Multiple sclerosis (MS) is a cell-mediated autoimmune disease directed against myelin antigens of the central nervous system (CNS) [1
]. The pathological process of MS includes demyelination, multifocal inflammation, blood–brain barrier disruption, encephalitogenic immune cells, reactive gliosis, oligodendrocyte loss, and axonal degeneration [2
]. Experimental autoimmune encephalomyelitis (EAE) is a widely used animal model of MS [3
], and the progression of EAE occurs in three consecutive steps: The first step is a peripheral immune response (the generation of anti-myelin antibodies and activation of monocytes, T lymphocytes, and B lymphocytes); the second is the disruption of the BBB, followed by the infiltration of encephalitogenic immune cells from the peripheral nervous system (PNS) to the CNS; and the third is the demyelination of nerve fibers [4
]. The infiltrate is mainly composed of T and B lymphocytes and macrophages [1
The molecular mechanisms that underlie these demyelinating diseases still remain unclear. Our lab recently demonstrated that the pathological disruption of zinc homeostasis during EAE is involved in demyelination and disease pathogenesis. Released zinc can induce the degradation of the extracellular matrix and the matrix metallopeptidase-9 (MMP-9)-dependent breakdown of the BBB, resulting in the migration of encephalitogenic immune cells and demyelinating antibodies. It also activates microglia and induces the release of proinflammatory cytokines, which cause damage to the myelin sheath [7
]. Our previous studies also demonstrated that the oral administration of clioquinol (CQ), a well-characterized zinc chelator, or the genetic deletion of zinc transporter 3 (ZnT3
)—which depletes zinc in synaptic vesicles—decreased the symptoms of and pathological changes in EAE [8
AMP-activated protein kinase (AMPK) is a serine/threonine kinase consisting of an α subunit and regulatory β and γ subunits. It serves as an integrator of energy balance and energy-dependent responses at the cell, tissue, and organism levels to facilitate context-specific responses to changes in the metabolic status [10
]. AMPK regulates many aspects of cellular metabolism. It is strongly induced by ATP depletion and other related stimuli to restore the cellular energy balance, but its overactivation is deleterious in pathological conditions, such as stroke [12
] and neurodegenerative diseases [13
]. 1H10 is a novel chemical inhibitor of AMPK that was discovered as a potential agent to protect against stroke-related injury [15
]. It has been reported that its administration protected against middle cerebral artery occlusion (MCAO)—induced brain injury and zinc-induced neurotoxicity. However, its protective effects have never been tested in MS.
-induced experimental autoimmune encephalomyelitis (EAE) as a model for MS [3
], we investigated the therapeutic potential of 1H10 to protect against disease progression and the pathological changes induced by zinc-mediated pathogenic mechanisms during EAE. Here, we found that 1H10 reduced the severity of EAE and attenuated demyelination, microglial activation, BBB disruption, MMP-9 activation, encephalitogenic immune cell infiltration, and the formation of abnormal zinc patches. Our findings highlight that a new zinc-chelating agent, 1H10, could be a promising therapeutic tool for MS treatment.
Using MOG35-55-induced EAE mice and a new zinc chelator, 1H10, we intended to validate the previously reported role of zinc in EAE disease progression. In the present study, we found a reduction in the clinical signs, disease progression, and pathological changes, such as demyelination, microglia/macrophage activation, BBB disruption, MMP-9 activation, encephalitogenic immune cell infiltration, and aberrant zinc patch formation in the spinal cord white matter, in 1H10-treated mice. 1H10 treatment also reduced AMPK phosphorylation in CD8+ T lymphocytes that infiltrated into the spinal cords of EAE mice. Thus, the effect of 1H10 in blocking EAE disease progression occurs mainly through zinc chelation and AMPK inhibition.
It has been increasingly recognized that zinc homeostasis has a major impact on the pathophysiological processes of MS, although the precise mechanism is not known. Zinc is one of the most abundant trace elements, essential for proper CNS function [20
]. However, the zinc synaptically released into the extracellular space can reach toxic levels during pathological conditions, such as seizures [21
], ischemia [22
], traumatic brain injury [23
], hypoglycemia [25
], and multiple sclerosis [7
]. The cytoplasmic influx of synaptically released zinc stimulates the activation of NADPH oxidases, mainly via protein kinase C (PKC) [26
]. Since there are zinc finger structures that are critical for the enzymatic function of PKC, zinc can regulate its activity [27
]. It has been reported that zinc induces the PKC-dependent activation of NADPH oxidase in cortical neurons and astrocytes, by Noh and Koh et al. [28
], and reactive oxygen species (ROS) production from NADPH oxidase leads to DNA damage and poly(ADP-ribose) polymerase-1 (PARP-1) activation in the nucleus, followed by neuronal death [29
]. Our previous studies sought to evaluate whether vesicular zinc is an important player in EAE-induced damage to the spinal cord white matter. We provided evidence that EAE induces vesicular zinc release from the synaptic terminals and increases the formation of aberrant zinc patches in the spinal cord, which is prevented by the administration of a zinc chelator [8
] or NADPH oxidase inhibitor [7
]. In addition, ZnT3
gene deletion, which specifically depletes vesicular zinc in the CNS, also leads to a reduction in the EAE-induced formation of aberrant zinc patches and demyelination in the white matter of the spinal cord [9
]. These results strongly suggest that white matter pathology following EAE is induced by a specific sequence of events, such as the liberation of vesicular zinc, followed by subsequent generation of NADPH oxidase derived-ROS, and the generation of a proinflammatory feedback loop. Here, we also found that EAE-induced aberrant zinc patch formation and demyelination were decreased by 1H10 administration. Our findings suggest that the zinc chelation by 1H10 showed protective effects against EAE-induced myelin sheet degeneration.
BBB permeability is affected by highly specialized complexes, such as neurons, pericytes, astrocytic foot processes, and the extracellular matrix (ECM) [31
]. The breakdown of the BBB is known to occur in a murine EAE model of MS [32
]. Additionally, zinc is a key contributing factor for the activation of MMP-9, a class of zinc-dependent endopeptidases that can degrade the ECM. MMP-9 plays a role in MS, implying the possibility that its activity may regulate the migration of encephalitogenic immune cells through the subendothelial basement membrane. Moreover, MMP-9 can cause demyelination through its proteolytic activity against MBP [33
]. Thus, the zinc released from synaptic terminals or damaged neuronal tissue can activate MMP-9, thereby causing BBB disruption and demyelination in the spinal cord of EAE mice. Persistent BBB permeability contributes to progressive demyelination by allowing infiltrating encephalitogenic immune cells or circulating factors, such as fibrinogen, to cross the BBB and attack antigens present on the myelin of the CNS [34
]. Furthermore, failed remyelination may be a consequence of persistent BBB permeability-mediated astrogliosis—preventing oligodendrocyte precursors (OPCs) from accessing the destroyed myelin—rather than an actual deficiency of OPCs, in late-stage lesions [35
]. The present study found that 1H10 treatment reduced EAE-induced pathological outcomes, such as MMP-9 activation, BBB breakdown, encephalitogenic immune cell infiltration, astrogliosis, and myelin sheath damage.
It has been reported that zinc affects the function of immune cells, such as microglia, macrophages, and T and B lymphocytes [37
]. Our previous studies have demonstrated that zinc is also linked to T cell-mediated autoimmune diseases and that the release of endogenous zinc is an upstream event triggering microglial activation. Zinc chelation by CQ attenuated microglial activation and encephalitogenic immune cell infiltration, suggesting that zinc may also be involved in encephalitogenic immune cell-mediated disease progression [8
]. The present study shows that 1H10 administration also reduced the activation of M1 microglia, mediators of proinflammatory responses, in the spinal cord of EAE mice. Therefore, 1H10 may be able to prevent further inflammatory processes within the spinal cord by silencing microglial cells, reducing the exposure of myelin antigens, infiltration of T lymphocytes, release of proinflammatory cytokines, and recruitment of other encephalitogenic immune cells.
In a murine EAE model of MS, autoreactive T lymphocytes recognize their target autoantigens—such as MBP, MOG, or proteolipid protein—as peptide fragments presented by major histocompatibility complex (MHC) molecules, become activated, and move to the CNS parenchyma, thereby causing myelin sheath destruction. The activated cytotoxic T lymphocytes produce high levels of proinflammatory cytokines, such as tumor necrosis factor-alpha (TNFα) and interferon-gamma (IFNγ). It has been reported that cytotoxic CD8+
T lymphocytes reactive to MBP have been shown to induce EAE [40
T lymphocytes can directly recognize and kill antigen-expressing cell types. Activated CD8+
T lymphocytes can also mediate the killing of target cells by the Fas-Fas ligand (FasL) pathway, or by the release of cytotoxic granules at the effector/target cell junction [43
]. Moreover, AMPK is well known to be a fundamental regulator of T cell metabolism, preserving cellular energy homeostasis [46
]. AMPK is activated in CD8+
T lymphocytes in response to energy stress, such as during infection and inflammation [47
]. AMPK activation is required for the survival of CD8+
T lymphocytes during infection and in the tumor microenvironment [49
]. In light of this evidence, we confirmed the extent of AMPK phosphorylation with vehicle and 1H10 treatment in MOG35-55
-induced EAE mice. The present study found that 1H10 administration diminished the infiltration of CD4+
T and CD20+
B lymphocytes in the spinal cords of the mice. In addition, the CD8+
T lymphocytes that infiltrated into the spinal cord white matter had significantly increased AMPK phosphorylation in the EAE mice, which was prevented by administration of 1H10. The effect of 1H10 could also be due to a reduced activation of the immune cells in the periphery during the EAE induction and it could explain why the EAE severity, the BBB disruption, and the immune cell infiltration are decreased.
Altogether, this study demonstrates that vesicular zinc and AMPK activation is involved in several steps of MS pathogenesis. The attenuation of EAE’s severity by 1H10 suggests that zinc chelation and AMPK inhibition have great therapeutic potential for treating multiple sclerosis.
4. Materials and Methods
C57BL/6 female mice, aged 8 weeks (18–22 g), were purchased from Daehan Biolink (DBL, Chungcheongbuk, Korea). Mice were housed in a temperature- and humidity-controlled environment (22 ± 2 °C and 55% ± 5% relative humidity under a 12 h light/12 h dark cycle) and supplied with the Purina diet (Purina, Gyeonggi, Korea) and water ad libitum. Animal use and relevant experimental procedures were approved by the Institutional Animal Care and Use Committee, Hallym University (Protocol # Hallym 2014-89; Date of approval: February 11, 2015). This study was written up in accordance with the ARRIVE (Animal Research: Reporting In Vivo Experiments) guidelines [50
4.2. EAE Induction and Clinical Evaluation
EAE was induced as previously described [7
]. Briefly, mice were immunized by the subcutaneous injection of 200 μL of a mixture of recombinant myelin oligodendrocyte glycoprotein 35-55 (MOG35-55
, AnaSpec, Fremont, CA, USA) in a mixture of incomplete Freund’s adjuvant (IFA, Sigma-Aldrich, St. Louis, MO, USA) and Mycobacterium tuberculosis
H37RaA (Difco Laboratories, Detroit, MI, USA). Pertussis toxin (PT, List Biological Laboratories, Campbell, CA, USA) was intraperitoneally injected at a dose of 400 ng on post-immunization days 0 and 2. A booster injection of MOG35-55
was given on day 7. The clinical signs of EAE were followed and scored daily on a 0–5 scale, on which 0 = no deficit; 0.5 = partial loss of tail tone or slightly abnormal gait; 1.0 = complete tail paralysis or both a partial loss of tail tone and mild hind limb weakness; 1.5 = complete tail paralysis and mild hind limb weakness; 2.0 = tail paralysis with moderate hind limb weakness (evidenced by frequent foot dropping between the bars of the cage top while walking); 2.5 = no weight-bearing on hind limbs (dragging) but with some leg movement; 3.0 = complete hind limb paralysis with no residual movement; 3.5 = hind limb paralysis with mild weakness in forelimbs; 4.0 = complete quadriplegia but with some movement of head; 4.5 = moribund; and 5.0 = dead.
4.3. 1H10 Administration and Experimental Design
1H10 was dissolved in DMSO and diluted with saline. 1H10 was intraperitoneally injected once per day at a dose of 5 ug/kg for the entire experimental course. Mice were divided into four groups for histological evaluation on day 21 post-immunization: (1) Sham without 1H10 (vehicle only, n = 5), (2) sham with 1H10 (1H10 only, n = 5), (3) EAE without 1H10 (EAE + Vehicle, n = 8), and (4) EAE with 1H10 (EAE + 1H10, n = 7). They were also divided into four groups to evaluate EAE-induced motor deficits on day 45: (1) Sham without 1H10 (vehicle only, n = 5), (2) sham with 1H10 (1H10 only, n = 5), (3) EAE without 1H10 (EAE + Vehicle, n = 8), and (4) EAE with 1H10 (EAE+1H10, n = 6).
4.4. Tissue Preparation and Cresyl Violet Staining
Mice were deeply anesthetized with urethane (1.5 g/kg, intraperitoneal) in sterile 0.9% NaCl at a volume of 0.01 mL/g body weight. Toe pinch was used to evaluate the effectiveness of the anesthesia. The mice were transcardially perfused with 0.9% NaCl and then with 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS). The spinal cords were post-fixed with 4% PFA in PBS for 1 h and then immersed in 30% sucrose for cryo-protection. Thereafter, the spinal cord was frozen and coronally sectioned with a cryostat microtome at a 30-μm thickness. To evaluate the infiltration of mononuclear cells, the thoracic spinal cord sections were stained with cresyl violet. For the quantification of the cresyl violet staining, five sections from each mouse were examined from five mice per group. These sections were coded and given to a blinded experimenter, who manually counted the numbers of infiltrating mononuclear cells in the spinal cords. Data were expressed as the average numbers of cresyl violet-positive cells.
To block endogenous peroxidase activity, sections were immersed in 1.2% hydrogen peroxide for 15 min at room temperature. After washing in PBS, the sections were incubated with primary antibodies in PBS containing 0.3% Triton X-100 at 4 °C overnight as follows: Monoclonal rat anti-CD4 (diluted 1:50, BD Bioscience, San Jose, CA, USA) and CD8 (diluted 1:50, BD Bioscience, San Jose, CA, USA) antibodies or a polyclonal goat anti-CD20 antibody (diluted 1:50, SantaCruz Biotechnology, Dallas, TX, USA). After washing in PBS, the sections were incubated in biotinylated anti-rat IgG (diluted 1:250; Vector, Burlingame, CA, USA) to detect CD4 and CD8 antibody, biotinylated anti-goat IgG (diluted 1:250; Vector, Burlingame, CA, USA) to detect CD20 antibody, or biotinylated anti-mouse IgG to detect endogenous IgG, for 2 h at room temperature. Thereafter, sections were immersed in avidin-biotin-peroxidase complex (Vector, Burlingame, CA, USA) for 2 h at room temperature. Between incubations, the sections were washed with PBS. The immune reaction was visualized with 3,3′-diaminobenzidine (Sigma-Aldrich, St. Louis, MO, USA) in 0.01 M PBS containing 0.015% H2O2, and the sections were mounted on gelatin-coated slides and coverslipped with Canada Balsam (Junsei chemical Co., Ltd., Chuo-ku, Tokyo, Japan). The immunoreactions were observed under an Olympus IX70 inverted microscope. To evaluate nonspecific effects, a few sections in every experiment were incubated in a buffer without any primary antibodies. This procedure always resulted in a complete lack of immunoreactivity. To quantify CD4, CD8, CD20, and IgG immunoreactivity, five coronal sections were analyzed by a blinded experimenter using ImageJ (National Institute of Health, Bethesda, Rockville, MD, USA). The immunofluorescence intensity and area of CD4, CD8, CD20, and IgG were expressed as the mean gray value and % area, respectively.
4.6. Immunofluorescence Analysis
Immunofluorescence labeling was performed as per routine immunostaining protocols, such as those referenced above. The primary antibodies used in this study were as follows: Rabbit anti-myelin basic protein (MBP; diluted 1:500; Invitrogen, Carlsbad, CA, USA); goat anti-Iba-1 (Iba-1; diluted 1:500; Abcam, Cambridge, UK); rat anti-CD68 (diluted 1:100; Bio-Rad Laboratories, Hercules, CA, USA); rat anti-CD8 (diluted 1:50; BD Bioscience San Jose, USA); rat anti-F4/80 (diluted 1:100; eBioscience, San Diego, CA, USA); rabbit anti-phospho-AMPKα 1/2 (diluted 1:200; Abcam); mouse anti-CD31 (diluted 1:200; Millipore, Cambridge, UK); rabbit anti-MMP-9 (diluted 1:200; Abcam, Cambridge, UK); rabbit anti-GFAP (diluted 1:500; Abcam, Cambridge, UK); mouse anti-synaptophysin (diluted 1:200; Cell Signaling Technology, Danvers, MA, USA); and rabbit anti-ZnT3
(diluted 1:200; Synaptic Systems, Göttingen, Germany). For double labeling, primary antibodies were simultaneously incubated and further processed for each antibody. For visualization of the primary antibody binding, fluorescent-conjugated secondary antibodies were applied: Alexa 488 and 594 (diluted 1:250; Invitrogen, Carlsbad, CA, USA). Sections were counterstained with DAPI (4,6-diamidino-2-phenylindole; diluted 1:1000; Invitrogen, Carlsbad, CA, USA). Fluorescence-stained sections were mounted on gelatin-coated slides and coverslipped with dibutylphthalate polystyrene xylene (DPX, Sigma-Aldrich, St. Louis, MO, USA). Fluorescence signals were detected using a Zeiss LSM 710 confocal microscope (Carl Zeiss, Oberkochen, Germany) with a sequential scanning mode for DAPI and Alexa 488 and 594. Stacks of images (1024 × 1024 pixels) from consecutive slices of 0.5–0.8 μm in thickness were obtained by averaging 15 scans per slice and were processed using ZEN 2 (blue edition, Carl Zeiss, Oberkochen, Germany). Images were taken from the thoracic spinal cord section. The quantification of the mean intensity and colocalization experiments was performed using the ZEN 2 software (blue edition, Carl Zeiss, Oberkochen, Germany). The overlap coefficient (Manders’ coefficient) was used as the colocalization coefficient. The area of immunoreactivity was measured using ImageJ (National Institute of Health, Bethesda, Rockville, MD, USA) and expressed as % area. In addition, five sections from each mouse were scored by a blinded experimenter to quantify microglial activation. The criteria for microglial activation were based on the number of F4/80 immunoreactive cells and their morphology [37
4.7. Zinc Staining (TSQ Method)
-(6-methoxy-8-quinolyl)-para-toluenesulfonamide (TSQ) histochemical method was used as previously described [51
]. Briefly, mice were sacrificed on day 21 post-immunization by decapitation under 5% of isoflurane anesthesia, and the brains were removed and frozen in powdered dry ice. The frozen unfixed spinal cords were coronally sectioned at a 20-μm thickness in a −15 °C cryostat and then thawed on gelatin-coated slides and air-dried. The sections were immersed in a solution of 4.5 μm TSQ (Molecular Probes, Eugene, OR, USA) in 140 mM sodium barbital and 140 mM sodium acetate buffer (pH 10.5–11) for 60 s, and then rinsed for 60 s in 0.9% saline. TSQ binding was imaged with a fluorescence microscope (Olympus upright microscope, epi-illuminated with 360 nm UV light) and photographed through a 500-nm long-pass filter using an INFINITY3-1 CCD cooled digital color camera (Lumenera Co., Ottawa, ON, Canada) with the INFINITY Analyze software (the release version 6.0). The intensity of TSQ was measured using ImageJ (NIH, Bethesda, MD, USA) and expressed as mean gray values.
4.8. Cell Culture
Both neuronal and mixed glial neuronal cultures were prepared from embryonic mice at 13–14 days as described previously [53
]. In brief, the growth medium consisted of Dulbecco’s Modified Eagle Medium (DMEM, GibcoBRL, Grand Island, NY, USA) with 2 mM glutamine, 5% fetal bovine serum, and 5% horse serum. Minced cerebral cortices combined with growth medium were seeded onto a poly-D-lysine (Sigma, St. Louis, MO, USA) pre-coated plate at 8–9 hemispheres per 24-well plate. The cultures were incubated at 37 °C in a humidified 5% CO2
atmosphere. All cultures were used at 10–14 days in vitro. These experiments were performed under the guidelines for the care and use of mice in research and under protocols approved by the Animal Care and Use Committee of Sejong University.
4.9. Detection of Zinc-Chelating Capacity in Cell Cultures or in Test Tubes
To detect the amount of intracellular free zinc in mouse neuronal cultures, we pre-loaded 5 μM FluoZin-3 (Kd[Zn2+] = 15 nM, Molecular Probes, Eugene, OR, USA) in Eagle’s Minimal Essential Medium (MEM, GibcoBRL) for 30 min. Then, 300 μM ZnCl2 in Hank’s balanced salt solution (HBSS, Biowest, MO, USA; supplemented with 1.8 mM CaCl2, 1.22 μM MgSO4, 3.15 μM MgCl2, and 1.94 mM glucose) was treated to cortical cultures for 15 min. Subsequently, the cortical cultures were washed with HBSS to eliminate extracellular zinc, and then 1H10 (10, 20, or 40 μM) was used for the post-treatment of the cultures. The relative fluorescence units (RFUs) of FluoZin-3 were measured with a fluorescence microplate reader at Ex/Em = 494/516 nm (Molecular Devices, Sunnyvale, CA, USA) at 10-min intervals for up to 1 h. The capacity for zinc chelation was measured using Newport Green DCF (dipotassium salt, Kd[Zn2+] = 1 μM, Molecular Probes) in test tubes. For this, 0.1 μM Newport Green DCF and 20 μM zinc chloride were mixed—with or without the varied concentrations of 1H10, clioquinol (positive control) or ionomycin (negative control)—in HEPES-buffered saline (135 mM NaCl, 5 mM KCl, 1.5 mM CaCl2, 10 mM HEPES (pH 7.4), 6 mM glucose). The RFUs were measured with a fluorescence microplate reader at Ex/Em = 505/535 nm (Molecular Devices) after 30 min. Log–phase graphs showing the dose-dependent zinc-chelating capacity and half-maximal inhibitory concentration (IC50) values were calculated by using a commercial scientific 2-D graphing and statistics program (Prism 5; GraphPad Software, La Jolla, CA, USA).
4.10. Statistical Analysis
All data were reported as mean ± SEM. Repeated measure ANOVAs were conducted to investigate differences in the clinical scores over time among groups using SPSS ver.21. Other comparisons between vehicle- and 1H10-treated groups were performed with a two-tailed unpaired Student’s t-test. In order to compare the values among 4 groups, the remaining data were analyzed by the Kruskal-Wallis test with post-hoc analysis using Bonferroni correction. p-values less than 0.05 (p < 0.05) were considered to be statistically significant.