The Role of Glia in Canine Degenerative Myelopathy: Relevance to Human Amyotrophic Lateral Sclerosis

Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease characterized by progressive degeneration of motor neurons and grim prognosis. Over the last decade, studies on neurodegenerative diseases pointed on the role of glia in supporting the proper function of neurons. Particularly, oligodendrocytes were shown to be essential through myelin production and supplying axons with energy metabolites via monocarboxylate transporters (MCT). We have used dogs with naturally occurring degenerative myelopathy (DM) which closely resembles features observed in human ALS. We have performed two types of analysis of spinal cord tissue samples: histology and molecular analysis. Histology included samples collected from dogs that succumbed to the DM at different disease stages, which were compared to age-matched controls as well as put in the context of young spinal cords. Molecular analysis was performed on spinal cords with advanced DM and age-matched samples and included real-time PCR analysis of selected gene products related to the function of neurons, oligodendrocytes, myelin, and MCT. Demyelination has been detected in dogs with DM through loss of eriochrome staining and decreased expression of genes related to myelin including MBP, Olig1, and Olig2. The prominent reduction of MCT1 and MCT2 and increased MCT4 expression is indicative of disturbed energy supply to neurons. While Rbfox3 expression was not altered, the ChAT production was negatively affected. DM in dogs reproduces main features of human ALS including loss of motor neurons, dysregulation of energy supply to neurons, and loss of myelin, and as such is an ideal model system for highly translational studies on therapeutic approaches for ALS.


Introduction
Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disorder with an inevitable progression, extremely poor prognosis, and no cure till now. While ALS is relatively rare, its occurrence increases with age, so the burden of this deadly disease will be growing, as societies age [1]. In the USA, there are approximately 5600 people diagnosed with ALS every year. ALS affects both upper and lower motor neurons (UMN, LMN). Progressive damage of these neurons leads to lack of muscle stimulation and atrophy. The involvement of respiratory muscles usually leads to death [2]. Historically, ALS has been considered as a disorder of motor neurons, and accordingly, a majority of research efforts focused on this aspect. Currently, there is a growing appreciation for the role of the microenvironment and all neural phenotypes in the brain [3]. The studies conducted during the last decade have raised the supportive role of glia in the homeostasis and survival of neurons. Oligodendrocytes, due to their unique spatial arrangement and large contact surface with motor neurons, seem to be extremely important. Oligodendrocytes are well known for their role in producing myelin, which ensheaths and insulates axons in the central nervous system (CNS), facilitating rapid and energy-efficient conduction of action potentials. Recent studies have demonstrated that the role of oligodendrocytes extends beyond myelination in the maintenance and long-term survival of neurons [4]. The energy demands of firing neurons are supplied along the entire length of axons, which, in the case of motor neurons, spans up to 1 m. Condensed insulation of the axon by the myelin membrane, except for the nodes of Ranvier, creates a significant barrier for energy substrate transfer to the axon, which precipitates the need for a mechanism to facilitate local feeding of axons by oligodendrocytes. As shown by Lee and others [4], oligodendroglia do, indeed, participate in the transportation of energy substrates into axons through specialized monocarboxylate transporters (MCT).
Extensive activation of the most abundant glial component-astrocytes-was seen both in ALS patients and in dogs with degenerative myelopathy (DM) [5,6]. Astrocytes are significant contributors to neuronal dysfunction and death in ALS pathology. Moreover, it was reported that co-culture of astrocytes from ALS patients, both familial and spontaneous cases, is toxic and leads to the death of motor neurons [7]. Furthermore, astrocytes isolated from sporadic ALS (sALS) patients and transplanted to a mouse model resulted in degeneration of the motor neurons and non-motor neurons of the host [8]. These observations about non-cell autonomous pathology are largely derived from transgenic animal models, which are clearly useful tools; however, there is an increasing criticism that artificially created animal models of human disease are not reliable for the evaluation of treatment outcomes, as they do not address the disease pathophysiology and variability existing in human population. Canine DM is an inherited, progressive, adult-onset neurodegenerative disease with clinical, histopathologic, and genetic parallels to human ALS, including presence of mutated superoxide dismutase 1 gene (SOD1) [9]. The similarities between the canine and human nervous systems, exposition to similar environmental factors as human, the homogeneity in the onset and clinical progression of disease, and the ability to longitudinally analyze and collect samples from affected dogs with DM represent a uniquely valuable large animal disease model for therapeutic development.
Our objective was to evaluate whether glia, and their MCTs may play a significant role in the degeneration of motor neurons in canine DM similarly to previously reported for transgenic SOD1 mice [4]. This study was focused on the assessment of the status of the glial component, including mature astrocytes, and myelinating oligodendrocytes, as well as MCT in the spinal cord to give a basis for using DM to investigate glia-focused therapeutic strategies in a highly translational setting of large animals with ALS-like disease (experimental design is shown in the Fig. 1).

Quantitative Analysis of Motor Neurons
Cresyl violet staining revealed characteristic large-sized motor neurons within the spinal cord (Fig. 2). We performed motor neuron counts in both old and young controls and did not observe significant differences between those two groups, thus excluding effect of the age. Quantitative analysis has shown that the number of motor neurons was reduced in dogs with DM, compared to both young and old control (unaffected) animals. The highest reduction of motor neurons was seen in most advanced, grade four DM (12.00 ± 0.84 in DM-affected animals vs. 37.10 ± 1.67 in old control animals). There was clear progression of the motor neuron loss with progression of the disease, and starting from DM grade 2, the difference was statistically significant when compared to age-matched control groups (p < 0.05). Moreover, we evaluated a number of motor neurons in the region of lamina IX, which contained somatic neurons. We were unable to determine any correlations between grade and number of motor neurons in this region. The highest reduction in number of motor neurons was in dogs with grade four of DM (9.412 ± 0.91).

The Assessment of Demyelination
Eriochrome staining enabled myelin visualization in the spinal cord. The dark blue color of eriochrome was prominent throughout white matter in control dogs. Regions of demyelination were seen as voids of the blue color primarily on the outer surface of the spinal cord, as shown in Fig. 3, in dogs with initial stages of disease. Whereas, the brightening of blue color was evident in dogs with advanced DM. Results are shown in Fig. 4. There was a significant decrease in the myelination level of the tissues of DM dogs compared to controls. There was a significant difference starting from grade 3 (65.57 ± 2.56) compared to old control dogs (86.69 ± 2.92).

Molecular Analysis of Matched Spinal Cord of Dogs Affected with DM
The expression of different genes in two groups (the unaffected-control and DM-affected) of spinal cords was measured. The DM-affected dogs had a higher expression of genes related to microglia, including AIF1 (p < 0.0001) and lower expression of genes related to microglia including MBP (p = 0.02), Olig1 (p = 0.025), and Olig2 (p = 0.048). There was no significant difference in the expression of GFAP, MAG, or the MOG gene (p > 0.05), but we observed a downward trend in the expression profiles of the MAG gene.
Disease did not influence Rbfox3 expression (Fig. 5a), but it markedly affected the expression of the ChAT gene ( Fig.  5a), which encodes an enzyme that catalyzes the biosynthesis of the neurotransmitter acetylcholine.
There were significant differences in the expression of MCT1 (p = 0.0005; Fig. 6a), MCT2 (p = 0.0012; Fig.  6a), MCT4 (p = 0.0075; Fig. 6a) responsible for coding a proton-linked monocarboxylate transporter that catalyzes the movement of many monocarboxylates, such as lactate and pyruvate, across the plasma membrane. Western blot analysis of proteins MCT1, -2, and -4 showed a prominent reduction in MCT1 and -2 proteins in samples from DM-affected dogs (Fig. 6b). We noticed an increase in the level of the MCT4 protein level compared to unaffected dogs.

Discussion
Over the last decade, multiple transgenic animal models of oligodendroglia injury have been investigated. All of these have demonstrated axonal pathology independently. This may suggest that myelin pathology causes axonal injury in ALS. This led us to evaluate the glia component in dogs that are affected by DM. Previous papers have outlined the histopathological changes in the spinal cords of dogs affected by DM [10][11][12][13]. Prior histopathologic studies of canine DM showed a complete loss of axon/ myelin cylinders replaced by prominent astrogliosis [6,12]. Our results confirmed demyelination and neuronal degeneration in the spinal cords of dogs affected by DM. Regions of demyelination were seen as the absence of blue EC staining (Fig. 3). Loss of blue staining was most prominent in the spinal cord of a dog with grade four DM. In spinal cords affected by DM, the lack of myelin was most noticeable in the periphery of the spinal cord, which corresponds to the location of the dorsal spinocerebellar pathways and is in line with the observations by March et al., 2009 [12]. Significant differences were present between grade three and four compared to control old unaffected dogs. No significant loss of myelin was observed in the spinal cords of the control dogs, which led us to hypothesis that such loss of myelin could be related to the grade of DM, and occurs in a correlation with axonal degeneration and secondary demyelination as shown with cresyl violet staining (Fig. 4). Along with the loss of myelin with the grade of DM, we also observed a loss of neurons in the spinal cord, even in grade 2 (Fig.  2). The reduction of neuron numbers in the gray matter in each grade of DM was statistically significant compared to controls [14]. ChAT gene expression was reduced corresponding to histological analysis. Moreover, ChAT activity was markedly reduced in analyzed spinal cords from ALS patients which confirms clinical relevance of our model [15,16]. Loss of myelin ensheathing in the spinal cord was confirmed with quantitative analyses based on the expression of genes MBP (myelin basic protein), Olig1, and Olig2. The examination of MCT provided a great opportunity to investigate whether primary myelin degeneration has any impact on the degeneration of motor neurons. A lower mRNA expression of MCT1 was found in a mouse model of ALS resulting in neuron cell loss,  [17]. An increase in the MCT4 which is expressed predominantly in astrocytes can be correlated with an increase in the number of both astrocytes and microglia within the spinal cord, which was previously shown by other investigators [6,11,18]. We confirmed statistically significant increases of microglia by AIF1 gene expression within spinal cord from dogs with DM, but we were not able to see significant increase of expression of GFAP gene in those samples. However, we have observed an increasing trend in expression of GFAP gene.
A decrease in MCT2 presumably can be linked with a decrease in the expression of the Olig1 and Olig2 genes. We hypothesized that degeneration of oligodendrocytes leading to poor myelination of axons was connected to neuronal degeneration and lower expression of MCT2. Further investigation should be focused on expression of these genes on different stages of disease. Our study supports the thesis that glial component as well as MCT are affected in this naturally occurring ALS-like disease in dogs. Evidence about the role of glia in DM is mounting [19], and it is essential to undertake further studies on glia replacement. While transplanted glial progenitors were shown to be highly efficacious in dysmyelinated mice [20], translation of this approach to humans largely failed [21]. Extreme differences in brain size between rodents and human make clinical translation challenging, and in this context, it is essential to perform future studies in

Sample Collection
We have received tissue samples courtesy of JRC (Dept. Veterinary Medicine and Surgery, College of Veterinary Medicine, University of Missouri, USA). Mid to lower thoracic spinal cord segments were collected immediately after euthanasia and frozen for total RNA and protein extraction, or fixed in 4% formalin for histological examination. This region of the spinal cord was selected because the histopathologic changes become apparent early in the disease process and progress in severity throughout the disease course [12]. Tissues were obtained from dogs euthanized due to a diagnosis of presumptive DM, or from older dogs (control) euthanized from other non-neurologic disease causes. Young (< 5 years) and older (> 9 years) dogs were included to evaluate for the effect of age on the number and condition of motor neurons in histological assessment. Control dogs were euthanized due to unrelated diseases (e.g., blood in stool, weight loss, orthopedic diseases, or chronic renal disease, or abdominal mass; or were euthanized as a part of another study as control animal). Samples from DM-affected dogs were included if the following criteria were met: (1) clinical history and neurologic examination consistent with DM, with assignment to disease stage based on a clinical grading scale (Table 1) [9]; (2) homozygosity for 118G > A SOD1 mutation; and (3) histopathologic diagnosis of DM that confirmed axonal degeneration and astroglial proliferation that was most severe in the dorsal portion of the lateral funiculus and in the dorsal funiculi of the caudal thoracic spinal cord. The severity of DM was rated according to a grading scheme [9]. Initially, clinical signs emerge as an asymmetric spastic paraparesis and general proprioceptive (GP) ataxia (stage I), progressing to nonambulatory paraparesis/paraplegia (stage II) with the emergence of pelvic limb lower motor neuron signs within 1 year of disease onset. Dog owners often elect euthanasia when their dog becomes non-ambulatory in the pelvic limbs. Neurologic deficits progress to include thoracic limb weakness (stage III) followed by flaccid tetraplegia, generalized muscle atrophy, and brainstem dysfunction with eventual respiratory failure (stage IV). In total, 17 dogs were included in histological assessment (12 DM dogs, 2 age-matched controls, 3 young control animals), and 9 dogs were included in molecular analysis (5 DM dogs and 4 age-matched controls). The number of animals used in the experiments is shown in Table 1. Outlined experimental design is shown in the Fig. 1.

Histological Assessment
Tissue samples protected in 4% formalin were cryoprotected in 30% sucrose and then frozen on dry ice. Next, samples were cryosectioned on a Hyrax C25 PLMC cryostat (Zeiss, Warsaw, Poland) in 10-μm sections, mounted on slides, and stored at − 20°C.

Eriochrome Staining
For eriochrome (EC) staining, sections were rehydrated in a graded series of ethanol (ETOH; 95%, followed by 70% ETOH and distilled water), 10 min each at room temperature. Next, sections were stained for 15 min using an eriochrome cyanine solution. Subsequently, sections were differentiated by alternating between exposure to 0.1% NH 4 OH for 3-7 s and rinsing in distilled water for~30 s until the blue background was reduced. The total time of exposure to NH 4 OH was approximately 20 s. After the last rinse in distilled water, sections were dehydrated in 70%, 95%, and 100% alcohol (two times for each), overexposed in xylene (three times for 10 min), and mounted using DPX mounting medium (Sigma-Aldrich, USA). Myelin stained with eriochrome cyanine R was visualized and assessed using bright-field microscopy. Images were captured under × 20 magnification using an Axio Scan.Z1 (Zeiss) and the myelination areas were analyzed using Adobe Photoshop CC software (San Jose, CA, USA). Areas of the spinal cord were outlined using the magnetic lasso tool; the pictures were converted to grayscale, and then measured. The myelination area was measured with the threshold at the level of 148 to analyze the percent of myelination.

Cresyl Violet Staining
For cresyl violet staining, sections were rehydrated in a graded series of ethanol dilutions (ETOH; 95%, followed by 70% ETOH and distilled water), for 3 min each at room temperature. Then, sections were rinsed in distilled water and transferred for one and 0.5 min to cresyl violet solution heated to 60°C. Sections were de hydra t ed in 70%, 95%, and 100% alcoh ol, overexposed in xylene (two times for 10 min), and mounted using DPX mounting medium (Sigma-Aldrich, USA). Neurons stained with cresyl violet were visualized and assessed using bright-field microscopy. Images were captured under × 20 magnification using an Axio Scan.Z1 (Zeiss), and the number of neurons is analyzed using ZEN software (Zeiss). For the statistical analysis, the total number of neurons within gray matter and motor neurons within Rexed lamina IX were used and compared between animals.

RNA Extraction and Reverse Transcription
Total RNA was extracted from spinal cord samples using Total RNA Plus concentrator according to the manufacturer's protocol (A&A Biotechnology, Poland). The RNA purity and concentration were measured using a NanoDrop 1000 spectrophotometer (Thermo Fisher Scientific Inc., USA). All sample purity ratio of 260 nm and 280 nm was between 1.9 and 2.0. The next step was reverse transcription of RNA to cDNA using a High-Capacity RNA-to-cDNA™ Kit (Thermo Fisher Scientific Inc., USA). Two types of controls were used: one without complementary DNA (cDNA), and the second, in the absence of the reverse transcriptase. The obtained cDNA was diluted in nuclease-free water and stored at − 80°C until analysis.

Real-time PCR
Real-time quantitative PCR was achieved using Applied Biosystems 7900HT Fast Real-time PCR System (Applied Biosystems, USA). PCR conditions included the following: an initial denaturation step (10 min at 95°C), followed by   Table 2, Thermo Fisher Scientific Inc., USA) filled to 10 μl with nuclease-free water. Expression levels were normalized to GAPDH, and relative expression was calculated using the comparative threshold cycle method (ΔCT).

Data Analysis
Statistical analyses were performed using GraphPad Prism7 (MDF Co., Ltd., Japan). Data are expressed as the mean ± SEM. Data from real-time PCR and Western blot were analyzed using the Student's t test. Additionally, we used oneway ANOVA to analyze results from histological staining. Differences between groups were considered statistically significant when the p value was < 0.05.