Neuronal Loss and Decreased GLT-1 Expression Observed in the Spinal Cord of Pembroke Welsh Corgi Dogs With Canine Degenerative Myelopathy

Abstract

Canine degenerative myelopathy (DM) is a progressive neurodegenerative disease that is frequently found in Pembroke Welsh Corgi (PWC) dogs. Canine DM is potentially a spontaneous animal model for human amyotrophic lateral sclerosis (ALS) because of similar lesions and the involvement of superoxide dismutase 1 (SOD1) mutation. However, the ventral horn lesion in DM has not been characterized in detail. Glutamate excitotoxicity due to deficiency of the glutamine-glutamate cycle has been implicated in neuron death in ALS. Thus, we examined 5 PWC dogs with an SOD1 mutation that were affected by DM, 5 non-DM PWC dogs, and 5 Beagle dogs without neurologic signs to assess the neuronal changes and the expression levels of 2 glial excitatory amino acid transporters (glutamate transporter 1 [GLT-1] and glutamate/aspartate transporter [GLAST]). The number of neurons in the spinal ventral horns of the DM dogs was significantly decreased, whereas no change was found in the cell size. Chromatolysis, lipofuscin-laden neurons, and marked synapse loss were also observed. GLT-1 expression was strikingly decreased in DM dogs, whereas GLAST expression showed no significant change. The results indicate that excitotoxicity related to the reduced expression of GLT-1, but not GLAST, may be involved in neuron loss in DM, as in human ALS, whereas intraneuronal events may differ between the 2 diseases.

Keywords

amyotrophic lateral sclerosis dog excitatory amino acid transporter excitotoxicity immunohistochemistry neurodegenerative disease spinal cord

Canine degenerative myelopathy (DM) is a progressive neurodegenerative disease that causes advancing ataxia and muscle atrophy.¹ Affected dogs initially exhibit hindlimb weakness and paralysis, followed by forelimb ataxia and culminating in paralysis, dysuria, and dyschezia.^7,19 Recently, a superoxide dismutase 1 (SOD1) mutation including c.118G>A was shown to be associated with canine DM in several breeds, including the Pembroke Welsh Corgi (PWC), Boxer, and German Shepherd,² followed by a report of c.52A>T mutation in a Bernese Mountain Dog.⁴⁰ PWC dogs have been reported to be a favorite breed for studying DM in recent years.^7,19,23 In fact, all cases of canine DM diagnosed at the University of Tokyo between 2009 and 2011 occurred in this breed and were associated with an SOD1 mutation. Histiocytic sarcoma¹⁴ and inflammatory myopathy³⁸ also commonly occur in PWC in Japan and produce similar clinical signs, which often complicates diagnosis until histopathology is performed.

More than 100 SOD1 gene mutations are involved in human familial amyotrophic lateral sclerosis (ALS), and the pathologic lesions of ALS are similar to those of canine DM, making canine DM a potentially useful animal model of ALS. We previously reported that the overproduction of nitric oxide through inducible nitric oxide synthase (iNOS) and/or the abnormal accumulation of denatured proteins might be involved in the pathogenesis of DM in PWC dogs,²³ as in human ALS. Since lesions of the spinal gray matter are not necessarily required for the diagnosis of canine DM, they have not been examined in sufficient detail, even though loss of the ventral motor neurons is the major event of ALS pathology.

Glutamate is the predominant excitatory amino acid in the central nervous system (CNS). Excitatory amino acid transporters (EAATs) on the cell membrane, which reuptake extracellular glutamate into the cell, modulate homeostasis of synaptic transmission. Of these transporters, astrocytic glutamate transporter 1 (GLT-1; EAAT2) and glutamate/aspartate transporter (GLAST; EAAT1) are responsible for most of the glutamate transport. Glutamine synthetase (GS), which catalyzes the production of glutamine from glutamate and ammonia, is found in astrocytes and oligodendrocytes⁸ and may be important for maintenance of the glutamate concentration.

Excessive glutamate exposure is toxic to neurons through massive Ca²⁺ entry into the cell. It is known that motor neurons are vulnerable to excitotoxicity,^5,39 and it is thought that excitotoxicity and oxidative stress play a major role in motor neuron loss in patients with ALS.

It is also reported that the downregulation of GLT-1 is involved in the disease progression in patients with ALS and in some types of SOD1-transgenic ALS model mice.^{4,12,24,28,29} To our knowledge, no studies have examined the involvement of excitotoxicity in canine DM.

The purpose of the present study is to elucidate the severity and morphologic features of neuronal loss and the involvement of GLT-1, GLAST, and GS in canine DM.

Materials and Methods

Animals and Tissue Samples

Spinal cords of 5 PWC dogs pathologically diagnosed with DM (age, 9–14 years; mean, 12.3 years), as well as those of 5 PWC dogs without significant lesion (age, 6–11 years; mean, 8.9 years) and 6 Beagle dogs as control dogs without neurologic signs (age, 1–14 years; mean age unknown), were used. Information on the dogs is summarized in Table 1. Dog Nos. 12 and 13 were already described as DM cases in our previous report.⁶ Representative specimens from the cervical (C5–C7), thoracic (T5), and lumbar (L3–L5) cords were used. Only T13 to L3 were available in dog No. 15, T5 to T12 in dog No. 11, and C1 to T11 in dog No. 5. Examined regions are indicated in Table 1.

Table 1.

Dogs Examined in the Present Study.

Dog No.	Breed	Sex	Age	Onset to Death	Clinical Signs of DM	Sample Examined			Major Pathologic Diagnosis at Necropsy	SOD1 Mutation
Dog No.	Breed	Sex	Age	Onset to Death	Clinical Signs of DM	C	T	L	Major Pathologic Diagnosis at Necropsy	SOD1 Mutation
1	Beagle	F	ND	–	–	+	+	+	None	None
2	Beagle	F	ND	–	–	+	+	+	None	None
3	Beagle	F	14 y	–	–	+	+	+	ND	None
4	Beagle	F	6 y	–	–	+	+	+	None	None
5	Beagle	F	6 y	–	–	+	+	–	None	None
6	Beagle	F	8 y	–	–	+	+	+	Glomerulonephritis	None
7	PWC	SF	6 y, 11 mo	–	–	+	+	+	Inflammatory myopathy	Homozygous
8	PWC	SF	10 y, 4 mo	–	–	+	+	+	Inflammatory myopathy	Homozygous
9	PWC	SF	11 y, 2 mo	–	–	+	+	+	Histiocytic sarcoma (brain)	Heterozygous
10	PWC	M	6–7 y	–	–	+	+	+	None	Homozygous
11	PWC	CM	11 y, 3 mo	–	–	–	+	–	T-cell lymphoma	Heterozygous
12	PWC	M	12 y, 6 mo	3 y	HLW, FLW, DU	+	+	+	DM	Homozygous
13	PWC	CM	14 y, 6 mo	2 y, 5 mo	TP, DU, RD	+	+	+	DM	Homozygous
14	PWC	SF	12 y, 2 mo	2 y, 7 mo	TP, DU, RD	+	+	+	DM	Homozygous
15	PWC	SF	9 y, 6 mo	2 y, 9 mo	TP, DU, RD	–	–	+	DM	Homozygous
16	PWC	M	13 y	4 mo	HLW	+	+	+	DM, B-cell lymphoma (spleen)	Homozygous

Abbreviations: C, cervical 5–7; CM, castrated male; DM, degenerative myelopathy; DU, dysurea; F, female; FLW, forelimb weakness; HLW, hindlimb weakness; L, lumbar 3–5; ND, no data; M, male; PWC, Pembroke Welsh Corgi; RD, respiratory dysfunction; SF, spayed female; T, thoracic 5; TP, tetraplegia.

Genotyping for SOD1 c.118G>A Mutation

Genomic DNA was extracted from paraffin-embedded tissue of each dog, using a commercial kit (DNA Isolator PS Kit; Wako Pure Chemical Industries, Osaka, Japan) according to the manufacturer’s instructions. The genotypes were determined using a real-time polymerase chain reaction (PCR) method reported previously.⁶

Histopathology, Immunohistochemistry, and Immunofluorescence

Tissue samples from spinal cords were fixed in 10% formalin, processed routinely, and embedded in paraffin wax. Then, 4-μm-thick paraffin sections were subjected to hematoxylin and eosin (HE), Luxol fast blue (LFB)–HE, and Kluver-Barrera stains and used for immunohistochemical analysis.

Immunohistochemical analysis was performed using the EnVision polymer method (Dako, Glostrup, Denmark). Table 2 lists the primary antibodies used. Deparaffinized sections were first autoclaved at 120°C for 10 minutes in 10 mM citrate buffer (pH 6.0 or 9.0) for antigen retrieval. Then, the tissue sections were treated with 3% hydrogen peroxide (H₂O₂)–methanol at room temperature for 5 minutes and incubated in 8% skimmed milk–Tris-buffered saline with 0.2% Tween 20 (TBST) at 37°C for 1 hour to avoid nonspecific reactions. The sections were then incubated at 4°C overnight with one of the primary antibodies (Table 2). After being washed 3 times in TBST, the sections were incubated with Envision horseradish peroxidase (HRP)–labeled polymer anti–mouse or anti–rabbit immunoglobulin G (IgG) (DAKO, Glostrup, Denmark) at 37°C for 50 minutes. Then, the sections were washed with TBS and the reaction visualized with 0.05% 3,3′-diaminobenzidine and 0.03% H₂O₂ in TBS. Counterstaining was performed with Mayer’s hematoxylin.

Table 2.

Primary Antibodies Used.

Antibody	Immunogen	Type	Marker for	Antigen Retrieval	Dilution	Source, Cat. No.
GFAP	GFAP isolated from cow spinal cord	Rabbit, p	Astrocyte	none	1:1000	DAKO, Glostrup, Denmark, Z0334
GFAP	Intermediate filament cytoskeleton from cultured human glioma cells	Mouse, m (GF 12.24)	Astrocyte	Citrate buffer, pH 6.0	Prediluted	PROGEN Biotechnik, Heidelberg, Germany, 65011
Iba1	C-terminus of Iba1	Rabbit, p	Microglia	Citrate buffer, pH 6.0	1:500	Wako Pure Chemical Industries, Osaka, Japan, 019-19741
NF	Neurofilament purified from human brain	Mouse, m (2F11)	Axon (NF-L)	none	Prediluted	Dako, N1591
βIII-tubulin	Peptide corresponding to the C-terminus of βIII-tubulin	Mouse, m (5G8)	Neuron	Citrate buffer, pH 6.0	1:2000	Promega, Madison, WI, G712A
Synaptophysin	Crude fractions of coated vesicles from bovine brain	Mouse, m (SY38)	Synaptic vesicle	Target retrieval solution, pH 9.0	1:30	DAKO, M0776
Olig2	Recombinant mouse Olig2	Rabbit, p	Oligodendrocyte	Target retrieval solution, pH 9.0	1:200	Millipore, Billerica, MA, AB9610
Ubiquitin	Ubiquitin isolated from cow erythrocytes	Rabbit, p	—	none	1:200	DAKO, Z0458
EAAT1 (GLAST)	Synthetic MAP of C-terminal region of canine GLAST	Rabbit, p	—	Citrate buffer, pH 6.0	1:300 (IHC), 1:75 (IF)	Sato et al, 2000³³
EAAT2 (GLT-1)	C-terminal amino acid residue of rat GLT	Rabbit, p	—	Citrate buffer, pH 6.0	1:1000 (IHC), 1:300 (IF)	Wako Pure Chemical Industries, 015-16421
GS	Sheep glutamine synthetase amino acids 1–373	Mouse, m (6/GS)	—	Citrate buffer, pH 6.0, or target retrieval solution, pH 9.0	1:300	BD Biosciences, San Jose, CA, 610518

Abbreviations: GFAP, glial fibrillary acidic protein; Iba1, ionized calcium-binding adapter molecule 1; NF, neurofilament; Olig2, oligodendrocyte transcription factor 2; EAAT, excitatory amino acid transporter; GLAST, glutamate/aspartate transporter; GLT-1, glutamate transporter 1; GS, glutamine synthetase; m, monoclonal; p, polyclonal; IF, immunofluorescence; IHC, immunohistochemistry; MAP, mitogen-activated protein; — not the cell type marker.

Ventral motor neurons were defined as βIII-tubulin–positive structures with morphologic features of a neuron, including a minor axis of more than 15 μm in the ventral horn, not including intermediolateral nuclei. Counts of neurons were performed using representative sections.

For double immunofluorescence staining, sections were incubated with a primary antibody and then with another primary antibody (Table 2). The sections were incubated with a cocktail of fluorescein isothiocyanate (FITC)–conjugated anti–mouse IgG (1:100; Invitrogen Life Technologies, Paisley, UK) and Alexa 594–conjugated anti–rabbit IgG (1:100; Vector Laboratories, Burlingame, CA) at room temperature for 1 hour. After being mounted with mounting medium (Vector Laboratories), the specimens were observed and photographed using a DMI3000B microscope (Leica Microsystems, Wetzlar, Germany).

Data Analysis

Macroscopic images of the transverse section of spinal cords were captured using the Nikon COOLSCAN IV ED (Nikon Imaging, Tokyo, Japan). Microscopic images were captured using a Nikon DXM1200F digital camera and ACT-1 software (Nikon Instruments, Tokyo, Japan). Image analysis was conducted using the National Institutes of Health (NIH) Image J software (NIH, Bethesda, MD). Briefly, the color channel was split to RGB 8-bit images, and then the positive area was measured as a limited area with a threshold set 0 to 200. Immunopositive areas for synaptophysin or GLT-1 were measured on a macroscopic image and expressed as the ratio of positive area/total gray matter area. For GLAST, the ratio of positive areas was measured in 6 randomly selected squares (3 in each side) from the ventral horn, dorsal horn, and lateral funiculus on a microscopic image. For GS, the number of positive cells in the random 6 squares from the ventral horn was counted. Then, the average of the 6 areas was calculated.

Data are expressed as the mean ± SEM. For statistical analyses of the data, 1-way analysis of variance (ANOVA) followed by Scheffé’s multiple comparison procedure was used. Differences between groups were considered statistically significant when P < .05.

Results

Among the dogs assessed in the present study, all DM-affected dogs and 3 of 5 non-DM PWC dogs had a c.118G>A homozygous mutation in the SOD1 gene. The other PWC dogs (dog Nos. 9 and 11) had a heterozygous mutation. Beagle dogs did not have the mutation (Table 1).

Histopathologic features of dog Nos. 12 and 13 were described in detail in our previous report.²³ Axonal degeneration and/or loss, as well as demyelination of the white matter, were commonly observed in the spinal cord of dog Nos. 14 to 16 (Figs. 9 and 10), as well as dog Nos. 12 and 13. The lesions were most severe in the thoracic spinal cord of dog Nos. 12, 13, 14, and 16, especially in the dorsolateral funiculi. The severity of these lesions differed among the dogs; the severest were in dog Nos. 14 and 15, whereas it was mild in dog No. 16. The cervical cord of dog No. 16 was almost intact. Proliferation of glial fibrillary acidic protein (GFAP)–positive astrocytes was observed to be correlated with the severity of the axonal loss and demyelination (Fig. 11). Proliferation of calcium-binding adapter molecule 1 (Iba1)–positive microglia throughout the white and gray matter was observed in DM dogs, especially in dog Nos. 12 and 15 (Fig. 12). Myelin loss and axonal degeneration were significant in the spinal ventral roots of dog Nos. 14 and 15, whereas these changes were relatively mild in the other DM dogs.These lesions were not observed in other two groups of dogs (Figs. 1-8).

Figure 1. Spinal cord, T5; Beagle dog No. 4. Normal staining. Luxol fast blue and hematoxylin-eosin (LFB-HE). Figure 2. Spinal cord, dorsolateral funiculi, T5; Beagle dog No. 4. Normal neurofilament (NF) immunoreactivity. 3,3′-Diaminobenzidine (DAB) and hematoxylin. Figure 3. Spinal cord, dorsolateral funiculi, T5; Beagle dog No. 4. Glial fibrillary acidic protein (GFAP) immunoreactivity. DAB and hematoxylin. Figure 4. Spinal cord, boundary of the gray and white matter, T5; Beagle dog No. 2. Normal immunoreactivity for ionized calcium-binding adapter molecule 1 (Iba1). DAB and hematoxylin. Figure 5. Spinal cord, T5; Pembroke Welsh Corgi (PWC), non–degenerative myelopathy (DM) dog No. 10. Similar staining to that in the Beagle (Fig. 1). LFB-HE. Figure 6. Spinal cord, dorsolateral funiculi, T5; PWC, non-DM dog No. 10. NF immunoreactivity similar to that in the Beagle (Fig. 2). DAB and hematoxylin. Figure 7. Spinal cord, dorsolateral funiculi, T5; PWC, non-DM dog No. 8. GFAP immunoreactivity similar to that in the Beagle (Fig. 3). DAB and hematoxylin. Figure 8. Spinal cord, boundary of the gray and white matter, T5; PWC, non-DM dog No. 8. Iba1 immunoreactivity similar to that in the Beagle (Fig. 4). DAB and hematoxylin. Figure 9. Spinal cord, T5; PWC, DM dog No. 14. Marked loss of LFB staining was observed in the white matter compared with the Beagle (Fig. 1) and non-DM PWC (Fig. 5), especially in the dorsolateral funiculi. LFB-HE. Figure 10. Spinal cord, dorsolateral funiculi, T5; PWC, DM dog No. 14. Severe loss and degeneration (swelling) of NF-positive axons were observed. DAB and hematoxylin. Figure 11. Spinal cord, dorsolateral funiculi, T5; PWC, DM dog No. 15. Increased GFAP-positive area and appearance of GFAP-positive swollen cells. DAB and hematoxylin. Figure 12. Spinal cord, boundary of the gray and white matter, T5; PWC, DM dog No.15. Increased number of Iba1-positive cells was observed. DAB and hematoxylin.

A large amount of lipofuscin was observed frequently in the cytoplasm of neurons in all DM dogs and, to some degree, in aged non-DM PWC dog Nos. 8, 9, and 10 and Beagle dog No. 3. Chromatolysis of large motor neurons was observed in the ventral horn of the cervical, thoracic, and lumbar cords of DM-affected dogs, especially in dogs Nos. 12 and 14. No cytoplasmic inclusion bodies or apoptotic bodies were detected on HE sections in any dog in the present study. Ubiquitin-positive structures were not observed in the surviving neurons, but ubiquitin positivity was identified in degenerated neurons and/or in the neuropil.

There were fewer βIII-tubulin–positive neurons in the ventral horn in DM-affected dogs than in dogs in the other 2 groups (Figs. 13–16, Table 3). The mean numbers of such neurons in the cervical, thoracic, and lumbar cords from DM, non-DM PWC, and Beagle dogs are shown in Table 3. The decrease was statistically significant in thoracic and lumbar cords compared with both non-DM PWC and Beagle dogs. Both large neurons and small neurons survived in the DM sections (Fig. 16). No significant difference in the average size of neuronal soma between DM, non-DM PWC, and Beagle dogs was detected (Fig. 16, Table 3).

Figure 13 Spinal cord, ventral horn, T5; Beagle dog No. 2. βIII-tubulin–positive neurons. 3,3′-Diaminobenzidine (DAB) and hematoxylin. Figure 14. Spinal cord, ventral horn, T5; Pembroke Welsh Corgi (PWC), non–degenerative myelopathy (DM) dog No. 9. βIII-tubulin–positive neurons. DAB and hematoxylin. Figure 15. Spinal cord, ventral horn, T5; PWC, DM dog No. 14. Fewer numbers of βIII-tubulin–positive neurons were observed compared with Beagle (Fig. 13) and non-DM PWC (Fig. 14). DAB and hematoxylin. Figure 16. The scatter plots of the shorter diameter (minor axis) of the βIII-tubulin–positive neuronal soma in the ventral horn of the cervical, thoracic, and lumbar cords of dog Nos. 1 to 6 (Beagle; light gray dots), 7 to 11 (non-DM PWC; gray dots), and 12 to 16 (DM; black dots). Horizontal line represents the average size. ND, no data.

Table 3.

Mean Size and Number ± SEM of βIII-tubulin–Positive Neurons of Beagle, Non–Degenerative Myelopathy (DM) Pembroke Welsh Corgi (PWC), and DM Dogs.

Characteristic	Beagle	Non-DM PWC	DM
Cervical
Size, μm	25.0 ± 1.5	25.1 ± 0.2	25.4 ± 2.3
Number	76.5 ±14.6	85.5 ± 13.2	52.0 ± 11.1
Thoracic
Size, μm	21.9 ± 0.7	22.7 ± 1.1	23.7 ± 1.4
Number	25.5 ± 4.7	31.2 ± 5.5	12.5 ± 3.4^a,b
Lumbar
Size, μm	27.0 ± 0.8	25.0 ± 0.7	25.6 ± 1.7
Number	83.2 ± 18.8	99.5 ± 10.8	32.1 ± 8.3^a,b

P < .05, DM vs ^anon-DM PWC or ^bBeagle.

The ratios of synaptophysin-positive area in the gray matter of DM-affected dogs were significantly decreased to approximately half of those in non-DM PWC dogs or Beagle dogs in the cervical, thoracic, and lumbar cords (Figs. 17–20). The reduction was most prominent in the ventral horn of all regions. Synaptophysin immunoreactivity could not be detected in dog Nos. 1, 2, 7, and 12, presumably because the sample tissues had not been freshly fixed.

Figure 17

Spinal cord, T5; Beagle dog No. 2. (a) Normal synaptophysin immunoreactivity. (b) Higher magnification of (a), ventral horn. 3,3′-Diaminobenzidine (DAB) and hematoxylin. Figure 18. Spinal cord, T5; Pembroke Welsh Corgi (PWC), non–degenerative myelopathy (DM) dog No. 8. (a) Synaptophysin immunoreactivity similar to that in the Beagle (Fig. 17a). (b) Higher magnification of (a), ventral horn. DAB and hematoxylin. Figure 19. Spinal cord segment T5; PWC, degenerative myelopathy dog No. 14. (a) Synaptophysin immunoreactivity is notably reduced compared with the Beagle (Fig. 17a). (b) Higher magnification of (a), ventral horn. DAB and hematoxylin. Figure 20. The ratio of synaptophysin-positive area to the gray matter area in the Beagle (light gray bars), non-DM PWC (gray bars), and DM (black bars) dogs. Synaptophysin immunoreactivity was significantly decreased in DM dogs. *P < .05.

GLT-1 immunoreactivity was distributed throughout the gray matter of the spinal cord of non-DM PWC and Beagle dogs, with the highest expression in the substantia gelatinosa of the dorsal horn and around large ventral motor neurons (Figs. 21–24). Compared with both non-DM PWC and Beagle dogs, the GLT-1 expression in DM dogs was significantly decreased (Fig. 24) in all regions examined, especially in the thoracic and lumbar cords. Results of double immunofluorescence staining indicated that the GLT-1–positive area was mostly localized between the neuron and GFAP-positive process (Figs. 25, 28, 31), whereas no GLT-1–positive staining was observed on the surface of GFAP-positive cells with a large cytoplasm in the gray matter of the DM dogs (Fig. 25). The GLT-1–positive staining was localized especially around βIII-tubulin–positive large neurons (Figs. 26, 29, 32) and synaptophysin-positive synaptic vesicles (Figs. 27, 30, 33), whereas the expression levels were decreased in DM dogs (Figs. 32, 33).

Figure 21 Spinal cord, T5; Beagle dog No. 5. (a) Normal glutamate transporter 1 (GLT-1) immunoreactivity. (b) Higher magnification of (a), ventral horn. 3,3′-Diaminobenzidine (DAB) and hematoxylin. Figure 22. Spinal cord, T5; Pembroke Welsh Corgi (PWC), non–degenerative myelopathy (DM) dog No. 8. (a) GLT-1 immunoreactivity similar to that in the Beagle (Fig. 21a). (b) Higher magnification of (a), ventral horn. DAB and hematoxylin. Figure 23. Spinal cord segment T5; PWC, DM dog No. 12. (a) GLT-1 immunoreactivity is remarkably reduced compared with the Beagle (Fig. 21a) and non-DM PWC (Fig. 22a). (b) Higher magnification of (a), ventral horn. DAB and hematoxylin. Figure 24. The ratio of GLT-1–positive area to the gray matter area in Beagle (light gray bars), non-DM PWC (gray bars), and DM (black bars) dogs. GLT-1 immunoreactivity was significantly decreased in DM dogs. *P < .05.

Figure 25 Spinal cord, ventral horn; Beagle dog No. 6. Glutamate transporter 1 (GLT-1) immunoreactivity is mostly localized between large neurons (*) and glial fibrillary acidic protein (GFAP)–positive processes (arrowhead). Double immunofluorescence for GLT-1 (red, Alexa 594) and GFAP (green, fluorescein isothiocyanate [FITC]). Figure 26. Spinal cord, ventral horn; Beagle dog No. 6. GLT-1 immunoreactivity surrounds the βIII-tubulin–positive large ventral neurons. Double immunofluorescence for GLT-1 (red, Alexa 594) and βIII-tubulin (green, FITC). Figure 27. Spinal cord, ventral horn; Beagle dog No. 6. GLT-1 immunoreactivity surrounds the synaptophysin-positive structure around the large ventral neuronal soma (*). Double immunofluorescence for GLT-1 (red, Alexa 594) and synaptophysin (green, FITC). Figure 28. Spinal cord, ventral horn; Pembroke Welsh Corgi (PWC), non–degenerative myelopathy (DM) dog No. 9. GLT-1 immunoreactivity localization is similar to that in the Beagle dog (Fig. 25, arrowhead). Asterisks (*) indicate neuronal soma. Double immunofluorescence for GLT-1 (red, Alexa 594) and GFAP (green, FITC). Figure 29. Spinal cord, ventral horn; PWC, non-DM dog No. 9. GLT-1 immunoreactivity localization is similar to that in the Beagle dog (Fig. 26). Double immunofluorescence for GLT-1 (red, Alexa 594) and βIII-tubulin (green, FITC). Figure 30. Spinal cord, ventral horn; PWC, non-DM dog No. 9. GLT-1 immunoreactivity localization is similar to that in the Beagle dog (Fig. 27). Double immunofluorescence for GLT-1 (red, Alexa 594) and synaptophysin (green, FITC). Figure 31. Spinal cord, ventral horn; PWC, DM dog No. 14. GLT-1 immunoreactivity is mostly localized between large neurons (*) and GFAP-positive processes (arrowhead). No GLT-1 immunoreactivity is observed on the surface of GFAP-positive cells with a large cytoplasm (arrows). Double immunofluorescence for GLT-1 (red, Alexa 594) and GFAP (green, FITC). Figure 32. Spinal cord, ventral Figure 32 (continued). horn; PWC, DM dog No. 14. The GLT-1 expression is decreased, whereas the distribution is similar to that in the Beagle (Fig. 26) and non-DM PWC dogs (Fig. 29). Double immunofluorescence for GLT-1 (red, Alexa 594) and βIII-tubulin (green, FITC). Figure 33. Spinal cord, ventral horn; PWC, DM dog No. 14. The GLT-1 expression is decreased, whereas the distribution is similar to that in the Beagle (Fig. 27) and non-DM PWC (Fig. 30). Red lumps with intrinsic fluorescence are erythrocytes. Double immunofluorescence for GLT-1 (red, Alexa 594) and synaptophysin (green, FITC).

A nonsignificant change of the GLAST expression level in the ventral and dorsal horn was exhibited in DM dogs compared with that in non-DM PWC or Beagle dogs (Figs. 34–37). By double staining, GLAST expression was shown to be co-localized with GFAP in the gray and white matter of DM and the 2 other groups of dogs (data not shown). Astrocytes included in the gliosis lesions in the degenerated white matter or reactive astrocytes in the gray matter showed no significant GLAST immunoreactivity.

Figure 34

Spinal cord, ventral horn, T5; Beagle dog No. 4. Normal glutamate/aspartate transporter (GLAST) immunoreactivity. 3,3′-Diaminobenzidine (DAB) and hematoxylin. Figure 35. Spinal cord, ventral horn, T5; Pembroke Welsh Corgi (PWC), non–degenerative myelopathy (DM) dog No. 7. GLAST immunoreactivity similar to that in the Beagle (Fig. 34). DAB and hematoxylin. Figure 36. Spinal cord, ventral horn, T5; PWC, DM dog No. 13. GLAST immunoreactivity similar to those in the Beagle (Fig. 34) and non-DM PWC (Fig. 35). Asterisks (*) indicate neuronal soma. DAB and hematoxylin. Figure 37. The ratio of GLAST-positive area in the dorsal horn (dh), ventral horn (vh), and lateral funiculus (lf). GLAST immunoreactivity was not significantly changed in Beagle (light gray bars), non-DM PWC (gray bars), and DM (black bars) dogs.

GS-positive cells were observed throughout the gray matter, and a few were found in the white matter, mostly in the boundary area with the gray matter (Figs. 38–41). By double immunofluorescence, GS immunoreactivity was detected in mostly GFAP-negative and olig2-positive cells with scant cytoplasm (oligodendrocytes) and partly GFAP-positive astrocytes (data not shown). Compared with non-DM PWC or beagle dogs, the average number of GS-positive cells was slightly decreased in DM dogs, but this difference was not statistically significant (Fig. 41).

Figure 38.

Spinal cord, ventral horn, T5; Beagle dog No. 4. Normal glutamine synthetase (GS) immunoreactivity. 3,3′-Diaminobenzidine (DAB) and hematoxylin. Figure 39. Spinal cord, ventral horn, T5; Pembroke Welsh Corgi (PWC), non–degenerative myelopathy (DM) dog No. 9. GS immunoreactivity similar to that in the beagle (Fig. 38). DAB and hematoxylin. Figure 40. Spinal cord, ventral horn, T5; PWC, DM dog No. 13. GS immunoreactivity is slightly decreased compared with those in the Beagle (Fig. 38) and non-DM PWC (Fig. 39). DAB and hematoxylin. Figure 41. The number of GS-positive cells per unit area in the ventral horns of the Beagle (light gray bars), non-DM PWC (gray bars), and DM (black bars) dogs. GS expression in the DM dogs was slightly lower than those of the 2 other groups of dogs but not significant.

Discussion

Considering the fact that a high percentage of PWC dogs examined in the present study had an SOD1 mutation, an inherited predisposition to DM might be frequently present in the breed in Japan, irrespective of whether loss or gain of function of the molecule is directly involved in the etiology. Histopathologic lesions observed in the spinal cords of dog Nos. 12 to 16 affected by DM are consistent with those outlined in previous reports describing canine DM,^7,19,21 and the distribution of the spinal lesions was almost identical to that of previous reports on DM in PWC.^7,19 The severity of the pathologic lesions was closely associated with the severity of the clinical signs (Table 1). Whereas no sufficient examination was done on the tissues other than CNS in this study, the ventral root lesions are consistent with the recent reports on peripheral lesions of DM.³⁴

The pathologic changes in the ventral neurons in DM dogs included chromatolysis, particularly in dog Nos. 12 and 14, and deposition of lipofuscin, which was observed in all DM dogs. Loss of the ventral neurons was more severe in dog Nos. 13 and 15. Chromatolytic or lipofuscin-bearing neurons have been generally observed in motor neurons of patients with ALS.^18,30 Lipofuscin deposition was also observed in aged non-DM PWC dogs, presumably because of normal aging. Because of the difference of ages of the Beagle, non-DM PWC, and DM dogs in this study, the results do not necessarily indicate that the lipofuscin deposition is a consequence of the DM itself.

The loss of synaptophysin immunoreactivity in the present cases of DM indicates that a synaptic loss occurs in DM during the process of neuronal degeneration, which is similar to what occurs in ALS.^{15,16,31,32,44} Similar to our previous report²³ and unlike in ALS, inclusion bodies mimicking Bunina bodies, Lewy body–like inclusions, and/or aggregated ubiquitin-positive skein-like inclusions/structures were absent in the surviving neurons of DM-affected dogs.¹¹ Formation of such aggregations or inclusion bodies indicates the dysfunction of 2 major intracellular degradative pathways (ubiquitin-proteasome system and autophagy system); abnormalities of these systems are thought to be involved in neurodegeneration and ALS.^17,41 Considering the absence of these features, the mechanisms of neuronal death in canine DM are not likely to be the same as in ALS.

The result of size measurement indicates ventral neurons larger than 30 or even 50 μm were still surviving in DM cases. In addition, there was no difference in the average size of neurons between DM dogs and non-DM PWC or Beagle dogs. These facts indicate nonselective (not only large neuron) degeneration of the ventral horn. These findings also suggest that the ventral neuronal loss in the DM-affected spinal cord may not occur through apoptosis, since apoptotic cells usually undergo shrinkage. Moreover, no apoptotic body was observed in the DM dogs in the present study. Although the role of apoptosis in neuronal death in ALS is still controversial,^20,36,42,43 the results of this study suggest that apoptosis is unlikely to be the mechanism of neuronal death in canine DM.

The distribution pattern of GLT-1 in the spinal cords of the present non-DM PWC or Beagle dogs is consistent with that previously outlined in rat, mouse, and human spinal cords.^10,22,26 GLT-1 downregulation is reported in patients with ALS and ALS model mice after disease onset.^4,12,29 The decreased GLT-1 expression level in DM dogs suggests that the deficiency of glial reuptake of excitatory amino acids may play a similar role in DM pathogenesis as it does in patients with ALS and ALS mouse models.²⁷ Absence of a significant difference between non-DM PWC dogs with the SOD1 mutation and Beagle dogs without the mutation also indicates that the decrease of GLT-1 expression is a secondary event to the disease onset and may be involved in disease progression rather than disease onset, as both a consequence and an exacerbating factor of the disease. In addition, the fact that there was no significant change of the GLAST expression level in any region of the DM dogs in the present study may also support the hypothesis that GLT-1 is responsible for excitotoxicity of the spinal motor neurons in canine DM. There are reports^13,25 indicating that astrocytic nitric oxide production through iNOS is associated with glutamate uptake activity of GLT-1 or glutamate-induced excitotoxicity. Considering our previous finding that reactive astrocytes express iNOS in the gray matter of DM dogs,²³ it is likely that an interaction between NO production and glutamate uptake occurs in DM.

GS localization in astrocytes and oligodendrocytes in canine spinal cords examined in this study coincides with previous reports on cats and rats.^8,37 Although the number of GS-positive cells was slightly decreased in DM dogs compared with those in the 2 other groups, no statistically significant difference among the 3 groups was detected. Alterations in GS levels have been documented in other neurodegenerative disorders,³⁵ but reports of GS expression in patients with ALS are very limited.³ In a type of SOD1 transgenic mice, GS expression and/or activity appears to be unchanged in the spinal cord.⁹ The role of GS expression in both ALS and canine DM remains undefined, and further studies using a larger number of samples are needed to clarify this issue.

In conclusion, excitotoxicity via the deficiency of the glutamine/glutamate cycle due to reduced expression of astrocytic GLT-1, but not GLAST, may contribute to the striking neuron loss in the ventral horn of DM dogs during disease progression, as in ALS. On the other hand, intraneuronal events such as degradative dysfunction during disease progression seem to differ between DM and ALS.

Footnotes

Acknowledgements

We are grateful to the owners of the examined dogs who participated in this research.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Averill

Jr . Degenerative myelopathy in the aging German Shepherd dog: clinical and pathologic findings. J Am Vet Med Assoc. 1973;162(12):1045–1051.

Awano

Johnson

Wade

. Genome-wide association analysis reveals a SOD1 mutation in canine degenerative myelopathy that resembles amyotrophic lateral sclerosis. Proc Natl Acad Sci U S A. 2009;106(8):2794–2799.

Bos

Hoogland

Meine Jansen

. Increased glutamine synthetase but normal EAAT2 expression in platelets of ALS patients. Neurochem Int. 2006;48(4):306–311.

Bruijn

Becher

Lee

. ALS-linked SOD1 mutant G85R mediates damage to astrocytes and promotes rapidly progressive disease with SOD1-containing inclusions. Neuron. 1997;18(2):327–338.

Carriedo

Yin

Weiss

. Motor neurons are selectively vulnerable to AMPA/kainate receptor-mediated injury in vitro. J Neurosci. 1996;16(13):4069–4079.

Chang

Kamishina

Mizukami

. Genotyping assays for the canine degenerative myelopathy–associated c.118G>A (p.E40K) mutation of the SOD1 gene using conventional and real-time PCR methods: a high prevalence in the Pembroke Welsh Corgi breed in Japan [published online January 18, 2013]. J Vet Med Sci. doi: 10.1292/jvms.12-0451.

Coates

March

Oglesbee

. Clinical characterization of a familial degenerative myelopathy in Pembroke Welsh Corgi dogs. J Vet Intern Med. 2007;21(6):1323–1331.

D’Amelio

Eng

Gibbs

. Glutamine synthetase immunoreactivity is present in oligodendroglia of various regions of the central nervous system. Glia. 1990;3(5):335–341.

Fray

Dempster

Williams

. Glutamine synthetase activity and expression are not affected by the development of motor neuronopathy in the G93A SOD-1/ALS mouse. Mol Brain Res. 2001;94 (1–2):131–136.

10.

Fray

Ince

Banner

. The expression of the glial glutamate transporter protein EAAT2 in motor neuron disease: an immunohistochemical study. Eur J Neurosci. 1998;10(8):2481–2489.

11.

Hara

Nakamura

Matsui

. Suppression of basal autophagy in neural cells causes neurodegenerative disease in mice. Nature. 2006;441(7095):885–889.

12.

Howland

Liu

She

. Focal loss of the glutamate transporter EAAT2 in a transgenic rat model of SOD1 mutant-mediated amyotrophic lateral sclerosis (ALS). Proc Natl Acad Sci U S A. 2002;99(3):1604–1609.

13.

Ida

Hara

Nakamura

. Cytokine-induced enhancement of calcium-dependent glutamate release from astrocytes mediated by nitric oxide. Neurosci Lett. 2008;432(3):232–236.

14.

Ide

Uchida

Kagawa

. Pathological and immunohistochemical features of subdural histiocytic sarcomas in 15 dogs. J Vet Diagn Invest. 2011;23(1):127–132.

15.

Ikemoto

Kawanami

Llena

. Immunocytochemical studies on synaptophysin in the anterior horn of lower motor neuron disease. J Neuropathol Exp Neurol. 1994;53(2):196–201.

16.

Ince

Slade

Chinnery

. Quantitative study of synaptophysin immunoreactivity of cerebral cortex and spinal cord in motor neuron disease. J Neuropathol Exp Neurol. 1995;54(5):673–679.

17.

Kato

. Amyotrophic lateral sclerosis models and human neuropathology: similarities and differences. Acta Neuropathol. 2008;115(1):97–114.

18.

Kusaka

Imai

Hashimoto

. Ultrastructural study of chromatolytic neurons in an adult-onset sporadic case of amyotrophic lateral sclerosis. Acta Neuropathol. 1988;75(5):523–528.

19.

March

Coates

Abyad

. Degenerative myelopathy in 18 Pembroke Welsh Corgi dogs. Vet Pathol. 2009;46(2):241–250.

20.

Martin

. Mitochondrial pathobiology in ALS. J Bioenerg Biomembr. 2011;43(6):569–579.

21.

Miller

Barber

Porter

. Degenerative myelopathy in two Boxer dogs. Vet Pathol. 2009;46(4):684–687.

22.

Milton

Banner

Ince

. Expression of the glial glutamate transporter EAAT2 in the human CNS: an immunohistochemical study. Mol Brain Res. 1997;52(1):17–31.

23.

Ogawa

Uchida

PARK

. Immunohistochemical observation of canine degenerative myelopathy in two Pembroke Welsh Corgi dogs. J Vet Med Sci. 2011;73(10):1275–1279.

24.

Pardo

Wong

Benson

. Loss of the astrocyte glutamate transporter GLT1 modifies disease in SOD1G93A mice. Exp Neurol. 2006;201(1):120–130.

25.

Rao

Weiss

. Excitotoxic and oxidative cross-talk between motor neurons and glia in ALS pathogenesis. Trends Neurosci. 2004;27(1):17–23.

26.

Rauen

Kanner

. Localization of the glutamate transporter GLT-1 in rat and macaque monkey retinae. Neurosci Lett. 1994;169 (1–2):137–140.

27.

Rothstein

Dykes-Hoberg

Pardo

. Knockout of glutamate transporters reveals a major role for astroglial transport in excitotoxicity and clearance of glutamate. Neuron. 1996;16(3):675–686.

28.

Rothstein

Martin

Kuncl

. Decreased glutamate transport by the brain and spinal cord in amyotrophic lateral sclerosis. N Engl J Med. 1992;326(22):1464–1468.

29.

Rothstein

Van Kammen

Levey

. Selective loss of glial glutamate transporter GLT-1 in amyotrophic lateral sclerosis. Ann Neurol. 1995;38(1):73–84.

30.

Sasaki

Iwata

. Ultrastructural study of synapses in the anterior horn neurons of patients with amyotrophic lateral sclerosis. Neurosci Lett. 1996;204 (1–2):53–56.

31.

Sasaki

Maruyama

. Decreased synaptophysin immunoreactivity of the anterior horns in motor neuron disease. Acta Neuropathol. 1994;87(2):125–128.

32.

Sasaki

Maruyama

. Synapse loss in anterior horn neurons in amyotrophic lateral sclerosis. Acta Neuropathol. 1994;88(3):222–227.

33.

Sato

Inaba

Suwa

, et al. Inherited defects of sodium-dependent glutamate transport mediated by glutamate/aspartate transporter in canine red cells due to a decreased level of transporter protein expression. J Biol Chem. 2000;275(9):6620–6627.

34.

Shelton

Johnson

O’Brien

. Degenerative myelopathy associated with a missense mutation in the superoxide dismutase 1 (SOD1) gene progresses to peripheral neuropathy in Pembroke Welsh Corgi and Boxers. J Neurol Sci. 2012;318:55–64.

35.

Smith

Carney

Starke-Reed

. Excess brain protein oxidation and enzyme dysfunction in normal aging and in Alzheimer disease. Proc Natl Acad Sci U S A. 1991;88(23):10540–10543.

36.

Sobue

Hashizume

Yasuda

. Phosphorylated high molecular weight neurofilament protein in lower motor neurons in amyotrophic lateral sclerosis and other neurodegenerative diseases involving ventral horn cells. Acta Neuropathol. 1990;79(4):402–408.

37.

Tansey

Farooq

Cammer

. Glutamine synthetase in oligodendrocytes and astrocytes: new biochemical and immunocytochemical evidence. J Neurochem. 1991;56(1):266–272.

38.

Toyoda

Uchida

Matsuki

. Inflammatory myopathy with severe tongue atrophy in Pembroke Welsh Corgi dogs. J Vet Diagn Invest. 2010;22(6):876–885.

39.

Van Den Bosch

Vandenberghe

Klaassen

. Ca2-permeable AMPA receptors and selective vulnerability of motor neurons. J Neurol Sci. 2000;180 (1–2):29–34.

40.

Wininger

Zeng

Johnson

. Degenerative myelopathy in a Bernese Mountain Dog with a novel SOD1 missense mutation. J Vet Intern Med. 2011;25(5):1166–1170.

41.

Winslow

Rubinsztein

. Autophagy in neurodegeneration and development. Biochem Biophys Acta. 2008;1782(12):723–729.

42.

Won

Kim

Gwag

. Cellular and molecular pathways of ischemic neuronal death. J Biochem Mol Biol. 2002;35(1):67–86.

43.

Yamazaki

Esumi

Nakano

. Is motoneuronal cell death in amyotrophic lateral sclerosis apoptosis? Neuropathology. 2005;25(4):381–387.

44.

Zang

Lopes

Cheema

. Loss of synaptophysin-positive boutons on lumbar motor neurons innervating the medial gastrocnemius muscle of the SOD1G93A G1H transgenic mouse model of ALS. J Neurosci Res. 2005;79(5):694–699.