Mutations linked to transcription factors Sox9 and Sox10 have long been related to diseased states in humans and mice.16,17 However, much less is known about the role of Sox8 in disease. In 2013 the rs2744148 SNP, located downstream of the Sox8 gene was associated with a higher risk to develop MS.13 Moreover, Sox8 is expressed in Ol, the myelinating cells of the CNS, where it influences myelin maintenance and myelin gene expression.11 Due to its relevance in Ol physiology and connection to MS risk, we examined whether lack of Sox8 may have an effect in the cuprizone model of toxic demyelination.
Sox8−/− mice are viable and do not show any severe developmental phenotype.18 The most obvious phenotype of these mice lies in smaller fat deposits, that result in ~ 30% lighter weight than WT animals.18,19 We observed that Sox8 deficient mice lose significantly more weight over the course of the cuprizone diet. These results suggest that not only do Sox8−/− mice experience adipose tissue degeneration in adulthood 19, but they seem to be more prone to weight loss under the cuprizone diet. Whether this tendency to lose weight expands to other stress conditions would need to be evaluated in future studies. In addition, we described a motor deficit in these mice upon cuprizone exposure. To the best of our knowledge, the motor skills of the Sox8−/− line have not been analyzed in detail. Notably, we identified a slight motor phenotype in naïve Sox8 deficient mice. Even though we did not assess the muscle tissue in this study, the presence of motor deficits opens an interesting venue of research for Sox8. Previous research has linked Sox8 to myogenic differentiation and it has been recently identified as a genetic variant of interest for severe adolescent idiopathic scoliosis and familial essential tremor.20–22 Altogether, the behavioral findings of this study further characterized this mouse line and open compelling questions regarding the role of Sox8 in disease.
SoxE transcription factors are closely entangled in Ol physiology. Sox9 is essential for proper Ol specification and Sox10 for terminal differentiation.5,6 Due to the redundancy with its partners, it has been challenging to identify a unique function for Sox8.7,8 Only recently, Turnescu et al. have proposed that Sox8 is necessary for myelin maintenance and myelin gene expression.11 In the present study, we challenged Sox8 deficient mice with the cuprizone model of toxic demyelination which allows the analysis of both de- and remyelinating stages. This approach may elucidate whether Sox8, by itself, may be relevant in a demyelinating condition. Interestingly, Sox8 deficient mice showed a delay in the remyelinating phase. A similar hiatus has been previously reported in these mice in early developmental stages, where Ol terminal maturation was reduced up to 50% resulting in a transient delay in myelination.8 However, this mouse line has not been challenged in a disease model before and it is the first time that Sox8 is reported to have an effect in the adult diseased brain.
It is becoming clear that many developmental mechanisms, such as myelination, are to some degree recapitulated during adult regeneration or insult.23–25 Here, we observe a similar delay in the regenerative process of remyelination as occurs during developmental myelination in the absence of Sox8. To shed light on the cause of the remyelination delay we analyzed different oligodendroglial sub-populations that are essential for this process. Recent studies have shown that newly generated Ol, derived from OPC, are very efficient during remyelination, in contrast to the modest contribution of surviving mature Ol.26–28 Hence, we investigated the changes in proliferating OPC during cuprizone. In accordance with the literature,29 we detected an increase in proliferating glia in CC at the peak of demyelination, an increment that was significantly hindered in the Sox8 deficient group. Additionally, we reported a failure to replenish the mature Ol population during the remyelination stage in Sox8−/− mice. In the cuprizone model, mature Ol numbers decrease dramatically until week five. Afterwards, they start to repopulate the demyelinated area until they resemble the original density.29 Such reduction and eventual recovery of mature oligodendroglia is precisely what we observed in the WT group. The lack of Sox8 might impair OPC proliferation and consequently the proper replacement of mature Ol. They recover slowly in the Sox8−/− group indicating that this failure represents a delay more than a permanent damage, resembling the phenotype observed in Sox8 deficient mice during development. Future studies, implementing either further time points or chronic cuprizone may help to fully answer this question. While we observed significantly reduced numbers of mature Ol by wk 6, we detected deficits in remyelination in Sox8−/− mice as early as wk 5.5. Therefore, remyelination deficits may not be solely based on a lack of mature Ol, but Sox8 may have a separate effect on remyelination, aside from proliferation and differentiation of oligodendroglial cells.
Local gliosis is characteristic of the cuprizone model, hence we analyzed both astrocytic and microglial response. The described changes seem to be independent of the local astrogliosis since we did not observe significant differences in the level of astrocytic activation. Microglia rapidly gets activated upon cuprizone diet and it steadily decreases during remyelination.29,30 Unexpectedly, we observed a tendency for microglia to remain activated in the Sox8 group in early stages of remyelination. In recent years the ample role that OPC play in neuroinflammation has been recognized, as well as the capacity of OPC to impact other inflammatory cell types, such as microglia.31,32 OPC have suppressive capacity over microglia and the ablation of this population leads to overactivation of microglia in a neuroinflammatory milieu.32 It is tempting to speculate that the deficient OPC proliferation observed at wk 5 might be involved in the impaired control of the microglia response. Further studies would be necessary to link the two observed changes.
Altogether, our report enhances the characterization of the Sox8 knockout mouse line, and it highlights its utility to understand the role of Sox8 in disease context. Until now, Sox8 has been given little attention due to the lack of unique functions independently from its partners Sox9 and Sox10. Although Sox8 appears to be dispensable during Ol development, it seems to have greater impact in adult mice and to be specially involved in pathological pathways. In the present study we showed that lack of Sox8 results in increased motor deficits, weight loss and remyelination delay in the cuprizone model. Additionally, we report a defect in OPC proliferation which may lead to a failure to replenish mature Ol populations and may be possibly involved in deficient microgliosis reduction. In conclusion, our report opens many questions regarding Sox8, not only in MS but other degenerative diseases. The so long shunned member of the SoxE family seems to have more critical functions in pathological conditions than previously considered. It would be of most interest to fully dissect its contributions to disease and analyze its potential as a therapeutic target.