In recent decades, studies on the etiopathogenesis of SRMA mostly focused on the role of lymphocytes [3, 21, 28, 29]. Indeed, for many years the adaptive immune system has been believed to play the most important role in triggering an inappropriate immune response. However, more recently, the innate immune system aroused much interest for its ability to modify the adaptive immune response, particularly in autoimmunity and immune-mediated diseases [30, 31]. TLRs are important components of the innate immune system: their ability to initiate and propagate inflammation protects the organism from infectious diseases . On the other hand, an excessive activation of these receptors may lead to immune disorders [10, 30, 33, 34]. The ambivalent role of these receptors makes them interesting candidates for immune pathological studies in SRMA patients, especially because CSF cells can be studied by systematic flow cytometric studies in this large animal model. The activation profile of TLRs in SRMA was tested to support the hypothesis that these receptors are stimulated by infectious antigens or endogenous proteins (self-antigens). It was hypothesized that they are key factors for the initiation of the inflammatory process and provide an indirect hint of the etiology of the disease. Therefore, the expression of TLRs in dogs affected with SRMA was measured and compared to infectious diseases or other neurological conditions.
The hypothesis that SRMA is maintained by a continuous bacterial infection had to be rejected by the current study. The comparison of TLRs expression profiles of SRMA dogs with dogs affected with bacterial/pyogenic infection failed to show clear similarities between the two groups. In addition, the clear response to long-term treatment with glucocorticosteroids does not support a suspected classical bacterial infection [35–37]. However, triggering of an autoimmune reaction by bacteria cannot be ruled out completely in SRMA.
Indeed, TLR4 was statistically more frequently expressed on monocytes of dogs with untreated SRMA and dogs with pyogenic infections. TLR4 recognizes not only lipopolysaccharides, but also some endogenous ligands, such as heat shock proteins (HSP60, HSP70), fibronectin, hyaluronic acid, fibrinogen and heparan sulfate . In the current study, the triggering factor for increased TLR4 expression, a self or non-self antigen, was not examined. A recent study showed that HSP 70 is elevated in SRMA . Therefore, it seems to be very likely that the triggering protein in SRMA might derive from a self-antigen such as the HSP 70.
The role of TLR4 in human patients with sepsis, but also in non-infectious diseases, such as inflammatory bowel disease and rheumatoid arthritis, were studied . In the canine patient the role of TLR4 during sepsis and the related systemic inflammatory response syndrome has still not been investigated. However, a potential role of TLR4 in dogs with osteoarthritis  and chronic enteropathies  was proven.
The role of TLR4 for neutrophil recruitment into the CNS was demonstrated in a murine model of systemic inflammation . Similar mechanisms might lead to the invasion of neutrophils into the subarachnoidal space in SRMA and explain the exorbitant neutrophilic pleocytosis in acute cases. A recent study on human large vessel vasculitides suggested that TLR4 is causing transmural panarteritis . Clinical and histopathological findings in dogs affected with SRMA include neutrophilic leukocytosis, neutrophilic pleocytosis and systemic vasculitis. Treatment with glucocorticosteroids reduces these pathological processes and the expression of TLR4 on monocytes declines significantly contemporarily (P = 0.019). These findings strongly suggest that TLR4 plays an important role in triggering the described pathological findings in SRMA. Additionally, it opens the discussion for new treatment modalities, such as anti-TLR4 antibodies and TLR4 antagonist; some compounds from the latter class are already under clinical trials for treatment of sepsis in human patients [39, 42].
TLR9 seems to be constantly increased on PB leukocytes in almost every disease examined in the current study. However, patients with SRMA and dogs affected with meningoencephalitides of unknown etiology showed the highest expression of TLR9, suggesting a potential role of this TLR in inflammatory CNS diseases with a possible autoimmune component. TLR9 is primarily involved in the recognition of bacterial DNA . In human medicine, TLR9 also seems to play an important role in class-switching to pathogenic autoantibody production in systemic autoimmunity , suggesting the possibility that an autoimmune reaction can maintain the inflammatory process in SRMA. However, the role of TLR9 inducing autoimmunity is controversially discussed. TLR9 can have a more regulatory function  or enhance pathologic processes  in the same animal model of multiple sclerosis. Further studies are needed to elucidate the role of TLR9 in SRMA.
Dendritic cells, once activated by TLR4 and TLR9, produce interleukin-23 (IL-23), which subsequently activate CD4+ T cells. The last are known to shift towards Th17-differentiation under the effect of IL-6 and transforming growth factor beta 1 (TGFβ1) [30, 44]. This relatively new class of T helper cells is believed to be involved in triggering aberrant immune responses and recruiting neutrophils [45, 46]. A recent study of our research group showed the concomitant increased intrathecal production of IL-6 and TGF beta-1 in dogs affected with SRMA , suggesting a new hypothesis: that the aberrant immune response in SRMA might be associated with Th17 cells maintaining the autoimmune reaction.
Although most TLR responses lead to inflammation, there are studies suggesting an important role of TLRs in homeostasis [43, 48]. The role of certain TLRs has been validated in different human diseases resulting in a wide research area focusing on possible new treatment strategies . In case of such multivalent receptors, the great challenge is to reduce the unnecessary inflammation without affecting regulatory functions of TLRs. For example, many efforts attempt to find partial TLR4 agonists, rather than antagonists, and some compounds are already currently available for human use [39, 42]. So far in companion animals, the interest has been limited to TLR ligands for developing new vaccines , but considering the rapid progress in human medicine, a similar breakthrough is expected soon in veterinary medicine. SRMA would be an ideal model to study such treatment strategies.
Limitations of this study include the absence of true controls, the number of patients in some groups and the variation of the time of sampling in the course of the disease. We tried to overcome the limitations comparing a broad spectrum of different neurological diseases. Including only a few cases of a relatively uncommon disease in dogs, bacterial meningoencephalitis, might be the reason TLRs in CSF samples were not statistically comparable. We tried to overcome this limitation by including dogs with pyogenic infections not affecting the nervous system, since SRMA is considered to be a systemic inflammatory disorder .