Deficiency of microglial Hv1 channel is associated with activation of autophagic pathway and ROS production in LPC-induced demyelination mouse model

All animal procedures were approved by the Institute of Animal Care Committee of Tongji Medical College, Huazhong University of Science and Technology, China. Mice were housed in the specific pathogen free (SPF) animal facility with water and food supplied ad libitum. They were kept in an alternating 12-h periods of light and dark cycle at the standard conditions of 22 °C temperature and relative humidity of 55–60%. Adult C57BL/6 male and female mice (wild type, WT; 20–25 g; 10-12 weeks old) were obtained from Hunan SJA Laboratory Animal Co. Ltd., Hunan, China. The murine strain Hv1^−/− (Hv1 knockout, Hv1 KO, Jackson Laboratory, Bar Harbor, ME, USA) was used wherever mentioned [13].

Based on the research of Qianqian Luo et al. [15] with some modifications, mice were anesthetized with isoflurane (induced at 3%, and maintained at 1.2–1.6%), and positioned in a stereotaxic frame. Corpus callosum demyelination was induced by stereotaxic injection of 2 μL of 1% LPC (Sigma) in 0.9% NaCl solution at the rate of 0.5 μL/min using a 32-gage, two-inch needle attached to a 5 μl Hamilton syringe at two points. The first injection site was 1.0 mm lateral to the bregma, 1.1 mm anterior, and 2.4 mm deep. The second injection site was 1.0 mm lateral to the bregma, 0.6 mm anterior, and 2.1 mm deep. After injection, the needle was kept in each position for an additional 10 min. The day of injection was regarded as day 0. Mice were kept for a period of 5, 10, or 28 days (5 dpi, 10 dpi, and 28 dpi) and subsequently harvested for immunofluorescence staining or Western blot analysis (Fig. 1a).

Morris water maze (MWM) was performed according to the experiment of Charles V Vorhees [16] and was conducted to assess the spatial learning and memory abilities of the animals. The Morris water maze consisted of a white cylindrical tank (120 cm in diameter and 38 cm in height) and a hidden platform 10 cm in diameter, which submerged 1 cm below the surface of the water with a depth of 30 cm and a temperature of 25 °C. The pool was divided into four quadrants, and the platform was placed at the midpoint of a random quadrant. Each quadrant had a geometric pattern. Each mouse received seven consecutive days of training beginning on day 2 or day 20, then followed by 1 day of rest, and the test was conducted on day 10 or day 28. Before the first training, each mouse was placed on a platform for 15 s, and was allowed to swim freely for 30 s. Then it was assisted to rest on the platform for another 15 s. In each experiment, the mice were placed in four quadrants of water. The amount of time each mouse took to locate the hidden platform and the motion trail was recorded by digital camera and ANY-maze software (Stoelting, CO, USA). If it did not find the platform within 90 s, it was placed on the platform for 10 s at the end of the experiment, and the escape delay was recorded as 90 s. The platform was removed on day 10 or 28, and the animals were placed in the relative platform quadrant. The time each mouse spent was recorded in the target quadrant in 90 s. After the test, the animals were dried with a towel and kept warm.

At 5, 10, and 28 dpi, mice were sacrificed under 5% isoflurane. For histologic staining and immunofluorescence, they underwent cardiac perfusion with 30 ml 0.1% phosphate buffer (PBS), followed by 30 ml 4% paraformaldehyde (PFA). PBS and PFA were precooled to 4 °C. After perfusion, the brain was removed, post-fixed in 4% PFA overnight (4 °C), and completely dehydrated in 30% sucrose. Serial 20 mm coronal sections were cut on a constant temperature (−20 °C) frozen slicer (1950 cm, Leica, Germany). For Western blot, the mice were perfused with ice-cold PBS containing 1 U/ml of heparin. The brain was extracted, rapidly frozen in liquid nitrogen-cooled isopentane, and stored at −80 °C for later use.

As previously described [17], histological changes to myelin of the corpus callosum were observed and graded by Luxol fast blue (LFB) staining. In brief, the brain slices were placed in LFB dye (G1030, Servicebio, Ltd., Wuhan, China) and heated at 60 °C for 6-8 h. After rinsing, the brain slices were differentiated alternately in a lithium carbonate solution and 70% ethanol, then dehydrated with 75%, 90%, and 100% ethanol. Finally, the slices were soaked in xylene for 5-10 min and sealed with neutral resin. Images were taken by with an optical microscope (DP 50, Olympus, Japan). White matter lesions were assessed in the medial part of the corpus callosum (CC) region. The severity of white matter lesions was classified as normal (grade 0), disordered arrangement of nerve fibers (grade 1), obvious cavitation formation (grade 2), and disappearance of myelin sheath fibers (grade 3). Data is presented as median ± quartile. Mann-Whitney U test was used for comparison between groups.

Mice were anesthetized, and intracardiac perfusion was performed with 2.5% glutaraldehyde in 4% paraformaldehyde. The brain was removed and 1 mm thick coronal sections were cut. The corpus callosum was cut into 1 mm³ pieces, then post-fixed with 2.5% glutaraldehyde. The samples were observed under routine electron microscopy (Hitachi HT7700) at 200 kV, ×1700 magnification. The ratio of myelinated axon thickness to axon diameter (G-ratio) of 280 fibers (70 axons per mouse, 4 mice per group) was measured by Image J (National Institutes of Health, Bethesda, MD, USA). The G-ratio was directly proportional to axon diameter, which directly reflected the degree of myelin sheath formation around the axon. In addition, microglia under an electron microscope were defined by the following characteristics to determine the existence of autophagosome of microglia. The cytoplasm of microglia is electron-dense and the nucleus is lenticular, which can be distinguished from other types of cells. Microglia shows distinct heterochromatin pattern. There is a deep band of heterochromatin with high electron density near the nuclear envelope and a dense heterochromatin network in the whole nucleu s[18].

For immunofluorescence staining, brain slices were permeabilized by 0.25% Triton-X100 in PBS, then blocked with 10% bovine serum albumin (Sigma–Aldrich) for 1 h at 37 °C. Slides were incubated with primary antibodies overnight at 4 °C and washed three times with PBS for 10 min. Slides were incubated with secondary antibody 1 h in the dark at room temperature. Then the sections were washed three times for 10 min each in PBS before taking pictures. Primary antibodies used for immunofluorescence were ionized calcium-binding adapter molecule1 (Iba-1,1:500,Wako Pure Chemical Industries); Fc RII/III receptor (CD16/32, 1:120, 553142, BD Pharmingen); macrophage mannose receptor (CD206, 1:100, AF2535, R&D system); myelin basic protein (MBP,1:200, 10458-1-AP, Proteintech); adenomatous polyposis coli (APC, 1:200, OP80, Millipore); cluster of differentiation 68 (CD68, 1:500, MCA1957, Bio-Rad); 8-hydroxyguanosine (8-OHG, 1:200, ab48508, Abcam); oligodendrocyte lineage transcription factor (Olig2, 1:200, 13999-1-AP, Proteintech); LC3B (1:200, A7198, ABclonal); P62/ SQSTM1(1:200, P0067, Sigma). Secondary antibodies labeled with FITC, AlexaFluor488, AlexaFluor594, or Cy3 were purchased from Jackson ImmunoResearch Laboratories. 4′,6-Diamidino-2-phenylindole (DAPI, 1 μg/ml, Thermo Fisher Scientific) was used for nuclear staining. Sections were imaged by confocal microscopy (FV1200, Olympus, Japan). Image J was used to analyze the resulting images. For immunoreactivity of 8-OHG, image J was used to analyze the average optical density of 8-OHG positive regions under the same conditions of immunofluorescence staining and the exposure intensity. Imaris was used to reconstruct microglia morphology in 3D [19, 20].

The corpus callosum injury lesion of the mouse brain was extracted for Western blot analysis as previously described [21]. White matter tissue was then dissected with the RIPA buffer (Beyotime, China) containing PMSF and cocktail. Bovine serum albumin (BSA, Beyotime, China) was used as the standard for protein concentration determinations. Protein content in each sample (30 μg) was separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) of 8-12%. After electrophoresis, protein was transferred to NC membrane and incubated overnight with primary antibody at 4 °C. The membrane was then incubated with a secondary antibody for 1 h at room temperature. Images were collected using an Odyssey CLx Imager (LI-COR Biosciences). Finally, the integrated optical density (OD) of the signals was semi-quantified with Image J. β-actin was used as an internal control. Major Western blot antibodies included MBP (1:1000, 10458-1-AP; Proteintech); LC3B (1:1000, A7198, ABclonal); P62/SQSTM1(1:1000, P0067, Sigma); Beclin-1(1:1000, 11306-1-AP, Proteintech); β-actin (1:8000, 66009-1-lg, Proteintech). DelightTM 800 was combined with anti-mouse (H + L) and anti-rabbit (H + L) (1:4000, ABclonal).

All values and error bars in the quantitative figures are expressed as mean ± SD except for LFB staining. SPSS 19.0 was used for statistical analysis. Significance was compared between groups by repeated analysis of two-factor analysis of variance (ANOVA), Mann-Whitney U test, or repeated measures of univariate analysis of variance. P < 0.05 was considered as statistically significant.

The demyelination of brain white matter induced by LPC is a classic model, which is limited to the injection area, accompanied by spontaneous myelination regeneration. It is a simple, stable, and easy to repeat demyelination model. LPC mainly leads to the cleavage of myelin membrane and the integrity of axon. As the immune response of demyelination induced by LPC is mild, most axons are not damaged [15, 22]. In this study, the demyelination lesion, caused by a stereoscopic injection of LPC, was manifested by partial myelin loss within the corpus callosum. LFB staining was used to evaluate the degree of myelin loss in WT and Hv1^−/− mice. In WT mice, obvious demyelination was observed at 5, 10, and 28 days after injection (dpi), indicating stable and persistent demyelination due to the two-point LPC injection. Compared with the WT group, Hv1^−/− mice showed less demyelination (Fig. 1b and c). In order to further assess the loss of myelin, myelin basic protein (MBP) was analyzed at each time point. We evaluated the area of MBP loss in the medial part of the CC. Obvious MBP loss was observed in the WT demyelination group. However, MBP loss in the Hv1^−/− demyelination group was reduced (Fig. 1d and e), in accordance with the above LFB staining. We also used Western blot to analyze MBP expression. In line with the immunofluorescence, Hv1 deficiency resulted in less MBP degradation (Fig. 1f and g).

APC and Olig2 are markers of mature oligodendrocytes and total oligodendrocytes lineage cells, respectively. Differentiation and maturation of oligodendrocytes is a key part of myelin sheath recovery following injury. The ratio of APC⁺/Olig2⁺ as determined by immunofluorescence could reflect the extent of oligodendrocytes maturation. In the WT group, the APC⁺/Olig2⁺ ratio was significantly reduced after LPC treatment. But in the Hv1^−/− group, the ratio was significantly higher (Fig. 1h and i). Therefore, these results suggested that the microglial Hv1 proton channel deficiency reduced LPC-mediated demyelination by promoting myelin repair.

It has been shown that microglia turn toward a classically activated phenotype in MS [23]. Our previous studies have found that this pro-inflammatory microglial polarization exacerbates white matter injury [10]. As such, it was worth considering that the function of microglia and whether Hv1 can influence the activation of microglia in LPC-induced demyelination. Consequently, we next investigated microglial activation and morphological change in our model. Staining for myeloid cell marker ionized calcium-binding adapter molecule 1 (Iba-1) showed that microglia significantly aggregated in and around the CC injection area. However, in Hv1^−/− mice, there was considerably fewer aggregated microglia when compared to WT mice after LPC exposure (Fig. 2a and b). Morphological analysis of microglia in the injection area revealed a more ramified phenotype, with larger soma area in WT mice. Yet at the anaphase of demyelination (28 dpi), the soma of microglia Hv1^−/− mice were smaller than in WT mice (Fig. 2c and d). This suggests that Hv1 knockout inhibits microglial activation in the late stage of the disease.

Hv1 deletion has been shown to increase the number of anti-inflammatory microglia and attenuate brain damage following photothrombotic ischemic stroke [11]. Therefore, we investigated whether Hv1^−/− also has a protective role in demyelination. Fc RII/III receptor (CD16/32) and macrophage mannose receptor (CD206) are ubiquitous markers for classically activated microglia and alternatively activated microglia, respectively [24]. In WT demyelination mice, the percentage of CD16/32 and Iba-1 co-localization significantly increased compared to the WT sham mice. Meanwhile, a significantly smaller percentage was observed in the Hv1^−/− demyelination group (Fig. 2e and f). In parallel, CD206 showed an upward trend in immunofluorescence in Hv1^−/− mice, revealing that microglia transitioned toward a protective phenotype (Fig. 2g and h). Thus, these results indicate that Hv1 might facilitate the classical activation of microglia.

Hv1 is known to be involved in microglial NOX-dependent ROS production [13, 25]. Classically activated microglia promoted myelin breakdown via ROS and other substances that exacerbate MS, which manifests as increased demyelination and nerve injury [26]. Our previous studies have shown that Hv1 knockout attenuated microglia-derived ROS production in cuprizone-induced demyelination and ischemic stroke [11, 27]. We assumed that ROS played a role in microglia phenotypic transformation and autophagy activation. An oxidative stress marker, 8-hydroxyguanosine (8-OHG) was detected in the corpus callosum area. At 5, 10, and 28 dpi, we quantified the immunofluorescence intensity of 8-OHG, and it increased significantly compared with that in the WT sham group. Interestingly, the immunoreactivity of 8-OHG was substantially reduced in Hv1^−/− mice when compared with WT mice after LPC injection (Fig. 3a and b). Thus, these results indicate that microglial Hv1 deficiency suppressed ROS production in the corpus callosum after LPC injection.

Previous studies have shown that autophagy is involved in pro-inflammatory activation of microglia [10]. Additionally, accumulation of ROS was reported to induce autophagy [28, 29]. We hypothesized that Hv1 affects microglial autophagy via ROS inhibition. Microtubule-associated protein 1A/1B light chain 3 (LC3) is a soluble protein, which is widely distributed in mammalian tissues and cultured cells. At the same time, a cytoplasmic form of LC3 (LC3-I) binds to phosphatidylethanolamine to form a LC3-phosphatidylethanolamine conjugate (LC3-II), which is recruited into the autophagosome membrane. The flip of the autophagosome marker LC3-I to LC3-II reflects the increased autophagy activity [30, 31]. Furthermore, poly-ubiquitin-binding protein sequestosome 1 (SQSTM1, P62) tends to accumulate when autophagy is inhibited and to decrease when autophagy is induced. Monitoring LC3 and P62 levels can be used as markers of autophagy [32‐34]. Beclin-1 also plays a key role in autophagy, helping to initiate autophagy [35]. Western blot analysis revealed that LPC treatment increased the expression of LC3 II and beclin-1, and downregulated p62 expression in the CC of WT. Hv1 knockout suppressed LPC-induced conversion of LC3 I to LC3 II, inhibited beclin-1 upregulation and increased p62 protein expression (Fig. 4a-d). Moreover, we also observed that Hv1 deficiency inhibited the upregulation of LC3 and iba1 double-positive microglia after LPC injection and increased P62 puncta (Fig. 4e and f). This indicates that LPC-induced demyelination could activate autophagy in microglia and Hv1 deletion suppressed this process.

Based on the above results, we next explored the ultrastructural changes of the myelin sheath at the peak of inflammation. Consistently, we found that destruction of myelinated structures was detected in WT demyelination mice, presenting as cavitation and delamination, while intact layered myelin sheaths were observed in the WT sham group. We used a quantitative analysis of the G-ratio to confirm the damage of myelin sheath ultrastructure. G-ratio was significantly reduced in Hv1^−/− demyelination mice compared with that of WT mice, suggesting that Hv1 deletion reduced myelin damage (Fig. 4g and h). In addition, electron microscopy showed that the formation of autophagic vesicles in microglia increased after LPC injection and decreased in the Hv1^−/− demyelination group (Fig. 4i). Therefore, we hypothesize that the decrease in autophagy in Hv1^−/− mice microglia might be related to the reduction of ROS.

To analyze the spatial memory ability of LPC injection in WT and Hv1^−/− mice, we monitored Morris water maze performance at 10 and 28 dpi. Consistent with previous studies [15], we found that spatial memory was significantly impaired in demyelination mice, as represented by the percentage of time spent swimming in the target quadrant, which decreased when the platform was absent. However, Hv1^−/− mice displayed remarkable improvement in the memory test (Fig. 5a-d). Overall, these results indicate Hv1 deficiency rescued the spatial memory impairment.