This study investigated whether canine MUE and EAE models display abnormally hyperphosphorylated tau. S396 + p-tau was observed in glial cells in the inflammatory lesions of all dogs with EAE and one of the three dogs with MUE. Moreover, positive staining of S396 + p-tau was found in the background of lesions in dogs with EAE. These findings suggest that hyperphosphorylated tau might be involved in the pathological mechanism in canine EAE and MUE similar to human MS.
Previously comparable findings of clinical signs, MRI, and histopathology were noted between the canine EAE model and MUE, and these features were also identified in this study [25]. In the T1-weighted images, the intracranial lesions of all dogs with EAE and dogs with MUE were confirmed from hypo- to iso-intensity and in T2-weighted and FLAIR images, they were identified as hyperintensity. The lesions were mainly observed in the white matter and/or the cortex of the brain. Identical findings were reported in the MRI scan in human MS [27].
Histopathologically, perivascular inflammatory cell infiltration and parenchymal cell accumulations of inflammatory cells were mainly characterized in both canine EAE and MUE lesions [25, 28, 29]. These features were also identified in this study. Further, swelling of vascular endothelial cells and astrocytic gliosis were confirmed. In the previous studies, the results of IHC in the canine EAE model and MUE revealed CD3-positive macrophages and T cells predominantly [20, 25]. MS is also a typical inflammatory demyelinating disease similar to MUE in relation to T cell-mediated inflammatory pathogenesis and CD3-positive macrophages [28, 30]. Glial fibrillary acidic protein (GFAP), used as a marker of disease progression in human MS, was also responsive to astrocytes in MUE [29, 31, 32]. Correspondingly, GFAP-positive astrocytes were present within and around the inflammatory lesions of the canine EAE model [25]. Based on these findings, histopathological characteristics suggest that MUE, EAE, and MS are comparable to each other.
Hyperphosphorylated tau is expressed in glial cells such as astrocytes and oligodendrocytes in the mouse EAE model and MS [15]. It triggers the loss of microtubule stability and collapse of microtubules [6]. Hyperphosphorylated tau also aggregates into pathological tau oligomers, ultimately forming pathological insoluble neurofilament tangles [6]. Similar to previous studies, the IHC distribution of S396 + p-tau in the lesion of the canine EAE model was identified in this study. Therefore, the canine EAE, which had the S396 + p-tau expressed lesion in this study could be speculated to have tau pathology similar to that found in previous research [15, 16, 33]. In addition, S396 + p-tau was expressed not only in glial cells but also in the background of the lesion. Past studies have reported that p-tau seems to disrupt glial cells, cause inflammation stimuli, propagate to other cells, and repeat its pathology [15, 16, 33]. Thus, background S396 + p-tau lesion could be the cause or result of an inflammatory lesion of immune-mediated meningoencephalitis in this study. Although further studies should be conducted, hyperphosphorylated tau pathology appears to play a role in canine neuroinflammatory diseases similar to that in human MS.
EAE-1 and EAE-2 had similar tau expression patterns, whereas EAE-3 did not. Tau expression was strong in EAE-1 and EAE-2, but moderate in EAE-3. This could be related to a variation in the brain that was used to induce EAE. Hyperphosphorylated tau protein may seed like a prion, according to a previous study [34]. It was possible that hyperphosphorylated tau in the brain, which was used as the source, had an effect. EAE-2 was induced using the same brain as EAE-1, while EAE-3 was induced using the brain of EAE-1. As a result, variations in tau expression in this study could be attributed to differences in the brain samples used. Experiments in dogs with EAE induced utilizing the same brain should be conducted to confirm this.
The results of IHC showed positive labeling of p-tau in all dogs with EAE and one dog with MUE (MUE-1). Contrary to dogs with EAE, S396 + p-tau was not identified in the other two dogs with MUE. The absence of S396 + p-tau in dogs with MUE may be due to several factors. As mentioned above, although MS, EAE, and MUE share various histopathological characteristics, there are also subtle differences [22]. The pathological hallmark of MS and EAE is a lesion consisting of perivascular infiltration of inflammatory cells, with subsequent demyelination [28, 35, 36]. Demyelination also occurs in MUE, but rather necrosis is a key pathology of MUE [22, 29]. Thus, MUE may have a partially different pathology than MS; therefore, the tau pathology involved in the pathological course might also be different.
Phosphorylation of tau protein may occur at several amino acid residues [6]. In humans, the anti-phospho-tau (pS202/pT205) monoclonal antibody (AT8) has been mainly used in tau pathology studies, as it has been shown to be involved in the oligomerization of tau in Alzheimer’s and MS in humans [16, 37]. Although studies on tau pathology in veterinary medicine have been limited, one study reported that the S396 epitope is more involved in the pathology of canine cognitive dysfunction syndrome than the AT8 epitope, which is associated with neurofibrillary tangles [26]. Compared with AT8, the S396 epitope is hyperphosphorylated at an early stage of the disease [38–40]. The only dog with MUE with the result of S396 + p-tau was necropsied at 165 days after confirmation of onset, while the other dogs with MUE were necropsied at 304 and 1165 days, respectively, after the first signs were identified. Due to the prolongation of MUE, it is likely that a late epitope rather than an early epitope was mainly involved or that other tau protein epitopes were involved in MUE pathology. Therefore, further studies are needed for other p-tau epitopes, such as AT8 and AT180.
In this study, all dogs with MUE were treated with immunosuppressants, such as prednisolone and MMF, from diagnosis to death. Previous studies reported that hyperphosphorylation of tau was induced by hyperactivating the mammalian target of rapamycin (mTOR) pathway [41, 42]. Further, it is well known that glucocorticoids, such as prednisolone, inhibit the mTOR pathway [43, 44]. In addition, a study reported decreased p-tau expression in the group containing the mouse EAE model which was treated with prednisolone [14]. Altogether, it was suspected that p-tau was not expressed in dogs with MUE due to the long-term immunosuppressive treatment. Additionally, it was presumed that a little amount of p-tau was confirmed for MUE-1 because the treatment period was relatively short compared with that of other dogs with MUE.
According to a previous NME study [20], NME lesions could be divided into three stages depending on the extent of tissue necrosis and the severity of the inflammatory response. Compared with MUE-1, moderate tissue necrosis and severe inflammatory changes were noted in the histopathologic examination of MUE-3. In MUE-2, severe malacic changes and cavitations were dominant, although inflammatory changes were less compared with MUE-1 and MUE-3. Based on the histopathologic features and the survival time, MUE-2 and MUE-3 were considered to be in the late stages of NME compared with MUE-1. EAE models were euthanized within two days after symptom confirmation; hence, the disease was in its infancy, and necrosis hardly progressed. Considering that accumulation of S396 + p-tau was mainly found inside glial cells in MS and that most of these glial cells were in a necrotic state in MUE cases, p-tau could not be normally confirmed. To confirm this, further studies on MUE brains in the early stage of the disease are needed.
This study had several limitations. First, the sample group was too small to generalize the relationship between MUE and tau protein. The number of healthy controls, in particular, was insufficient, as was the number of samples by the type of MUE. Additional studies should be carried out with a larger number of samples. Second, various types of p-tau antibodies were not utilized. As mentioned above, hyperphosphorylation could occur at several amino acid residues. In this study, only the anti-tau (phospho-S396) antibody was used for IHC. The possibility of expressing p-tau of different isotypes could not be confirmed. Therefore, studies including other p-tau antibodies such as AT8 and AT180 should be conducted. Finally, there were no validations in untreated dogs with MUE and testing in dogs with MUE at an initial stage. Significant discrepancies in the outcomes could be caused by differences in immunosuppressive treatment and the extent of brain tissue necrosis. Brain samples from an initial MUE were not available. Therefore, further research is required with samples that fulfill these criteria.