Main

Amyotrophic lateral sclerosis (ALS) is an adult onset neurodegenerative disease specific to motor neurons. Both cell-autonomous and non-cell-autonomous mechanisms significantly contribute to the progression and the onset of disease according to the previous studies using ALS model mice.1, 2, 3, 4, 5, 6 Thus, not only motor neurons but also glial cells, including astrocytes and microglia, are involved in the pathogenesis. Although some agents have been proven effective for combating the pathogenesis of ALS in mice, the clinical trials have shown negative results, and no therapeutic agents of ALS have yet been developed.7

Microglia polarization is sometimes categorized into classical (M1) and alternative (M2) activation. Although this categorization might be an oversimplification, microglia can be polarized into an activation state that is intermediate between a neuro-harmful and a protective state. Lipopolysaccharide (LPS) is known as a representative M1 polarization inducer, and M1 microglia express proinflammatory molecules that include tumor necrosis factor-α (TNF-α), interleukin-1β (IL-1β), interferon-γ (IFN-γ), and nitric oxide (NO) as well as cell surface markers, CD86 and CD68. On the other hand, IL-4 induces M2 polarization.8, 9, 10 M2 microglia express different molecules, such as IL-4, arignase1, Ym1, CD206, and IL-10, and show neuroprotective effects.8, 9, 11, 12 It has recently been reported that M2 microglia markers are transiently enhanced at the presymptomatic phase of ALS, whereas M1 microglia markers are gradually increased during the disease progression.13 However, the mechanisms regulating microglial polarity remain elusive.

Minocycline is one of the tetracycline antibiotics and can penetrate into the central nervous system. Presymptomatic administration of minocycline delays disease onset and progression in a mouse model of ALS via the inhibition of inflammation and glial activation.14, 15, 16, 17, 18 But, treatment with minocycline after the disease onset does not prolong the survival of individuals with ALS in either a mouse model or human clinical cases.19, 20 Nevertheless, minocycline is a commonly used, strong inhibitor of microglial activation. Thus, this drug is a useful tool to investigate the mechanisms underlying microglial polarization and the pathogeneses of many diseases accompanied by microglial activation.

Here, we demonstrate the selectivity of minocycline in the inhibition of microglial polarization in an ALS mouse model. Minocycline selectively inhibited the microglial polarization into M1, but not M2, both in vivo and in vitro.

Results

Minocycline administration delays the pathogenesis of ALS in a mouse model

As it was previously reported that minocycline delayed the pathogenesis of ALS in a mouse model,15 we first examined the effect of minocycline in SOD1G93A mice, a mouse ALS model. Minocycline (33 mg/kg) was intraperitoneally administered five times a week from 8 weeks of age to the end stage. Lifespan was significantly prolonged by 26.9 days in mice receiving minocycline (control, 168.2±11.3 days versus minocycline, 195.1±15.5 days; Figure 1a). Disease onset determined by a decrease in rotarod performance at 15r.p.m. was delayed by 20.6 days (control, 140.4±13.1 days versus minocycline, 161.0±18.4 days; Figure 1b), whereas disease duration did not differ between the control and minocycline-treated groups (Figure 1c).

Figure 1
figure 1

Minocycline treatment suppressed the disease in SOD1G93A mice. (a) The lifespan of SOD1G93A mice was 168.2±11.3 days (n=19) and that of minocycline-treated ones was 195.1±15.5 days (n=20; P<0.01). (b) The disease onset of SOD1G93A mice, as determined by a decrease in rotarod performance at 15 r.p.m., was 140.4±13.1 days. The disease onset of minocycline-treated mice was delayed to 161.0±18.4 days (P<0.01). (c) The disease duration of SOD1G93A mice, as determined by 15 r.p.m. rotarod performance, was 27.8±2.9 days. That of minocycline-treated SOD1G93A mice was 33.7±3.9 days. Black line, SOD1G93A mice; gray line, minocycline-treated SOD1G93A mice. Open column, SOD1G93A mice; closed column, minocycline-treated SOD1G93A mice. Error bars, S.E. NS, not significant

Full size image

Minocycline attenuates microglial activation in the spinal cord

Immunohistochemical analysis was performed to examine the morphological changes of microglia at 24 weeks of age. Microglia showed ameboid shapes and increased in number in the spinal cord of SOD1G93A mice (Figure 2a). The size of microglial soma also became larger. Minocycline treatment reduced the number and soma size of microglia (Figure 2a). These data indicate that microglia were activated in SOD1G93A mice, and minocycline suppressed this activation. Microglial activation and deactivation in the spinal cord were confirmed by flow cytometry and western blotting. Thus, flow cytometry showed that the expression level of CD11b in individual microglia was increased in SOD1G93A mice, and was suppressed by minocycline administration (Figures 2b and c). Western blotting analysis demonstrated that the protein expressions of Iba1 and CD68 were increased in SOD1G93A mice, and diminished by minocycline administration (Figure 2d).

Figure 2
figure 2

Minocycline inhibited microglial activation in SOD1G93A mice. (a) The spinal cords of non-Tg, SOD1G93A, and minocycline-treated SOD1G93A mice at 24 weeks (wk) were stained with anti-Iba1 antibody. Bars, 20 μm. (b) A representative profile of CD11b expression at 24 wks. Blue line: non-Tg mice; red line: SOD1G93A mice; orange line: minocycline-treated SOD1G93A mice. (c) Quantitative data on the mean fluorescence intensity (MFI) of CD11b (n=3). Error bars, S.E. **P<0.01, *P<0.05. (d) Lumbar spinal cord lysates from non-Tg, SOD1G93A, and minocycline-treated SOD1G93A mice at 24 wks were subjected to western blotting against Iba1 and CD68 (n=2). β-Actin was used as the internal control

Full size image

Minocycline diminishes the expression of M1, but not M2, markers

The expression of M1 markers was gradually increased as the disease progressed; significant differences compared with the values at 9 weeks of age were observed at 18 weeks to the end stage (24 weeks) in the expressions of CD68, CD86, and IFN-γ, and 15 weeks to the end stage in the expressions of TNF-α and IL-1β in untreated SOD1G93A mice (Figures 3a–e). These increases were significantly attenuated by the treatment with minocycline (Figures 3a–e). Interestingly, the significant differences observed between the control and minocycline groups differed temporally depending on the markers; for example, IL-1β showed a difference from 15 weeks to the end stage, whereas IFN-γ showed a difference at the end stage (Figures 3a–e). In contrast, the expression of M2 markers (CD206, arginase1, IL-4, IL-10, and Ym1) was transiently enhanced at the early phase (12–15 weeks), and this expression was not changed between the control and minocycline-treated groups (Figures 4a–e).

Figure 3
figure 3

Minocycline inhibited the expression of M1 markers in the progressive disease phase. RNA was extracted from the lumbar spinal cords. The temporal mRNA expression profiles of M1 markers ((a) CD68, (b) CD86, (c) TNF-α, (d) IL-1β, and (e) IFN-γ) were examined by quantitative RT-PCR. Open columns, SOD1G93A mice (n=4); closed columns, minocycline-treated SOD1G93A mice (n=3). Error bars, S.E. **P<0.01, *P<0.05. #P<0.05, ##P<0.01 compared with 9 weeks mice

Full size image
Figure 4
figure 4

Minocycline did not affect the transient enhancement of M2 markers in early phase pathogenesis. RNA was extracted from the lumbar spinal cords. The temporal mRNA expression profiles of M2 markers ((a) CD206, (b) arginase1, (c) IL-4, (d) IL-10, and (e) Ym1) were examined by quantitative RT-PCR. Open columns, SOD1G93A mice (n=4); closed columns, minocycline-treated SOD1G93A mice (n=3). Error bars, S.E. All expression profiles were not significant in two-way ANOVA. #P<0.05, ##P<0.01 compared with 9 weeks mice

Full size image

M1 marker CD86 is expressed in Iba1-positive microglia

The selective effect of minocycline on M1 polarization prompted us to examine M1 marker expression in microglia during the pathogenesis of ALS in SOD1G93A mice. CD86 was expressed in a subpopulation of Iba1-positive cells in the lumbar spinal cord of SOD1G93A mice at 24 weeks (Figure 5a). Flow cytometric analysis for CD86 expression demonstrated that the fluorescence intensity was significantly increased in microglia in SOD1G93A mice, and that the augmented fluorescence intensity was decreased after the minocycline administration (Figures 5b and c). Furthermore, CD86 or CD68 expression was detected in a subpopulation of microglia in SOD1G93A mice as early as 12 weeks of age (Figures 5d and e).

Figure 5
figure 5

CD86 and CD68 were expressed in a subpopulation of microglia. (a) The SOD1G93A mouse lumbar spinal cords at 24 weeks were stained with anti-CD86 (M1 marker) and anti-Iba1 (microglia marker). Bars, 50 μm. Arrows indicate the co-localized cells. (b) A representative profile of CD86 expression in CD11b-positive cells at 24 weeks. Blue line: non-Tg mice; red line: SOD1G93A mice; orange line: minocycline-treated SOD1G93A mice. (c) Quantitative data of MFI of CD86 (n=3). Error bars, S.E. **P<0.01, *P<0.05. (d and e) The SOD1G93A mouse lumbar spinal cords at 12 weeks were stained with anti-CD86 or anti-CD68 and anti-Iba1. Bars, 50 μm. Arrows indicate the co-localized cells, and arrow heads indicate the Iba1 single positive cells

Full size image

In vitro expression of M1 markers induced by LPS is selectively inhibited by minocycline

According to the temporal expression profile of M1 and M2 markers in vivo, we speculated that minocycline affected M1 polarization, but not M2 polarization. To address this hypothesis, we examined whether or not the expression of M1 markers induced by LPS was inhibited by the treatment with minocycline using the primary cultured microglia. The expression of M1 markers (TNF-α, IL-1β, IFN-γ, and CD86) was significantly upregulated by LPS, and attenuated by the co-treatment with minocycline (Figures 6a–d, third and fourth columns). The expression of M1 markers was not increased by the treatment with IL-4 (Figures 6a–d, fifth columns). Consistent with these mRNA expressions, the production of NO and the protein expression of the inflammatory cytokines (TNF-α and IL-β) were also upregulated by LPS, and these effects were significantly inhibited by minocycline (Figures 6e–g). In contrast, the expression of M2 markers (IL-4, IL-10, arginase1 and CD206) was significantly increased by IL-4 treatment, but the expression was not attenuated by the co-treatment with minocycline (Figures 6h–k, fifth and sixth columns).

Figure 6
figure 6

Minocycline selectively inhibited the LPS-induced M1 polarization of microglia. Primary cultured microglia were treated with LPS, IL-4, minocycline, or a combination of these agents. The expressions of M1 (TNF-α (a), IL-1β (b), IFN-γ (c), CD86 (d)) were examined by quantitative RT-PCR. (e) The concentration of nitric oxide induced by LPS in the culture medium was measured after the treatment with minocycline. The concentration of inflammatory cytokines (IL-1β (f) and TNF-α (g)) induced by LPS was measured by ELISA and M2 (IL-4 (h), IL-10 (i), arginase1 (j), and CD206 (k)) markers were examined by quantitative RT-PCR. **P<0.01,*P<0.05. NS, not significant (n=5).

Full size image

Minocycline inhibits the upregulation of NF-κB expression in vivo and in vitro

LPS stimulates the NF-κB pathway.21, 22, 23 Activated NF-κB upregulates the expression of inflammatory genes. To investigate the mechanism of minocycline action, we examined the LPS–NF-κB axis. The NF-κB expression was upregulated by LPS and inhibited by the co-treatment of minocycline in vitro (Figure 7a). The mRNA expression of NF-κB in primary cultured microglia was also induced by LPS, and diminished by minocycline (Figure 7b). The NF-κB protein and mRNA expressions were induced in the spinal cord of SOD1G93A mice, and were inhibited by minocycline treatment (Figure 7c and d).

Figure 7
figure 7

The LPS-induced upregulation of NF-κB was inhibited by the treatment with minocycline. (a) A representative image of western blotting. Primary cultured microglia were treated by LPS, IL-4, minocycline, or a combination of these agents. Whole-cell lysates were subjected to western blotting. β-Actin was used as the internal loading control. (b) The mRNA expression of NF-κB was analyzed by quantitative RT-PCR in primary cultured microglia. The protein (c) and mRNA (d) expression of NF-κB in the spinal cords of non-Tg, SOD1G93A, and minocycline-treated SOD1G93A mice at 24 weeks were analyzed by western blotting and quantitative RT-PCR, respectively

Full size image

Discussion

We found that minocycline inhibited the expression of cell surface markers of M1-polarized microglia (CD86 and CD68) as well as the production of inflammatory cytokines (IL-1β, TNF-α, and IFN-γ) in vivo and in vitro (Figures 3 and 6). However, M2 marker expression was not affected. This is the first study to demonstrate the selectivity of the action of minocycline in a subpopulation of microglia.

Minocycline administration after disease onset does not prolong the survival of ALS mice, although presymptomatic treatment with minocycline is effective.19 As it is known that the M2 microglial response starts from 8-week-old of SOD1G93A mice,13 we wanted to see whether the antibiotic has any effects on M2 polarization in SOD1G93A mice. However, we found that minocycline administered after 8 weeks did not affect M2 marker expression (Figure 4), and exerted a potent suppressive effect on M1 marker expression, which began around 15–18 weeks (Figure 3). These data collectively suggest that the dormant M1 polarization may precede the enhanced expression of known M1 markers, which were used in this study, and minocycline may inhibit this polarization. Future analyses of the precise mechanism of minocycline action will reveal and distinguish mechanisms underlying two types of polarization, M1 and M2.

Our data suggest that minocycline may work in the early disease phase by suppressing the microglia activation polarized to M1, leading to the suppression of the pathogenesis of disease. However, minocycline could not completely inhibit the expression of M1 markers in vivo and in vitro (Figures 3 and 6). This insufficient inhibition may have caused the progression of ALS in a mouse model and human patients. On the other hand, we also found that minocycline did not affect the transient enhancement of M2 markers (IL-4, IL-10, arginase1, CD206, and Ym1) in vivo (Figure 4). Therefore, the combination of an M1 suppressor and an agent that could enhance or sustain M2 polarization after the disease onset could have potential for improving the survival of ALS patients.

It has been reported that minocycline partially suppresses the production of inflammatory molecules (IL-6, TNF-α, and IL-1β) induced by LPS in peripheral monocytes through inhibition of the nuclear translocation of NF-κB.24 We found that the protein and mRNA expression of NF-κB were affected by minocycline in microglia (Figure 7). The data were reproducible in LPS-induced microglia activation and in the ALS mouse model, suggesting that minocycline may inhibit not only the nuclear translocation but also transcription of NF-κB. Interestingly, IL-4 treatment did not induce the upregulation of NF-κB expression. These differences might have caused the different expression profiles of M1 and M2 markers by the treatment of minocycline in vivo and in vitro.

The G93A-mutated SOD1 protein activates microglia that mediate toxicity for motor neurons.25, 26 This activation can be blocked by neutralizing antibodies against CD14, TLR2, and TLR4.26 Thus, the G93A-mutated SOD1 protein induces microglial activation for toxicity via CD14 and TLR pathways in ALS pathogenesis. NF-κB is downstream these pathways. Furthermore, the NF-κB pathway is activated in ALS, and NF-κB is a key transcription factor affected in spinal cord microglia in SOD1 mutant mice.27, 28 How minocyline is effective on M1 polarization in SOD1G93A spinal cords could be explained by its inhibitory effects on activation of TLR pathways induced by the mutated SOD1 in microglia.

Materials and Methods

Mice and minocycline administration

The animal experiments described in this article were performed in accordance with protocols approved by the institutional animal committee. All animals were treated and cared for in accordance with the Nagoya University School of Medicine Guidelines pertaining to the treatment of experimental animals. SOD1G93A transgenic mice, which carry the G93A mutant form of the human SOD1 (B6.Cg-Tg [SOD1-G93A] 1Gur/J line), were purchased from the Jackson Laboratory (Bar Harbor, ME, USA). The protocols for the maintenance and genotyping of these mice were described previously.29 **The transgenic mice were randomly divided into minocycline hydrochloride (Nichi-iko Pharmacceutical Co. Ltd., Toyama, Japan)-treated and untreated groups. Minocycline (33 mg/kg) was administered intraperitoneally five times a week from 8 weeks after birth to end stage.

Disease onset and lifespan experiments

Mice were observed daily for survival. An investigator who was blinded to the genotype of mice measured the body weight and motor performance. Both measurements were started at the age of 63 days, and the body weight of mice was measured for every 5 days. Testing of motor function using a Rotarod Treadmill for mice (MK-610A; Muromachi, Tokyo, Japan) was performed once a week. Each weekly session consisted of three trials on the constantly rotating setting at a speed of 15 r.p.m. The time remaining on the rotor during a 3-min period was recorded. Mice were judged to have failed the test when, on average in three separate trials, they fell off the rotarod before 2 min. The time of disease onset was determined by the time when mice began to fail the rotarod test. Lifespan was defined by the duration between birth and the end stage. Disease duration was defined by the duration between the onset and the end stage.

Immunohistochemistry

Mice were deeply anesthetized with diethylether and transcardially perfused with phosphate-buffered saline (PBS) followed by 4% paraformaldehyde in 0.1 M phosphate buffer (Wako,Osaka,Japan). Isolated lumbar spinal cords were fixed with 4% paraformaldehyde overnight, and cryoprotected by 20% sucrose in 0.1 M PBS at 4 °C over the subsequent night. The tissue samples were embedded in Tissue-Tek O.C.T. compound (Sakura Finetek, Torrance, CA, USA) and quickly frozen by liquid nitrogen. Frozen tissues were cut into 20 μm sections on a cryostat (CM1800; Leica Instruments, Heidelberg, Germany) and collected on MAS-coated glass slides (Superfrost; Matsunami Glass, Osaka, Japan). The sections were washed in PBS. After blocking with 3% bovine serum albumin in PBS containing 0.1% Triton X-100 for 30 min at room temperature, the sections were incubated at 4 °C overnight with the following primary antibodies: rabbit anti-Iba1 (1 : 500; WAKO), rat anti-mouse CD86 (1 : 200; BD Biosciences, San Jose, CA, USA), and rat anti-mouse CD68 (1 : 100; AbD Serotec, Oxford, UK). After three rinses with PBS, the sections were incubated for 60 min at room temperature with the following secondary antibodies: Alexa Fluor 594-conjugated donkey anti-rat IgG (Invitrogen, Carlsbad, CA, USA) and Alexa Fluor 647-conjugated donkey anti-rabbit IgG (Invitrogen). Subsequently, the sections were rinsed in PBS, mounted with Fluorsave (Calbiochem, San Diego, CA, USA), and observed under a BZ-9000 fluorescence microscope (Keyence, Osaka, Japan).

Western blotting

Mice were deeply anesthetized with diethylether and perfused transcardially with PBS. Isolated lumbar spinal cords were lysed with lysis buffer (1% Triton X-100 and 1% protease inhibitor cocktail (Nacalai Tesque, Kyoto, Japan) in PBS) by sonication for 30 s at 4 °C. The homogenates were centrifuged at 20 000 × g for 10 min, and the concentration of the soluble proteins was measured by Bradford protein assay using a Protein Quantification Kit (Dojindo, Kumamoto, Japan). The soluble protein solutions were mixed with 4 × sample buffer (0.25 M Tris-HCl, 20% mercaptoethanol, 8% SDS, 20% sucrose, 0.008% bromophenol blue; pH 6.8) and boiled for 5 min. SDS-PAGE was performed in 6, 10, and 15% gels. After transferring the proteins to a polyvinylidene difluoride membrane (Hybond-P; GE Healthcare, Uppsala, Sweden) and blocking it with PBS-T containing 5% skim milk for 1 h at room temperature, the membranes were incubated with rabbit anti-Iba1 (1 : 1000; WAKO), rat anti-mouse CD68 (1 : 500; AbD Serotec), mouse anti-NF-κB (1 : 1000; Cell Signaling Technology, Beverly, MA, USA), and mouse anti-β-actin (1 : 10 000; Sigma, St Louis, MO, USA). After wishing with 5% skim milk in PBS-T, the membranes were incubated with the HRP-conjugated secondary antibodies for 60 min at room temperature. Binding antibodies were visualized by using ECL and ECL Plus kits (GE Healthcare).

Flow cytometry

The spinal cords obtained from non-Tg, SOD1G93A, and minocycline-treated SOD1G93A mice at 24 weeks were dissociated with 5 mg/ml collagenase in HBSS at 37 °C for 45 min, and filtered by 70 μm cell strainers (Becton Dickinson, San Jose, CA, USA). Immune cells were separated by centrifugation using 38% Percoll in PBS at 2000 × g for 20 min. The immune cells were suspended with PBS containing 0.5% bovine serum albumin and 2 mM EDTA and incubated with CD11b affinity beads (Milteny Biotech GmbH, Bergisch Gladbach, Germany) at 4 °C for 15 min. Fcγ receptors on CD11b-enriched cells were blocked by incubation with anti-CD16/CD32 antibody (1 : 400; BD Biosciences) at 4 °C for 15 min. Then, the cells were stained with Alexa 488-conjugated rat anti-mouse CD11b and PE-Cy 7 conjugated rat anti-mouse CD86 (2 μg/sample; BD Biosciences) antibodies at 4 °C for 30 min. Data analysis was performed with FACS Canto II, FACSDiva (Becton Dickinson) and FlowJo (Tree Star, Ashland, OR, USA).

Quantitative RT-PCR

Total RNA was extracted from the lumbar spinal cords using an RNeasy Lipid Tissue kit (Qiagen, Valencia, CA, USA) according to the manufacturer’s recommendations. The cDNA was prepared from 1 μg of total RNA by using a Transcriptor First Strand cDNA Synthesis Kit (Roche Diagnostics, Mannheim, Germany) following the standard protocols. Quantitative PCR was performed on a M × 3000P (Agilent Technologies, Santa Clara, CA, USA) using the synthetic primers and SYBR Green (Agilent Technologies). Samples were subjected to 40 cycles of amplification at 95 °C for 15 s and 60 °C for 1 min, after holding at 50 °C for 2 min and 95 °C for 10 min. Relative expression was calculated using the 2-(Ct experimental sample – Ct internal control sample (GAPDH)) method. The sequences of primers used are listed in Table 1.

Table 1 The primer sequences using quantitative RT-PCR of M1 and M2 markers
Full size table

Primary culture

Primary cultured glial cells were prepared from the cortexes of postnatal day 1 wild-type mice. The brain cortexes were removed aseptically from the skulls and the meninges were excised carefully under a dissecting microscope. The cortexes were fully dissociated by pipetting and the small pieces of tissues were cultured in poly-L-lysine-coated flasks in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum and 1% Zell Shield (Minerva Biolabs, Berlin, Germany), then incubated at 37 °C in a humidified atmosphere containing 5% CO2. The culture media were changed every 7 days. The mixed glia cultured for 2 weeks was shaken at 120 r.p.m. on a gyratory shaker for 15 min at 37 °C. The floating cells were reseeded in the fresh PLL-coated dishes and cultured for the subsequent night. More than 95% of the attached cells were found to be Iba1-positive. The culture media were replaced with serum-free DMEM and stimulated with 100 ng/ml LPS (Sigma), 20 ng/ml IL-4 (Peprotech, Rocky Hill, NJ, USA), and 20 μM minocycline for 24 h.

NO and cytokine assay

The NO assay was performed using the previously reported protocols. Briefly, 100 μl of culture supernatants was mixed with the same volume of Griess assay regent (1% sulfanilamide (Sigma), 0.1% naphthylethylenediamine dihydrochloride, and 2.5% phosphoric acid). Ten minutes later, the 560-nm absorbance was measured by Multiskan JX (Thermo Labsystems, Helsinki, Finland). Sodium nitrate solution was used as the standard. The cytokines were measured using Quantkine ELISA kits (R&D Systems, Minneapolis, MN, USA) according to the manufacturer’s protocols.

Statistical analysis

The Kaplan–Meier method (log-rank test) was used for comparing the lifespan and disease onset of mice of each genotype. Mann–Whitney’s U-test was used for analyzing disease duration. Two-way ANOVA was used to evaluate the difference in the temporal expression profile of M1 and M2 markers between two groups. We subsequently performed an unpaired Student’s two-tailed t-test at each week. In all statistical analyses, a value of P<0.05 was considered to indicate significance. The statistical analysis was performed using SAS 9.3 (SAS Institute Inc., Cary, NS, USA) and SPSS (SPSS Inc., Chicago, IL, USA) software. The investigators performing the statistical analyses were blinded to the group assignments in all procedures.