta 1685 (2004) 48–62 51

2.2. Meganeurite formation and ectopic dendritogenesis

The use of the Golgi staining method to evaluate

neuronal morphology in a cortical biopsy from a child with

an undiagnosed and progressive form of mental retardation

in 1975 led to a remarkable discovery. Golgi impregnations

of these cortical neurons revealed the presence of elongated

swellings at axon hillocks from which projected dendritic

spines as well as longer, processes which themselves

possessed spines [25]. The swelling at the axon hillock

came to be known as a meganeurite, and the associated

abnormal dendritic spines and processes as bectopicdendritogenesisQ, due to its peculiar and inappropriate

location on the cell. Thus it appeared that these postnatal,

mature neurons were undergoing a recapitulation of

dendritogenesis normally seen only in early development,

and as a result an entirely new basilar dendritic system was

being generated below the normal basilar dendrites of the

perikaryon. The child in this case was subsequently

diagnosed with a rare form of GM2 gangliosidosis (AB

variant) [26] in which the degradative enzyme (h-hexosa-minidase) for GM2 ganglioside was normal, but failed to

function as a consequence of a deficiency in its activator

protein [27]. Meganeurites and accompanying neuritic and

synaptic changes were subsequently reported in a variety of

storage diseases in children, including NPC disease [28,29].

Here meganeurites were often enormous in size, dwarfing

adjacent neuronal perikarya, and exhibiting elaborate

dendritic processes complete with normal-appearing spines

(Fig. 2).

The discovery of ectopic dendritogenesis in a primary

ganglioside storage disease was believed to indicate that

gangliosides contributed in some manner to dendritic

growth regulatory mechanisms and that the resulting altered

Fig. 2. Camera lucida reconstructions of Golgi-impregnated layer III

cortical pyramidal neurons from (A) feline (10 weeks), (B) human (3.5

years) and (C) murine (10 weeks) NPC1 disease. A and B are from

neocortex, C is from the neocortical–piriform cortex junction. Note the

presence of an enlarged, dendritic spine/neurite covered axon hillock in A

(arrows), extensive dendritic neurites on a large meganeurite in B (arrows),

and a meganeurite with ectopic spines in C (arrows). Arrowheads indicate

axons. Calibration bar equals 25 AM.

synaptic connectivity was a basis for brain dysfunction,

most notably mental retardation [25,30]. However, it

required the discovery and analysis of storage diseases in

large animal models (see below) to fully address the

underlying reasons for this renewal of dendritogenesis in

lysosomal disorders. To this day the phenomenon of ectopic

dendritogenesis remains a finding unique to disorders of the

endosomal–lysosomal system [31].

2.3. Formation of axonal spheroids

Axonal spheroids [neuroaxonal dystrophy (NAD)] are

focal swellings of axons, indicating focal sites of damage.

They can be found following acute axonal injury such as

Wallerian degeneration as well as in varieties of chronic

neurodegenerative disorders such as infantile neuroaxonal

dystrophy and Sandhoff disease. However, as pointed out by

Elleder et al. [32], formation of axonal spheroids in early-

onset NPC disease is a very conspicuous pathological

feature. On routine H and E preparations spheroids appear

as homogenous or finely granular focal axonal swellings

occurring throughout the brain, in particular in the cerebral

and cerebellar white matter, brainstem and posterior

columns of the spinal cord. At the ultrastructural level, an

accumulation of 10-nm filaments is found within the

homogeneous spheroids whereas granular spheroids are

composed of masses of apparent degenerated organelles,

pleiomorphic bodies and amorphous dense substances (Fig.

3). Both patterns have been reported as mixed to various

proportions in human NPC disease [32]. The ultrastructural

features of spheroids are similar to those of dystrophic

axons described in various neurological disorders, including

other storage disorders, as well as in experimental studies

[33,34]. Axonal spheroids are not as numerous in late-onset

cases of NPC disease, in which neurofibrillary tangles are

conspicuous, and it appears that axonal spheroids have

inverse relation to the presence of neurofibrillary tangles

[35]. Spheroids in human NPC disease have also been found

to express positive immunoreactivity with antibodies to

ubiquitin and phosphorylated neurofilaments.

2.4. Neurofibrillary tangles (NFTs)

Neurofibrillary tangles consist of paired helical filaments

(PHFs) formed by phosphorylated microtubular protein,

PHFtau, and are considered to be one of the hallmarks in

Alzheimer disease (AD). In NPC disease, NFTs are found in

almost all cases of juvenile/adult cases, although distribu-

tion and numbers of NFT-expressing neurons vary consid-

erably case by case [35,36]. The youngest NPC case with

NFTs so far reported in the literature was a 10-year-old child

[37] but more recently one of us (KS) has personally

observed a few NFTs in the cerebral cortex of a 3-year 10-

month-old patient who was previously reported as negative

[37]. Immunocytochemically and also by Western blot

analysis using a number of different monoclonal antibodies

S.U. Walkley, K. Suzuki / Biochimica et Biophysica Acta 1685 (2004) 48–6252

(T14, PHF1, AT180, AT8, AT270, 12E8) that bind to AD-

brain-derived PHFtau, PHFtau in NPC and AD brains was

indistinguishable, suggesting similar mechanisms may play

a role in formation of NFTs [38]. Ultrastructural features of

NFTs in NPC disease consist of PHFs identical to those of

Alzheimer disease (Fig. 4). Apolipoprotein (Apo) E is a

major cholesterol transport molecule in the brain and in

Fig. 4. Human NPC1 disease exhibits the presence of NFTs. (A)

Bielschowsky silver stain showing the presence of cerebral cortical neurons

with tangles (arrows), and a comparable section (C) showing that these

neurons are also immunoreactive for Alz50 (arrows). (B) Ultrastructure of

twisted PHFs in affected neurons (arrows). Calibration bar in C equals 50

Am for A, 0.15 Am for B, and 40 Am for C.

Fig. 3. Ultrastructure of spheroids in myelinated axons of the CNS from

feline (A), human (B) and murine (C) NPC1 disease. Note the presence of

tubulovesicular profiles and dense bodies in each. B also shows a mass of

electron dense granular material, possibly degenerating neurofilaments, and

in C there is an accumulation of inclusions in the distal process of an

oligodendrocyte. (�12,000).

Alzheimer disease Apo Ee4 allele is associated with an

earlier onset of tangle formation and increased tangle load.

Dist et al. [39] recently reported that mean levels of free

cholesterol in tangle-bearing neurons were higher than in

tangle-free neurons. Tangles are a particularly conspicuous

feature of NPC disease patients who are homozygous for the

ApoE q4 allele, in whom deposition of h-amyloid is also

detectable [40]. In one of those cases, Lewy bodies were

detected in the substantia nigra. The Lewy bodies are the

pathological hallmark of Parkinson disease and dementia

with Lewy bodies, and the major protein constituent is a-

synuclein. More recently, however, Lewy bodies have been

reported in Alzheimer disease [41–43] and in other

tauopathies [44]. Saito et al. [45] have found expression

of phosphorylated a-synuclein as well as tau in neurons in

many juvenile NPC cases. In some neurons these two

proteins were co-localized. Co-localization of tau and a-

synuclein epitopes has been reported in cases of dementia

Fig. 5. The Golgi apparatus in NPC1�/� neurons shows fragmentation

when stained with the MG160 antibody. (A, C) Normal cortical pyramidal

neuron (human) and cerebellar Purkinje cell (mouse), respectively,

illustrating normal morphology of the Golgi apparatus. (B, D) NPC1-

deficient cortical neurons (human) and cerebellar Purkinje cell (mouse),

respectively, showing fragmented appearance in disease. Calibration bar in

A equals 25 Am and applies to each.

S.U. Walkley, K. Suzuki / Biochimica et Biophysica Acta 1685 (2004) 48–62 53

with Lewy bodies [46]. Thus, NPC neurons appear to

share common mechanisms of neurodegeneration with

other brain disorders. Recently, NFTs and axonal sphe-

roids were reported to occur in an NPC patient with a

mutation in NPC2 [20]. NFTs have never been detected in

the cerebellum of NPC patients, or indeed even in the

cerebellum of individuals with AD, although several early

markers of NFT formation such as hyperphosphorylated

tau and altered cell cycle regulators have been detected

immunocytochemically in the cerebellum of NPC1

patients [47].

2.5. Fragmentation of Golgi apparatus

Cholesterol is an essential structural component of the

plasma membrane as well as of intracellular membranes

such as endosomes, lysosomes, Golgi and trans-Golgi

network (TGN). In NPC patients (as well as the NPC1

mouse model, as described later), we have found, by using

immunocytochemical staining, apparent fragmentation of

the Golgi/TGN in cerebellar Purkinje cells and cerebral

cortical neurons (Fig. 5). The antibody used for these studies

recognizes a conserved membrane sialoglycoprotein of the

medial cistern of the Golgi apparatus, MG160 [48–50]

[unpublished data (KS)]. Fragmentation of the Golgi

apparatus was first demonstrated in anterior horn cells and

later in Betz cells in amyotrophic lateral sclerosis (ALS),

and in neurons in AD [51]. Fragmentation of Golgi

apparatus has been also detected in the mouse models of

ALS prior to the onset of clinical signs [52]. Similar

fragmentation has been recognized following depolymeri-

zation of microtubules following drug treatment [53] or

physiological disassembly of microtubules during mitosis

[54,55]. Thus, the fragmentation of the Golgi apparatus is

unlikely to represent a specific pathological process but may

be related to a more general degenerative process such as,

for example, disease-induced disorganization of micro-

tubules. Its presence/absence in other forms of neuronal

storage disorders has not yet been explored.

2.6. Neurodegeneration (loss of neurons)

The brains of patients with end-stage NPC disease are

often atrophic, suggesting that there has been considerable

loss of neuronal elements. Rare TUNEL-positive neurons

can be found in the brain of NPC disease patients

[unpublished observation (KS)] but necrotic/dying neurons

as seen in cerebral ischemia are not reported. How neurons

degenerate and what causes neurodegeneration are not well

defined in NPC or in other storage disorders. Like many

other chronic neurodegenerative disorders, apoptotic death

is a likely possibility and further investigation is needed to

understand the mechanism of neuronal death in these

disorders. As in many other chronic neurodegenerative

diseases of humans, it is difficult to assess the mechanism of

neuronal death in the postmortem brains because of many

factors such as postmortem duration, status of terminal

disease state, and so forth. Well-controlled studies with

animal models are needed.

3. NPC1 loss of function in cats

Five surviving kittens from a litter of seven were donated

to the New York State College of Veterinary Medicine in

1988 due to the presence of unusual neurological symptoms

in two of the animals. Subsequent analysis revealed the

presence of a neurovisceral storage disease resembling NPC

in the affected cats and cultured fibroblasts exhibited

decreased ability to esterify exogenous cholesterol [13]. A

colony of these animals was established by heterozygote

breeding and additional affected kittens have been success-

fully produced for a variety of studies. Following identi-

fication of the NPC1 gene in the human and mouse forms of

NPC disease, complementation analysis of the human and

feline disorders revealed involvement of the same gene [56],

Fig. 6. Ultrastructure of storage bodies in feline (A) and murine (B) NPC1-

deficient neurons. Note the presence of polymorphous and membranous

inclusions consistent with glycolipid storage. (�21,000).

S.U. Walkley, K. Suzuki / Biochimica et Biophysica Acta 1685 (2004) 48–6254

thus documenting the validity of this important large animal

model.

Cats deficient in NPC1 exhibit striking clinical signs

[29,57]. Subtle intention tremors first appear at 8–12

weeks of age. Motor signs become progressively con-

spicuous over the ensuing weeks and involve truncal

ataxia, head tremor and dysmetria. Ambulation is compro-

mised by 4–5 months of age although animals remain alert

and interested in their environment. By 6–7 months the

animals spend more time in sternal recumbency and

exhibit severe head and neck tremors. Positional nystag-

mus and opisthotonus are evident by 9–10 months of age.

Few animals survive beyond 11 months.

3.1. Neuronal storage

Cats with NPC disease exhibit accumulation of excess

lactosylceramide, glucosylceramide, phospholipids, and

unesterified cholesterol in liver and GM2 and GM3

gangliosides in brain [13,23]. Storage within the CNS is

extensive and essentially involves all neurons and other cell

types beginning during the early postnatal period. Ganglio-

side accumulation in neurons has been documented using

antibodies to GM2 and GM3, and storage, particularly of

GM2, is a near-universal feature of the disease in neurons

(Fig. 1). Storage bodies are distributed throughout the

perikarya of neurons and often extend out into proximal

dendrites. Occasionally, large collections of storage inclu-

sions can be found in dendritic profiles. Storage in some

types of neurons typically generates the formation of

meganeurites at axon hillocks. Ultrastructurally, storage

material in neurons consists of heterogeneous populations of

vacuoles exhibiting loose membranous swirls and stacks as

well as abnormal-appearing multi-vesicular profiles. As

such, they appear remarkably similar to the polymembra-

nous cytoplasmic bodies described in human NPC disease

(above) and in the mouse models (Fig. 6). The presence of

membranous storage materials is consistent with the

presence of significant glycolipid storage.

3.2. Meganeurites and ectopic dendritogenesis

The use of Golgi staining and other morphological

techniques in cats with NPC1 disease, as well as in a variety

of other types of neuronal storage disorders [Niemann–Pick

type A, mucopolysaccharidosis (MPS) type I, a-mannosi-

dosis, GM1 and GM2 gangliosidosis], has revealed that

meganeurite formation and ectopic dendritogenesis are a

widespread feature of lysosomal diseases [58]. Meganeur-

ites have been found to occur just proximal to the axonal

initial segment and therefore represent deformations of the

axon hillock/lower perikarya region of neurons. This

finding, coupled with ultrastructural differences, established

that meganeurites and axonal spheroids are distinctly

different structures [59]. Meganeurites themselves are found

to either be covered with dendritic spines and ectopic

dendrites (as seen in the disorders listed above) or to be

smooth-surfaced and to lack new spines, neurites or

synapses. The presence of the latter so-called aspiny

meganeurites on some types of neurons and in some

diseases (e.g., the Batten disorders or neuronal ceroid

lipofuscinoses) has been interpreted to indicate that mega-

neurites themselves are volume-accommodators following

intraneuronal storage, whereas the growth of new dendritic

S.U. Walkley, K. Suzuki / Biochimica et Biophysica Acta 1685 (2004) 48–62 55

membrane is likely due to a disturbance in some specific

metabolic process controlling dendritic initiation [58]. For

some neurons undergoing storage, particularly in feline

models of storage diseases (including NPC1 disease),

intraneuronal storage appears to be insufficient to generate

prominent meganeurites but axon hillocks nonetheless

exhibit striking growth of new dendritic processes (Fig.

2). Ectopic dendrites have been documented as forming

synaptic contacts with local axons, with most appearing

asymmetrical in form and therefore likely excitatory in

function [60].

All lysosomal diseases exhibiting ectopic dendritogene-

sis have revealed near identical patterns of involvement,

with the same cell types (cortical pyramidal neurons,

multipolar cells of amygdala and claustrum) appearing

vulnerable to this change in each disease. The presence of

ectopic dendrites and altered connectivity in cortical and

subcortical and limbic system structures is compatible with

this change contributing to both cognitive and behavioral

deterioration in affected individuals. Overall, when compar-

ing different storage diseases in the feline models, only the

relative frequency of neurons in a given brain area varied,

e.g., from 95% of cortical pyramidal neurons in GM2

gangliosidosis to 40% in Niemann–Pick type C and to

smaller percentages in non-primary glycosphingolipidoses

[61]. This was believed to argue that features intrinsic to

certain types of neurons, coupled with specific metabolic

abnormalities in common between different storage dis-

eases, likely were coalescing to generate this unusual

pathological feature. The availability of NPC and other

storage diseases in a single species (i.e., cats) has allowed

for systematic analysis of the hypothesis that ectopic

dendritogenesis was being triggered by abnormal sequestra-

tion of gangliosides. These studies revealed that one

particular ganglioside (GM2) was consistently present as

part of the lysosomal storage process in all neurons

characterized by ectopic dendritic sprouting [61]. These

studies, coupled with related work showing heightened

elevation of this same ganglioside during early development

in normal brain when pyramidal neurons sprout dendrites,

led to the hypothesis that GM2 ganglioside is functionally

allied with mechanisms controlling dendritogenesis in both

normal and storage disease-affected pyramidal neurons

[62–64].

3.3. Formation of axonal spheroids

Neuroaxonal dystrophy, as described for human NPC

disease above, is also a major feature of the feline model.

Indeed, many of the storage diseases found in this species

[GM1 and GM2 gangliosidosis (Sandhoff), Niemann–Pick

disease type A, and a-mannosidosis] exhibit this phenom-

enon in abundance whereas only a few (e.g., mucopoly-

saccharidosis type I) do not [58]. Ultrastructurally, most

spheroids in cats appear granular in nature and are found in

both gray and white matter regions of brain. While they are

evident as PAS-negative profiles in PAS-stained plastic

sections of brain, their frequency of occurrence typically

appears unimpressive unless labeled by specific antibodies.

For example, labeling with antibodies to glutamic acid

decarboxylase (GAD), the synthetic enzyme for the

inhibitory neurotransmitter, g-aminobutyric acid (GABA),

revealed that spheroids were remarkably abundant and

specifically affected GABAergic neurons in many brain

regions [65]. Only rarely were spheroids labeled signifi-

cantly with antibodies to neurofilament proteins, a finding

consistent with the paucity of neurofilaments in most

spheroids in this species.

The overall distribution and incidence of axonal sphe-

roids in different regions of the CNS was found to closely

correlate with the type and severity of clinical neurological

signs in affected cats. Cats with NPC1 disease were found to

exhibit spheroids in all areas of brain containing GABAer-

gic neurons, including cerebral cortex, basal ganglia,

cerebellum and brainstem. Spheroids on cerebellar Purkinje

cells were particularly evident when labeled with antibodies

specific for this type of neuron (calbindin, parvalbumin) as

well as with GAD [29]. Spheroids in Purkinje cell axons in

NPC disease appeared to form earliest in distal axons and

over time enlarged in size and became more conspicuous in

proximal portions of axons. Storage disorders in cats that

lacked significant spheroid formation, such as MPS I, also

lacked significant neurological signs (e.g., motor system

dysfunction) in spite of widespread neuronal storage, again

consistent with spheroid formation being a major generator

of clinical disease.

As mentioned above, spheroids in all types of storage

diseases consisted of essentially identical accumulations of

tubulovesicular and multivesicular profiles, dense bodies,

and mitochondria (Fig. 3). They rarely contained significant

amounts of neurofilamentous material, as is characteristic

of spheroids in toxic neuropathies and some human storage

diseases (see earlier discussion). Importantly, they also do

not contain tertiary lysosomes resembling the storage

bodies within neuronal perikarya, which differ disease by

disease according to the specific lysosomal hydrolase that

is defective. Axonal spheroids in feline models of storage

diseases also did not appear to be bretraction bulbsQ of

dying axons. Rather, they are swellings of significant size

that occur along the length of an axon with axonal

continuities clearly visible on both proximal and distal

sides [65]. These structural characteristics of spheroids

suggest that they may represent defects in axoplasmic

transport. Accumulations of similar heterogeneous organ-

elles have been reported to occur distal to crush or low

temperature lesions in axons of experimental animals,

indicating that these types of materials are characteristic

of a block in retrograde transport [66]. More recent studies

using murine NPC disease, reviewed below, suggest that

defects in cholesterol transport in axons and/or abnormal-

ities in oligodendroglia may play a role in the generation of

axonal spheroids.

S.U. Walkley, K. Suzuki / Biochimica et Biophysica Acta 1685 (2004) 48–6256

3.4. Cytoskeletal abnormalities and Golgi fragmentation

Cats with NPC1 disease have been evaluated for NFTs

but the phenomenon has not been observed in this species

[T. Mitchell and colleagues, personal communication].

Studies designed to analyze the integrity of the Golgi/

TGN have not yet been carried out in cats.

3.5. Neurodegeneration (loss of neurons)

Studies of the feline model of NPC1 disease have not

reported significant death of neurons in the CNS of affected

animals apart from that of cerebellar Purkinje cells.

However, systematic assessment of this issue has not been

rigorously pursued and, given the significant amount of

neuroaxonal dystrophy found throughout the CNS (as a

potential forerunner of neuronal demise), it is conceivable

that populations of neurons other than just Purkinje cells are

undergoing premature degeneration. In the cerebral cortex

there is an indication for thinning of the gray matter and,

interestingly, the incidence of ectopic dendritogenesis on

cortical pyramidal neurons appears less over time (unlike

GM1 and GM2 gangliosidosis in cats in which the incidence

increases significantly over the first year of life [67]). One

explanation for reduced incidence of these cells in NPC

disease, given that ectopic dendrites themselves may not,

once formed, be a reversible feature [68], is that they are

dying. Cerebellar Purkinje cell death itself is both con-

spicuous and extensive in NPC disease. Axonal spheroid

formation is the earliest abnormality seen on these cells

(apart from intraneuronal storage) and is extensive in more

proximal axonal regions by 5 months of age, a time when

Purkinje cell loss could be clearly documented [29]. By 7

months of age Purkinje cell loss was extensive but spheroid

formation less common, presumably because most Purkinje

cells and their axons by this stage had degenerated and been

removed from the tissue.

4. NPC1 and NPC2 loss of function in mice

Two murine models of NPC1 disease, BALB/c npcnih

(NPC1�/�) and C57BL/KsJ (sphingomyelinosis or SPM)

mice, are known [14,15]. In NPC1�/�, the murine ortholog

of NPC1 is mutated [69]. SMP mice are found to be allelic

with the NPC1�/� mice by crossbreeding experiments [70]

but the exact gene mutation has not yet been identified.

Recently, Sleat et al. [12] have reported generating a murine

NPC2 hypomorph that expresses 0–4% residual protein in

different tissues. In examining its phenotype, it was found to

closely resemble the NPC1�/� mouse.

Pathological changes in the NPC1 and NPC2 murine

models resemble those observed in late infantile NPC

disease in humans. The BALB/c npcnih (NPC1�/�) mice

were initially reported as a lysosomal lipid storage disorder

characterized by excessive tissue deposition of cholesterol

and phospholipid, largely in sphingomyelin content [14].

Major neuropathological changes in these mice are neuronal

storage, axonal spheroids, hypomyelination and loss of

neurons (neurodegeneration). Unlike human cases, how-

ever, no NFTs have been reported in the mouse models of

NPC. Hypomyelination is reported in the corpus callosum

and anterior commissure [71] but myelination is well

advanced in the internal capsules, cerebellar white matter,

brain stem and spinal cord. Myelin degeneration becomes

apparent after 60 days of age, possibly secondary to axonal

degeneration [72].

4.1. Neuronal storage

Finely granular storage materials are seen in the neuronal

perikarya throughout the central and peripheral nervous

system of both the NPC1 and NPC2 mouse models.

Biochemical analysis of brain has revealed marked accu-

mulation of GM2 and GM3 gangliosides and several neutral

glycolipids [12,71,73]. The localization of GM2 and GM3

gangliosides in storage neurons has been demonstrated

immunocytochemically (Fig. 1) [12,23]. Interestingly, when

double-labeled with fluorescent markers and examined by

confocal microscopy, the GM2 and GM3 gangliosides were

found to largely occur in separate vesicle populations,

possibly suggesting independent mechanisms to account for

their storage [74]. Intracellular storage in NPC mice is most

conspicuous in large pyramidal neurons, Purkinje cells, and

neurons in the lateral thalamus and brainstem. Many

Purkinje cells are lost in mice older than 60 days of age

and degenerative changes in dendritic arbors and focal

swellings of axons are well recognized in remaining

Purkinje cells. These structural changes of Purkinje cells

have been particularly well demonstrated using immunocy-

tochemical staining for calbindin. Storage materials in

neuronal perikarya as well as distended processes can also

be immunostained with the antibody for ubiquitin [72]. The

neuronal inclusions consist of concentrically arranged

lamellar and multivesicular structures, similar to those of

human and feline NPC1 disease (Fig. 6) [12,72,74–76].

Inclusions are also detected in microglia/macrophages,

astrocytes and oligodendrocytes [71,72,77]. These inclu-

sions are structurally similar to each other but show some

different cell-specific profiles. In addition, as noted in the

liver, spleen and lung [78], bcholesterol cleftsQ indicating thepresence of cholesterol are well recognized in macrophages

in the brain [72]. bCholesterol cleftsQ have not been detected

in the neurons. However, filipin staining of brain tissue has

revealed numerous positive granules in neurons and other

cell types [12,23], which are believed to represent choles-

terol-filled endosomes/lysosomes (Fig. 1). Studies using

cultured murine NPC sympathetic neurons by Karten et al.

[79] have demonstrated altered cholesterol distribution

between neuronal cell bodies and axons. In agreement with

a previous in vitro study with embryonic striatal neurons

[80], they found no differences in cholesterol content in

S.U. Walkley, K. Suzuki / Biochimica et Biophysica Acta 1685 (2004) 48–62 57

neurons as a whole between NPC1-deficient and wild-type

mice. However, by filipin staining, cholesterol was found

sequestered in neuronal perikarya and the cholesterol

content of the distal axons was significantly less, suggesting

defective transport of cholesterol in NPC1-deficient neurons

[79]. Intraneuronal storage of free cholesterol has been

demonstrated in vivo in the mouse model by filipin staining

[23,81] as well as by a novel cholesterol-binding reagent

known as BC-theta [82]. In the latter study, the authors

reported that neuronal accumulation of cholesterol occurred

as early as 9–10 days of age and preceded evidence of

detectable accumulation of GM1 and other gangliosides

analyzed. In other studies of ganglioside elevations in the

NPC mice, neuronal accumulation of GM2 was found to

precede that of GM3 and to be detectable during this same

time frame (1–2 weeks of age) (Gondre-Lewis and Walkley,

unpublished data).

There is evidence for significant interaction between

cholesterol and glycosphingolipids in neurons, most notably

through co-localization in specialized membrane micro-

domains known as lipid rafts [83]. The significance of

glycosphingolipid accumulation in NPC disease and its

relationship to NPC1 and NPC2 function, and to the

sequestration of cholesterol, are controversial. Glycosphin-

golipids residing in lipid rafts in the plasma membrane are

believed to be imported into the endosomal–lysosomal

system for degradation and recycling via the Golgi/TGN.

One recent finding suggests that the high level of

accumulating cholesterol in the perikarya in NPC1-deficient

neurons blocks the endosome/lysosome to Golgi transport

and this event then leads to an accumulation of glyco-

sphingolipids [84,85]. This would argue, consistent with

long-held beliefs about NPC1 function, and more recently

about NPC2 as a cholesterol-binding protein, that ganglio-

side storage occurs secondary to cholesterol sequestration.

However, studies in which the synthesis of complex

gangliosides in NPC1 mice was limited by producing

double mutant animals lacking both NPC1 and GalNAc-

transferase, the enzyme responsible for synthesis of GM2,

GD2 and higher order gangliosides, suggest otherwise [9].

While, as expected, double mutant mice accumulated GM3

ganglioside in a manner equivalent to NPC1�/� mice and

lacked storage of GM2, they also exhibited dramatic

reduction in storage of unesterified cholesterol. Neurons

that did store cholesterol consistently showed storage of

GM3, whereas some GM3-accumulating neurons lacked

detectable staining for cholesterol. Neurons exhibiting

cholesterol accumulation but lacking evidence of GM3

staining were not observed, an important finding that strongly

suggests that cholesterol accumulation in NPC1�/� neurons

is in some manner dependent on the storage of complex

gangliosides [9]. These findings are consistent with the

possibility that NPC1 plays a key role in ganglioside

metabolism and/or trafficking, and possibly in the homeo-

static control of glycosphingolipid synthesis in a manner

originally proposed for cholesterol [23]. The generation of

NPC2�/� mice [12] means that the function of NPC2 can be

similarly tested in genetic crosses with mice lacking specific

enzymes in the glycosphingolipid synthetic pathway.

4.2. Meganeurite formation and ectopic dendritogenesis

Golgi studies of murine NPC1 disease have revealed the

presence of small meganeurites decorated with occasional

dendritic spines on some cortical pyramidal neurons [23].

However, the extensive ectopic neuritogenesis and large

dendritic spine/ectopic dendrite-covered meganeurites

observed in feline and human NPC1 disease, respectively,

have not been observed in mice (Fig. 2). Meganeurites

without any evidence of dendritic spines or processes

(baspinyQ types) are often observed in affected mice

revealing the presence of significant storage but suggesting

an absence of dendritogenic factors (or sensitivity to such

factors) that in affected cat and human brain must be

prominent. Interestingly, Golgi staining of a wide variety of

neuronal storage diseases in mice provides similar findings

[62]. That is, ectopic dendritogenesis is not a conspicuous

feature even in disorders already documented as exhibiting

this phenomenon in humans (e.g., Tay–Sachs disease). The

degree of ectopic spines and processes found in NPC1 mice,

in fact, is the most observed to date in any murine storage

disease model. One interpretation of the lack of significant

ectopic dendritogenesis in murine storage disease models is

that postnatal mechanisms of dendritic plasticity in rodent

brain may be more limited than those in carnivore (cat) and

primate (human). That murine NPC1 disease shows the

greatest degree of ectopic dendritic membrane production

and is also the earliest onset and most rapidly progressive of

the known murine storage diseases is consistent with this

view. Determining why ectopic dendritogenesis is limited in

mouse neurons may eventually be revealing as to important

regulatory factors controlling postnatal dendritogenesis in

all mammalian species.

4.3. Axonal spheroids

Focal swellings within axons (spheroids) are very

conspicuous in NPC1�/� mice just as in young human

NPC patients and in the feline NPC1 disease [23,76]. They

are recognized throughout the brain as early as 15 days of

age and are gradually increased in number with progression

of the disease process. Predilection sites include internal

capsule, substantia nigra, medial lemniscus, cerebellar

nuclei and white matter, pontine base, dorsal nuclei in

medulla and spinal white matter. Topographically, the extent

of neuronal storage does not necessarily coincide with the

presence of axonal spheroids. The topographic distribution

of axonal spheroids appears to be influenced by the

background strain. In the NPC�/� mice that are backcrossed

to C57BL/6, heavy concentrations of axonal spheroids are

observed in the fimbria [unpublished observation by KS].

They can be stained with antibodies to amyloid precursor

S.U. Walkley, K. Suzuki / Biochimica et Biophysica Acta 1685 (2004) 48–6258

protein, parvalbumin, ubiquitin and phosphorylated or non-

phosphorylated neurofilaments. At the ultrastructural level,

three types of spheroids have been identified in the NPC1

mouse model: (1) focally swollen axons with centrally

displaced bundles of neurofilaments and peripherally dis-

placed inclusions consisting of abnormal mitochondria,

lamellar and or dense bodies, (2) enlarged inner tongue of

oligodendrocytes with numerous inclusions ensheathing

apparently normal axons, and (3) accumulation of abnormal

inclusions in both axolemma and inner tongue processes

[72]. Bu et al. [86] have reported increases of cyclin-

dependent kinase 5 (cdk5) and its potent activator p25, a

proteolytic fragment of p35, in the axonal spheroids,

together with hyperphosphorylated cytoskeletal protein.

They suggested that focal deregulation of cdk5/p25 in

axons leads to cytoskeletal abnormalities and eventual

neurodegeneration in NPC disease. Cholesterol deficiency

in cultured neurons is reported to result in hyperphosphor-

ylation of microtubular protein tau accompanied by focal

axonal degeneration [87], and cultured sympathetic neurons

of NPC1�/� mice showed deficient cholesterol in distal

axons [79]. Axonal spheroids in NPC1�/� mice are

generally negative for filipin or BC-theta stain [unpublished

observation by KS]. These studies are supportive of the

hypothesis by Bu and co-workers in that distal axonal

degeneration due to deficient cholesterol may be the initial

event of neurodegeneration in NPC. As will be described

below, Fas ligand (Fas-L), the molecule involved in Fas/

Fas-L mediated cell death pathway, is expressed in macro-

phages in areas where axonal spheroids are numerous.

Axonal spheroids are commonly found together with the

degeneration of distal oligodendroglial processes. Degener-

ation of distal oligodendroglial cell processes has been well

documented as distal oligodendrogliopathy in myelin

disorders [88,89], suggesting that perturbed cholesterol

trafficking in oligodendrocytes may be an additional

contributory factor for the formation of axonal spheroids.

4.4. Cytoskeletal abnormalities

As reported above, NFTs are a hallmark of Alzheimer

disease and consist of hyperphosphorylated aggregations of

microtubular protein tau, PHFtau. NFTs are commonly

associated neuropathological changes in the brain of human

patients with NPC [35–37] but not detected histologically in

the mouse model of NPC [90]. However, site-specific

hyperphosphorylation of tau (at Ser-396 and Ser-404)

accompanied by an activation of MAP kinase has been

reported in the brain of NPC1�/� mice and some of the

cerebellar granular neurons and cerebral cortical neurons

were reported to be immunostained with antibody PHF-1

[91]. Bu et al. [86] found that hyperphosphorylation of

neurofilaments (NF-M isoform), MAP2 and tau was

associated with a significant increase in activity of the

cyclin-dependent kinase 5 (cdk5) and some PHF-1 immu-

noreactive neurons. Immunoreactivity to phosphorylated

neurofilaments is detected in many axonal spheroids as

stated above.

4.5. Golgi fragmentation

As noted in human NPC1 disease, Golgi fragmentation

was also detected in neurons in the cerebral cortex as well as

in cerebellar Purkinje cells in NPC1�/� mice (Fig. 5). The

numbers of Purkinje cells with fragmented Golgi apparatus

appear to increase prior to disappearance of the cells.

Fragmented Golgi apparatus was observed in the brainstem

neurons in NPC1�/� mice but also in the brainstem neurons

in twitcher mice, a murine model of globoid cell leukodys-

trophy [unpublished observation by KS]. In contrast,

fragmentation of the Golgi apparatus was not detected in

the GM1 gangliosidosis model mice. Thus, further inves-

tigation is needed to clarify the significance of this

intriguing phenomenon.

4.6. Neurodegeneration (loss of neurons)

The brains of NPC1�/� and SPM mice are small

(approximately 2/3 of wild type at 40 days) and become

atrophic with loss of neurons, most pronounced in the

cerebellar Purkinje cells and thalamus [75,76,90,92]. Clin-

ical course and neurodegeneration could not be prevented

by pharmacological treatments and genetic manipulations

aimed at alleviation of cholesterol storage [93,94]. However,

reduction of glycosphingolipid storage in NPC1�/� mice

through the use of a glycosphingolipid synthesis inhibitor

delayed Purkinje cell death and clinical onset [95]. Similar

increases in longevity were observed in some (though not

all) double mutant mice lacking both NPC1 and GalNAc-

transferase [9] (see also Ref. [96]). Genetic cross studies

which removed low-density lipoprotein receptor [97] or that

introduced the human Bcl-2 transgene into NPC1�/� mice

[98] failed to reveal clinical benefit. The loss of Purkinje

cells was most dramatically abated by introduction of wild-

type NPC1 protein to the NPC1�/� mice, proving that

mutation of NPC1 gene is responsible for neurodegenera-

tion in NPC1�/� mice [99].

Loss of Purkinje cells in NPC1�/� mice follows a well-

defined pattern [76,100]. Zebrin II-negative Purkinje cells

disappear first and degeneration of other Purkinje cells

follows this event. At the terminal stage, the majority of

Purkinje cells disappear with exception of those in lobules

IX and X of the posterior vermis. Surviving cells expressed

ectopic tyrosine hydroxylase and the small heat shock

protein HSP25 [100]. This orderly pattern of Purkinje cell

loss in NPC1�/� mice suggests that the loss of neurons in

NPC1 is not resulting from an accidental (necrotic) event

but is consistent with programmed (apoptotic) cell death.

Scattered TUNEL-positive cerebral cortical neurons and

cerebellar Purkinje cells and also the presence of active

caspase 3-immunoreactive Purkinje cells are consistent with

cell death being due to apoptosis [101]. Since overexpres-

S.U. Walkley, K. Suzuki / Biochimica et Biophysica Acta 1685 (2004) 48–62 59

sion of Bcl-2 failed to prevent neurodegeneration [98], cell

death in NPC1�/� mice is unlikely to be through Bcl-2

mediated mitochondrial cell death pathway. Numerous Fas-

L immunoreactive macrophages/microglia appeared in

NPC1 mice with close association to axonal spheroids

[101]. Since Fas/Fas-L induced apoptosis cannot be blocked

by Bcl-2 [102], involvement of the Fas/Fas-L signaling

pathway is a possibility in the cell death in NPC1�/� mice.

Using amino cupric silver stain, Ong and co-workers

reported widespread degeneration of nerve fibers followed

by degeneration of nerve cell bodies as early as day 9. They

suggested that the earliest structure involved was the nerve

terminal and axons and functional disruption of perisynaptic

astrocytes may contribute to neurodegeneration [103]. The

pattern of Purkinje cell loss in NPC2 mice has recently been

reported to closely resemble that observed in NPC1 disease

[12]. Indeed, double mutant mice lacking both NPC1 and

NPC2 exhibited the same pattern of cerebellar pathology in

a time frame equivalent to NPC1 disease.

4.7. Function of NPC1 vs. NPC2 in mice

Development of a murine model of NPC2 disease

created the opportunity to cross the two disease models to

produce NPC1;NPC2 double mutants [12]. It was reasoned

that if the NPC1 and NPC2 proteins functioned in

independent metabolic pathways, the loss of both proteins

in the same animal would likely lead to a more severe

phenotype than that observed with either alone. If

however, both proteins co-operated in carrying out the

same function, the disease phenotype would likely

resemble that of the more severe of the two mutants,

i.e., the NPC1 mice. Results showed the latter, that is,

double mutants appeared essentially indistinguishable

phenotypically from NPC1 mice in terms of disease onset

and progression, pathology, neuronal storage and bio-

chemistry of lipid accumulation [12]. These findings

strongly suggest that NPC1 and NPC2 proteins function

in concert to facilitate the intracellular retro-transport of

lipids from endosomes/lysosomes to other cellular sites.

5. Summary

Neuropathological consequences of NPC1 and NPC2

deficiency in neurons of humans and in feline and murine

models of this important lysosomal disease are briefly

reviewed. Basic structural alterations as they occur in NPC

disease in the three species are remarkably similar. Several

important interspecies distinctions, however, were observed.

The phenomenon of ectopic dendritogenesis, which is a

conspicuous feature of many cases of human NPC disease,

shows attenuated involvement in the feline disease and is

even less evident in the mouse models. This may reflect

important species differences in postnatal mechanisms of

dendritic plasticity as opposed to actual differences in the

function(s) of NPC1 and NPC2 in each species. Likewise,

neurons in many cases of human NPC disease, but not

mouse or cat, exhibit NFT similar to AD. This again may

possibly reflect species differences in disease response (e.g.,

the effects of cholesterol dysregulation on specific tau

isoforms in neurons) as opposed to actual differences in the

function of NPC1/NPC2 proteins. In future studies it will be

important to expand the analysis of NPC1 and NPC2 loss of

function to neurons in other species, including Drosophila

and C. elegans, to more closely analyze the ultimate role of

these proteins in retroendocytic lipid trafficking and other

possible functions. While genotype–phenotype analysis and

detailed neuropathological studies in humans with NPC

disease should be pursued, determining the consequences of

NPC1 and NPC2 loss of function in simpler cells and

organisms can provide critical insights into cell biological

mechanisms of NPC disease presently unrecognized in

humans, and may ultimately hold the key to delineating

pathogenic cascades and developing new and successful

therapeutic measures.

Acknowledgements

The authors would like to thank our many colleagues as

well as current and past members of our laboratories who

have so generously helped us in the study of Niemann–Pick

disease type C. The murine NPC2 studies referred to were

part of a collaborative effort between one of us (SUW) and

Drs. D. Sleat and P. Lobel, who are gratefully acknowl-

edged. We thank the National Institutes of Health and the

Ara Parseghian Medical Research Foundation for grant

support.

References

[1] A.C. Crocker, S. Farber, Niemann–Pick disease: a review of eighteen

patients, Medicine 37 (1957) 1–95.

[2] C. Patterson, M.T. Vanier, K. Suzuki, J.A. Morris, E. Carstea, E.B.

Neufeld, J.E. Branchette-Mackie, P.G. Pentchev, Niemann–Pick

Disease type C: a lipid trafficking disorder: Part 16. Lysosomal

disorders, Chap. 145, in: C.R. Scriver, A.L. Beaudet, W.S. Sly, D.

Valle (Eds.), The Metabolic and Molecular Basis of Inherited Disease,

eighth edition, McGraw-Hill, New York, 2001, pp. 3611–3634.

[3] H.G. Hers, Progress in gastroenterology: inborn lysosomal diseases,

Gastroenterology 48 (1965) 625–633.

[4] R.O. Brady, J.N. Kanfer, M.B. Mock, D.S. Fredrickson, The

metabolism of sphingomyelin: II. Evidence of an enzymatic defi-

ciency in Niemann–Pick disease, Proc. Natl. Acad. Sci. U. S. A. 55

(1966) 366–369.

[5] P.G. Pentchev, M.T. Vanier, K. Suzuki, M.C. Patterson, Niemann–

Pick disease type C: a cellular cholesterol lipidosis, in: C.R. Scriver,

A.L. Beaudet, W.S. Sly, D. Valle (Eds.), The Metabolic and

Molecular Basis of Inherited Disease, seventh edition, McGraw-

Hill, New York, 1995, pp. 2625–2639.

[6] E.D. Carstea, J.A. Morris, K.G. Coleman, S.K. Loftus, D. Zhang, C.

Cummings, J. Gu, M.A. Rosenfeld, W.J. Pavan, D.B. Krizman, J.

Nagle, M.H. Polymeropoulos, S.L. Sturley, Y.A Ioannou, M.E.

Higgins, M. Comly, A. Cooney, A. Brown, C.R. Kaneski, E.J.

S.U. Walkley, K. Suzuki / Biochimica et Biophysica Acta 1685 (2004) 48–6260

Blanchette-Mackie, N.K. Dwyer, E.B. Neufeld, T. Chang, L. Liscum,

J.F. Strauss III, K. Ohno, M. Zeigler, R. Carmi, J. Sokol, D. Markie,

R.R. O’Neill, O.P. van Diggelen, M. Elleder, M.C. Patterson, R.O.

Brady, M.T. Vanier, P.G. Pentchev, D.A. Tagle, Niemann–Pick C1

disease gene: homology to mediators of cholesterol homeostasis,

Science 277 (1997) 228–231.

[7] J.P. Davies, F.W. Chen, Y.A. Ioannou, Transmembrane molecular

pump activity of Niemann–Pick C1 protein, Science 290 (2000)

2295–2298.

[8] K. Malathi, K. Higaki, A.H. Tinkelenberg, D.A. Balderes, D.

Almanzar-Paramio, L.J. Wilcox, N. Erdeniz, F. Redican, M.

Padamsee, Y. Liu, S. Khan, F. Alcantara, E.D. Carstea, J.A.

Morris, S.L. Sturley, Mutagenesis of the putative sterol-sensing

domain of yeast Niemann–Pick C-related protein reveals a

primordial role in subcellular sphingolipid distribution, J. Cell

Biol. 164 (2004) 547–556.

[9] M.C. Gondre-Lewis, S.U. Walkley, Cholesterol accumulation in

NPC1-deficient neurons is ganglioside-dependent, Curr. Biol. 13

(2003) (2003) 1324–1329.

[10] S. Naureckiene, D.E. Sleat, H. Lackland, A. Fensom, M.T. Vanier,

R. Wattiaux, M. Jadot, P. Lobel, Identification of HE1 as the

second gene of Niemann–Pick C disease, Science 290 (2000)

2298–2301.

[11] N. Friedland, H.L. Liou, P. Lobel, A.M. Stock, Structure of a

cholesterol-binding protein deficient in Niemann–Pick type C2

disease, Proc. Natl. Acad. Sci. U. S. A. 100 (2003) 2512–2517.

[12] D.E. Sleat, J.A. Wiseman, M. El-Banna, S.M. Price, L. Verot, M.M.

Shen, G.S. Tint, M.T. Vanier, S.U. Walkley, P. Lobel, Genetic

evidence for non-redundant functional co-operativity between NPC1

and NPC2 in lipid transport, Proc Natl Acad Sci U. S. A. 101 (2004)

5886–5891.

[13] A.C. Lowenthal, J.F. Cummings, D.A. Wenger, M.A. Thrall, P.A.

Wood, A. de Lahunta, Feline sphingolipidosis resembling Niemann–

Pick disease type C, Acta Neuropathol. 81 (1990) 189–197.

[14] M.D. Morris, C. Bhuvaneswaran, H. Shio, S. Fowler, Lysosome lipid

storage disorder in NCTR-BALB/c mice. I. Description of the

disease and genetics, Am. J. Pathol. 108 (1982) 140–149.

[15] S. Miyataki, H. Yoshida, S. Mitsuoka, H. Enomoto, S. Ikehara, A

mouse model for Niemann–Pick disease, J. Heredity 77 (1986)

379–384.

[16] M. Sym, M. Basson, C. Johnson, A model for Niemann–Pick type C

disease in the nematode Caenorhabditis elegans, Curr. Biol. 10

(2000) 527–530.

[17] J.K. Fink, M.R. Filling-Katz, J. Sokol, D.G. Cogan, A. Pikus, B.

Sonies, B. Soong, P.G. Pentchev, M.E. Comly, R.O. Brady, N.W.

Baron, Clinical spectrum of Niemann–Pick disease type C,

Neurology 39 (1989) 1040–1049.

[18] M.T. Vanier, G. Millat, Niemann–Pick disease type C, Clin. Genet.

64 (2003) 269–281.

[19] G. Millat, C. Marcais, C. Tomasetto, K. Chikh, A.H. Fenson, K.

Harzer, D.A. Wenger, K. Ohno, M.T. Vanier, Niemann–Pick C1

disease: correlations between NPC1 mutations, levels of NPC1

protein, and phenotypes emphasize the functional significance of the

putative sterol-sensing domain and of the cysteine-rich luminal loop,

Am. J. Hum. Genet. 68 (2001) 1373–1385.

[20] H.H. Klunemann, M. Elleder, W.E. Kaminski, K. Snow, J.M. Peyser,

J.F. O’Brien, D. Munoz, G. Schmitz, H.E. Klein, W.W. Pendlebury,

Frontal lobe atrophy due to a mutation in the cholesterol binding

protein HE1/NPC2, Ann. Neurol. 52 (2002) 743–749.

[21] K. Suzuki, K. Suzuki, Lysosomal disease, in: D.I. Graham, P. Lantos

(Eds.), Greenfield’s Neuropathology, seventh edition, Arnold,

London, 2002, pp. 666–670 (Chap. 11).

[22] M.T. Vanier, Lipid changes in Niemann–Pick disease type C brain:

personal experience and review of the literature, Neurochem. Res. 24

(1999) 481–489.

[23] M. Zervas, K. Dobrenis, S.U. Walkley, Neurons in Niemann–Pick

disease type C accumulate gangliosides as well as unesterified

cholesterol and undergo dendritic and axonal alterations, J. Neuro-

pathol. Exp. Neurol. 60 (2001) 49–64.

[24] R.D. Terry, M. Weiss, Studies in Tay–Sachs disease: II. Ultra-

structure of the cerebrum, J. Neuropathol. Exp. Neurol. 22 (1963)

18–55.

[25] D.P. Purpura, K. Suzuki, Distortion of neuronal geometry and

formation of aberrant synapses in neuronal storage diseases, Brain

Res. 116 (1976) 1–21.

[26] C.M. de Baecque, K. Suzuki, I. Rapin, A.B. Johnson, D.L.

Whethers, K. Suzuki, GM2 gangliosidosis, AB variant. Clinico-

pathological study of a case, Acta Neuropathol. (Berl) 33 (1975)

207–226.

[27] K. Sandhoff, T. Kolter, T Topology of glycosphingolipid degrada-

tion, Trends Cell Biol. 6 (1996) 98–103.

[28] H. Braak, E. Braak, H.H. Goebel, Isocortical pathology in type C

Niemann–Pick disease. A combined Golgi-Pigmentarchitectonic

study, J. Neuropathol. Exp. Neurol. 42 (1983) 671–687.

[29] P.A. March, M.A. Thrall, D.E. Brown, T.W. Mitchell, A.C.

Lowenthal, S.U. Walkley, GABAergic neuroaxonal dystrophy and

other cytopathological alterations in feline Niemann–Pick disease

type C, Acta Neuropathol. 94 (1997) 164–172.

[30] D.P. Purpura, S.U. Walkley, Aberrant neurite and spine generation in

mature neurons in the gangliosidoses, in: M.M. Rapport, A. Gorio

(Eds.), Gangliosides in Neurological and Neuromuscular Function,

Development and Repair, Raven Press, NY, 1981, pp. 1–16.

[31] S.U. Walkley, Pathogenic cascades and brain dysfunction, in: F.M.

Platt, S.U. Walkley (Eds.), Lysosomal Disorders of the Brain, Oxford

University Press, 2004, pp. 290–324 (Chap. 12).

[32] M. Elleder, A. Jirasek, F. Smid, J. Ledvinova, G.T.N. Besley,

Niemann–Pick disease Type C: study on the nature of the cerebral

storage process, Acta Neuropathol. (Berl.) 66 (1985) 325–336.

[33] P.W. Lampert, A comparative electron microscopic study of reactive,

degenerating, regenerating and dystrophic axons, J. Neuropathol.

Exp. Neurol. 26 (1967) 345–368.

[34] M.B. Graeber, W.F. Blakemore, G.W. Kreutzberg, Cellular pathology

of the central nervous system, in: D.I. Graham, P. Lantos (Eds.),

Greenfield’s Neuropathology, seventh edition, Arnold, London,

2002, pp. 134–135.

[35] K. Suzuki, C.C. Parker, P.G. Pentchev, D. Katz, B. Ghetti, A.N.

D’Agostino, E.D. Carstea, Neurofibrillary tangles in Niemann–Pick

disease type C, Acta Neuropathol. (Berl.) 89 (1995) 227–238.

[36] S. Love, L.R. Bridges, C.P Case, Neurofibrillary tangles in

Niemann–Pick disease type C, Brain 118 (1995) 119–129.

[37] K. Suzuki, C.C. Parker, P.G. Pentchev, Niemann–Pick disease type C:

neuropathology revisited, Dev. Brain Dysfunct. 10 (1997) 306–320.

[38] I.A. Auer, M.L. Schmidt, V.M.-Y. Lee, R. Curry, K. Suzuki, R.W.

Shin, P.G. Pentchev, E.D. Carstea, J.Q. Trojanowski, Paired

helical filament tau (PHFtau) in Niemann–Pick type C disease

is similar to PHFtau in Alzheimer’s disease, Acta Neuropathol. 90

(1995) 547–551.

[39] R. Dist, V. Meske, T.G. Ohm, Tangle-bearing neurons contain more

free cholesterol than adjacent tangle-free neurons, Acta Neuropathol.

101 (2001) 547–554.

[40] Y. Saito, K. Suzuki, E. Nanba, T. Yamamoto, K. Ohno, S. Murayama,

Niemann–Pick type C disease: accelerated neurofibrillary tangle

formation and amyloid h deposition associated with apolipoprotein E

q4 homozygosity, Ann. Neurol. 52 (2002) 351–355.

[41] C.F. Lippa, H. Fujiwara, D.M. Mann, et al., Lewy bodies contain

altered a synuclein in brains of many familial Alzheimer’s disease

patients with mutations in presenilin and amyloid precursor protein

genes, Am. J. Pathol. 153 (1998) 1365–1370.

[42] R.L. Hamilton, Lewy bodies in Alzheimer’s disease: a neuro-

pathological review of 145 cases using a-synuclein immunohisto-

chemistry, Brain Pathol. 10 (2000) 378–384.

[43] Y. Trembath, C. Rosenberg, J.F. Ervin, et al., Lewy body pathology

is a frequent co-pathology in familial Alzheimer’s disease, Acta

Neuropathol. (Berl.) 105 (2003) 484–488.

S.U. Walkley, K. Suzuki / Biochimica et Biophysica Acta 1685 (2004) 48–62 61

[44] Y. Saito, M. Kawai, K. Inoue, et al., Widespread expression of a-

synuclein and tau immunoreactivity in Hallervorden–Spatz syn-

drome with protracted clinical course, J. Neurol. Sci. 177 (2000)

48–59.

[45] Y. Sato, K. Suzuki, C.M. Hulette, S. Murayama. Aberrant

phosphorylation of a-synuclein in human Niemann–Pick type C1

disease, J. Neuropathol. Exp. Neurol. 63 (2004) 323–328.

[46] T. Ishizawa, P. Mattila, P. Davies, D. Wang, D.W. Dickson,

Colocalization of tau and alpha-synuclein epitopes in Lewy bodies,

J. Neuropathol. Exp. Neurol. 62 (2003) 389–397.

[47] B. Bu, H. Klunemann, K. Suzuki, J. Li, T. Bird, L.-W. Jin, I. Vincent,

Niemann–Pick disease type C yields possible clue for why cerebellar

neurons do not form neurofibrillary tangles, Neurobiol. Dis. 11

(2002) 285–297.

[48] N.K. Gonatas, A. Stieber, Z. Mourelatos, Y. Chen, J.O. Gonatas,

Fragmentation of the Golgi apparatus of motor neurons in

amyotrophic lateral sclerosis, Am. J. Pathol. 140 (1992) 731–737.

[49] Z. Mourelatos, A. Yachnis, L. Rorke, J. Mikol, N.K. Gonatas, The

Golgi apparatus of motor neurons in amyotrophic lateral sclerosis,

Ann. Neurol. 33 (1993) 608–615.

[50] N.K. Gonatas, The Golgi apparatus in disease, in: E.G. Berger, J.

Roth (Eds.), The Golgi apparatus, Birkhðuser Verlag, Basel/Switzer-

land, 1997, pp. 247–273.

[51] A. Stieber, Z. Mourelatos, N.K. Gonatas, In Alzheimer’s disease

the Golgi apparatus of a population of neurons without neuro-

fibrillary tangles is fragmented and atrophic, Am. J. Pathol. 148

(1996) 415–426.

[52] Z. Mourelatos, N.K. Gonatas, A. Stieber, M.E. Gurney, M.C. Dal

Canto, The Golgi apparatus of spinal cord motor neurons in

transgenic mice expressing mutant Cu, Zn superoxide dismutase

becomes fragmented in early, preclinical stages of the disease, Proc.

Natl. Acad. Sci. U. S. A. 93 (1996) 5472–5477.

[53] W. Yang, B. Storrie, Scattered Golgi elements during microtubule

disruption are initially enriched in trans-Golgi proteins, Mol. Biol.

Cell 9 (1998) 191–207.

[54] A. Colanzi, T.J. Deerinck, M.H. Ellisman, V. Malhotra, A specific

activation of the mitogen-activated protein kinase kinase 1 (MEK 1)

is required for Golgi fragmentation during mitosis, J. Cell Biol. 149

(2000) 331–339.

[55] W.J. Nelson, W(h)ither the Golgi during mitosis? J. Cell Biol. 149

(2000) 243–248.

[56] K.L. Somers, D.A. Wenger, M.A. Royals, E.D. Carstea, H.E.

Connally, T. Kelly, R. Kimball, M.A. Thrall, Complementation

studies in human and feline Niemann–Pick type C disease, Mol.

Genet. Metab. 66 (1999) 117–121.

[57] K.R. Munana, P.J. Luttgen, M.A. Thrall, T.W. Mitchell, D.A.

Wenger, Neurological manifestations of Niemann–Pick disease type

C in cats, J. Vet. Intern. Med. 8 (1994) 117–121.

[58] S.U. Walkley, Cellular pathology of lysosomal storage disorders,

Brain Pathol. 8 (1998) 175–193.

[59] S.U. Walkley, A.P. Pierok, Ferric ion-ferrocyanide staining in

ganglioside storage disease establishes that meganeurites are of

axon hillock origin and distinct from axonal spheroids, Brain Res.

382 (1986) 379–386.

[60] S.U. Walkley, S. Wurzelmann, Alterations in synaptic connectivity in

cerebral cortex in neuronal storage diseases, Ment. Retard. Dev.

Disabil. Res. Rev. 1 (1995) 183–192.

[61] D.A. Siegel, S.U. Walkley, Growth of ectopic dendrites on cortical

pyramidal neurons in neuronal storage diseases correlates with

abnormal accumulation of GM2 ganglioside, J. Neurochem. 62

(1994) 1852–1862.

[62] S.U. Walkley, M. Zervas, S. Wiseman, Gangliosides as modulators

of dendritogenesis in storage disease-affected and normal pyramidal

neurons, Cereb. Cortex 10 (2000) 1028–1037.

[63] S.U. Walkley, Neurobiology and cellular pathogenesis of glycolipid

storage diseases, Trans. Royal Soc. B 358 (2003) 893–904.

[64] S.U. Walkley, Secondary increases in ganglioside expression in

lysosomal disorders, Seminars in Cell and Developmental Biology

15 (2004) 433–444.

[65] S.U. Walkley, H.J. Baker, M.C. Rattazzi, M.E. Haskins, J.-Y. Wu,

Neuroaxonal dystrophy in neuronal storage disorders: Evidence for

major GABAergic neuron involvement, J. Neurol. Sci. 104 (1991)

1–8.

[66] R.G. Parton, K. Simons, C.G. Dotti, Axonal and dendritic endocytic

pathways in cultured neurons, J. Cell Biol. 119 (1992) 123–137.

[67] S.U. Walkley, H.J. Baker, M.C. Rattazi, Initiation and growth of

ectopic neurites and meganeurites during postnatal brain develop-

ment in ganglioside storage disease, Dev. Brain Res. 51 (1990)

167–178.

[68] S.U. Walkley, S. Wurzelmann, D.A. Siegel, Ectopic axon hillock-

associated neurite growth is maintained in metabolically-reversed

swainsonine-induced neuronal storage disease, Brain Res. 410

(1987) 89–96.

[69] S.K. Loftus, J.A. Morris, E.D. Carstea, J.Z. Gu, C. Cummings, A.

Brown, J. Ellison, K. Ohno, M.A. Rosenfeld, D.A. Tagle, P.G.

Pentchev, W.J. Pavan, Murine model of Niemann–Pick C disease:

mutation in a cholesterol homeostasis gene, Science 277 (1997)

232–235.

[70] T. Yamamoto, K. Iwasawa, T. Tokoro, Y. Eto, K. Maekawa, A

possible same genetic defect in two Niemann–Pick disease model

mice, No To Hattatsu 26 (1994) 318–322 (article in Japanese).

[71] H. Weintraub, A. Abramovici, U. Sandbank, A.D. Booth, P.G.

Pentchev, B. Sela, Dysmyelination in NCTR-Balb/C mouse mutant

with a lysosomal storage disorder. Morphological survey, Acta

Neuropathol. (Berl.) 74 (1987) 374–381.

[72] Y. Higashi, P.G. Pentchev, S. Murayama, K. Suzuki, Pathology of

Niemann–Pick type C: studies of murine mutant, in: F. Ikuta

(Ed.), Neuropathology in Brain Research, Elsevier, Amsterdam,

1991, pp. 85–102.

[73] H. Weintraub, A. Abramovici, U. Sandbank, P.G. Pentchev, R.O.

Brady, M. Sekine, A. Suzuki, B. Sela, Neurological mutation

characterized by dysmyelination in NCTR-Balb/C mouse with

lysosomal lipid storage diseases, J. Neurochem. 45 (1985) 665–672.

[74] J. Litrenta, F. Adeshuko, S. Zhou, G. Stephney, J. Wiseman, K.

Dobrenis, D. Sleat, P. Lobel, S.U. Walkley, Glycosphingolipid-

cholesterol abnormalities in NPC2-deficient neurons appears equiv-

alent to those occurring in neurons lacking NPC1, suggesting NPC1–

NPC2 molecular cooperation, Abstr.-Soc. Neurosci. 537.15 (2003).

[75] J. Tanaka, H. Nakamura, S. Miyawaki, Cerebellar involvement in

mouse sphingomyelinosis: a new model of Niemann–Pick disease,

J. Neuropathol. Exp. Neurol. 47 (1988) 291–300.

[76] Y. Higashi, S. Murayama, P.G. Pentchev, K. Suzuki, Cerebellar

degeneration in the Niemann–Pick type C mouse, Acta Neuropathol.

85 (1993) 175–184.

[77] D.C. German, C.L. Liang, T. Song, U. Yazudani, C. Xie, J.M.

Dietschy, Degeneration of neurons and glia in the Niemann–Pick C

mouse is unrelated to the low-density lipoprotein receptor, Neuro-

science 109 (2002) 437–450.

[78] H. Shio, S. Fowler, C. Bhuvaneswaran, M.D. Morris, Lysosomal

lipid storage disorder in NCTR-BALB/c mice: II. Morphologic and

cytochemical studies, Am. J. Pathol. 108 (1982) 150–159.

[79] B. Karten, D.E. Vance, R.B. Campenot, J.E. Vance, Cholesterol

accumulates in cell bodies but it decreased in distal axons of

Niemann–Pick C1 deficient neurons, J. Neurochem. 83 (2002)

1154–1163.

[80] L.P. Henderson, L. Lin, A. Prasad, C.A. Paul, T.Y. Chang, R.A.

Maue, Embryonic striatal neurons from Niemann–Pick type C mice

exhibit defects in cholesterol metabolism and neurotrophin respon-

siveness, J. Biol. Chem. 275 (2000) 20179–20187.

[81] S. Treiber-Held, R. Dist, V. Meske, F. Albert, T.G. Ohm, Spatial and

temporal distribution of intracellular free cholesterol in brains of

NiemannPick type C mouse model showing hyperphosphorylated

S.U. Walkley, K. Suzuki / Biochimica et Biophysica Acta 1685 (2004) 48–6262

tau protein. Implications for Alzheimer’s disease, J. Pathol. 200

(2003) 95–103.

[82] P.C. Reid, N. Sakashita, S. Sugii, Y. Ohno-Iwashita, Y. Shimada,

W.I. Hickey, T.Y. Chang, A novel cholesterol stain reveals early

neuronal cholesterol accumulation in the Niemann–Pick type C1

mouse brain, J. Lipid Res. 45 (2004) 582–591.

[83] K. Simons, D. Toomre, Lipid rafts and signal transduction, Nat. Rev.

1 (2000) 31–41.

[84] V. Puri, R. Watanabe, M. Dominguez, X. Sun, C.L. Wheatley, D.L.

Marks, R.E. Pagano, Cholesterol modulates membrane traffic along

the endocytic pathway in sphingolipid storage diseases, Nat. Cell

Biol. 1 (1999) 386–388.

[85] R.E. Pagano, V. Puri, M. Dominguez, D.L. Marks, Membrane traffic

in sphingolipid storage diseases, Traffic 1 (2000) 807–815.

[86] B. Bu, J. Li, P. Davies, I. Vincent, Deregulation of cdk5,

hyperphosphorylation, and cytoskeletal pathology in Niemann–Pick

type C mutant model, J. Neurosci. 22 (2002) 6515–6525.

[87] Q.-W. Fan, W. Yu, T. Senda, K. Yanagisawa, M. Michikawa,

Cholesterol-dependent modulation of tau phosphorylation in cultured

neurons, J. Neurochem. 76 (2001) 391–400.

[88] S.K. Ludwin, E.S. Johnson, Evidence for a bDying BackQ gliopathyin demyelinating disease, Ann. Neurol. 9 (1981) 301–305.

[89] M. Rodriguez, B.W. Scheithauer, G. Forbes, P.J. Kelly, Oligoden-

drocyte injury is an early event in lesions of multiple sclerosis, Mayo

Clin. Proc. 68 (1993) 627–636.

[90] D.C. German, E.M. Quintero, C.L. Liang, B. Ng, S. Punia, C. Xie,

J.M. Dietschy, Selective neurodegeneration, without neurofibrillary

tangles, in a mouse model of Niemann–Pick C disease, J. Comp.

Neurol. 7 (2001) 415–425.

[91] N. Sawamura, J.-S. Gong, W.S. Garver, R.A. Heidenreich, H.

Ninomiya, K. Ohno, K. Yanagisawa, M. Michikawa, Site-specific

phosphorylation of tau accompanied by activation of mitogen-

activated protein kinase (MAPK) in brains of Niemann–Pick type C

mice, J. Biol. Chem. 276 (2002) 10314–10319.

[92] A. Yamada, M. Ukita, Y. Shinoda, M. Taniguchi, K. Higaki, H.

Ninomiya, K. Ohno, Progressive neuronal loss in the ventral

posterior lateral and medial nuclei of thalamus in Niemann–Pick

disease type C mouse brain, Brain Dev. 5 (2001) 288–297.

[93] M.C. Patterson, A.M. Di Bisceglie, J.J. Higgins, R.B. Abel, R.

Schiffmann, C.C. Parker, C.E. Argoff, R.P. Grewal, K. Yu, P.G.

Pentchev, R.O. Brady, N.W. Barton, The effect of cholesterol-

lowering agents on hepatic and plasma cholesterol in NiemannPick

disease type C, Neurology 43 (1993) 61–64.

[94] R.P. Erickson, W.S. Garver, F. Camargo, G.S. Hossain, R.A.

Heidenreich, Pharmacological and genetic modification of somatic

cholesterol does not substantially alter the course of CNS disease in

Niemann–Pick C mice, J. Inherit. Metab. Dis. 23 (2000) 54–62.

[95] M. Zervas, K.L. Somers, M.A. Thrall, S.U. Walkley, Critical role for

glycosphingolipids in Niemann–Pick disease type C, Curr. Biol. 11

(2001) 1283–1287.

[96] Y. Liu, Y.-P. Wu, R. Wada, E.B. Neufeld, K.A. Mullin, A.C. Howard,

P.G. Pentchev, M.T. Vanier, K. Suzuki, R.L. Proia, Alleviation of

neuronal ganglioside storage does not improve the clinical course of

the Niemann–Pick C disease mouse, Hum. Mol. Genet. 12 (2000)

1087–1092.

[97] C. Xie, D.K. Burns, S.D. Turley, J.M. Dietschy, Cholesterol is

sequestered in the brains of mice with Niemann–Pick type C disease

but turnover is increased, J. Neuropathol. Exp. Neurol. 59 (2000)

1106–1117.

[98] R.P. Erickson, O. Bernard, Studies on neuronal death in the mouse

model of Niemann–Pick C disease, J. Neurosci. Res. 68 (2002)

738–744.

[99] S.K. Loftus, R.P. Erickson, S.U. Walkley, M. Bryant, A. Incao, R.A.

Heidenreich, W.J. Pavan, Rescue of neurodegeneration in Niemann–

Pick-C mice by a prion-promoter Npc1 cDNA transgene, Hum. Mol.

Genet. 11 (2002) 3107–3114.

[100] J.R. Sarna, M. Larouche, H. Marzban, R.V. Sillitoe, D.E. Rancoort, R.

Hawkes, Patterned Purkinje cell degeneration in mouse models of

Niemann–Pick type C disease, J. Comp. Neurol. 456 (2003) 279–291.

[101] S. Takikita, T. Yagi, T. Fukuda, K. Suzuki, Fas-L expression of

activated microglia in NPC mouse brain, J. Neuropathol. Exp.

Neurol. 62 (2003) 554.

[102] D.C.S. Huang, M. Hahne, M. Schroeter, K. Frei, A. Fontana, A.

Villunger, K. Newton, J. Tschopp, A. Strasser, Activation of Fas

by FasL induces apoptosis by a mechanism that cannot be

blocked by Bcl-2 or Bcl-xL, Proc. Natl. Acad. Sci. 96 (1999)

14871–14876.

[103] W.-Y. Ong, U. Kumar, R.C. Switzer, A. Sidhu, G. Suresh, C.-Y. Hu,

S.C. Patel, Neurodegeneration in Niemann–Pick type C disease

mice, Exp. Brain Res. 141 (2001) 218–231.