Fasudil-Triggered Phagocytosis of Myelin Debris Promoted Meylin Regeneration via the Activation of TREM2/DAP12 Signaling Pathway in Cuprizone-Induced Mice

The inammation and demyelination of the central nervous system (CNS) are mainly involved in multiple sclerosis (MS), in which the disorder of myelin regeneration leads to continual neurologic impairment. Fasudil, one of the ROCK inhibitors, has been shown protective functions in some models of demyelinating diseases. In this study, Fasudil treatment ameliorated the behavioral performance and myelin loss in CPZ-fed mice. Here, we demonstrated a new role of Fasudil, which triggered microglia to uptake myelin debris in both cell and animal experiments. This increased phagocytosis was associated with the polarization of M2 microglia. Furthermore, we found that Fasudil enhanced the expression of triggering receptor expressed on myeloid cells 2 (TREM2) and DNAX-activating protein of 12 kDa (DAP12), which regulated microglial phagocytosis and M2 polarization. The silence of TREM2 effectively blocked Fasudil-triggered phagocytic capacity, suggesting that Fasudil-triggered phagocytosis depends on TREM2 signaling pathway. Based on these evidences that TREM2 regulates microglial M2 polarization and phagocytosis, future studies targeted Fasudil as a therapy for demyelinating and neurodegenerative diseases are warranted.


Introduction
Demyelinating disease is a kind of immune-mediated diseases in central nervous system (CNS), mainly characterized by multifocal in ammatory demyelination, including multiple sclerosis (MS) and neuromyelitis optica (NMO) (Höftberger et al. 2017). In recent years, the incidence rate of demyelinating diseases has been increasing worldwide, which seriously affects the life quality and physical/mental health of patients (Dobson et al. 2019;Hor et al. 2020). Although the etiology of demyelinating diseases is still unclear, its histological feature mainly includes T/B cells in ltration, microglial activation, neuroin ammation, oligodendrocytes (OLs) death, subsequent demyelination and neuronal death (Reich et al. 2018). In terms of MS, with the development of disease, the treatment becomes increasingly complex. At present, disease-modifying therapies can reduce the frequency of relapse and the severity of the MS (Bross et al. 2020). Considering that some new drugs may be associated with potentially serious but rare adverse events, it is necessary to maximize the bene tial pro le and minimize the risk to patients.
At present, the available drugs for MS are predominantly immune regulation, which do not directly target demyelination. Myelin regeneration is regarded as a critical approach for the therapy of demyelinating disorders, since the increase of myelin regeneration contributes to the recovery of neuronal function and clinical symptoms (Plemel et al. 2017;Luchetti et al. 2018). Recently, studies have been interested in myelin regeneration, which is a process of myelination around the axons and has been recorded in animal models and MS patients (Lubetzki et al. 2020;He et al. 2021). Therefore, myelin regeneration is a crucial way to treating MS, since myelin is an important element to protect myelin-axon unit (Stadelmann et al. 2019). For this reason, myelin regeneration from OLs is the main mechanism of natural repair against demyelination.
Myelin regeneration is the natural regeneration to counter demyelination (Franklin et al. 2008), but the reasons for the disorder or incompleteness of myelin regeneration during MS are not fully elucidated (Goldschmidt et al. 2009). The process of myelination is subjected to both positive and negative regulation (Plemel et al. 2017). The failure of remyelination can be segregated into at least two distinct stages: the mobilization of OPCs in the lesions is impaired, and/or the differentiation/maturation of OLs is obstructed (Plemel et al. 2017). In progressive MS patients, although OPCs are present in the demyelinating lesion, mature OLs are almost completely de cient (Chang et al. 2002;Kuhlmann et al. 2008). Therefore, the reason for the failure of remyelination may be partially due to the lack of migration/proliferation of OPCs, and more likely to the failure of OPC differentiation, especially due to the imbalance between inhibitory and stimulant signals in the lesion area.
Furthermore, the failure of remyelination may also be associated with the presence of myelin debris in demyelinating regions, which signi cantly obstruct the formation of mature OLs (Lloyd et al. 2019). The accumulation of degraded debris not only causes in ammatory response, but also impedes the restoration of myelin sheath in the CNS. Studies show that microglial phagocytosis is bene cial for tissue repair and play an critical role in the progress of demyelinating disorders (Neumann et al. 2009;Pinto et al. 2020). Therefore, skewing strategies to enhance the phagocytosis of microglia offer a promising alternative in the future treatment of MS.
Fasudil, a potent ROCK inhibitor, has been shown its bene cial effects in demyelinating diseases possibly through different mechanisms, including anti-in ammation, anti-oxidation, anti-apoptosis and so on (Yan et al. 2019). In this study, we evaluated the therapeutic effectiveness of Fasudil in CPZ-fed mice by microglial phagocytosis of myelin debris and explored the underlying cellular and molecular mechanisms of the action.

Animals
Male C57BL/6 mice (9-10 weeks), weighing 21-22 g, were obtained from Vital River Laboratory Animal Technology Co. Ltd. (Beijing, China). Mice were fed in pathogen-free conditions with constant temperature, humidity and light/dark cycle for seven days before experimental manipulation.
Then mice were divided into three groups (N = 12/group) as follows: normal diet group, CPZ diet plus saline treatment group (CPZ + NS) and CPZ diet plus Fasudil treatment group (CPZ + Fasudil). Mice were fed with 0.2% (w/w) CPZ diet for 6 weeks and weighed every other day. After fed with a standard diet for It has been showed that CPZ-induced mice have the abnormalities of anxiety-and depression-like behaviors (Sen et al. 2019;Mohamed et al. 2019). Therefore, elevated plus maze (EPM), and forced swimming (FS) tests were used respectively to measure anxiety and depression of mice (Lister et al. 1990;Naserzadeh et al. 2019). EPM test: each mouse were carefully put into the center of the plus-maze and the total distance in open arm was recorded during the limited time (10 min). FS test: each mouse was put into a cylinder lled with a depth of 20 cm water (25 ± 1℃). The mean swimming speed was calculated by SMART V3.0 software.

Tissue processing
Half of mice underwent transcardial infusion using saline and 4% paraformaldehyde (PFA). Then brains were dehydrated using graded sucrose solutions and embedded with OCT. Brain coronal sections were sliced for uorescent myelin and immuno uorescence staining. The other half of the mice was perfused with saline and brains were stored at -80℃ for subsequent expriments.

Fluorescent myelin staining
Brain sections were rehydrated with 0.2% TRITON X-100 and the FluoroMyelin Green staining solution (Thermo, USA) was prepared. Then, the sections were ooded with staining solution at RT for 20 min. After washed three times, the brain sections were observed under microscopy and quantitatively measured by Image-Pro Plus 6.0 software.

Immuno uorescent staining
After blocking with 1% BSA/PBS, brain slides were added with anti- Iba1

Microglial culture
The BV2 mouse microglia obtained from ShenKe Biological Technology Co., Ltd. and cultured in complete medium (Dulbecco's Modi ed Eagle Medium with 10% fetal bovine serum, 100 µg/ml streptomycin and 100 U/ml penicillin) at a constant incubator. BV2 microglia were incubated for further experiments.
Puri cation and labeling of myelin debris Myelin debris was extracted as previously reported (He et al. 2019). Brie y, mice brain was homogenized in sucrose and washed with PBS. Then the pellet was obtained using gradient density centrifugation and labeled in carboxy uorescein succinimidyl ester (CFSE) solution. After centrifugated, uorescein-labeled myelin debris (FMD) was washed with PBS and resuspended to 100 mg/ml. Transfection BV2 cells were cultured in 24-well plates overnight. According to the instructions of Lipofectamine 2000 transfection reagent (Thermo, USA), 1 µl Lipofectamin 2000 was added into 50 µl Opti-MEM (Gibco, USA) in each well and incubated for 5 min. Then appropriate amounts of normal control (NC), TREM2-siRNA1, TREM2-siRNA2 and TREM2-siRNA3 were diluted respectively with 50 µl Opti-MEM for 5 min. The two mixtures were mixed thoroughly for 20 min and added to the the wells containing the basic medium. After Flow cytometry analysis 1) Phagocytic assay of BV2 cells after Fasudil intervention. To determine whether Fasudil promotes the microglial phagocytosis of FMD, BV2 microglia were plated in 24-well plates and groups were set as follows: PBS group (PBS), FMD plus PBS group (Myelin + PBS) and FMD plus Fasudil group (Myelin + Fasudil). Myelin + PBS and Myelin + Fasudil groups were incubated with 5 mg/ml FMD. According to our previous study, Myelin + Fasudil group was intervented with 15 µg/ml Fasudil . FMD + Microglia were analyzed by ow cytometry.
2) Phenotype analysis of BV2 cells. PBS group (PBS), Fasudil group (Fasudil), FMD plus PBS group (Myelin + PBS) and FMD plus Fasudil group (Myelin + Fasudil) were set for the subsequent experiments. After BV2 microglia were cultured in 24-well plates and 10 cm culture dishes, Myelin + PBS and Myelin + Fasudil groups were incubated with FMD. Fasudil and Myelin + Fasudil groups were added with 15 µg/ml Fasudil. BV2 microglia in four groups were obtained after 48 h culture.
Phagocytic assay by uorescence reader and microscopy BV2 microglia were cultured in 24-well plates and intervented with 5 mg/ml FMD for 48 h.
Unphagocytosed FMD was washed away, and level of phagocytosis was analyzed under uorescence microscopy and multifunctional Microporous Plate Reader using uorescence excited light (485 nm).

Immunocytochemistry
After cultured and intervened in 24-well plates with coverslips, BV2 microglia in four groups was xed with 4% PFA and incubated with 1% BSA/PBS. Then these cells were added with the following antibodies: anti-iNOS, anti-Arg-1, anti-TREM2 and anti-DAP12 at 4℃ overnight, followed by secondary antibodies for 1.5 h. Data were analyzed by Image-Pro Plus 6.0 software.

Data analysis
All experiments were replicated three times, and Graphpad Prism 8.0 software was used for statistical analysis. Signi cance of the results was analyzed using one-way analysis of variance (ANOVA) followed by Tukey's multiple comparisons. Data were expressed as the mean ± SD. The statistically signi cant effects are indicated by asterisks ( * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001).

Results
Fasudil ameliorated behavior disorder and protected the myelin sheath in CPZ-induced mice It is well-known that the CPZ-induced demyelinating mice has been widely used to study myelin regeneration in the CNS. The selective loss of OLs can be observed after 2-week CPZ feeding, and obvious demyelination found at six weeks . The design of animal experiment and change of body weight were showed in Fig. 1a and b. The weight of CPZ-diet mice was lower than that of normalfed mice, but there was no difference compared with Fasudil-treated mice.
Some clinical manifestations, such as anxiety-and depression-like behaviors, have hinted the demyelinating lesions in the brain. In this study, EPM and FS tests were performed to analyze these behavior disorders. The results demonstrated that mice fed with CPZ for 6 weeks showed more anxiety and depression than those of normal-fed mice (Fig. 1c, P < 0.0001, respectively). However, those behavioral abnormalities were improved effectively by Fasudil intervention (Fig. 1c, P < 0.0001 and P < 0.001, respectively).
Histopathologically, uorescent myelin staining was used to observe the severity of myelin sheath injury. The results revealed that the loss of myelin sheath in the corpus callosum and striatum of CPZ diet mice was more than those of normal-fed mice (Fig. 1d, P < 0.0001). Fasudil intervention obviously increased the intensity of uorescent myelin staining (Fig. 1d, P < 0.01). These results suggested that CPZ diet mice showed demyelination, which was obviously reversed by Fasudil intervention.
Fasudil enhanced the phagocytosis of myelin debris by microglia Myelin debris which is formed after demyelination prevented OPC recruitment and differentiation from remyelinating OLs (Neumann et al. 2009). Therefore, removing myelin debris is particularly important for myelin regeneration. Firstly, we observed that some MBP staining was co-located with Iba1 + microglia in CPZ-induced mice (Fig. 2a), and Fasudil intervention increased Iba1 + microglia co-staining MBP compared with CPZ diet mice (Fig. 2a, P < 0.05), indicating that microglia can phagocytize myelin debris and Fasudil may promote microglial phagocytosis. Secondly, in the following in vitro cell experiments, our results showed that the uorescence intensity of phagocytic debris in BV2 microglia increased signi cantly by uorescence scanning of microplate reader after Fasudil intervention compared with the PBS treatment group (Fig. 2b, P < 0.01). The co-localization of FMD and BV2 cells under phase contrast of uorescence microscope was shown in enlarged image (Fig. 2b). Finally, the analysis from ow cytometry further showed that the CFSE + cells representing debris phagocytosis were signi cantly increased after Fasudil intervention (Fig. 2c, P < 0.0001). In short, these results indicated that Fasudil intervention signi cantly promoted the clearance of toxic debris by microglia.

The phagocytosis was accompanied by microglial M2 polarization
The enhancement of microglial phagocytosis is related to M2 phenotypic transformation (Li et al. 2019). In order to further con rm whether Fasudil triggers the microglial M1 phenotype toward M2 phenotype under myelin debris phagocytosis, we analyzed the M1 markers (CD16/32, iNOS) and the M2 markers (CD206, Arg-1) in vitro experiments. The results demonstrated that the expression of iNOS mRNA/protein and Arg-1 mRNA was increased after myelin debris stimulation, especially the expression of iNOS mRNA/protein ( Fig. 3a and b, P < 0.01, respectively). Fasudil intervention downregulated the expression of iNOS mRNA (Fig. 3a, P < 0.05), but obviously increased the expression of Arg-1 mRNA (Fig. 3a, P < 0.01). At the same time, myelin debris stimulation increased the expression of iNOS protein (Fig. 3b, P < 0.01), which was effectively inhibited by Fasudil intervention (Fig. 3b, P < 0.01). Conversely, the addition of Fasudil induced the expression of Arg-1 protein (Fig. 3b, P < 0.01).

Fasudil-triggered phagocytosis of myelin debris was dependent on TREM2 signaling pathway
Recent studies has reported that microglial TREM2/DAP12 pathway participates in engulfment of myelin debris and inhibits neuroin ammation of CNS (Konishi et al. 2018;Cignarella et al. 2020). Using immunocytochemistry staining, we observed the expression of TREM2 and DAP12 in vitro experiments. The results revealed that, compared with Myelin + PBS group, the expression of TREM2 and DAP12 was signi cantly upregulated in Myelin + Fasudil group (Fig. 6a and b). To further con rm whether Fasudil-triggered phagocytosis of myelin debris is dependent on TREM2 signaling pathway, we downregulated TREM2 via transduction with the TREM2-siRNA vector in vitro. We screened 3 TREM2-siRNA vectors and used TREM2-siRNA1 to perform the following experiments (Fig. 6c). As shown in Fig. 6d, TREM2 knockdown inhibited the engulfment of FMD induced by Fasudil treatment (P < 0.0001). In brief, these results indicated that microglial clearance of myelin debris activated by Fasudil is mediated through TREM2/DAP12 signaling pathway.
Next, we need to con rm whether Fasudil intervention can induce the expression of TREM2 and DAP12 in CPZ demyelinating model. Consistent with in vitro results, the immuno uorescent staining revealed that TREM2 + Iba1 + and DAP12 + Iba1 + cells in the corpus callosum of CPZ + Fasudil mice was elevated than that in control and CPZ + NS mice ( Fig. 7a and b). Similarly, western blot assay also de ned that Fasudil treatment upregulated the expression of TREM2 and DAP12 compared with that of CPZ diet mice (Fig. 7c, P < 0.0001, respectively).

Fasudil facilitated the generation of OPCs in CPZ-fed mice
The mobilization and maturation of OPCs are two key steps during myelin regeneration. After the phagocytosis of myelin debris by microglia, we wondered how about the OPCs in the demyelinating area? The results of following experiments showed that compared with CPZ + NS mice, Fasudil treatment upregulated the NG2 + OPCs (Fig. 8a), which expressed Ki67 (Fig. 8b), indicating that these OPCs are proliferating. Western blot assay also suggested that Fasudil treatment upregulated the expression of NG2 protein compared with CPZ-fed mice (Fig. 8c, P < 0.01).

Discussion
Myelin debris, a toxic product aggregated in the lesion areas, triggers in ammatory responses in the demyelinating brain of experimental models and MS patients (Clarner et al. 2012). Some studies have demonstrated that myelin debris not only upregulates the expression of in ammatory moleculars (TNF-α, IL-1β, IL-6), but also downregulates the production of anti-in ammatory mediators (IL-4 and TGF-β) (Sun et al. 2010;Yang et al. 2011). In addition, some inhibitory factors (NogoA, oligodendrocyte-myelin glycoprotein and myelin-associated glycoprotein) in myelin debris hinders axonal regeneration and further activates the immune system to cause myelin degeneration (McKerracher et al. 1994;Chen et al. 2000). Abundant studies have shown that the accumulation of myelin debris delay the e ciency of myelin regeneration (Kotter et al. 2006;Lampron et al. 2015), in which myelin debris particularly impedes OPC differentiation, indicating that the phagocytosis of myelin debris is a precondition before myelin regeneration can be started (Shields et al. 1999;Kotter et al. 2006).
The functions of microglia in demyelinating disorders, such as MS and its animal model, is still controversial. Studies have shown that microglia encompass a large amount of harmful factors, such as releasing proteases, in ammatory cytokines and free radicals, as well as promoting T lymphocyte However, previous studies have indicated that microglia have bene cial effect on neurodegenerative diseases, particularly on the recovery of experimental autoimmune encephalomeylitis (EAE) progression (Du et al. 2017;Yan et al. 2019). Numerous reports hold that microglia plays a crucial role in phagocytizing myelin debris. In demyelinating models, the disfunction of microglial phagocytosis slows the elimination of toxic debris and postpones the process of myelin regeneration (Ruckh et al. 2012;Voss et al. 2012;Marteyn et al. 2016). TREM2, a lipid sensor on microglia, binds to myelin debris and promotes microglial phagocytosis (Poliani et al. 2015). All these observations suggest that myelin debris should be inhibitory to myelin regeneration. Therefore, the elimination of myelin debris plays a bene cial role in axonal remyelination.
Fasudil is an intracellular calcium antagonist and the only clinically approved ROCK inhibitor up to now. Our studies have demonstrated that Fasudil possesses multiple functions, including anti-in ammation, immunomodulation, microglial M2 polarization, synaptogenesis and promoting secretion of neurotrophic factors in the CNS (Yan et al. 2019;Ding et al. 2021). Here, our results indicated that Fasudil enhanced microglial engulfment of myelin debris in cell and animal experiments, accompanied by the upregulation of TREM2/DAP12 and polarization of M2 microglia. It is speculated that Fasudil intervention can improve behavioral abnormality and myelin loss in CPZ-induced mice, which may be related to accelerating phagocytosis and clearance of myelin debris.
In this study, Fasudil effectively upregulated the expression of TREM2 and DAP12 on microglia. TREM2, a phagocytic associated receptor, is involved in microglial clearance of apoptotic neurons, toxic debris and β-amyloid (Fu et al. 2014;Cignarella et al. 2020). DAP12, a signaling adapter protein that pairs with TREM2, is essential for mediation of TREM2 signaling (Yao et al. 2019). The blockade of TREM2 during the progression of EAE led to disease deterioration with more in ammatory reactivity and demyelination in the CNS (Piccio et al. 2007). In CPZ-induced demyelinating mice, the aggregation of myelin debris and axonal injury was aggravated in TREM2 knock-out mice (Cantoni et al. 2015;Poliani et al. 2015). Recent study showed that TREM2 activation on microglia promoted the elimination of myelin debris in the CNS of CPZ-induced mice, resulting in the increase of OPCs and the formation of OLs in lesion areas (Cignarella et al. 2020). Additionally, previous studies indicated that TREM2 not only mediated myelin uptake, but also promoted debris degradation through the phagolysosomal pathway (Cantoni et al. 2015). As a TREM2-mediated signal is conducive to regulation of microglia for debris clearance, it is thought to be required for myelin regeneration (Lampron et al. 2015).
Besides, TREM2 plays an critical role in microglial M2 polarization. Knockdown of TREM2 in BV2 cells impeded M2 polarization and resulted in the cascade ampli cation of M1 microglial in ammatory responses; Conversely, overexpression of TREM2 strengthened M2 polarization and alleviated microgliamediated in ammation (Zhang et al. 2018). TREM2 esiRNA decreased the secretion of protective cytokines and the expression of M2 markers, indicating that TREM2 regulated microglial M2 polarization ). In accordance with the above results, Fasudil induced the upregulation of TREM2/DAP12 in microglia, which, on one hand, could regulate the engulfment of myelin debris, and on the other hand, induce the polarization of M2 microglia. Consequently, Fasudil intervention enhanced the proliferation of OPCs and accelerated the differentiation of OLs, which may be related to the activation of TREM2/DAP12 signaling pathway, resulting in the increase of myelin debris clearance by M2-polarized microglia.
Undoubtedly, we need more investigations to further explore the above-mentioned ndings in the following study. Firstly, the biological effects of microglia had not been dynamically observed after the phagocytosis of myelin debris. Secondly, we only found the high expression of TREM2 and DAP12, but did not understand how Fasudil affect this signaling pathway. Thirdly, we did not clarify whether there is an intrinsic relationship between the upregulation of phagocytic receptors and microglia polarization to M2 phenotype.
In conclusion, we demonstrated that Fasudil ameliorated the behavioral performance and myelin loss in CPZ-induced mice, enhanced the phagocytosis of myelin debris and polarization of M2 microglia, which depends on TREM2/DAP12 signaling pathway. As targeting microglial phagocytosis and M2 polarization, TREM2/DAP12 signaling pathway is upregulated by Fasudil, which should be related to the improvement of microenvironment in the brain. Therefore, future studies on the possibility to use Fasudil as a therapeutic for myelin regeneration are warranted.

Declarations
Author contributions ZBD, QXH and LJS designed the study, carried out the tests. BGX and CGM conceived the study, participated in its design and coordination and helped draft the manuscript. QW, GYH and GGC participated in its design and revised the manuscript. CGM revised and nalized the manuscript. YQL, ZC and JZY participated in the statistical analysis. All authors read and approved the nal manuscript. Compliance with ethical standards

Con icts of interest
The authors declared that the research was conducted in the absence of any commercial or nancial relationships that could be construed as a potential con ict of interest.    Fasudil enhanced phagocytosis by polarizing microglia in vitro. BV2 microglia were stimulated with 5 mg/ml FMD in the absence/presence of 15 μg/ml Fasudil for 48 h. Then cells were stained respectively with microglial M1 markers (CD16/32, iNOS and IL-12) and M2 markers (CD206, Arg-1 and IL-10). (a) Non-phagocytic and phagocytic M1 microglia were analyzed using ow cytometry, (b) non-phagocytic and phagocytic M2 microglia were analyzed using ow cytometry. Quantitative results are mean±SD. **P<0.01, ***P<0.001, ****P<0.0001.

Figure 5
Fasudil shifted M1 to M2 phenotype in CPZ-fed mice. (a) Immuno uorescence staining with anti-Iba1 and anti-iNOS in the regions of corpus callosum, (b) immuno uorescence staining with anti-Iba1 and anti-Arg-1 in the regions of corpus callosum, (c) the expression of Arg-1 and iNOS protein in the extract of brains by western blot, (d) ideograph of mouse brain for the observation. Quantitative results are mean±SD. *P<0.05, **P<0.01, ***P<0.001.