Development of cerebellar pathology in the canine model of mucopolysaccharidosis type IIIA (MPS IIIA)

Development of cerebellar pathology in the canine model of mucopolysaccharidosis type IIIA (MPS IIIA)

Molecular Genetics and Metabolism 113 (2014) 283–293 Contents lists available at ScienceDirect Molecular Genetics and Metabolism journal homepage: w...

3MB Sizes 0 Downloads 14 Views

Molecular Genetics and Metabolism 113 (2014) 283–293

Contents lists available at ScienceDirect

Molecular Genetics and Metabolism journal homepage: www.elsevier.com/locate/ymgme

Development of cerebellar pathology in the canine model of mucopolysaccharidosis type IIIA (MPS IIIA) Sofia Hassiotis a,⁎, Robert D. Jolly b, Kim M. Hemsley a a b

Lysosomal Diseases Research Unit, South Australian Health and Medical Research Institute, PO Box 11060, Adelaide, South Australia 5001, Australia Institute of Veterinary, Animal and Biomedical Sciences, Massey University, Palmerston North 4442, New Zealand

a r t i c l e

i n f o

Article history: Received 20 August 2014 Received in revised form 14 October 2014 Accepted 14 October 2014 Available online 22 October 2014 Keywords: Lysosomal storage disorder Mucopolysaccharidosis Sanfilippo syndrome Purkinje cell Cerebellum Dog

a b s t r a c t The temporal relationship between the onset of clinical signs in the mucopolysaccharidosis type IIIA (MPS IIIA) Huntaway dog model and cerebellar pathology has not been described. Here we sought to characterize the accumulation of primary (heparan sulfate) and secondary (GM3) substrates and onset of other changes in cerebellar tissues, and investigate the relationship to the onset of motor dysfunction in these animals. We observed that Purkinje cells were present in dogs aged up to and including 30.9 months, however by 40.9 months of age only ~12% remained, coincident with the onset of clinical signs. Primary and secondary substrate accumulation and inflammation were detected as early as 2.2 months and axonal spheroids were observed from 4.3 months in the deep cerebellar nuclei and later (11.6 months) in cerebellar white matter tracts. Degenerating neurons and apoptotic cells were not observed at any time. Our findings suggest that cell autonomous mechanisms may contribute to Purkinje cell death in the MPS IIIA dog. © 2014 Elsevier Inc. All rights reserved.

1. Introduction Mucopolysaccharidosis type IIIA (MPS IIIA; MIM 252900) is a neurodegenerative lysosomal storage disorder (LSD) characterized by a deficiency in N-sulfoglucosamine sulfohydrolase (SGSH; EC 3.10.1.1), a lysosomal enzyme required to catabolize the glycosaminoglycan, heparan sulfate [1]. SGSH deficiency causes primary storage of partiallydegraded heparan sulfate, the accumulation of which may inhibit other lysosomal enzymes [2], potentially leading to secondary accumulation of gangliosides GM2 and GM3. The principal site of pathology is the central nervous system. Clinical symptoms include speech delay, sleep disturbance, hyperactivity, aggression and anxiety, with progressive neurological and behavioral deterioration and early death [3]. The Australian clinical prevalence of MPS IIIA is 1/114,000 live births [4] and no treatment is currently available.

Abbreviations: GAD65/67, glutamic acid decarboxylase 65/67; GlcNS-UA, glucosamineN-sulfate-uronic acid; INCL, infantile neuronal ceroid lipofuscinosis; LC–ESI-MS/MS, liquid chromatography–electrospray ionization tandem mass spectrometry; LSD, lysosomal storage disorder; MPS IIIA, mucopolysaccharidosis type IIIA; MPS IIIB, mucopolysaccharidosis type IIIB; NP, Niemann–Pick; PKC, Purkinje cell(s); RCA-1, ricinus communis agglutinin-1; rTdT, recombinant terminal deoxynucleotidyl transferase; SGSH, N-sulfoglucosamine sulfohydrolase; TUNEL, terminal dUTP nick end labeling. ⁎ Corresponding author at: Research Scientist, CNS Therapeutics Lab, Lysosomal Diseases Research Unit, PO Box 11060, Adelaide, South Australia 5001, Australia. E-mail addresses: Sofi[email protected] (S. Hassiotis), [email protected] (R.D. Jolly), [email protected] (K.M. Hemsley).

http://dx.doi.org/10.1016/j.ymgme.2014.10.008 1096-7192/© 2014 Elsevier Inc. All rights reserved.

The MPS IIIA Huntaway dog is a naturally-occurring large animal model of the human disease, in which the clinical signs are predominantly those of a movement disorder, developing from approximately 24 months of age and attributed to lesions in the cerebellum [5,6]. It results from an insertion mutation of adenosine at amino acid position 708–709, leading to a frameshift and early termination of SGSH at position 228 [7]. There is no detectable SGSH activity [8], with consequent accumulation of heparan sulfate and secondary substrates such as gangliosides [9]. MPS IIIA dogs are often euthanized early because of the progressive nature of clinical signs, which include ataxia, hypermetria and wide-base stance of limbs. Intention tremor is not observed. Description of the nature of cerebellar lesions in the dogs is mainly limited to late-stage disease, including accumulation of storage cytosomes in Purkinje cells (PKC) as well as PKC loss [6]. Storage cytosomes are complex and ultrastructurally, appear mostly comprised of membranous materials (i.e. lipids), with accumulation of granular material (postulated to be precipitated glycosaminoglycan) observed in macrophages and visceral organs [6]. The storage cytosomes have an orange/yellow autofluorescence as well as variable staining characteristics with general histological stains such as Sudan black, periodic acid Schiff and luxol fast blue [5,6]. Mild to moderate loss of granule cells, mild proliferation of Bergmann glia including hypertrophy of Bergmann glial processes and the occasional spheroid in the cerebellar white matter tracts have also been reported [6]. Neuroinflammation has been observed in the cerebellum and other brain regions [9]. PKC are the sole source of cerebellar signaling and their loss has been implicated in cerebellar dysfunction in other animal models of LSD, e.g.

284

S. Hassiotis et al. / Molecular Genetics and Metabolism 113 (2014) 283–293

canine mucopolysaccharidosis type IIIB (MPS IIIB) [10] and fucosidosis [11,12], as well as α-mannosidosis guinea pigs [13] and murine models of MPS IIIB [14], infantile neuronal ceroid lipofuscinosis (INCL) [15] and Niemann–Pick (NP) types A, B and C [16–19]. In α-mannosidosis guinea pigs and some of the murine models, the loss of PKC occurs in an anterior to posterior direction, correlating with a loss of calbindin-D28K, a calcium-binding protein involved in calcium homeostasis [13,15,16,18, 19]. Whilst calcium homeostasis has not been examined in the MPS IIIA dog, dilated stacks of smooth endoplasmic reticulum have been noted in hypertrophied PKC [6]. As the endoplasmic reticulum is a site of calcium ion storage, this suggests perturbed calcium flux [6]. Using both histological and biochemical methods, this retrospective study on archival material examined the development of cerebellar lesions in relation to the onset of clinical signs in the canine model of MPS IIIA.

Table 1 Animal groups and number of cerebellar lobules available for the study. F = female; M = male; a = frozen sections of posterior lobe of cerebellum available for GM3 immunohistochemistry. Genotype & gender

Unaffected M M F F F F MPS IIIA Pre-symptomatic (b12 months)

2. Materials and methods 2.1. Animals and tissue collection Archival material obtained from unaffected (heterozygous) and affected (MPS IIIA) dogs bred and housed in a research kennel facility, was used in this study. All dogs had been genotyped using published methods [7] and had undergone routine neurological examinations (personal communication Dr Neil Marshall). Tissues were collected for studies approved by the Animal Ethics Committee of Massey University (Palmerston North, New Zealand) and adhered to ‘The Code of Ethical Conduct for the Use of Live Animals for Teaching and Research’ (New Zealand ‘Animal Welfare Act 1999’). Approval was also obtained by the Animal Ethics Committee of the Children, Youth and Women's Health Service (Adelaide) and adhered to the ‘Australian Code of Practice for the Care and Use of Animals for Scientific Purposes’, 7th edition, 2004. At the time of euthanasia most of the animals' brain (unaffected n = 6; MPS IIIA n = 10) and cerebellar (unaffected n = 2; MPS IIIA n = 7) weights were recorded. The brain was then bisected along the sagittal midline, producing a left and right hemi-cerebellum. Samples available for the study included fresh frozen tissue from the cerebellar mid-line dorsal lobules 4/5 of the right hemi-cerebellum and formalin-fixed, paraffin-embedded or snap-frozen tissue from sagittal mid-line, medial or lateral left hemi-cerebellum. Unaffected and MPS IIIA animals were then grouped according to age and genotype and on the basis of the outcome of neurological exams carried out prior to euthanasia, the MPS IIIA group was further divided into a ‘presymptomatic’ cohort (where no clinical signs referable to the disease were observed) or a ‘symptomatic’ group (where clinical signs of ataxia, hypermetria and wide-based stance were present). The groups are outlined in Table 1. 2.2. Reagents for staining Mouse monoclonal antibodies to ganglioside GM3 (Clone GMR6; #370695) and calbindin-D28K (#C9848) were purchased from Seikagaku Corporation (Tokyo, Japan) and Sigma-Aldrich (St. Louis, MO, USA), respectively. Polyclonal rabbit antibodies to ubiquitin (#Z0458) and glutamic acid decarboxylase 65/67 (GAD65/67; #G5163) were obtained from DAKO (Glostrup, Denmark) and Sigma-Aldrich (St. Louis, MO, USA) respectively. A biotinylated lectin from ricinus communis agglutinin 1 (RCA-1; #b1085) was purchased from Vector Laboratories (Burlingame, CA, USA). FluoroJade B (#AG310) was purchased from Millipore (Temecula, CA, USA). Recombinant terminal deoxynucleotidyl transferase (rTdT; #M1875) and biotin-16-dUTP (#11093070910) for the terminal dUTP nick end-labeling (TUNEL) assay were obtained from Promega Corporation (Wisconsin, USA) and Roche Diagnostics (Mannheim, Germany), respectively. Biotinylated donkey anti-rabbit IgG (#711-065-152) and biotinylated donkey anti-

Symptomatic (N30 months)

Age (months)

Total (n)

Presence of cerebellar lobules 1

2

3

4

5

6

7

8

9

10

● ●

● ● ●

● ● ●

● ● ● ● ● ●

● ● ● ● ● ●

● ● ● ● ●

● ● ● ● ●

● ● ●

● ● ● ● ● ● ● ● ● ●

● ● ● ● ● ● ● ● ● ●

● ● ● ● ● ● ● ● ● ●

6 2.2a 3.6a 33.0 33.0 39.0 67.5a









● ● ● ● ●





● ● ● ● ●

● ● ● ● ● ●

● ● ● ● ● ● ●







10 M F F F M M M F F F

2.2a 4.3 6.0a 6.0a 6.0a 11.6a 30.9a 40.9a 43.7 46.9



● ● ●

● ●

● ● ●







mouse IgG (#715-065-150) secondary antibodies were purchased from Jackson Immunoresearch Laboratories (West Grove, PA, USA). Biotinylated goat anti-mouse IgM (#BA2020) was obtained from Vector Laboratories (Burlingame, CA, USA). 2.3. Immunohistochemistry and histochemical methods Six-micron thick paraffin sections of sagittal cerebellum, cut using a rotary Leica microtome (RM2235), were stained in batches. Briefly, for immunohistochemistry, sections were de-waxed and rehydrated prior to pre-treatment with 0.05% trypsin (Sigma; #T8128) for calbindinD28K. Pre-treatment was not required for ubiquitin and GAD65/67. Endogenous peroxidases were quenched with 3% hydrogen peroxide; following blocking of non-specific proteins with 10% normal serum (specific to the secondary antibody), primary antibodies (1/1000 calbindin-D28K; 1/8000 ubiquitin; and 1/4000 GAD65/67) were incubated overnight at room temperature. Sections were then incubated with the appropriate biotinylated secondary antibody, donkey anti-mouse IgG (1/800) or donkey anti-rabbit (1/800). To visualize staining, sections were then conjugated with avidin (Vector ABC Elite kit; PK6100; Vector Laboratories) and developed using a DAKO liquid DAB chromogen kit (DAKO; #K346811). Histochemical staining of microglia with biotinylated lectin RCA-1 was performed using a modified method from Kiatipattanasakul et al. [20], involving pre-treatment with 0.05% Trypsin and quenching of endogenous peroxidases as described above, followed by overnight incubation of biotinylated lectin RCA-1 (1/2500) with visualization as described above. The sections stained with calbindinD28K, ubiquitin, GAD65/67 and RCA-1 were lightly counterstained with hematoxylin. Using a cryostat (Thermo Electron Corporation, Cheshire, UK), sixmicron thick frozen sections of sagittal posterior cerebellar lobe were cut for immunohistochemical detection of GM3 (1/750) using published methods [21,22]. Samples available for GM3 staining are indicated in Table 1. Frozen tissue from one additional MPS IIIA (25 months) and two unaffected dogs (6 months) became available during the study. The presence of neuronal degeneration in sagittal cerebellar paraffin sections was assessed by FluoroJade B staining using previously published methods [23,24]. Detection of apoptosis via TUNEL assay utilized a method modified from Gavrieli et al. [25] and Portera-Cailliau et al. [26]. Modifications made included blocking endogenous peroxidases with 1% hydrogen peroxide in 10 mM Tris buffer, pH 8.0 and using a reaction solution containing 0.03 units of rTdT and 9 nmol of biotinylated

S. Hassiotis et al. / Molecular Genetics and Metabolism 113 (2014) 283–293

2.4. Quantification of the total number of PKC and PKC expressing calbindin-D28K Identification of cerebellar lobules visualized in hematoxylin and eosin-stained sections of cerebellum was performed in an operatorblinded manner with the aid of a dog brain atlas [27]. An image of the entire cerebellum from all dogs was taken using an Olympus BX61 motorized microscope fitted with an Olympus Colorview III camera and AnalySIS Lifescience Research Cell P software (build 2640; Olympus, Australia). Using the measurement tool, a line was drawn along the entire PKC layer of each cerebellar section from each dog, to determine its length in micrometers (μm). The number of PKC and PKC expressing calbindin-D28K was determined via manual counting by an examiner blinded to the dog age/ genotype. In the PKC layer any large cell with/without a nucleus was interpreted and therefore counted as a PKC. Any PKC exhibiting calbindin-D28K-positive immunoreactivity was deemed calbindinD28K-positive. Data are expressed as the average total number of PKC/mm of PKC layer in entire cerebellar section of each dog. To examine the pattern of PKC loss, and the number of PKC expressing calbindinD28K in each cerebellar lobule, the number of PKC, and the number of calbindin-D28K-positive PKC per mm of PKC layer in each cerebellar lobule was determined, respectively. A descriptive assessment of the distribution and pattern of calbindin-D28K staining was also performed. 2.5. Assessment of other pathology A sample of cerebellar lobule 6 was present in all dogs and therefore chosen to assess disease progression (Table 1). GM3 immunostaining in the PKC layer region, RCA-1 histochemistry in the molecular layer and GAD65/67 immunostaining in the deep cerebellar nuclei were quantified in a blinded manner. Under 200× magnification, ten-fifteen fields per dog for GM3 and RCA-1 staining and two-nine fields per dog for GAD65/67 staining were imaged using an Olympus BX41 microscope fitted with a UC50 color camera. Thresholding was applied to determine the optical density of positive GM3 staining using Olympus AnalySIS Lifescience Cell B software (version 2.8, build 1235; Olympus, Australia). Manual counts of the number of activated RCA-1-positive microglia and number of GAD65/67-positive axonal spheroids (≥ 5 μm in diameter) were performed using touch counting. Data were logtransformed and reported as the average % thresholded area of GM3 per mm2, average number of activated RCA-positive cells per mm2 and average number of GAD65/67-positive axonal spheroids per mm2. Descriptive assessments of GAD65/67, ubiquitin, FluoroJade B and TUNEL staining were made. 2.6. Liquid chromatography–electrospray ionization tandem mass spectrometry (LC–ESI-MS/MS) analysis of a heparan sulfate-derived disaccharide The amount of protein in homogenized cerebellar tissue from cerebellar lobules 4/5 was determined using a micro-BCA Assay kit (#23235; Thermo Scientific, Rockford, Il, USA). To determine the relative level of a heparan sulfate-derived disaccharide, glucosamine-Nsulfate-uronic acid (GlcNS-UA), 100 μg of total protein equivalent was derivatized and analyzed by LC–ESI-MS/MS using published methods [28] with all experimental samples run in one batch. GlcNS-UA was expressed as a relative ratio of the peak area of the disaccharide and the internal standard/microgram of total protein. The coefficient of variation was determined by including eight replicates of MPS IIIA mouse brain tissue. The intra-batch coefficient of variation was calculated to be 3.8%.

2.7. Statistics Where grouped data was assessed, the data were log-transformed and examined using one-way analysis of variance (ANOVA) with posthoc testing using a Bonferroni correction to adjust for multiple group comparisons. Graph Pad Prism (versions 5.01 and 6) software was used. Data are shown as mean ± standard error of the mean and p b 0.05 was determined to be statistically significant. 3. Results 3.1. Cerebellar weights At the time of post-mortem, one symptomatic MPS IIIA dog aged 43.7 months, had visually-apparent atrophy of the cerebellum (circled in Fig. 1). The weights of the cerebella from symptomatic MPS IIIA dogs appeared somewhat lower than those from the unaffected and pre-symptomatic MPS IIIA groups, but this did not reach statistical significance. 3.2. Purkinje cell morphology and loss Pale, granular, slightly translucent eosinophilic storage material was first noted in the somas of some PKC of MPS IIIA animals at 4.3 months of age and was easily identifiable in most PKC in dogs N 11.6 months. Substrate-filled cytosomes within the cytoplasm of the PKC increased in number with disease progression, and at times resulted in the nucleus of the cell being displaced to one side of the cytoplasm (Fig. 2A). PKC somas became distended and hypertrophied dendrites were apparent. Although PKC were observed in MPS IIIA dogs up to and including 30.9 months of age, dogs N 40 months of age exhibited a disorganized interface between the molecular and granule cell layers, and PKC were often absent (Fig. 2C). Proliferation of Bergmann glia was apparent. Increased numbers of engorged glia were seen in the molecular layer in MPS IIIA dogs aged N 40 months (c.f. unaffected dogs) (Figs. 2E, F). The average number of PKC in unaffected and MPS IIIA dog cerebellar lobes is shown in Fig. 2G. We observed a striking loss of PKC in MPS IIIA dogs between 30.9 and 40.9 months of age, with few PKC remaining in the three dogs N 40 months of age. The MPS IIIA dog euthanized at 30.9 months had been noted to exhibit milder ataxia than the other three dogs euthanized N 40 months of age. The number and range of lobules able to be examined were inconsistent between animals, thus a comparison between the number of PKC in anterior (lobules 1–5) and posterior (lobules 6–10) lobes could not be performed. However, we note that there was prominent PKC loss in both anterior and posterior lobes in the MPS IIIA animal aged 46.9 months (Figs. 3C, D; arrows). There was no anterior lobe available for assessment in the other MPS

10 Weight of Cerebellum (expressed as % of total brain weight)

16-dUTP/ml of TdT buffer. Staining was visualized by the method described above using both Vectastain ABC Elite and DAKO liquid DAB chromogen kits. Sections were lightly counterstained with hematoxylin.

285

8 6 4 2 0 Unaffected Pre-symptomatic Symptomatic Heterozygous MPS IIIA MPS IIIA (>30 months) (<12 months) n=2

n=3

n=4

Fig. 1. Cerebellum weight as a percentage of total brain weight. Circled data point indicates an MPS IIIA dog (aged 43.7 months) with visible brain atrophy at post-mortem.

286

S. Hassiotis et al. / Molecular Genetics and Metabolism 113 (2014) 283–293

MPS IIIA

Unaffected

A

B

30.9 mo

33 mo

C

D ML

ML

PKC

PKC GCL

GCL

L

40.9 mo

39 mo

E

F

40.9 mo

39 mo

G 20

Average #PKC/mm of PKC layer in entire cerebellar section

Onset of Clinical Signs

15

10

5

0 2.2 3.6 33

33

39 67.5

2.2 4.3

6

Unaffected

6

6

11.6 30.9 40.9 43.7 46.9

MPS IIIA Age (months)

Fig. 2. Morphology of Purkinje cells (PKC) in hematoxylin and eosin stained sections of cerebellar cortex (A–F) and PKC counts (G). (A) shows a high-powered image of a PKC in a symptomatic MPS IIIA dog aged 30.9 months (mo). Note the accumulation of translucent, highly eosinophilic inclusions (arrowhead) in the distended PKC, which displaces the nucleus (star). (B) PKC in the brain of an unaffected dog aged 33 months. Note the prominent nucleolus (black arrow). (C) Cerebellar cortex of MPS IIIA dog aged 40.9 months reveals a lack of PKC and disorganization of the PKC layer interface (arrow). (D) Same brain area in an unaffected animal aged 39 months, with distinct molecular layer (ML), granule cell layer (GCL) and PKC layer. Arrows depict gaps in the Purkinje cell layer. (E) Engorged glia (arrows) containing storage material in the molecular layer of a symptomatic MPS IIIA dog aged 40.9 months. (F) Same region in an unaffected animal aged 39 months. (G) Quantification of the total number of PKC in the entire cerebellar section of each animal. PKC are present in all unaffected dogs and in MPS IIIA dogs up to and including 30.9 months of age. There is a striking loss of PKC in the MPS IIIA dogs between 30.9 and 40.9 months of age.

S. Hassiotis et al. / Molecular Genetics and Metabolism 113 (2014) 283–293

Posterior Lobe

Anterior Lobe

Number of PKC/mm of PKC layer in each cerebellar lobule

UNAFFECTED

A

287

20

B

15

10

5

Number of PKC/mm of PKC layer in each cerebellar lobule

MPS IIIA

C

Lobule 2

Lobule 3

Lobule 4

Lobule 5

Lobule 6

20

X

Lobule 7

Lobule 8

Lobule 9

X X

X

2.2 3.6 33 33 39 67.5

Lobule 1

X

2.2 3.6 33 33 39 67.5

X

2.2 3.6 33 33 39 67.5

X

2.2 3.6 33 33 39 67.5

X

2.2 3.6 33 33 39 67.5

X

2.2 3.6 33 33 39 67.5

X

2.2 3.6 33 33 39 67.5

X

2.2 3.6 33 33 39 67.5

X X

2.2 3.6 33 33 39 67.5

Cerebellar Lobule #

X

2.2 3.6 33 33 39 67.5

0 X X X X

Age (mo)

Lobule 10

D

15

10

5

Cerebellar Lobule #

X

XX

XX

X

Lobule 1

Lobule 2

Lobule 3

Lobule 4

Lobule 5

X

X

X

X X XX X

2.2 4.3 6 6 6 11.6 30.9 40.9 43.7 46.9 2.2 4.3 6 6 6 11.6 30.9 40.9 43.7 46.9 2.2 4.3 6 6 6 11.6 30.9 40.9 43.7 46.9 2.2 4.3 6 6 6 11.6 30.9 40.9 43.7 46.9 2.2 4.3 6 6 6 11.6 30.9 40.9 43.7 46.9

XX

2.2 4.3 6 6 6 11.6 30.9 40.9 43.7 46.9 2.2 4.3 6 6 6 11.6 30.9 40.9 43.7 46.9 2.2 4.3 6 6 6 11.6 30.9 40.9 43.7 46.9 2.2 4.3 6 6 6 11.6 30.9 40.9 43.7 46.9 2.2 4.3 6 6 6 11.6 30.9 40.9 43.7 46.9

0 X X X X X X XX X X X X X X XX X X X

Age (mo)

Lobule 6

Lobule 7

Lobule 8

Lobule 9

Lobule 10

Fig. 3. Number of PKC in the anterior (A, C; lobules 1–5) and posterior (B, D; lobules 6–10) lobes of the cerebellum of unaffected (A, B) and MPS IIIA dogs (C, D). PKC loss (indicated by arrows) was seen in both the anterior and posterior lobes of MPS IIIA dogs (X indicates no cerebellar lobule available for assessment; mo = months).

IIIA animals aged 40.9 and 43.7 months, however they exhibited PKC loss in the posterior lobes (Fig. 3D; arrows). 3.3. Calbindin-D28K staining of Purkinje cells In both the unaffected and MPS IIIA dogs aged 2.2 months, calbindin-D28K staining of PKC was most intense in the outer (gyral) region of the lobules, with darker staining of dendrites compared to somas. However, PKC in the inner (sulcal) regions exhibited minimal or no calbindin-D28K staining. In unaffected dogs aged 33 and 39 months, calbindin-D28K reactivity was intense in both dendrites and somas of PKC throughout the cerebellar folia, with more dendritic staining noted in sulcal regions compared to gyral regions (Figs. 4A, B). The intensity of calbindin-D28K immunostaining in PKC in the unaffected dog aged 67.5 months was decreased overall, both gyrally and sulcally. In contrast, in the seven MPS IIIA dogs ≤ 30.9 months of age, the amount of calbindin-D28K present in the soma and dendrites of PKC was variable, but generally low, and located predominantly in the gyri with little or no staining in sulcal regions (Figs. 4C, D). Significant PKC loss precluded assessment calbindin-D28K staining in MPS IIIA dogs aged N40 months. Quantification of the number of PKC expressing calbindin-D28K relative to those that did not, is shown in supplementary Fig. S1. Despite variations in the number of lobules available for assessment,

unaffected dogs appeared to reach a point (around 33–39 months of age) where most PKC expressed calbindin-D28K. A subsequent decline in the number of immunoreactive cells then occurred (67.5 months). In contrast, just one six-month-old MPS IIIA dog exhibited calbindinD28K-reactivity in all PKC in cerebellar lobules 5, 6 and 10. All other MPS IIIA dogs never achieved complete PKC expression of calbindinD28K.

3.4. GAD65/67 staining in the cerebellum PKC and other GABAergic neurons in the cerebellum were also stained with antibodies to GAD65/67. GAD65/67-positive axon terminals of (presumptive) basket cells, synapsed on PKC somas and cytoplasmic staining in PKC appeared variable in unaffected dogs aged ≤3.6 months. Intense cytoplasmic staining was noted in unaffected dogs aged 33 and 39 months (Figs. 4E, F) but was observed to decrease at 67.5 months. GAD65/67 immunoreactivity in pre-symptomatic MPS IIIA dogs resembled that in unaffected dogs ≤3.6 months of age. The MPS IIIA dog aged 30.9 months did not exhibit the same level of cytoplasmic staining seen in unaffected animals aged 33 and 39 months of age (Fig. 4G), and in dogs N40 months of age, rows of GAD65/67-positive pinceaus, formed by terminating axons of basket cells were observed (Fig. 4H).

288

S. Hassiotis et al. / Molecular Genetics and Metabolism 113 (2014) 283–293

Calbindin-D28K

B

Calbindin-D28K

Unaffected

A

33 mo Calbindin-D28K

D

Calbindin-D28K

MPS IIIA

C

33 mo

30.9 mo GAD65/67

F

GAD65/67

Unaffected

E

30.9 mo

33 mo GAD65/67

H

GAD65/67

MPS IIIA

G

39 mo

30.9 mo

40.9 mo

Fig. 4. Distribution and staining intensity of calbindin-D28K (A-D) and GAD65/67 (E-H) in the cerebellar cortex. Representative images of calbindin-D28K staining in the PKC layer of cerebellar lobule 6, from the outer (gyral) and inner (sulcal) regions in an unaffected dog aged 33 months (mo) (A, B respectively) and in an MPS IIIA dog aged 30.9 months (C, D respectively). Note the decreased level of calbindin-D28K staining in the MPS IIIA dog (C, D). GAD65/67 staining in the PKC layer of unaffected dogs aged 33 (E) and 39 months (F). Reduced GAD65/67 staining was apparent in the PKC layer of an MPS IIIA dog aged 30.9 months (G). Loss of PKC in the MPS IIIA dog aged 40.9 months resulted in rows of GAD65/67-stained pinceaus (arrows; H).

3.5. Spheroidal inclusions in the cerebellum The presence of spheroid-like inclusions was assessed by ubiquitin, calbindin-D28K and GAD65/67 immunohistochemistry. Minimal numbers of small ubiquitin-reactive inclusions were noted in the deep

cerebellar nuclei and white matter tracts of all unaffected dogs. The ubiquitin staining pattern in the cerebellum of MPS IIIA dogs aged ≤6 months was similar to that in unaffected dogs, with a slight increase in the number of inclusions noted in these areas in dogs aged 11.6 and 30.9 months (data not shown). In N 40 month old MPS IIIA animals,

S. Hassiotis et al. / Molecular Genetics and Metabolism 113 (2014) 283–293

only presumptive macrophage-type glia were found to contain ubiquitin. In the deep cerebellar nuclei of unaffected animals, calbindin-D28Kand GAD65/67-positive axons were noted with varying numbers of focal swellings and varicosities (Figs. 5A, C). However, in MPS IIIA dogs, calbindin-D28K- and GAD65/67-positive spheroid-like inclusions were observed in varying sizes and amounts in animals aged from 4.3 to 30.9 months (Figs. 5B,D). In the white matter tracts proximal to

289

the deep cerebellar nuclei, calbindin-D28K-positive spheroid-like inclusions were numerous in dogs aged 11.6 and 30.9 months, but GAD65/67-positive spheroids were only abundant in the dog aged 11.6 months. Occasional torpedo-shaped structures were noted in the white matter tracts of the MPS IIIA dog aged 11.6 months. Very few calbindin-D28K- and GAD65/67-positive spheroids were observed at N 40 months, presumably due to PKC loss. Quantification of the mean number of GAD65/67-positive axonal spheroids/mm2 in the deep

Unaffected

MPS IIIA

B

Calbindin-D28K

A

33 mo

30.9 mo

C

GAD65/67

D

33 mo

30.9 mo

C lin ic a l S ig n s

Log Y+1

2 .0

1 .5

1 .0

2

s p h e r o id s /m m in t h e d e e p c e r e b e lla r n u c le i

A v e r a g e # G A D 6 5 /6 7 - p o s it iv e a x o n a l

2 .5

0 .5

X

0 .0

A g e (m o n th s )

2 .2

3 .6

33

3 9 6 7 .5

U n a ffe c te d H e te r o z y g o u s

X 2 .2

X 4 .3

6

1 1 .6 3 0 .9 4 0 .9 4 3 .7 4 6 .9 M P S IIIA

Fig. 5. Assessment of axonal spheroidal inclusions in the deep cerebellar nuclei region. Calbindin-D28K and GAD65/67 staining in the deep cerebellar nuclei region of an unaffected dog aged 33 months (mo) (A, C respectively) showing immunoreactive focal swellings and varicosities (arrows), and an MPS IIIA dog aged 30.9 months (B, D respectively), showing the presence of immunopositive axonal spheroids (thick arrows). (E) Quantification of the average number of GAD65/67-positive axonal spheroids/mm2 in the deep cerebellar nuclei (X — tissue unavailable for investigation).

290

S. Hassiotis et al. / Molecular Genetics and Metabolism 113 (2014) 283–293

cerebellar nuclei, revealed a peak at 11.6 months of age in MPS IIIA dogs, and a gradual decline thereafter (Fig. 5E). At their peak, spheroids reached ~ 18 μm diameter. Tissue from one unaffected dog aged 67.5 months and two MPS IIIA dogs aged 2.2 and 40.9 months was not available for this investigation. 3.6. Neurodegeneration and cell death in the cerebellum No degenerating or apoptosing Purkinje cells were observed using FluoroJade B and TUNEL staining (respectively; data not shown). FluoroJade B-positive staining of the vasculature and leptomeningeal tissues was observed in all dogs examined. Astrocyte-like cells in the white matter tracts of all MPS IIIA dogs were positively-stained, as were the radial processes of Bergmann glia in the molecular layer of dogs aged N 40 months. A few TUNEL-positive cells were noted in the granule cell layer and even fewer in the molecular layer of all dogs b6 months of age. TUNEL-positive fragmented DNA was noted in presumptive macrophage-type cells, scattered throughout the granule cell layer of MPS IIIA dogs N 40 months of age. 3.7. Characterization of the time-course of substrate storage in the cerebellum The heparan sulfate-derived disaccharide, GlcNS-UA, was undetectable in unaffected dog cerebellum (Fig. 6A). In contrast, we measured A

Clinical Signs

Relative Disaccharide Peak Area/mg protein in cerebellar homogenate

14

high (relative) levels of GlcNS-UA in the cerebellum of MPS IIIA dogs at the earliest age assessed (2.2 months); this was sustained with disease progression. There was minimal staining of GM3 ganglioside observed in unaffected dog tissue (n = 5; Fig. 6B). In contrast, GM3-positive puncta were detected in the dendrites and somas of PKC, in addition to cells of the molecular and granule cell layers, in MPS IIIA dogs as young as 2.2 months of age (Fig. 6D). Quantification revealed GM3 levels to be significantly elevated in the MPS IIIA dog PKC layer, in both presymptomatic (n = 3; p b 0.05) and symptomatic (n = 3; p b 0.05) animals (c.f. unaffected). Tissues from the MPS IIIA dog aged 25 months exhibited notably less GM3 staining, when compared to the 30.9 and 40.9 month old MPS IIIA dog tissues; anecdotally, this dog had also had less severe ataxia. Two MPS IIIA dogs (aged 2.2 and 11.6 months) were excluded from quantification as cerebellar lobule 6 was unidentifiable in the sections. 3.8. Characterization of the time-course of neuroinflammatory changes in the cerebellum Resting microglia exhibiting small somas with long processes were predominant in the cerebellum of the unaffected dogs examined (n = 6; Figs. 6E, F). In contrast, in MPS IIIA dogs as young as 2.2 months, activated microglia with enlarged amoeboid-like somas and short processes were occasionally observed in the molecular layer (Fig. 6G). At

Unaffected

12

M PS IIIA

10 8 6 4 2 0 0

10

20

30

40

50

60

70

80

Age (months)

B Average %TH Area GM3 in Purkinje cell layer Log Y+1

0.6

Clinical Signs

Unaffected

C

GM3

D

GM3

MPS IIIA

0.5 0.4 0.3 0.2 0.1

2.2 mo

2.2 mo

0.0 0

10

20

30

40

50

60

70

80

Age (months) C lin ic a l S i g n s

3

U n a ffe c te d

F

RCA-1

G

RCA-1

M P S IIIA

2 Log Y+1

A v e r a g e # a c t iv a t e d R C A - + v e

c e ll s /m m 2 i n M o le c u l a r L a y e r

E

1

33 mo

30.9 mo

0 0

10

20

30

40

50

60

70

80

A g e (m o n th s )

Fig. 6. Time-course of substrate storage and neuroinflammatory changes in the cerebellum of unaffected and MPS IIIA dogs. (A) Relative amount of a heparan sulfate-derived disaccharide, glucosamine-N-sulfate-uronic acid (GlcNS-UA)/mg of protein in the cerebellum of unaffected and MPS IIIA dogs. (B) Quantification of average amount of GM3 ganglioside in the PKC layer in unaffected and MPS IIIA dogs. Representative images of GM3 staining in the PKC layer of an unaffected and an MPS IIIA dog aged 2.2 months (mo) (C, D respectively). A PKC (arrowhead) and other cell types (arrows) are highlighted. (E) Quantification of the average number of activated RCA-positive microglia in the molecular layer. Representative images showing RCA-1 staining of microglia in an unaffected dog aged 33 months (F) and an MPS IIIA dog aged 30.9 months (G). Insets: resting and activated microglia (F and G, respectively).

S. Hassiotis et al. / Molecular Genetics and Metabolism 113 (2014) 283–293

291

Onset of Clinical Signs

11.6 months of age, increased numbers were observed in this area, with occasional activated microglia also noted in both the purkinje and granule cell layers. Statistical analysis revealed that symptomatic MPS IIIA dogs exhibited significantly more activated microglia in the molecular layer than pre-symptomatic dogs (p b 0.05).

GlcNS-UA GM3

4. Discussion This study aimed to determine the timing and pattern of PKC loss and describe the appearance of other cerebellar pathology, in relation to the emergence of clinical signs in the MPS IIIA dog. For the first time, we demonstrate that PKC remain present up to and including 30 months of age in MPS IIIA dogs, but are precipitously lost thereafter, with no more than ~ 12% remaining after 40 months of age. This coincides with the emergence of clinical symptoms of ataxia. Whilst PKC loss does occur in MPS III mice, when lifespan is taken into account, it appears to be a late-disease stage phenomenon (≥ 8 months; [14,29,30]). PKC loss has also been described in other LSD models: murine INCL [15] and NP-A/B/C [16–19], α-mannosidosis guinea pigs [13], and canine MPS IIIB [10] and fucosidosis [11,12]. Although additional data from MPS IIIA dogs aged 30–40 months is required, our findings indicate global, rather than an anterior–posterior gradient of PKC loss. The pattern of PKC loss was not examined in MPS IIIB dogs [10] or MPS IIIA/E mice [29,30], however both anterior and posterior lobes exhibited statistically significant decreases in PKC in the MPS IIIB mouse [14]. A gradient of PKC loss was observed in the INCL, NP-A/B/C and α-mannosidosis models and we note that PKC death can also occur randomly or posteriorly, as reported in the tambaleante and nervous mouse models, respectively [31]. In order to examine the mechanism of PKC death, TUNEL staining was applied, however TUNEL-positive PKC were not observed in MPS IIIA cerebellum. Apoptotic processes occur over a relatively short time frame (12–24 h in vivo; 2–3 h in vitro) and are dependent on tissue type [25,32], therefore whilst PKC were present in MPS IIIA dogs at 30.9 months, they were absent at 40.9 months so we may have missed the ‘window’ for TUNEL-reactivity. PKC undergoing apoptotic cell death have been observed in murine and human NP-C [33] and murine Sandhoff disease [34]. TUNEL-positive cerebellar granule cells were also seen in murine and human late-INCL [15,35]. The relationship between the appearance of various pathological lesions in MPS IIIA dog cerebellum, the loss of PKC and the onset of clinical signs is summarized in Fig. 7, which demonstrates the time-course of the change in the number of PKC, the accumulation of primary and secondary substrates and the appearance of microgliosis and axonal spheroids, in relation to the onset of clinical signs. We have utilized data from MPS IIIA mice [36,37], to extrapolate the GlcNS-UA and GM3 time-course prior to 2.2 months of age. The loss of PKC was preceded by the accumulation of both primary (heparan sulfate) and secondary (GM3) substrates, although the levels of both appeared to change little between pre- and post-symptomatic disease stages. Exogenous application of heparan sulfate fragments to microglia in culture [38], or ganglioside accumulation in the brain [39] , activates microglia and elicits an inflammatory response, mediated via the toll-like receptor 4 signaling pathway. The activation of both astrocytes (Bergmann glia) and microglia in the MPS IIIA dog cerebellum may ultimately result in non-cell autonomous PKC death. We note that there was a significant increase in the number of activated microglia seen in the molecular layer as dogs moved from pre- to clinical stages. In murine INCL, microglial activation only occurs after PKC begin to disappear [15], suggesting that the immune cell activation in this condition is a response to PKC death, not the cause of it. Further, PKC in NP-C conditional mutants (with NP-C deficiency in PKC only) died at approximately the same rate as global NP-C mutants, suggesting that glia are not involved in PKC cell death in this disorder either [16]. The contribution of microglial activation to PKC death in MPS IIIA therefore warrants further attention.

Microgliosis

Axonal Spheroids

Number of Purkinje cells

10

20

30 40 Age (months)

50

60

Fig. 7. Proposed time-course of pathological alterations with disease progression in the MPS IIIA dog cerebellum. Alterations in the number of PKC observed with age, changes in primary (heparan sulfate-derived disaccharide, glucosamine-N-sulfate-uronic acid (GlcNS-UA)) and secondary (ganglioside GM3) substrate levels and the number of activated microglia and axonal spheroids seen. Dotted lines represent interpretation of potential change after assessment of analogous mouse data.

Unaffected dogs appeared to reach an age when the majority of PKC became calbindin-D28K-positive, an event that did not occur in MPS IIIA PKC. Developmental changes in calbindin-D28K expression [40] also correlate with important stages in PKC maturation, such as synaptogenesis and cytoskeleton formation [41]. Therefore, calbindin-D28K deficiency in MPS IIIA dog cerebellum could have ramifications for normal cell development and function. Spinocerebellar ataxia type 1 transgenic mice with a calbindin-null mutation in their PKC exhibit exacerbated disease progression [42]. Similarly, loss of calbindin in cholinergic neurons in the basal forebrain increases vulnerability to phosphorylated tau accumulation, tangle formation, and degeneration resulting in Alzheimer's disease [43]. These findings suggest that calbindin loss in certain neuronal populations makes them more vulnerable to cell autonomous degeneration and death, which may be a factor contributing to the loss of PKC observed here. Further investigations in the MPS IIIA dog are required to examine (a) whether other calcium-binding proteins compensate for the lack of calbindin, and (b) whether PKC are electrophysiologically normal. The development of axonal spheroids may also have contributed to autonomous death of PKC. These swellings in the axon are often seen in the nodes of Ranvier and appear to begin in distal portion of the axon. Here, we noted them in the deep cerebellar nuclei of pre-symptomatic (4.3 month-old) MPS IIIA dogs; not only did the number and size of the lesions increase with disease progression, but lesions also began to appear in the cerebellar white matter tracts as the dogs aged. GAD65/67-positive axonal spheroids in PKC have been described in other LSD such as NP-C, mucolipidosis IV, GM1/2 gangliosidosis and α-mannosidosis [44–46], and these lesions, which appear to contain organelles, multi-vesicular structures and a variety of different proteins, likely result from deficits in axonal transport. As discussed in the informative review by Walkley and colleagues [47], GABAergic neurons appear particularly vulnerable to spheroid formation and once established, these structures are likely to further impair axonal trafficking. It is therefore conceivable that they contribute to declining PKC health and function and cell-autonomous death. This hypothesis requires further investigation. Most pathological changes occur much earlier than the onset of clinical signs. These observations in the MPS IIIA dog concur with those

292

S. Hassiotis et al. / Molecular Genetics and Metabolism 113 (2014) 283–293

reported for other neurodegenerative LSD e.g. α-mannosidosis in guinea pigs [13] and murine Niemann–Pick A [17], where substrate accumulation precedes the onset of clinical symptoms. Previously published studies of the MPS IIIA mouse model [36,37] also indicate that heparan sulfate levels in the mouse brain are significantly elevated at birth, but neurological symptoms do not appear in that model for several months thereafter. Whilst there may be clinical changes that occur early in the course of disease in the canine model of MPS IIIA e.g. subtle cognitive deficiencies which may only become evident when sensitive behavioral testing is employed, the accumulation of primary and secondary substrates, or other pathological lesions may also need to reach a certain ‘threshold’ before cellular dysfunction ensues. 5. Conclusion This is the first study correlating the temporal appearance of cerebellar lesions and clinical changes in the MPS IIIA dog. Near complete loss of PKC occurs between ~ 31-40 months of age, co-incident with the appearance of motor dysfunction. Primary and secondarily-stored substrate accumulation, inflammatory cell activation and the presence of axonal spheroids in PKC axons greatly precede the death of these neurons and potentially renders them vulnerable to cell autonomous cell death processes. We suggest that future studies exploring pathways to PKC death use animal models with targeted SGSH gene deletion in either PKC or glia in order to determine their respective contributions to PKC loss. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ymgme.2014.10.008. Funding and role of funding source The work presented here formed the basis of a Bachelor of Health Science Honors project by S. Hassiotis. The Lysosomal Diseases Research Unit funded this project, and received a contribution toward costs from the University of Adelaide. The University of Adelaide had no role in the study design, collection, analysis and interpretation of data, writing of the manuscript and the decision to submit the manuscript for publication. Conflict of interest The authors declare no conflict of interest. Acknowledgments We wish to thank Dr Neil Marshall for the care of and information on the MPS IIIA dogs; Professor Peter Blumbergs and Mr Jim Manavis for assistance in histology and providing TUNEL methodology; and Ms Leanne Winner and Dr Maria Fuller for derivatizing and running cerebellar samples for tandem mass spectrometry (respectively). References [1] E.F. Neufeld, J. Muenzer, The mucopolysaccharidoses, in: C.R. Scriver (Ed.), The metabolic and molecular bases of inherited diseases, McGraw-Hill, New York, 2001, pp. 3421–3452. [2] J.L. Avila, J. Convit, Inhibition of Leucocytic Lysosomal Enzymes by Glycosaminoglycans In Vitro, Biochem. J. 152 (1975) 57–64. [3] M.J. Valstar, S. Neijs, H.T. Bruggenwirth, R. Olmer, G.J. Ruijter, R.A. Wevers, O.P. van Diggelen, B.J. Poorthuis, D.J. Halley, F.A. Wijburg, Mucopolysaccharidosis type IIIA: clinical spectrum and genotype-phenotype correlations, Ann. Neurol. 68 (2010) 876–887. [4] P.J. Meikle, J.J. Hopwood, A.E. Clague, W.F. Carey, Prevalence of lysosomal storage disorders, JAMA 281 (1999) 249–254. [5] R.D. Jolly, F.J. Allan, M.G. Collett, T. Rozaklis, V.J. Muller, J.J. Hopwood, Mucopolysaccharidosis IIIA (Sanfilippo syndrome) in a New Zealand Huntaway dog with ataxia, N. Z. Vet. J. 48 (2000) 144–148. [6] R.D. Jolly, A.C. Johnstone, E.J. Norman, J.J. Hopwood, S.U. Walkley, Pathology of mucopolysaccharidosis IIIA in Huntaway dogs, Vet. Pathol. 44 (2007) 569–578.

[7] G. Yogalingam, T. Pollard, B. Gliddon, R.D. Jolly, J.J. Hopwood, Identification of a mutation causing mucopolysaccharidosis type IIIA in New Zealand Huntaway dogs, Genomics 79 (2002) 150–153. [8] K.M. Hemsley, E.J. Norman, A.C. Crawley, D. Auclair, B. King, M. Fuller, D.L. Lang, C.J. Dean, R.D. Jolly, J.J. Hopwood, Effect of cisternal sulfamidase delivery in MPS IIIA Huntaway dogs — a proof of principle study, Mol. Genet. Metab. 98 (2009) 383–392. [9] A.C. Crawley, N. Marshall, H. Beard, S. Hassiotis, V. Walsh, B. King, N. Hucker, M. Fuller, R.D. Jolly, J.J. Hopwood, K.M. Hemsley, Enzyme replacement reduces neuropathology in MPS IIIA dogs, Neurobiol. Dis. 43 (2011) 422–434. [10] N.M. Ellinwood, P. Wang, T. Skeen, N.J. Sharp, M. Cesta, S. Decker, N.J. Edwards, I. Bublot, J.N. Thompson, W. Bush, E. Hardam, M.E. Haskins, U. Giger, A model of mucopolysaccharidosis IIIB (Sanfilippo syndrome type IIIB): N-acetyl-alpha-Dglucosaminidase deficiency in Schipperke dogs, J. Inherit. Metab. Dis. 26 (2003) 489–504. [11] G.S. Kondagari, P. Ramanathan, R. Taylor, Canine fucosidosis: a neuroprogressive disorder, Neurodegener. Dis. 8 (2011) 240–251. [12] G.S. Kondagari, J. Yang, R.M. Taylor, Investigation of cerebrocortical and cerebellar pathology in canine fucosidosis and comparison to aged brain, Neurobiol. Dis. 41 (2011) 605–613. [13] A.C. Crawley, S.U. Walkley, Developmental analysis of CNS pathology in the lysosomal storage disease alpha-mannosidosis, J. Neuropathol. Exp. Neurol. 66 (2007) 687–697. [14] C.D. Heldermon, A.K. Hennig, K.K. Ohlemiller, J.M. Ogilvie, E.D. Herzog, A. Breidenbach, C. Vogler, D.F. Wozniak, M.S. Sands, Development of sensory, motor and behavioral deficits in the murine model of Sanfilippo syndrome type B, PLoS One 2 (2007) e772. [15] S.L. Macauley, D.F. Wozniak, C. Kielar, Y. Tan, J.D. Cooper, M.S. Sands, Cerebellar pathology and motor deficits in the palmitoyl protein thioesterase 1-deficient mouse, Exp. Neurol. 217 (2009) 124–135. [16] M.J. Elrick, C.D. Pacheco, T. Yu, N. Dadgar, V.G. Shakkottai, C. Ware, H.L. Paulson, A.P. Lieberman, Conditional Niemann–Pick C mice demonstrate cell autonomous Purkinje cell neurodegeneration, Hum. Mol. Genet. 19 (2010) 837–847. [17] S.L. Macauley, R.L. Sidman, E.H. Schuchman, T. Taksir, G.R. Stewart, Neuropathology of the acid sphingomyelinase knockout mouse model of Niemann–Pick A disease including structure-function studies associated with cerebellar Purkinje cell degeneration, Exp. Neurol. 214 (2008) 181–192. [18] J. Sarna, S.R. Miranda, E.H. Schuchman, R. Hawkes, Patterned cerebellar Purkinje cell death in a transgenic mouse model of Niemann–Pick type A/B disease, Eur. J. Neurosci. 13 (2001) 1873–1880. [19] J.R. Sarna, M. Larouche, H. Marzban, R.V. Sillitoe, D.E. Rancourt, R. Hawkes, Patterned Purkinje cell degeneration in mouse models of Niemann–Pick type C disease, J. Comp. Neurol. 456 (2003) 279–291. [20] W. Kiatipattanasakul, H. Nakayama, S. Nakamura, K. Doi, Lectin histochemistry in the aged dog brain, Acta Neuropathol. 95 (1998) 261–268. [21] R. McGlynn, K. Dobrenis, S.U. Walkley, Differential subcellular localization of cholesterol, gangliosides, and glycosaminoglycans in murine models of mucopolysaccharide storage disorders, J. Comp. Neurol. 480 (2004) 415–426. [22] A.A. Lau, H. Hannouche, T. Rozaklis, S. Hassiotis, J.J. Hopwood, K.M. Hemsley, Allogeneic stem cell transplantation does not improve neurological deficits in mucopolysaccharidosis type IIIA mice, Exp. Neurol. 225 (2010) 445–454. [23] L.C. Schmued, C. Albertson, W. Slikker Jr., Fluoro-Jade: a novel fluorochrome for the sensitive and reliable histochemical localization of neuronal degeneration, Brain Res. 751 (1997) 37–46. [24] L.C. Schmued, K.J. Hopkins, Fluoro-Jade B: a high affinity fluorescent marker for the localization of neuronal degeneration, Brain Res. 874 (2000) 123–130. [25] Y. Gavrieli, Y. Sherman, S.A. Ben-Sasson, Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation, J. Cell Biol. 119 (1992) 493–501. [26] C. Portera-Cailliau, C.H. Sung, J. Nathans, R. Adler, Apoptotic photoreceptor cell death in mouse models of retinitis pigmentosa, Proc. Natl. Acad. Sci. U. S. A. 91 (1994) 974–978. [27] M. Singer, The Brain of the Dog in Section, W.B. Saunders Company, Philadelphia and London, 1962. [28] K.M. Hemsley, A.J. Luck, A.C. Crawley, S. Hassiotis, H. Beard, B. King, T. Rozek, T. Rozaklis, M. Fuller, J.J. Hopwood, Examination of intravenous and intra-CSF protein delivery for treatment of neurological disease, Eur. J. Neurosci. 29 (2009) 1197–1214. [29] B. Kowalewski, W.C. Lamanna, R. Lawrence, M. Damme, S. Stroobants, M. Padva, I. Kalus, M.A. Frese, T. Lubke, R. Lullmann-Rauch, R. D'Hooge, J.D. Esko, T. Dierks, Arylsulfatase G inactivation causes loss of heparan sulfate 3-O-sulfatase activity and mucopolysaccharidosis in mice, Proc. Natl. Acad. Sci. U. S. A. 109 (2012) 10310–10315. [30] A. Ruzo, M. Garcia, A. Ribera, P. Villacampa, V. Haurigot, S. Marco, E. Ayuso, X.M. Anguela, C. Roca, J. Agudo, D. Ramos, J. Ruberte, F. Bosch, Liver production of sulfamidase reverses peripheral and ameliorates CNS pathology in mucopolysaccharidosis IIIA mice, Mol. Ther. 20 (2012) 254–266. [31] M. Wassef, C. Sotelo, B. Cholley, A. Brehier, M. Thomasset, Cerebellar mutations affecting the postnatal survival of Purkinje cells in the mouse disclose a longitudinal pattern of differentially sensitive cells, Dev. Biol. 124 (1987) 379–389. [32] A. Saraste, Morphologic criteria and detection of apoptosis, Herz 24 (1999) 189–195. [33] Y.P. Wu, H. Mizukami, J. Matsuda, Y. Saito, R.L. Proia, K. Suzuki, Apoptosis accompanied by up-regulation of TNF-alpha death pathway genes in the brain of Niemann– Pick type C disease, Mol. Genet. Metab. 84 (2005) 9–17. [34] Y.P. Wu, R.L. Proia, Deletion of macrophage-inflammatory protein 1 alpha retards neurodegeneration in Sandhoff disease mice, Proc. Natl. Acad. Sci. U. S. A. 101 (2004) 8425–8430.

S. Hassiotis et al. / Molecular Genetics and Metabolism 113 (2014) 283–293 [35] Y. Hachiya, M. Hayashi, S. Kumada, A. Uchiyama, K. Tsuchiya, K. Kurata, Mechanisms of neurodegeneration in neuronal ceroid-lipofuscinoses, Acta Neuropathol. 111 (2006) 168–177. [36] A.C. Crawley, B.L. Gliddon, D. Auclair, S.L. Brodie, C. Hirte, B.M. King, M. Fuller, K.M. Hemsley, J.J. Hopwood, Characterization of a C57BL/6 congenic mouse strain of mucopolysaccharidosis type IIIA, Brain Res. 1104 (2006) 1–17. [37] B. King, P. Savas, M. Fuller, J. Hopwood, K. Hemsley, Validation of a heparan sulfatederived disaccharide as a marker of accumulation in murine mucopolysaccharidosis type IIIA, Mol. Genet. Metab. 87 (2006) 107–112. [38] J. Ausseil, N. Desmaris, S. Bigou, R. Attali, S. Corbineau, S. Vitry, M. Parent, D. Cheillan, M. Fuller, I. Maire, M.T. Vanier, J.M. Heard, Early neurodegeneration progresses independently of microglial activation by heparan sulfate in the brain of mucopolysaccharidosis IIIB mice, PLoS One 3 (2008) e2296. [39] I. Jou, J.H. Lee, S.Y. Park, H.J. Yoon, E.H. Joe, E.J. Park, Gangliosides trigger inflammatory responses via TLR4 in brain glia, Am. J. Pathol. 168 (2006) 1619–1630. [40] S. Siso, S. Tort, C. Aparici, L. Perez, E. Vidal, M. Pumarola, Abnormal neuronal expression of the calcium-binding proteins, parvalbumin and calbindin D-28k, in aged dogs, J. Comp. Pathol. 128 (2003) 9–14. [41] A.M. Iacopino, W.B. Rhoten, S. Christakos, Calcium binding protein (calbindin-D28k) gene expression in the developing and aging mouse cerebellum, Brain Res. Mol. Brain Res. 8 (1990) 283–290.

293

[42] P.J. Vig, J. Wei, Q. Shao, M.E. Lopez, R. Halperin, J. Gerber, Suppression of calbindinD28k expression exacerbates SCA1 phenotype in a disease mouse model, Cerebellum 11 (2012) 718–732. [43] D. Riascos, D. de Leon, A. Baker-Nigh, A. Nicholas, R. Yukhananov, J. Bu, C.K. Wu, C. Geula, Age-related loss of calcium buffering and selective neuronal vulnerability in Alzheimer's disease, Acta Neuropathol. 122 (2011) 565–576. [44] 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 (Berl) 94 (1997) 164–172. [45] M.C. Micsenyi, K. Dobrenis, G. Stephney, J. Pickel, M.T. Vanier, S.A. Slaugenhaupt, S.U. Walkley, Neuropathology of the Mcoln1(−/−) knockout mouse model of mucolipidosis type IV, J. Neuropathol. Exp. Neurol. 68 (2009) 125–135. [46] 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. [47] S.U. Walkley, J. Sikora, M. Micsenyi, C. Davidson, K. Dobrenis, Lysosomal compromise and brain dysfunction: examining the role of neuroaxonal dystrophy, Biochem. Soc. Trans. 38 (2010) 1436–1441.