• Open Access

GM2 Gangliosidosis Variant 0 (Sandhoff-Like Disease) in a Family of Toy Poodles

Authors


Corresponding author: Osamu Yamato, DVM, PhD, Laboratory of Clinical Pathology, Department of Veterinary Clinical Sciences, Faculty of Agriculture, Kagoshima University, 1-21-24 Kohrimoto, Kagoshima 890-0065, Japan; e-mail: osam@agri.kagoshima-u.ac.jp.

Abstract

Background: GM2 gangliosidosis variant 0 (human Sandhoff disease) is a lysosomal storage disorder caused by deficiencies of acid β-hexosaminidase (Hex) A and Hex B because of an abnormality of the β-subunit, a common component in these enzyme molecules, which is coded by the HEXB gene.

Objective: To describe the clinical, pathological, biochemical, and magnetic resonance imaging (MRI) findings of Sandhoff-like disease identified in a family of Toy Poodles.

Animals: Three red-haired Toy Poodles demonstrated clinical signs including motor disorders and tremor starting between 9 and 12 months of age. The animals finally died of neurological deterioration between 18 and 23 months of age. There were some lymphocytes with abnormal cytoplasmic vacuoles detected.

Methods: Observational case study.

Results: The common MRI finding was diffuse T2-hyperintensity of the subcortical white matter in the cerebrum. Bilateral T2-hyperintensity and T1-hypointensity in the nucleus caudatus, and atrophic findings of the cerebrum and cerebellum, were observed in a dog in the late stage. Histopathologically, swollen neurons with pale to eosinophilic granular materials in the cytoplasm were observed throughout the central nervous system. Biochemically, GM2 ganglioside had accumulated in the brain, and Hex A and Hex B were deficient in the brain and liver. Pedigree analysis demonstrated that the 3 affected dogs were from the same family line.

Conclusions and Clinical Importance: The Sandhoff-like disease observed in this family of Toy Poodles is the 2nd occurrence of the canine form of this disease and the 1st report of its identification in a family of dogs.

Abbreviations:
CSF

cerebrospinal fluid

Hex

β-hexosaminidase

LFB

Luxol fast blue

MRI

magnetic resonance imaging

ORF

open reading frame

TLC

thin-layer chromatography

Lysosomal diseases are a group of rare genetic disorders of cellular catabolism.1,2 GM2 gangliosidoses comprise a group of lysosomal diseases caused by excessive accumulation of GM2 ganglioside and related glycolipids in lysosomes, especially in the lysosomes of neurons.3 These diseases are inherited as autosomal recessive traits. The major clinical signs are progressive motor and psychointellectual dysfunctions, startle response, and blindness. The hydrolysis of GM2 ganglioside is catalyzed by lysosomal acid β-hexosaminidase (Hex), but this hydrolysis requires that the ganglioside be complexed with a substrate-specific cofactor, the GM2 activator protein. There are 2 main isoenzymes of Hex: Hex A, a heterodimer of αβ subunits, and Hex B, a homodimetric structure ββ, and only Hex A can act on the complex of GM2 ganglioside and GM2 activator protein. Defects in any of the 3 related genes may lead to GM2 gangliosidosis: HEXA, which encodes the α-subunit of Hex A; HEXB, which encodes the β-subunit of Hex A and Hex B; and GM2A, which encodes the GM2 activator protein. As a result, there are 3 main forms of GM2 gangliosidosis: Tay-Sachs disease (B variant), resulting from mutations of the HEXA gene, is associated with deficient activity of Hex A but normal Hex B; Sandhoff disease (0 variant), resulting from mutations of the HEXB gene, is associated with deficient activity of both Hex A and Hex B; and AB variant or GM2 activator protein deficiency caused by mutations of the GM2A gene, characterized by normal Hex A and Hex B but the inability to form a functional complex of GM2 ganglioside and GM2 activator protein, required for hydrolysis of GM2 ganglioside.

In animals, naturally occurring GM2 gangliosidosis has been reported in dogs,4–8 cats,9–13 pigs,14,15 deer,16 and flamingos.17 In dogs, GM2 gangliosidosis corresponding to human Sandhoff disease has been identified only in a Golden Retriever.8 As other variants of canine GM2 gangliosidosis, an increase of total Hex activity was reported in a Japanese Spaniel6 and a mixed-breed dog,7 suggesting that these cases may have been affected by GM2 activator protein deficiency or AB variant. A form of canine GM2 gangliosidosis identified first in German Shorthair Pointers4 was suspected with B1 variant that is allelic with human Tay-Sachs disease or B variant.18,19

The present paper describes the clinical, biochemical, pathological, and magnetic resonance imaging (MRI) characteristics of canine Sandhoff disease observed in 3 dogs in a family of Toy Poodles. This is the 2nd report of canine Sandhoff disease, and the 1st identification of this disease in a family of dogs.

Materials and Methods

Cases

A 13-month-old, 2.6-kg, female Toy Poodle (animal 1) with a red coat was referred to Tamura Animal Clinic for evaluation of stiff gait in the hindlimbs and frequent falling starting at 12 months of age. The dog needed a long time to eat and vomited foamy stomach fluid once to several times a day. On neurological examination, wide-based stance in the hindlimbs, ataxia, intention tremor, and postural deficit in the hindlimbs were recorded. Hematological examination showed mild lymphocytosis (5226/μL). A blood smear stained with Wright-Giemsa stain demonstrated the presence of abnormal cytoplasmic vacuoles sometimes including azurophilic granules in approximately 10% of the lymphocytes (Fig 1A) and excessive vacuoles in the monocytes (Fig 1B). Whole body radiographic examination did not show any abnormality. Cerebrospinal fluid (CSF) was transparent and showed normal cell numbers (2/μL). At that time, a degenerative disease was suspected, but no treatment was attempted based on the owner's request. Eight months after the 1st presentation, animal 1 was brought back to our clinic at 21 months of age because of astasia, anorexia, kyoodling, and panophthalmitis. The dog was thin (1.5 kg) and stuporous with bilateral panophthalmitis. The dog was euthanized at the owner's request soon after the 2nd MR examination, and necropsy was performed. Histopathological examination was limited to the brain at the owner's request.

Figure 1.

 Vacuolated cytoplasm of a circulating lymphocyte (A) and monocyte (B) from a dog (animal 1) with Sandhoff-like disease. Wright-Giemsa stain, bar=10 μm.

A 15-month-old, 2.1-kg, female red Toy Poodle (animal 2) was referred to Tamura Animal Clinic for evaluation of frequent vomiting starting at 9 months of age and falling starting at 11 months of age. The dog needed a long time to eat. The dog also shook its head frequently. Neurological examination demonstrated drowsiness, wide-based stance in the hindlimbs, ataxia, intention tremor, postural deficit in all 4 limbs, decreased bilateral corneal reflex, and absence of bilateral menace response. Standard hematological examination demonstrated abnormal cytoplasmic vacuoles in the lymphocytes on blood smear. Whole body radiographic examination did not show any abnormality. At this 1st presentation, MRI was not performed because aspiration pneumonia was suspected because of frequent vomiting. At 16 months of age, animal 2 could not stand on a slippery floor, fell frequently, was blind, and had a reduced sense of smell. At 17–18 months of age, animal 2 was lethargic and could not drink water by itself because of severe tremor of the head and limbs. At 19 months of age, corneal ulcer developed in the left eye and a 3rd-eyelid flap was placed for treatment. Animal 2 could still recognize her owner at 21 months of age, but finally died of neurological deterioration in a recumbent posture at 23 months of age. Necropsy was performed with the owner's permission 6 hours after the death and fresh tissues for biochemical analysis were stored in a freezer.

A 9-month-old male red Toy Poodle (animal 3) was referred to the Animal Medical Center of Yamaguchi University for evaluation of head tremor. On neurological examination, ataxia and intention tremor were recorded. Standard hematological examination did not show any abnormality, and lymphocyte morphology on blood smear was not evaluated. This was the only examination and we were thereafter informed that animal 3 finally died at 18 months of age. Necropsy was not performed.

MRI Sequences

MRI of the brain was obtained with a 0.2 T systema in animal 1 at 13 months of age; a 0.3 T systemb in animal 1 at 21 months of age and animal 2 at 19 months of age; and a 0.2 T systemc in animal 3 at 9 months of age. Sequences included a fast spin-echo T2-weighted and spin-echo T1-weighted transverse and sagittal images. MRI was recorded under general anesthesia, and CSF was collected via the cistern magna while the animal was still under anesthesia.

Histopathological Study

Tissue samples of the brains and spinal cords from animals 1 and 2 and visceral organs from animal 2 were fixed in 10% formalin for histological study. Paraffin-embedded sections were prepared by the standard method and stained with hematoxylin and eosin, Luxol fast blue (LFB), and Sudan black B.

Biochemical Analysis

The ganglioside concentration was estimated with specimens of cerebrum from animals 1 and 2. Analysis of gangliosides was performed by a thin-layer chromatographic (TLC) method11 with a slight modification. The TLC plate was developed with chloroform-methanol-0.2% calcium chloride (45 : 55: 10, v/v/v). Measurements of lysosomal enzyme activities in the cerebrum and liver of animal 2 were performed by a fluorometric method with each artificial substrate. The β-galactosidase activity in tissue homogenates was measured with 4-methylumbelliferyl β-d-galactoside as a substrate.20 The Hex activity was measured with 4-methylumbelliferyl N-acetyl-β-d-glucosaminide as a substrate.21 The activity of the unheated sample gave the total Hex activity, while activity of the heated sample (50°C for 3 hours) gave that of Hex B. The activity of Hex A was calculated by determining the difference between these 2 values.

Pedigree Analysis

Pedigree analysis was performed based on information obtained from interviewing the breeder who produced animals 1 and 2, as well as the pedigree papers of animals 1–3 and the studbooks issued by the Japan Kennel Club.

Results

MRI Findings

In animal 1 at 13 months of age, signal intensity in the white matter of the cerebrum was higher than that in the gray matter on T2-weighted image (Fig 2A). At 21 months of age, the MRI showed dilated cerebral sulci and enlarged lateral ventricles in the transverse images (Fig 2B,C), and enlarged 4th ventricle and atrophy of the cerebellum on sagittal image (Fig 2D). Diffuse T2-hyperintensity in the subcortical white matter (Fig 2B,D,E), and T2-hyperintensity (Fig 2E) and T1-hypointensity in the nucleus caudatus (Fig 2F) were observed. In animal 2 at 19 months of age and animal 3 at 9 months of age, diffuse T2-hyperintensity of the subcortical white matter compared with the gray matter was also observed, but there were no apparent lesions in the nucleus caudatus (data not shown).

Figure 2.

 Magnetic resonance images from a dog (animal 1) with Sandhoff-like disease. (A) Transverse T2-weighted image (TR/TE=4,000/120 ms, 0.2 T system) at the level of the thalamus at 13 months of age, (B) transverse T2-weighted image (TR/TE=4,000/120 ms, 0.3 T system) at the level of the thalamus at 21 months of age, (C) transverse T1-weighted image (TR/TE=500/20 ms, 0.3 T system) at the level of the thalamus at 21 months of age, (D) sagittal T2-weighted image at 21 months of age, (E) transverse T2-weighted image at the level of the nucleus caudatus (arrows) at 21 months of age, and (F) transverse T1-weighted image at the level of the nucleus caudatus (arrows) at 21 months of age.

Pathological Findings

Grossly, there were no abnormalities in either the brain or visceral organs in animal 1 at 21 months of age or animal 2 at 23 months of age. Histopathological examination demonstrated that the neuronal cells were swollen like balloons throughout the central nervous system in animal 1 (Fig 3A–D). The swollen neurons demonstrated pale to eosinophilic granular materials in the cytoplasm, which were stained positively with LFB (Fig 3A,B) and Sudan black B (Fig 3C). These neuronal changes were also observed in the spinal ganglions (Fig 3D). Similar pathological lesions were also found in the nervous system of animal 2 (Fig 3E). In visceral organs other than the brain and spinal cord, hepatocytes showed many eosinophilic granules in their cytoplasm, and Kupffer cells were swollen with vacuoles in their cytoplasm (Fig 3F). The infiltration of vacuolated macrophages was observed in the white pulp of the spleen and lymphatic sinus of the lymph nodes (data not shown).

Figure 3.

 Histopathology of dogs with Sandhoff-like disease. (A) Cerebral cortex from animal 1, Luxol fast blue and hematoxylin and eosin (LFB-HE), bar=50 μm; (B) cerebellum from animal 1, LFB-HE, bar=50 μm; (C) medulla oblongata from animal 1, Sudan black B, bar=100 μm; (D) spinal ganglion from animal 1, HE, bar=100 μm; (E) cerebral cortex from animal 2, HE, bar=50 μm; and (F) the liver from animal 2, HE, bar=25 μm.

Biochemical Findings

Based on estimation of the ganglioside concentration by TLC analysis (Fig 4), GM2 ganglioside running faster than the other main gangliosides such as GM1, GD1a, GD1b, and GT1b gangliosides, were abundant in animals 1 and 2 whereas a band for GM2 ganglioside was not observed in an archived control sample that had been extracted from the brain of a normal 2-year-old Beagle dog. Activities of Hex A and Hex B in the cerebrum of animal 2 were markedly decreased, compared with those in the cerebrum of the control sample that had been previously measured for reference data20 (Table 1). Activities of Hex A and Hex B in the liver of animal 2 were also markedly decreased, compared with those in the liver of a control dog. In contrast, activities of β-galactosidase in the cerebrum and liver of animal 2 were moderately elevated, compared with those in a control dog.

Figure 4.

 Thin-layer chromatogram of gangliosides from the cerebrum in dogs (animals 1 and 2) with Sandhoff-like disease and a 2-year-old control dog (C) with 6 commercially available ganglioside standards (S).

Table 1.   Activities of β-hexosaminidase and β-galactosidase in the brain and liver of animal 2 and a 5-month-old control dog
 β-Hexosaminidase Aaβ-Hexosaminidase Baβ-Galactosidasea
CerebrumLiverCerebrumLiverCerebrumLiver
  • a

    The activities are expressed as nmol/h/mg of protein.

  • b

    b The values in parentheses show percentage of the control value.

Animal 236.0 (2.8)b56.8 (1.3)1.2 (0.6)2.6 (0.5)362 (351)226 (138)
Control1,2814,308216556103164

Pedigree Study

The pedigrees of animals 1–3 were studied (Fig 5). Animals 1 and 2 were from different litters of the same sire and dam. All sires and dams of animals 1–3 had a forefather in common.

Figure 5.

 Family pedigrees of Toy Poodles with Sandhoff-like disease. All the animals possessed a common ancestor.

Discussion

In the present paper, we reported the clinical, pathological, biochemical, and MRI findings of 3 red Toy Poodles from the same canine family (Fig 5). These animals showed similar clinical and MRI characteristics (Fig 2). The 2 animals that underwent histopathological examination showed similar clinico-pathological (Fig 1) and pathological findings (Fig 3), including vacuolated lymphocytes in the peripheral blood smear and swollen neurons with pale to eosinophilic granular cytoplasmic materials positively stained with LFB and Sudan black B throughout the central nervous system, suggesting that they were affected by one of the sphingolipidoses of lysosomal storage diseases. TLC analysis with specimens of cerebrum from two of the dogs demonstrated that GM2 ganglioside was accumulated in the brain (Fig 4). Enzymatic analysis with the cerebrum and liver from one of the dogs demonstrated that both Hex A and Hex B were deficient in the brain and liver (Table 1). In contrast, β-galactosidase activity was increased because of a compensatory reaction. These biochemical data strongly suggest that these 3 dogs from a single family were affected by GM2 gagnliosidosis variant 0, ie, Sandhoff-like disease. The Sandhoff-like disease observed in this family of Toy Poodles is the 2nd reported occurrence of this form of canine disease and its 1st identification in a family of dogs. The canine variant of this disease was originally identified in a Golden Retriever.8

The onset of Sandhoff-like disease in Toy Poodles was between 9 and 12 months of age. The affected dogs finally died between 18 and 23 months of age because of progression of neurological symptoms including vomiting, stiff gait, loss of balance, falling, ataxia, intention tremor, postural deficit, astasia, decreased menace response, visual defect, decreased corneal reflex, and decreased level of consciousness. Vomiting because of unknown cause was observed in 2 of the 3 affected dogs. These clinical courses and signs were similar to those recorded previously in a Golden Retriever with Sandhoff-like disease.8,22 The duration of neurological signs in the Golden Retriever from 11 to 15 months of age was moderately similar to that in the Toy Poodles. However, miosis and loss of palpebral and papillary light reflexes observed in the Golden Retriever were not seen in the Toy Poodles with Sandhoff-like disease. In addition, clinical signs observed in the Toy Poodles but not in the Golden Retriever consisted of frequent vomiting starting in the early stage of disease and decreased or absence of corneal reflex causing panophthalmitis and corneal ulcer. The duration and neurological signs in the Toy Poodles were moderately similar to those in German Shorthair Pointers with GM2 gangliosidosis, in which clinical signs start at approximately 6 months of age consisting of decreased ability to be trained and increased nervousness, and the affected animals were most likely to die before 2 years of age.23 However, the onset of Sandhoff-like disease in Toy Poodles (9–12 months of age) were earlier than that of suspected AB variant in a Japanese Spaniel and a mixed-breed dog (1.5 years of age).6,7

Abnormal cytoplasmic vacuoles in peripheral lymphocytes can facilitate a tentative practical diagnosis of many lysosomal storage diseases, although the differential diagnosis requires biochemical analyses to identify the storage materials and demonstrate the specific enzyme deficiency.11 In Toy Poodles in the present report, cytoplasmic vacuoles were observed in approximately 10% of peripheral lymphocytes (Fig 1A). However, there were not many vacuolated lymphocytes and the degree of vacuolation was not drastic, compared with those in feline Sandhoff-like disease11 and canine GM1 gangliosidosis.24 There were no vacuolated lymphocytes in the Golden Retriever with Sandhoff-like disease (O. Yamato, unpublished data), and this finding has not described in other canine GM2 gangliosidosis. Therefore, the lymphocyte vacuolation does not seem to be a highly frequent finding in canine GM2 gangliosidosis, and could be mild to moderate if this finding is present. In the present cases, excessive vacuoles in peripheral monocytes were observed (Fig 1B), but this finding is not a very useful contribution to the clinical diagnosis of this disease because monocyte vacuolation can also be seen under inflammatory conditions.

The common MRI finding in the 3 affected Toy Poodles was diffuse T2-hyperintensity of the subcortical white matter in the cerebrum (Fig 3), but this finding was not noted in a Golden Retriever with Sandhoff-like disease.22 In general, T2-hyperintensity of cerebral white matter is widely observed in sphingolipidoses in humans and animals, ie, GM2 gangliosidosis,25–27 GM1 gangliosidosis,28–30 globoid cell leukodystrophy,29 and metachromatic leukodystrophy.29 The main cause of the T2-hyperintensity in the cerebral white matter has been considered the primary hypoplasia of the myelin, delayed myelination, or both.28,29 In addition, in the present paper, the nucleus caudatus displayed bilateral T2-hyperintensity and T1-hypointensity in an animal at 21 months of age (Fig 2E,F), and a similar finding has been reported in a Golden Retriever with Sandhoff-like disease.22 In human Sandhoff disease, the thalamus and basal ganglion have been reported to exhibit focal T2-hyperintensity and T1-hypointensity, which presumably represent loss of axon and myelin, gliosis, and intralysosomal storage. Therefore, bilateral lesions around the thalamus may display a predilection toward Sandhoff disease in this area in both humans and animals. MRI findings suggesting atrophy of the cerebrum and cerebellum were also observed in an animal at 21 months of age, but brain atrophy is a nonspecific finding and sometimes seen in the late or terminal stage in animals affected with various lysosomal storage diseases.27,30

The molecular basis has not yet been clarified in any case of canine GM2 gangliosidosis whereas all causative mutations for feline GM2 gangliosidosis have already been identified. In feline Sandhoff-like disease, the mutations are as follows: an inversion of 25 base pairs at nucleotide position 1467–1491 in the open reading frame (ORF) of the feline HEXB gene in domestic shorthaired cats in the United States31; a deletion of cytosine at position 39 in the ORF in Korat cats32; a single nucleotide substitution from cytosine to thymine at position 667 in the ORF in Japanese domestic cats33; and a deletion of 15 base pairs at the 3′ end of intron 11 in European Burmese cats.13 Furthermore, the mutation of feline AB variant found in domestic shorthaired cats in the United States has been identified as a deletion of 4 base pairs around the 3′ end of the feline GM2A cDNA.12 Once a mutation is established for a disease like feline GM2 gangliosidosis, rapid and simple DNA tests can be used for diagnosis or genotyping. With DNA tests, preventive measures for the diseases can be implemented and animal model colonies can be established for utilization in the development of new therapeutic strategies for incurable human neurodenerative diseases. Actually, cat colonies of Sandhoff-like disease have been established and translational therapeutic studies in cats are being designed to prepare gene therapy for eventual treatment of humans with GM2 gangliosidosis.13 Canine Sandhoff-like disease should also become an important large animal model. Therefore, further studies are required to elucidate the molecular basis of the canine disease and establish a canine model of Sandhoff-like disease.

Footnotes

aMRP-20EX, Hitachi Medical Corporation, Tokyo, Japan

bAIRIS2-comfort, Hitachi Medical Corporation

cMRP-20, Hitachi Medical Corporation

Acknowledgments

The authors thank the breeders associated with animals 1, 2, and 3 for providing information about the pedigree and disease. This study was supported financially in part by grants (nos. 20380173, 20-08112, and 21658109, OY) from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

Ancillary