RESEARCH ARTICLE

Rho Kinase Inhibition Modulates Microglia Activation and Improves Survival in a Model of Amyotrophic Lateral Sclerosis Lars T€ onges,1 Rene G€ unther,1 Martin Suhr,1 Johannes Jansen,1 Alexander Balck,1 Kim-Ann Saal,1 Elisabeth Barski,1 Tobias Nientied,2 Alexander A. G€ otz,3 Jan-Christoph Koch,1 Bernhard K. Mueller,4 Jochen H. Weishaupt,5 Michael W. Sereda,2,6 Uwe-Karsten Hanisch,3 Mathias B€ ahr,1,7 and Paul Lingor1,7 Disease progression in amyotrophic lateral sclerosis (ALS) is characterized by degeneration of motoneurons (MN) and their axons, but is also influenced by neighboring cells such as astrocytes and microglial cells. The role of microglia in ALS is complex as it switches from an anti-inflammatory and neuroprotective phenotype in early disease to a proinflammatory and neurotoxic phenotype in later stages. Our previous studies in models of neurodegeneration identified rho kinase (ROCK) as a target, which can be manipulated to beneficially influence disease progression. Here, we examined the neuroprotective potential of the ROCK inhibitor Fasudil to target the central pathogenic features of ALS. Application of Fasudil to kainic acid-lesioned primary MN in vitro resulted in a strong prosurvival effect. In vivo, SOD1G93A mice benefited from oral treatment with Fasudil showing prolonged survival and improved motor function. These findings were correlated to an improved survival of motor neurons and a pronounced alteration of astroglial and microglial cell infiltration of the spinal cord under Fasudil treatment. Modeling a proinflammatory microglial phenotype by stimulation with LPS in vitro, Fasudil decreased the release of proinflammatory cytokines and chemokines TNFa, Il6, CCL2, CCL3, and CCL5 while CXCL1 release was only transiently suppressed. In sciatic nerve motor axons, neuromuscular junction remodeling processes were increased. In conclusion, we provide preclinical and neurobiological evidence that inhibition of ROCK by the clinically approved small molecule inhibitor Fasudil may be a novel therapeutic approach in ALS combining both neuroprotection and immunomodulation for the cure of this devastating disease. GLIA 2014;62:217–232

Key words: Fasudil, SOD1, transgenic, motoneuron, neuroprotection

Introduction

A

myotrophic lateral sclerosis (ALS) is a devastating disorder of the first and second motoneuron (MN), which results in progressive muscle weakness and usually death from respiratory failure within a few years following diagnosis. Most patients suffer from sporadic ALS, but about 5–10% of all ALS cases can be attributed to familial forms, which are

caused by mutations in the genes encoding for superoxide dismutase 1 (SOD1), Alsin, Fused in sarcoma, TAR DNAbinding protein 43, and many others (Ferraiuolo et al., 2011). Knowledge on familial ALS forms has permitted to generate transgenic (TG) animal models mimicking the disease, one of the examples being the SOD1G93A mouse model, which has been extensively used in preclinical studies on ALS

View this article online at wileyonlinelibrary.com. DOI: 10.1002/glia.22601 Published online December 6, 2013 in Wiley Online Library (wileyonlinelibrary.com). Received Sep 19, 2013, Accepted for publication Oct 29, 2013. Address correspondence to Lars T€ onges MD, Dept. of Neurology, University Medicine G€ ottingen, Robert-Koch-Str. 40, 37075 G€ ottingen, Germany. E-mail: [email protected] (or) Paul Lingor MD, Dept. of Neurology, University Medicine G€ ottingen, Robert-Koch-Str. 40, 37075 G€ ottingen, Germany. E-mail: [email protected] ottingen, Robert-Koch-Str. 40, G€ ottingen, Germany; 2Molecular and Translational Neurology, Max From the 1Department of Neurology, University Medicine G€ ottingen, Robert-KochPlanck Institute for Experimental Medicine, Hermann-Rein-Str. 3, G€ ottingen, Germany; 3Institute of Neuropathology, University Medicine G€ Str. 40, G€ ottingen, Germany; 4AbbVie Deutschland GmbH & Co KG, Global Pharmaceutical Research & Development, Knollstrasse 50, Ludwigshafen, Germany; 5 Department of Neurology, Ulm University, Albert-Einstein-Allee 11, Ulm, Germany; 6Department of Clinical Neurophysiology, University Medicine G€ ottingen, Robertottingen, Germany. Koch-Str. 40, G€ ottingen, Germany; 7Cluster of Excellence Nanoscale Microscopy and Molecular Physiology of the Brain (CNMPB), G€ Additional Supporting Information may be found in the online version of this article.

C 2013 Wiley Periodicals, Inc. 217 V

(Gurney et al., 1994; Turner and Talbot, 2008). Moreover, the analysis of a human phenotype of the SOD1G93A mutation was recently shown to closely resemble the typical phenotype of sporadic ALS, thus implicating comparable disease pathology of SOD1G93A ALS and sporadic ALS (Synofzik et al., 2010). The pathological mechanisms of the disorder likely involves protein aggregation, oxidative stress, excitotoxicity, and mitochondrial dysfunction (Ferraiuolo et al., 2011). These factors eventually lead to loss of neuromuscular junction (NMJ) integrity, retrograde axonal degeneration, and motoneuronal cell death. Therefore, ALS can be considered as a distal axonopathy (Fischer and Glass, 2007). Furthermore, glial cells in the vicinity of MN have been suggested to strongly contribute to the pathophysiology and their role thus has to be equally acknowledged in the development of therapeutic strategies (Philips and Robberecht, 2011). Even though numerous drugs have been tested in mouse models so far, only the glutamate-inhibitor riluzole, has been licensed as a disease-modifying treatment for ALS, but its effect is modest, prolonging the patients’ survival in the mean only by a few months (Stewart et al., 2001). Riluzole has been suggested to act as an anti-excitotoxic drug by inhibiting glutamate toxicity (Doble, 1996). Novel therapeutic approaches for ALS, however, have to target not only the demise of the MN soma, but most importantly, must address axonal degeneration and the role of glial cells in addition. Rho kinases (ROCK) are serine/threonine kinases, which are involved in the modulation of the cytoskeleton through phosphorylation of LIM kinases, myosin light chain Abbreviations

ANOVA BDNF bp CCL2 CCL3 CCL5 CNTF CXCL1 DAPI DIV DMEM Fas30 Fas100 GFAP Iba1 IL6 MTT PBS ROCK SMA SOD1G93A SOD1WT TBS TNFa Veh

218

Analysis of variance Brain derived neurotrophic factor Base pair Monocyte chemoattractant protein, MCP-1 Macrophage inflammatory protein, MIP-1a Regulated upon activation normal T-cell expressed and presumably secreted, RANTES Ciliary neurotrophic factor Chemokine (C-X-C motif) ligand 1 4,6-Diamidino-2-phenylindole Day in vitro Dulbecco’s modified Eagle’s medium Fasudil 30 mg/kg Fasudil 100 mg/kg Glial fibrillary acidic protein Ionized calcium binding adaptor molecule 1 Interleukin 6 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazoliumbromide Phosphate-buffered saline Rho kinase Spinal muscular atrophy Superoxide dismutase 1 G93A transgenic Superoxide dismutase 1 wild type Tris-buffered saline Tumor necrosis factor a Vehicle

or the ezrin/radixin/moesin protein complex (Mueller et al., 2005; Tonges et al., 2011). As our group and others have previously shown, the inhibition of ROCK using pharmacological small molecule inhibitors, such as Fasudil or Y-27632, could not only improve the regenerative response in the lesioned CNS, but also beneficially influence neuronal survival (Bermel et al., 2009; Lingor et al., 2007, 2008; Planchamp et al., 2008; Tonges et al., 2012). This was accompanied by the activation of intracellular cell survival signaling, such as the Akt pathway. In addition, improvement of neurological function or an increase of survival by ROCK inhibition was recently reported in animal models of Huntington’s disease and SMA, respectively (Bowerman et al., 2012; Li et al., 2009). Following up on these results, we have designed a study to evaluate the therapeutic potential of ROCK inhibition in a TG mouse model of ALS [B6SJL-TgN (SOD1-G93A) 1 Gur] through oral application of the small molecule inhibitor Fasudil. We quantified disease progression, motor function, survival and performed electrophysiology, immunohistology, and an analysis of microglia activation patterns. Because Fasudil is already in human use, the beneficial results of Fasudil treatment in this model suggest this ROCK inhibitor as a promising candidate molecule for human studies.

Materials and Methods MN and Astrocyte Cell Culture and Assays MN PREPARATION. Lumbar spinal cord mouse MN of E11.513.5 embryos of wild type (WT) mice (B6/SJL background) were generated using an antibody-based panning technique adapted from Wiese et al. (2010). Cells were collected and seeded on poly-L-ornithine/laminin coated cover slips in MN complete medium containing Neurobasal medium without (2)L-Glutamine (Invitrogen, Darmstadt, Germany), supplemented with 2% horse serum (Linaris, Wertheim, Germany), B27-supplement (Invitrogen, Darmstadt, Germany), BDNF (final concentration 10 ng/mL; Tebu Bio, Offenbach, Germany), and CNTF (final concentration 10 ng/mL; Tebu Bio, Offenbach, Germany) at a density of 25,000 cells/cm2. ASTROCYTE PREPARATION. Astrocytes of newborn mice (P0–P2) were grown in astrocyte medium (DMEM low glucose without L-glutamine; Lonza, Cologne, Germany) containing 20% inactivated fetal calf serum and 1% penicillin/streptomycin (both PAA Laboratories, Coelbe, Germany). Upon confluency cells were split and seeded in a density of 10,000 – 20,000 cells/cm2. MN SURVIVAL ASSAYS. MN were cultured for one day in MN complete medium. For MN monocultures, half of the medium was exchanged with astrocyte conditioned medium on day in vitro (DIV) 1, supplemented with Fasudil (final concentrations: 4 or 20 lM; LC Laboratories, Woborn, MA) or with a corresponding volume of vehicle (control). Additionally, BDNF and CNTF (final concentration 10 ng/mL, both) were supplemented. For MN-astrocyte co-cultures,

Volume 62, No. 2

T€ onges et al.: ROCK Inhibition in ALS

coverslips with MN were placed on DIV1 into wells containing astrocytes. Both, the mono- and the co-cultures were exposed to kainic acid (KA; 125 lM final concentration; Tocris, Bristol, UK) or to PBS on DIV6. Cell viability was assessed after 24 h by counting MTT positive MN in six random microscopic fields within a culture well one hour after addition of MTT (final concentration 500 lg/ mL; Sigma-Aldrich, St. Louis, Mo).

Interleukin 2-forward, CTAGGCCACAGAATTGAAAGATCT, Interleukin 2-reverse, GTAGGTGGAAATTCTAGCATCATCC). SOD1G93A female mice were used in all experiments. We administered Fasudil at an intermediate concentration of 30 mg/kg body weight per day (subsequently termed as Fas30) or at a high concentration of 100 mg/kg body weight per day (Fas100) in drinking water. Control groups received drinking water without supplementation (Veh).

Microglial Cell Culture and Assays

Clinical and Behavioral Animal Testing

MICROGLIAL CELL PREPARATION, CULTURING, AND TREATMENT. Primary microglia cultures were prepared from whole brains of newborn (P0) B6/SJL SOD1G93A and B6/SJL SOD1wt mice and cultured in complete Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum (both Invitrogen/ Gibco, Karlsruhe, Germany), 100 U/mL penicillin and 100 lg/mL streptomycin (both Biochrom, Berlin, Germany; Regen et al., 2011). Cells were plated either in 96-well plates with 100,000 cells/cm2 for ELISA or on PLL-coated coverslips in 24-well plates with 100,000 cells/cm2 for immunocytochemistry. Microglial cultures underwent five different treatments: (a) no treatment (controls); (b) stimulation with lipopolysaccharide (LPS, E.coli R515; Axxora/Appotech TLR ligand set 1, APO.54N-018) for 18 h (final concentration 10 ng/mL); (c) treatment with Fasudil (2, 10, or 50 lM) for 18 h; (d) stimulation with LPS and various concentrations of Fasudil (as above); (e) preincubation with Fasudil (50 lM) for 1, 4, or 12 h prior to LPS stimulation (in a continued presence of Fasudil). All experiments were quadruplicates of two separate culture preparations.

EXPERIMENTAL GROUPS. At day of life 50 (d50) mice were allocated to three treatment groups (Fas30, Fas100 and drinking water as Veh) in an investigator- and animal technician-blinded fashion. Groups were constituted to minimize intergroup variability by matching of animals with respect to body weight, and litter.

CYTOKINE AND CHEMOKINE MEASUREMENTS. Following microglial stimulation, released cytokines and chemokines were quantified in culture supernatants. Levels of IL-6, CCL2 (monocyte chemoattractant protein, MCP-1), CCL3 (macrophage inflammatory protein, MIP-1a), CCL5 (regulated upon activation normal T-cell expressed and presumably secreted, RANTES), and CXCL1 (KC, equivalent of the human GROa) were determined using DuoSet ELISA Development Kits (R&D Systems). TNFa (tumor necrosis factor a) was determined using MAXTM ELISA kit (Biolegend).

Animal housing, breeding, genotyping, and application of Fasudil All animal experiments were carried out according to the regulations of the local animal research council and legislation of the State of Lower Saxony. High-copy number B6/SJL-Tg(SOD1*G93A)1Gur/J (labeled as SOD1G93A in the following text; Gurney et al., 1994) were obtained from Jackson Labs (Stock Number 002726; Bar Harbor). The colony was maintained by crossing B6/SJL males harboring the transgene with WT B6/SJL females. Housing of animals was performed under a 12 h light/12 h dark dark cycle with free access to food and water. For genotyping, tail biopsies of 14-day-old pups (for in vivo studies) or of postnatal day 0 pups (for microglial in vitro studies) were subjected to standardized PCR protocols (Jackson Laboratory). Probe sequences were: hSOD1-forward, CATCAGCCCT AATCCATCTGA; hSOD1-reverse, CGCGACTAACAATCAAAGT GA;

February 2014

MONITORING DISEASE PROGRESSION. Neurological scores and body weight were assessed every three days for each mouse from 50 days of age. The neurological score employed a scale of 0 (worst) to 4 (best) according to Weydt et al. (2003). Mice showing an inability to right themselves 30 s after being placed on a side or having a 25% loss of body weight were scored as “dead,” and were euthanized using carbon dioxide. Disease duration was defined as time from disease onset (neurological score 3) until death (neurological score 0). ROTAROD TEST. After training for three consecutive times of at least 180 s at a constant speed of 15 r.p.m. the time for which an animal could remain on the rotating rod (Ugo-Basile, Comerio, Italy) was measured every three days. The mean of three trials was recorded for each animal at each time point. HANGING WIRE TEST. Animals were placed on a custom-made wire lid (0.8 cm spacing) and cautiously turned upside down, 60 cm above a straw-covered bottom. After training for three consecutive times of at least 180 s the latency to fall was measured every three days. The longest latency in three attempts was recorded at each time point.

Immunohistochemistry of Spinal Cord, Sciatic Nerve, and Gastrocnemius Muscle Sections Spinal cords, sciatic nerves, and gastrocnemius muscles were removed and processed as described before (Gunther et al., 2012). Spinal cord sections were immunolabeled with the primary antibodies (antiChAT, 1:100, Millipore; anti-GFAP, 1:200, DAKO; anti-Iba1, 1:500, WAKO, Osaka, Japan) and subsequently with the respective secondary antibody (Cy3 anti-goat or Cy2/Dylight anti-rabbit; both Dianova). Sciatic nerves were removed separately and sectioned with a microtome (Ultracut; Leica, Wetzlar, Germany) into semi-thin sections (380 nm). Staining was performed with Richardson-Solution [2:1:1 of 1% Azur-2-solution, 2% methyl blue solution, and 1% Borax-solution (Di-Natriumtetraborat-10-hydrat)]. Gastrocnemius muscle sections were labeled with primary antibodies (anti-neurofilament M, 1:1000, Millipore, and anti-VAChT, 1:1000, Sigma, or anti-GAP43, 1:300, abcam) and the respective secondary antibody (Alexa Fluor 488, 1:300, Invitrogen Molecular Probes) together with

219

a-bungarotoxin Alexa Fluor 594 conjugate (1:1000, Invitrogen Molecular Probes).

sented as mean 6 standard error of the mean. Significances are * P < 0.05; ** P < 0.01; *** P