Axonal degeneration is one of the earliest features of Parkinson’s disease pathology, which is followed by neuronal death in the substantia nigra and other parts of the brain. Inhibition of axonal degeneration combined with cellular neuroprotection therefore seem key to targeting an early stage in Parkinson’s disease progression. Based on our previous studies in traumatic and neurodegenerative disease models, we have identified rho kinase as a molecular target that can be manipulated to disinhibit axonal regeneration and improve survival of lesioned central nervous system neurons. In this study, we examined the neuroprotective potential of pharmacological rho kinase inhibition mediated by fasudil in the in vitro 1-methyl-4-phenylpyridinium cell culture model and in the subchronic in vivo 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine mouse model of Parkinson’s disease. Application of fasudil resulted in a significant attenuation of dopaminergic cell loss in both paradigms. Furthermore, dopaminergic terminals were preserved as demonstrated by analysis of neurite network in vitro, striatal fibre density and by neurochemical analysis of the levels of dopamine and its metabolites in the striatum. Behavioural tests demonstrated a clear improvement in motor performance after fasudil treatment. The Akt survival pathway was identified as an important molecular mediator for neuroprotective effects of rho kinase inhibition in our paradigm. We conclude that inhibition of rho kinase using the clinically approved small molecule inhibitor fasudil may be a promising new therapeutic strategy for Parkinson’s disease.
Parkinson’s disease is the most common neurodegenerative movement disorder with a median age of onset of ∼55 years and a prevalence of 3% in people >80 years (Strickland and Bertoni, 2004). The disorder is commonly diagnosed by its motor symptoms, but non-motor symptoms occur even before the onset of motor deficits and represent a major therapeutic problem (Lingor et al., 2011). All therapeutic approaches that are in clinical use today treat the symptoms, and there is an urgent need for the development of therapies that target the pathogenic mechanisms and act in a neuro-restorative manner (reviewed in Lees et al., 2009).
While the preferential loss of dopaminergic neurons in the substantia nigra pars compacta and the presence of Lewy bodies are considered as hallmarks of the disease (Braak et al., 2003), its initial pathology is likely to be characterized by neuritic and axonal degeneration (reviewed in Burke and O'Malley, 2012). Although Bernheimer et al. (1973) described that at the time of clinical diagnosis in human patients, ∼60% of substantia nigra pars compacta neurons have been lost and ∼80% of dopaminergic terminals are dysfunctional (Bernheimer et al., 1973), a recent comprehensive analysis of the literature suggests that the initial degeneration of the striatal innervation seems to be much more prominent. Burke and O’Malley (2012) conclude that at the time when first motor signs appear, there is only ∼30% loss of total substantia nigra pars compacta neurons, which is by far exceeded by the loss of striatal or putamenal dopaminergic markers that is estimated to be ∼70–80%. Thus, the extent of terminal loss in the striatum seems to be much more extensive than the amount of loss of substantia nigra pars compacta dopaminergic neurons, suggesting that the striatal dopaminergic nerve terminals are the primary target of the degenerative process and that neuronal death in Parkinson’s disease represents a ‘dying back’ pathology (Dauer and Przedborski, 2003). This hypothesis is also supported by neuropathological findings demonstrating that Lewy neurites outnumber Lewy bodies in early-stage Parkinson’s disease brains (Braak et al., 2003). Additionally, there is a vast body of evidence from genetic studies of Parkinson’s disease pointing to severe axonal pathology in this disease. For example, the greatest abundance of α-synuclein aggregates has been detected not in the cell bodies, but in the neuropil of human Lewy body disease brains, where it is entrapped within presynaptic terminals (Kramer and Schulz-Schaeffer, 2007). As demonstrated in a primary neuronal culture, exogenous recombinant α-synuclein fibrils can be taken up and form aggregates in the presynaptic terminal with subsequent propagation to the neuronal soma (Volpicelli-Daley et al., 2011). This axon-accentuated pathology is also seen in mouse models harbouring mutations of leucine-rich repeat kinase 2 (LRRK2), a protein that is frequently mutated in inherited forms of Parkinson’s disease (Greggio and Cookson, 2009).
To recapitulate Parkinson’s disease in animal models, several genetic and toxin-based disease paradigms have been developed. The 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-based mouse model of Parkinson’s disease has been studied for >20 years and is therefore well characterized in mice and monkeys (Langston et al., 1983; Cannon and Greenamyre, 2010). Using this model, several studies have clearly demonstrated cellular neuroprotective effects on dopaminergic cells with compounds such as monoamine oxidase B inhibitors, vitamin E, the mixed lineage kinase inhibitor CEP-1347, the anti-excitotoxic agent riluzole or coenzyme Q10 (Perry et al., 1985; Benazzouz et al., 1995; Beal et al., 1998; Saporito et al., 1999; Andringa et al., 2003). However, these findings were never successfully translated into the clinic (Lang, 2006; Olanow et al., 2006; Storch et al., 2007). This may be due to the fact that these experimental approaches have not adequately focused on prevention or restoration of axonal pathology in Parkinson’s disease, which is at least as important for the development of clinical disease as is the cellular damage of substantia nigra pars compacta neurons (Burke, 2010; Cheng et al., 2010). Interestingly, the MPTP model is an excellent tool for the study of axonal degeneration. Already within the first days after MPTP application, the number of tyrosine hydroxylase-immunoreactive fibres in the median forebrain bundle significantly decreases, whereas the dopaminergic neuron demise follows later (Li et al., 2009). In an elaborate study analysing synaptosomal preparations of MPTP-treated mice, chronically administered MPTP readily decreased the expression of the dopamine transporter at a time when cell bodies still remained intact (Muroyama et al., 2011). Remarkably, MPTP-lesioned monkeys exhibit a destruction of striatal terminals preceding the loss of substantia nigra pars compacta neurons (Herkenham et al., 1991), and the protection of striatal terminals alleviates the loss of substantia nigra pars compacta dopaminergic cells (Wu et al., 2003). The functionality of the nigrostriatal pathway depends both on cellular and axonal integrity, but these two compartments may suffer from mechanistically different types of degeneration as it is also the case in models of Wallerian degeneration (Coleman and Perry, 2002; Raff et al., 2002). Even more importantly, axonal degeneration may even precede and sometimes cause neuronal cell death (Coleman, 2005). In this study, we have examined a novel molecular target in Parkinson’s disease pathology that has been involved in the regulation of both cellular and axonal demise in neurodegenerative disease models.
The enzyme rho kinase has shown to be widely distributed in the mammalian CNS, with increasing expression with age (Hashimoto et al., 1999; Komagome et al., 2000). It is crucially responsible for the lack of axonal regeneration in the CNS (Mueller et al., 2005; Tonges et al., 2011b). Evaluating the role of rho kinase signalling in optic nerve lesion models, we have found that pharmacological inhibition of rho kinase is able to significantly enhance regeneration of optic nerve axons after lesion. Interestingly, rho kinase inhibition has also yielded cellular neuroprotective effects and has prevented apoptosis of retinal ganglion cells after optic nerve axotomy (Lingor et al., 2007, 2008). Because of these multifaceted axonal and cellular protective actions mediated by rho kinase inhibition, we have now examined the effects of the small molecule rho kinase inhibitor fasudil in the MPTP model of Parkinson’s disease. Fasudil has been extensively studied in animal models of cardiovascular disease due to its additional vasodilatator and anti-arteriosclerotic effects and is being evaluated currently for its clinical applicability in this context (Loirand and Pacaud, 2010).
Having detected a protective effect on perikaryal survival and neurite stability after fasudil treatment of primary midbrain neurons subjected to the MPTP metabolite 1-methyl-4-phenylpyridinium (MPP+) in vitro, we transferred these findings to the subchronic MPTP mouse model in vivo. Here, we were able to reproduce the neuroprotective response of pharmacological rho kinase inhibition with orally applied fasudil. The examination of neuroprotective pathways revealed an increased activation of Akt signalling after rho kinase inhibition, thus attributing rho kinase an important role in the regulation of both perikaryal survival and axonal integrity in the MPTP model of Parkinson’s disease.
Materials and methods
In vitro assays of dopaminergic cell survival and of neurite regeneration
Primary midbrain dopaminergic neurons were prepared under serum-free conditions as described previously (Lingor et al., 1999). Briefly, embryonic rat midbrain neurons from embryonic Day 14 Wistar rats were dissected and seeded on poly-l-ornithine/laminin-coated cover slips. On in vitro Days 1 and 3, cells received a treatment with phosphate buffered saline (PBS) that served as control or with different final concentrations of fasudil (4, 20 or 100 µM). Fasudil was purchased from LC Laboratories as a monohydrochloride salt [C14H17N3O2S.HCl; IUPAC name: 5-(1,4-diazepane-1-sulphonyl)isoquinoline], which is soluble in water. It has an half maximal inhibitory concentration (IC50) of 1.9 μM for the isoform 2 of rho kinase, which is predominantly present in the CNS (Tonges et al., 2011b).
To evaluate cell survival, MPP+ (Sigma) was added to the culture on in vitro Day 3 at a final concentration of 20 µM for 48 h. Survival of dopaminergic neurons was assessed after cell fixation and immunocytochemical processing by counting of tyrosine hydroxylase-positive immunostained cells in four randomly distributed visual fields of each cell culture. Additionally, the length of all tyrosine hydroxylase-immunopositive neurites of at least two randomly distributed visual fields of each cell culture was measured with ImageJ software to obtain the cumulative neurite length per visual field. This value was then divided by the number of tyrosine hydroxylase-positive cells per visual field to calculate neurite length per cell.
Neurite regeneration was evaluated with the cell scratch paradigm as described previously (Tonges et al., 2011a). On in vitro Days 1 and 3, cells were treated in the same manner as for cell survival studies with PBS control or with fasudil in different final concentrations (4, 20 or 100 µM). Neurite processes were mechanically transected with a thin silicone scraper on in vitro Day 3 and directly thereafter were treated with either PBS or MPP+ (final concentration 20 µM). After 48 h, cells were fixed, and the culture was immunolabelled for tyrosine hydroxylase. The length of the 10 longest tyrosine hydroxylase-immunopositive neurites from the scratch border was measured with ImageJ software in at least three randomly distributed visual fields of each cell culture.
For visualization of dopaminergic neurons, cells were fixed in 4% paraformaldehyde for 10 min at room temperature (22°C), permeabilized with 100% ice-cold acetone (AppliChem) for 10 min at −20°C, washed twice with PBS and blocked with 10% normal goat serum for 10 min at room temperature. Probes were incubated with the primary antibodies [rabbit anti-tyrosine hydroxylase 1:250 (Zytomed)] overnight at 4°C. After two washes with PBS, the appropriate Cy3-labelled secondary antibody (1:250, Dianova) was applied for 45 min at 37°C. Cell nuclei were then counter-stained with 4,6-diamidino-2-phenylindole (DAPI; Sigma) and mounted in Mowiol® (Hoechst). Fluorescence was observed and recorded on a Zeiss Axioplan-2 fluorescence microscope equipped with a CCD camera and AxioVision software (Carl Zeiss MicroImaging). For evaluation of dopaminergic survival and measurement of the neurites within the culture, pictures of two to four random visual fields per culture well were taken with a ×20 objective. For evaluation of dopaminergic regeneration at the scratch site, pictures of three random visual fields at the scratch site were taken with a ×20 objective.
To study the activation of cell survival pathways in midbrain dopaminergic neurons, cells were treated as described before for the in vitro assay of cell survival. 48 hours after MPP+ application, cells were lysed with a cell lysis buffer consisting of 10 mM HEPES (pH 7.2), 142 mM KCl, 5 mM MgCl2, 1 mM EGTA and 1% IGEPAL® plus protease inhibitors (‘Complete tablets’, Roche). The protein content of all cell samples was determined using the Bradford assay (Biorad), and equal amounts of protein (50 µg) were separated on a sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). Proteins were then electrotransferred onto a nitrocellulose membrane and blocked with 5% bovine serum albumin in Tris-buffered saline/Tween 20 (TBS-T) for 1 h. Membranes were then incubated with primary antibodies (anti-phosphorylated STAT3, anti-STAT3, anti-phosphorylated Akt, anti-Akt, anti-phosphorylated MAPK, anti-MAPK, anti-Bcl2, all 1:1000; all Cell Signaling Technologies; anti-growth associated protein-43 (GAP43), 1:100, Abcam) for 24 h at 4°C in TBS-T and 2% bovine serum albumin or with anti-GAPDH (1:5000, Biotrend) for 1 h at room temperature in TBS-T and 5% bovine serum albumin. This was followed by incubation with corresponding horseradish peroxidase-coupled secondary antibodies (1:1000 for 1 h at room temperature; Dianova). ECL Plus™ reagent (Amersham) was applied on the membrane, and the chemiluminescence was visualized on X-ray films (Kodak).
Subchronic MPTP mouse model of Parkinson’s disease and treatment with fasudil
For experiments with MPTP, 8- to 10-week-old male mice (C57BL/rehydration and 6JOla Hsd) were purchased from Harlan Winkelmann. Animals (n = 5–8 per experimental group) were housed in groups of five in individually ventilated cages (IVC, Tecniplast). They were kept under a 12-h light/12-h dark cycle with free access to food and water. For MPTP intoxication, mice received a total of five intraperitoneal injections of 30 mg/kg body weight MPTP (free base) dissolved in 0.9% saline on five consecutive days. The control group was injected with saline only. Appropriate guidelines were followed in handling MPTP and MPTP-injected mice (Przedborski et al., 2001). Fasudil was given at 100 mg/kg body weight as per oral gavage by feeding tubes for 2 weeks in the short treatment regimen. In the 6-week treatment period, it was applied at 30 mg or 100 mg/kg body weight in drinking water. The drinking behaviour was closely monitored to ensure an adequate dosage of fasudil. Treatment was started one day before the first MPTP application in the 2-week experiment, and it was started on the day of the first MPTP application in the 6-week experiment. Treatment was continued until animals were sacrificed 2 or 6 weeks after the first treatment application. All animal experiments were carried out according to the regulations of the local animal research council and legislation of the State of Lower Saxony.
Having completed the respective treatment periods, mice were sedated with chloral hydrate and perfused transcardially with saline, followed by 4% paraformaldehyde in PBS (pH 7.4). For the fasudil-treated animals treated for 2 weeks and the controls, the brains were quickly dissected, post-fixed in paraformaldehyde overnight at 4°C, cryoprotected in 30% sucrose in PBS for 24 h at 4°C, snap frozen and stored at −80°C until sectioning. For each group, frontal cryosections (25 µm) through the entire striatum and the substantia nigra were collected on object slides. After rehydration and several rinses in PBS, sections were blocked in 10% normal goat serum for 30 min at room temperature. This was followed by incubation with a polyclonal anti-tyrosine hydroxylase antibody (1:1000; RH-TH-15; Zytomed) or anti-NeuN antibody (1:1000; MAB377; Millipore) diluted in 0.1% Triton® X-100 and 1% normal goat serum in PBS for 24 h at 4°C. All brains were processed with the same antibody lot. After three rinses in PBS, sections were incubated with Cy3- or Cy2-conjugated affinity-purified IgG (goat-anti-rabbit; 1:200; Dianova) for 1 h at room temperature. Finally, sections were DAPI stained and mounted.
For fasudil-treated animals treated for 6 weeks and controls, brains were cryosectioned (30 µm) as described above and were then kept in 0.1% sodium azide. Using a free-floating method, sections were rinsed in 0.1 M Tris-buffered saline (pH 7.4) followed by a treatment with 10% methanol and 3% H2O2 in Tris-buffered saline for 5 min, then rinsed again three times in Tris-buffered saline and blocked with 5% normal goat serum in Tris-buffered saline for 60 min. Sections were incubated with a polyclonal anti-tyrosine hydroxylase antibody (1:1000; Chemicon) in 1% normal goat serum for 48 h and 4°C. After three rinses in Tris-buffered saline, the sections were treated with a secondary biotinylated antibody (1:200, Dianova) in Tris-buffered saline at room temperature for 60 min followed by rinsing in Tris-buffered saline again and incubation with VECTASTAIN® ABC Peroxidase (Standard Kit, PK-4000, Biozol) for 60 min at room temperature. After rinsing in Tris-buffered saline, the sections were treated with 3,3′-diaminobenzidine (DAB peroxidase substrate Kit, SK-4100, Vector Laboratories) for 4 min, rinsed again and collected on object slides. Having dried for ∼4 days, sections were dehydrated, and a Nissl staining procedure was performed by rinsing sections in water for 5 min, followed by an incubation in 1% thionine acetate (Sigma) for 7 min. Sections were rinsed with water again for 5 min, followed by a dehydration row with 70, 90 and 95% ethanol each for 5 min, followed by 5 min in 100% isopropanol and subsequent incubation at least three times in Xylol for 5 min. Finally, sections were mounted with DPX (Fluka Chemica).
For staining of striatal tyrosine hydroxylase-positive fibres, slices were incubated in chloroform–ethanol followed by dehydration, use of anti-tyrosine hydroxylase antibody (Zytomed) and DAB staining (Elite ABC Kit, Vector Laboratories), nickel intensification, formalin fixation and dehydration. Entellan® was used to mount the sections. Details are given in Schulz et al. (1995).
Stereological quantification of substantia nigra neurons and evaluation of striatal density
The number of tyrosine hydroxylase-, NeuN- or Nissl-positive cells in the substantia nigra was assessed using stereological methodology. Every fourth section through the substantia nigra was analysed using Stereo Investigator software (Stereo Investigator 9.0, MicroBrightField Inc.) and a Zeiss Axioplan microscope (Dehmer et al., 2004). A point grid was overlaid onto each section including the substantia nigra pars compacta only. Immunostained cells were counted by the optical fractionator method (×40 objective, counting frame 50 × 50 µm). The evaluation of NeuN and Nissl cells was performed to confirm loss of neurons rather than merely down-regulation of the tyrosine hydroxylase enzyme. Stereological counts were performed by a blinded investigator. Values represent one substantia nigra per animal.
Tyrosine hydroxylase-positive fibre density in the striatum was assessed in sections between Bregma +0.62 and −0.10 mm slightly modified from a previously published protocol (Szego et al., 2012). Every fourth striatal section was immunolabelled for tyrosine hydroxylase resulting in a total of six sections that were analysed per animal. In every section, four to six images were taken using a ×100 objective (Zeiss Axioplan) followed by image processing with background subtraction and binarization to optimally visualize tyrosine hydroxylase-immunopositive fibre areas. These were quantified as percentage of total area at 10 randomly distributed sites from the same image (ImageJ 1.43 u).
Neurochemical analysis of dopamine and metabolites
Twelve days after the first MPTP injection, animals were sacrificed and striata (n = 8 per experimental group) were immediately dissected on ice. Tissue samples were homogenized in a bead mill homogenizer (Precellys 24®, Peqlab) with 50 μl of 0.1 M perchloric acid per mg of striatal tissue. After centrifugation (5000g for 1 min followed by 10 000 g for 30 min, 4°C), 50 μl of supernatant was injected onto a C18 reverse-phase HR-80 catecholamine column (ESA). Dopamine, 3,4-dihydroxyphenylacetic acid and homovanillic acid were quantified by high-performance liquid chromatography (HPLC) with electrochemical detection. The mobile phase (pH = 2.9) consisted of 90% 75 mM sodium phosphate, 275 mg/l octane sulphonic acid solution and 10% methanol. Flow rate was 0.8 ml/min. Peaks were detected by an ESA Coulochem II with a model 5010 detector (E1 = 20 mV, E2 = 320 mV). Data were collected and processed using the Chromeleon computer system (Dionex).
MPP+ levels in the brain were determined by HPLC modified from the method described by Przedborski et al. (1996). Four mice of each experimental group were orally fed with either 100 mg/kg fasudil or the same volume of buffer (see above). Six hours later the mice were intraperitoneally injected with 30 mg/kg MPTP (free base). Mice were sacrificed 90 min later. Striata were quickly dissected on ice and frozen at −80°C until further processing. Tissue samples were homogenized in 50 µl of 0.1 mol/l perchloric acid per milligram tissue, and the debris removed by high-speed centrifugation. Supernatant (10 µl) was injected onto a reverse-phase column (Nucleosil 100-5 C18; Macherey-Nagel) and detected by fluorescence at an excitation wavelength of 295 nm and an emission wavelength of 375 nm (Fluorescence HPLC Monitor; Shimadzu). The mobile phase consisted of 650/1000 ml acetonitrile in phosphate buffer (pH 2.5). The flow rate was 0.4 ml/min.
The Cylinder test (Schallert et al., 2000) was adapted for use in mice to assess forelimb use during normal exploratory activity and was performed 2 or 6 weeks after the initial MPTP lesion in enlarged cohorts of animals (n = 14–15 per experimental group). Each mouse was placed in a transparent cylinder of 11.5 cm diameter and 25 cm in height for 5 min and videotaped during the test. A mirror was arranged behind the cylinder at an angle to enable the rater to record forelimb movements when the mouse turned away from the camera. Forelimb use of the first contact against the wall after rearing and during lateral exploration was recorded by the following criteria: (i) the first forelimb to contact the wall during a full rear was recorded as an independent wall placement for that limb; (ii) simultaneous use of both the left and right forelimb by contacting the wall of the cylinder during a full rear and for lateral movements along the wall was recorded as movement with ‘both’ forelimbs; (iii) full rears of the entire body without touching the wall with neither left nor right forelimb were recorded as ‘free’ rears; and (iv) when the mouse explored the wall laterally, alternating both forelimbs, it was recorded as movement with ‘both’ forelimbs. All movements were recorded during the 5-min test. The statistical analysis detailed the percentage of ‘both’, ‘right’, left’ or ‘free’ movements of all movements observed in the entire observation time.
The rotarod test was performed in the same cohorts subjected to the cylinder test 2 or 6 weeks after the initial MPTP lesion. To measure motor balance and coordination, mice were placed on a horizontal rotating rod with 3 cm diameter (rotarod for mice 47600, Ugo Basile). Five mice were tested simultaneously, separated by large disks. After the mice were placed on the rod at constant rotation speed of 5 rpm, the trial was started and rotation speed was automatically increased from 5 to 40 rpm within 5 min. The trial stopped when the mouse fell down, activating a switch that automatically stopped a timer, or when 5 min were completed. All mice were pre-trained on the rotarod on three consecutive days before MPTP treatment in order to reach a stable performance. The final test was performed in three sessions with an intertrial interval of 30 min to reduce stress and fatigue. Rotarod performance in all three runs was recorded, and the average time on the rotarod was compared between groups.
Statistical comparisons of values between two groups were carried out by unpaired Student’s t-test, and multiple group comparisons for in vitro experiments were carried out by two-way ANOVA followed by Tukey’s post hoc test. Statistical comparisons of values between two animal groups were carried out by unpaired Student’s t-test, and multiple group comparisons in animal experiments were conducted by one-way ANOVA with Tukey’s post hoc test. Striatal fibre density measurements were performed in a hierarchically nested design, the dependent variable being the fibre density, and the independent variables being vehicle or MPTP treatment, respectively. For statistical analyses, Kyplot software version 2.0 (KyensLab Incorporated), Statistica version 9.1 (StatSoft Inc.) and R software version 2.8.0 (R Development Core Team) were used. Data are represented as mean ± SEM or standard deviation (SD) as indicated. Significances are indicated with *P < 0.05; **P < 0.01; ***P < 0.001, unless stated otherwise.
Pharmacological rho kinase inhibition with fasudil protects midbrain dopaminergic neurons from MPP+ toxicity in vitro
We first studied the cell survival-promoting effects of the rho kinase inhibitor fasudil in the MPP+ neurotoxicity model in midbrain dopaminergic neuron cultures. Cell cultures that had only received a control treatment showed a significantly decreased number of surviving tyrosine hydroxylase-positive cells after MPP+ application (64.7 ± 4.9%) compared with PBS treatment, which was set to 100% (100 ± 2.6%). However, cell cultures that had been treated with an optimal concentration of fasudil (20 µM) and were subjected to MPP+ resulted in higher numbers of tyrosine hydroxylase-positive cells (85.9 ± 2.8%). A low or a high concentration of fasudil (4 and 100 µM, respectively) did not confer any cellular protection (72.3 ± 2.7% and 59.3 ± 3.0%). It is known that midbrain dopaminergic neurons that are not MPP+ lesioned and cultured under standard serum-free conditions still suffer from relevant cell loss due to their maintenance in a dissociated state. Interestingly, under these control conditions, rho kinase inhibition with fasudil also conferred a neuroprotective effect and maintained higher numbers of tyrosine hydroxylase-positive neurons with the optimal concentration of 20 µM (111.7 ± 1.9%). High concentrations of fasudil (100 µM), however, appeared to be toxic and resulted in reduced cell numbers (72.1 ± 3.4%). The two-way ANOVA revealed no statistical interaction between fasudil and MPP+, which indicates that the neuroprotective effects of fasudil are independent of MPP+ treatment (Fig. 1).
In the same cultures, which had been plated in a growth permissive environment on laminin, we evaluated the integrity of the tyrosine hydroxylase-immunopositive neuritic network with and without MPP+ stress. For this purpose, we measured tyrosine hydroxylase-positive neurites and quantified their length per cell in order to remain independent of cell survival effects. Compared with PBS control treatment, which was set to 100% (100 ± 5.5%), application of the optimal dose of 20 µM fasudil resulted in a significantly increased neurite length per cell (142.6 ± 7.8%). Interestingly, midbrain dopaminergic neurons that had been subjected to MPP+ only showed a trend of reduced neurite length (79.7 ± 9.5%) whereas 20 µM fasudil still maintained a larger neurite length with MPP+ (110.8 ± 8.2%) compared with the MPP+ control (Fig. 2A and C). If cumulative neurite length in culture was measured without normalization to cell numbers, this resulted in an increased neurite length with 20 µM fasudil. Compared with PBS control treatment (100 ± 9.4%), 20 µM fasudil led to a significant increase (165.9 ± 7.5%). In midbrain dopaminergic neurons that had been subjected to MPP+, 20 µM fasudil-treated cells obtained a cumulative outgrowth of 91.3 ± 12.1%, whereas the MPP+ controls achieved only 55.8 ± 7.3% (Fig. 2B and C).
Dopaminergic neuronal regeneration is increased by inhibition of rho kinase
To assess the regenerative neurite outgrowth stimulatory potential of fasudil, we performed a mechanical neurite transection of midbrain dopaminergic neurons that had been cultured for 3 days. Additionally, cells were either subjected to MPP+ or a control treatment at this time point. The evaluation of neurite outgrowth from the site of lesion was performed 2 days thereafter. Outgrowth of control-treated midbrain dopaminergic neurons was significantly reduced if cells were subjected to MPP+ instead of a PBS control that was set to 100% (72.6 ± 7.0% versus 100.0 ± 7.3%). Whereas a cell culture supplementation with fasudil 4 µM could not improve neurite outgrowth in comparison with the respective control in both PBS- and MPP+-treated cells (99.0 ± 4.1% and 78.0 ± 8.5%), supplementation with fasudil 20 µM was able to significantly increase the neurite length under both conditions (128.2 ± 7.5% and 101.3 ± 6.5%). The highest dose of fasudil 100 µM could improve neurite outgrowth only under PBS control conditions (134.7 ± 6.5%), but was not able to increase regeneration with concomitant MPP+ toxicity (54.6 ± 10.5%) (Fig. 3).
Fasudil protects mouse dopaminergic neurons from MPTP toxicity
To validate the effect of rho kinase inhibition in Parkinson’s disease models, we have chosen to take advantage of an in vivo model in which we applied MPTP in a subchronic systemic approach (MPTP injection of 30 mg/kg for five consecutive days). We found a substantial protection of tyrosine hydroxylase-immunopositive neurons in the substantia nigra 2 weeks after the first application of MPTP with daily feeding of fasudil at a dose of 100 mg/kg (7424 ± 291 cells) compared with vehicle-treated animals (5016 ± 375 cells). The non-MPTP treated control group exhibited the highest number of tyrosine hydroxylase cells (9141 ± 276 cells) (Fig. 4A and C). To verify that fasudil protection was not influenced by fasudil-dependent tyrosine hydroxylase down-regulation, we confirmed results by additional evaluation of NeuN-immunopositive cells in the substantia nigra (Fig. 4B).
The protective effect of fasudil on dopaminergic neurons was verified in a separate cohort of animals in which neuronal survival 6 weeks after the first MPTP lesioning stimulus was analysed. In this long-term treatment paradigm, both a fasudil dose of 30 mg/kg and of 100 mg/kg were applied. Compared with non-MPTP-treated controls, which exhibited highest numbers of tyrosine hydroxylase-immunopositive cells (12024 ± 1166 cells), there was no significant reduction in tyrosine hydroxylase cell numbers with MPTP treatment when fasudil was applied in both doses (fasudil 30 mg/kg: 10944 ± 522 cells; fasudil 100 mg/kg: 9755 ± 435 cells). In relation to MPTP-lesioned control animals (7949 ± 752 cells), the fasudil 30 mg/kg group showed a significantly higher number of tyrosine hydroxylase cells (Fig. 5A and C). Similar relations were found in the evaluation of Nissl-stained nigral neurons (Fig. 5B).
Fasudil prevents striatal dopamine depletion and attenuates reduction of striatal fibre density
Next, we asked whether fasudil, beyond morphological preservation of dopaminergic cell bodies, also preserves nigrostriatal innervating fibres after MPTP lesion. A 2-week treatment of fasudil (100 mg/kg) was able to significantly preserve the density of tyrosine hydroxylase-immunopositive fibres in the striatum (52.92 ± 0.89%), measured as tyrosine hydroxylase-immunopositive area in relation to the total area, in comparison with control-treated and MPTP-lesioned animals (42.72 ± 0.89%). The non-MPTP-lesioned control group exhibited the highest amount of tyrosine hydroxylase fibres (67.35 ± 0.64%) (Fig. 6A and B). To correlate striatal fibre morphology to a functional preservation of these axonal terminals, we quantified striatal dopamine and dopamine metabolite levels by HPLC. Although we found a decrease in striatal dopamine after MPTP intoxication in the vehicle groups using the same MPTP lesioning protocol as for the immunohistochemical analyses (3.56 ± 0.92 to 1.25 ± 0.18 ng/mg), dopamine levels in the fasudil and MPTP treatment group were significantly higher (2.05 ± 0.14 ng/mg) (Fig. 6C). A similar degree of preservation could be detected for homovanillic acid (vehicle control: 4.07 ± 0.65 ng/mg, MPTP control: 2.76 ± 0.46 ng/mg, fasudil MPTP: 3.96 ± 0.56 ng/mg) (Fig. 6E). Dihydroxyphenylacetic acid (DOPAC) levels were not significantly different between MPTP intoxication (13.46 ± 0.51 ng/mg) and controls (13.61 ± 0.67 ng/mg), whereas there was a trend towards higher levels in the fasudil treatment group (16.10 ± 0.95 ng/mg) (Fig. 6D). Moreover, the trend towards a lower metabolite ratio as calculated by ([homovanillic acid] + [DOPAC]) / [dopamine] in the fasudil treatment group (9.76 ± 3.14) versus MPTP control treatment (12.94 ± 3.64) indicates a reduced dopamine turnover and thus decreased cellular stress with fasudil (Fig. 6F). To exclude that fasudil by itself alters the MPTP metabolism, we measured MPP+ levels in the striatum without detecting a significant difference among both treatment groups (MPTP control: 57.50 ± 8.89 ng/ml versus fasudil MPTP: 52.40 ± 5.00 ng/ml) (Fig. 6G).
In a separate cohort of MPTP-lesioned animals, a fasudil treatment was applied for 6 weeks in doses of 30 mg/kg and 100 mg/kg. Compared with vehicle-treated animals (67.17 ± 0.76%), the MPTP-treated animals exhibited a persistent severe reduction in the density of tyrosine hydroxylase-immunopositive fibres in the striatum (33.79 ± 0.89%). Both treatments with fasudil resulted in a significantly higher tyrosine hydroxylase fibre density compared with MPTP control-treated animals (fasudil 30 mg/kg: 48.77 ± 0.85%; fasudil 100 mg/kg: 50.56 ± 0.88%) (Fig. 7A and B). Striatal dopamine and dopamine metabolite levels showed significantly higher levels in the 30 mg/kg fasudil treatment group compared with MPTP lesioned vehicle control (dopamine: fasudil 30 mg/kg MPTP: 6.46 ± 0.81 ng/mg versus MPTP control: 3.42 ± 0.69 ng/mg; DOPAC: fasudil 30 mg/kg MPTP: 2.90 ± 0.22 ng/mg versus MPTP control: 1.90 ± 0.14 ng/mg; homovanillic acid: fasudil 30 mg/kg MPTP: 1.28 ± 0.06 ng/mg versus MPTP control: 0.93 ± 0.09 ng/mg). With regards to DOPAC levels, the 100 mg/kg fasudil group was significantly higher than the MPTP control group (2.35 ± 0.23 ng/mg), and there was only a trend towards higher dopamine levels (6.16 ± 0.81 ng/mg). Metabolite ratios were lower for both fasudil 30 mg/kg and for 100 mg/kg, however, without reaching significance (Fig. 7C–F).
Behavioural performance is improved by fasudil treatment
The neuroprotective effects of 2-week fasudil (100 mg/kg) treatment could also be visualized in a behavioural analysis of motor functions. The cylinder test evaluates sensorimotor coordination and forelimb use. It is often used to compare side differences in unilateral lesion models, but can also provide information about whole body coordination, which is best reflected in rears without any paw use. In the vehicle control group, rears against the cylinder wall with both paws were most frequent (67.2 ± 4.0%). Free rears were observed less often but still present to a considerable extent (26.8 ± 3.2%). Touches against the wall with either right (3.2 ± 1.5%) or left (2.8 ± 1.1%) paw were infrequent as was also shown for the other groups [MPTP control right (2.9 ± 1.0%), left (1.5 ± 0.7%); fasudil MPTP right (3.0 ± 1.0%), left (2.4 ± 0.8%)]. Animals that had been MPTP lesioned showed a significantly elevated ratio of wall touches with both paws (88.0 ± 1.9%) and significantly less free rears (7.7 ± 1.7%), indicating impaired coordination performance. The MPTP group that had also received fasudil showed a behaviour that was similar to the non-MPTP-treated control group. Comparison with the MPTP lesion group indicated a significantly higher amount of free rears (16.5 ± 2.2%) and a reduced dependence of both paws touching the wall (78.1 ± 2.9%) (Fig. 8A). Another motor coordination test was included that evaluated average running performance on an accelerating rotarod. Here, the differences were less pronounced than in the cylinder test. Rotarod failure was earlier in the MPTP lesion group (143.2 ± 8.9 s) as in the vehicle control group (174.6 ± 11.4 s), and MPTP-treated animals that were also treated with fasudil showed a trend to longer time on the rotarod (163.8 ± 6.9 s), which was not significantly different from the vehicle control group (Fig. 8B).
The 6-week long-term treatment with fasudil 30 mg/kg or 100 mg/kg showed a maintained improvement of motor functions. In the cylinder test, the use of both paws was most common in the MPTP-lesioned control group (77.9 ± 5.4%), whereas a treatment with 30 mg/kg fasudil (49.0 ± 4.1%) or 100 mg/kg fasudil (59.2 ± 4.8%) reduced this behaviour in spite of the MPTP lesion. The more difficult free rears were observed more often in the 30 mg/kg fasudil group (39.8 ± 3.6%) than in the MPTP-lesioned control (19.8 ± 5.1%), whereas 100 mg/kg fasudil (34.3 ± 5.5%) did not result in a significant improvement (Fig. 9A). On the accelerated rotarod, MPTP-lesioned control animals did not show a difference in endurance (157.9 ± 9.8 s) compared with the vehicle control group (175.0 ± 12.8 s), and treatment with fasudil did not have a relevant influence on performance (fasudil 30 mg/kg: 149.8 ± 8.4 s; fasudil 100 mg/kg: 187.8 ± 9.4 s) (Fig. 9B).
Cellular survival after rho kinase inhibition is mediated through the Akt pathway
The molecular basis of the neuroprotective effect of fasudil was investigated in the MPP+intoxication paradigm of midbrain dopaminergic neurons in vitro. After 3 days in culture, cells were subjected to MPP+ or vehicle treatment for 48 h. In this experiment, we additionally included Y-27632 to evaluate a possible group effect by comparing the neuroprotective potential of this alternative rho kinase inhibitor with fasudil in midbrain dopaminergic neurons. Both compounds were applied in optimal final concentrations based on previous studies (20 µM fasudil and 10 µM Y-27632). Analysing activation of the signal transducer and activator of transcription (STAT)3, Akt and MAPK cell survival pathways, we could detect only an activation of Akt as demonstrated by elevated phosphorylated Akt levels under MPP+ lesioning conditions. This effect was significant in the fasudil group and showed a strong trend after application of Y-27632. Protein levels of Bcl2 and GAP43 were not changed (Fig. 10).
Current medical treatment options for Parkinson’s disease lack effectiveness in preventing clinical progression. Most experimental strategies have so far mainly focused on cellular conservation and neglected axonal protection, which may be a much earlier pharmacological target. We demonstrate in our preclinical study that pharmacological interference with rho kinase signalling is able to both preserve cellular and axonal function of lesioned dopaminergic cells. This is demonstrated both in the in vitro midbrain dopaminergic neuron MPP+ paradigm and in the subchronic MPTP mouse model of Parkinson’s disease in vivo.
Rho kinase inhibition protects from dopaminergic cell loss in toxicity-mediated lesion paradigms
Approaches that confer cellular protection in models of Parkinson’s disease have been the focus of research in preclinical models of Parkinson’s disease, as it was demonstrated that loss of dopaminergic cells is one of the major findings in post-mortem analysis of human patients with Parkinson’s disease (reviewed in Dauer and Przedborski, 2003). Because MPTP models recapitulate dopaminergic cell loss, they are among the most frequently used toxin-based mouse models of Parkinson’s disease in which cellular protective strategies have been examined. Nevertheless, only a few substances tested in these models have later been used in clinics, e.g. bromocriptine and pramipexole (Muralikrishnan and Mohanakumar, 1998; Zou et al., 2000), which may be due to the fact that MPTP models replicate only a very selective part of Parkinson’s disease pathology and do not model the largely unclear aetiological factors.
In our study, cultures of mouse ventral mesencephalon dopaminergic neurons were exposed to MPP+ doses that resulted in cell death of more than one-third of the population of tyrosine hydroxylase-immunopositive cells. Pharmacological treatment with the rho kinase inhibitor fasudil was able to rescue more than half of the dying dopaminergic cells if optimal pharmacological concentrations were used (Fig. 1). High concentrations of fasudil most likely result in off-target effects and influence the function of other kinases, as has been demonstrated for protein kinase N1 (PKN1) and protein kinase C (PKC) (Davies et al., 2000; Lingor et al., 2007). For our in vivo studies, we used two different fasudil concentrations, 30 mg/kg and 100 mg/kg, which have been successfully used in previous preclinical studies (Abe et al., 2004) and closely approach a human equivalent dose as calculated and applied for in clinical trials (Nohria et al., 2006; Reagan-Shaw et al., 2008). Based on these considerations, the neuroprotective potential of fasudil could also be shown in the MPTP model in vivo 2 and 6 weeks after initial MPTP lesion, where it attenuated nigral dopaminergic cell loss by >50% (Figs 4 and 5). This represents a robust protective effect compared with other therapeutically designed studies in this model (Nagel et al., 2008; Frank et al., 2012). Interestingly, the longer 6-week treatment with the fasudil dose of 30 mg/kg almost recovered tyrosine hydroxylase cell numbers to unlesioned control values, which indicates a sustained neuroprotective response over time, whereas fasudil 100 mg/kg showed only a trend to higher cell numbers. Our data therefore strongly suggest that fasudil, in vitro and in vivo, has a specific therapeutic dosage window where optimal neuroprotective effects can be expected.
Cellular protection mediated by pharmacological rho kinase inhibition has previously been demonstrated in other animal models. In a previous study, we were able to protect primary retinal ganglion cells from cell death induced by neurotrophin deprivation. Moreover, apoptotic cell death of axotomy-lesioned retinal ganglion cells in vivo was likewise reduced after intravitreal application of the rho kinase inhibitor Y-27632 (Lingor et al., 2008). If serum-deprived organotypic retinal cell cultures were treated with the rho kinase inhibitor dimethylfasudil, a protection from apoptotic cell death could be achieved (Tura et al., 2009). Interestingly, inflammatory processes were also attenuated in the latter study as was demonstrated by decreased levels of glial fibrillary acidic protein isoforms and reduced numbers of reactive astrocytes and microglia. The release of proinflammatory cytokines including tumour necrosis factor-α, interferon gamma and interleukin-6 was also shown to be reduced. Therefore, these parameters represent some of the most interesting targets for further analysis of fasudil-mediated neuroprotection in Parkinson’s disease.
Neuroprotection by fasudil is mediated by activation of cell survival pathways
Neuroprotective survival cascades seem to be differentially activated under rho kinase inhibition. In an analysis of neuroprotective effects in neurotrophin-deprived retinal ganglion cells, we found that rho kinase inhibition with Y-27632 and concomitant treatment with CNTF leads to an increase in the activation of MAPK and Akt survival signalling pathways. However, this was not the case for STAT3 signalling that, although stimulated by ciliary neurotrophic factor (CNTF) alone, was attenuated in combination with Y-27632 (Lingor et al., 2008). Evaluating cell survival responses after MPP+ application in primary midbrain dopaminergic neurons, we observed that Akt signalling is significantly activated after rho kinase inhibition with fasudil being more effective than Y-27632 (Fig. 10). This is in line with data from other studies suggesting that Akt plays a crucial role for the integrity of dopaminergic neurons, as it induces various trophic effects in murine models of Parkinson’s disease. The adeno-associated virus-assisted transduction of dopaminergic neurons in the substantia nigra with a gene encoding a myristoylated, constitutively active form of Akt resulted in pronounced trophic effects on dopamine neurons of adult and aged mice, including increases in neuron size, phenotypic markers and sprouting (Ries et al., 2006). Moreover, striatal dopaminergic axons were significantly preserved due to suppression of axon degeneration in the 6-hydroxydopamine lesion mouse model of Parkinson’s disease. This response has later been shown to be at least in part mediated by increased cellular mammalian target of rapamycin (mTor) activity and suppression of macroautophagy (Cheng et al., 2011). If the transducing adeno-associated virus contained a dominant negative form of Akt, a decreased phenotypic expression of tyrosine hydroxylase was observed in striatal dopaminergic axons and terminals both in aged and in α-synuclein overexpressing mice (Kim et al., 2011a). A putative downstream target of Akt is the Ras homolog enriched in brain/mammalian target of rapamycin (Rheb/mTor) pathway, which can foster axon growth and halt nigral cell death following striatal 6-hydroxydopamine lesion upon activation by adeno-associated virus transduction of substania nigra neurons with constitutively active Rheb (Kim et al., 2011b, 2012). The important role of Akt in Parkinson’s disease pathology is also supported by the examination of human post-mortem Parkinson’s disease brain dopaminergic neurons, which reveal a consistent depletion of phosphorylated Akt, but not of total Akt (Malagelada et al., 2008).
Fasudil attenuates dopaminergic neurite degeneration, fosters neurite outgrowth and preserves dopaminergic striatal innervation
Next to perikaryal protection we were most interested in the potential of fasudil to protect neuritic and axonal structures after toxic lesion in our models. It has been demonstrated not only for PC-12 or Ntera-2 cells (Zhang et al., 2006; Lingor et al., 2007) but also for primary neurons that rho kinase inhibition fosters neurite outgrowth (Lehmann et al., 1999; Borisoff et al., 2003). Furthermore, rho kinase inhibition has also been attributed an axon-stabilizing function because it was able to protect from lesion-induced axon degeneration in vitro, supposedly being promoted by RhoA/rho kinase signalling (Gallo, 2004). In an analysis of the dopaminergic neurite network in culture, we found that an optimal concentration of 20 µM fasudil was able to completely prevent MPP+-induced reduction of neurite length per cell. Under non-lesioning conditions, fasudil was even able to increase neurite length to >40% (Fig. 2A and C). When the entire neuritic network without normalization to cell numbers was quantified perikaryal survival effects elicited by fasudil additionally increase the cumulative, neurite length (Fig. 2B). Additionally, regenerative neurite outgrowth as evaluated in the scratch lesion model was significantly increased after fasudil treatment in an optimal dose range both in control and in MPP+ stress conditions. Interestingly, the high concentration of fasudil 100 µM was able to increase the length of the longest neurites only under non-MPP+ lesioning conditions, but failed if cells had been additionally subjected to MPP+. This could be due to a toxic effect of the high fasudil dose in addition to the MPP+ toxicity (Fig. 3). In view of all in vitro data, we conclude that fasudil is able to confer a strong perikaryal protection to dopaminergic neurons (Fig. 1). Furthermore, it has an additional, even more pronounced, protective effect on neuritic structures that is independent of perikaryal survival (Fig. 2). Similar to previous studies, fasudil reconfirmed its strong pro-regenerative properties (Fig. 3).
In the corresponding in vivo lesion models, the integrity of the striatal dopaminergic axonal terminal field was preserved by fasudil after MPTP treatment both morphologically and metabolically in the 2- and 6-week treatment paradigms (Figs 6 and 7). In contrast to the in vitro paradigm, a clear distinction between protection of perikarya and axons is more difficult to obtain, as both represent a functional entity. Treatment with fasudil conferred protection to perikarya and axons both in the 2- and 6-week evaluation. However, after 6-week treatment, when axonal pathology is strongest, fasudil (30 mg/kg) increased axonal structures by 44% whereas dopaminergic cells were increased by 38% only. This effect was also observed in the higher fasudil dose (100 mg/kg), which was less effective in protecting dopaminergic neurons after 6 weeks (23% increase of survival), whereas the striatal axon density was increased by 49%. These data suggest that the protective effect on dopaminergic striatal axons supercedes the pure neuroprotective effect of fasudil on dopaminergic neurons in the substantia nigra pars compacta. Concerning motor performance, fasudil 30 mg/kg best increased performance in the cylinder test, with almost reaching non-MPTP lesioned control values after 6 weeks (Figs 8 and 9). This indicates that there is an improvement of axonal terminal function with fasudil that is not evident in the axon structural analysis. Intriguingly, fasudil treatment showed a tendency towards increased striatal DOPAC levels after 2 weeks, and this effect was markedly significant after 6-week treatment (Figs 5 and 7). DOPAC is considered a primary intraneuronal metabolite of dopamine, whereas homovanillic acid rather represents a product of intra- and extraneuronal metabolism of released dopamine (Altar et al., 1992), and this increase could speak in favour of a pro-regenerative effect of fasudil on remaining dopaminergic striatal fibres.
Our examination of pharmacological rho kinase inhibition in the MPP+- and MPTP-based models of Parkinson’s disease demonstrates that fasudil not only increased dopaminergic cell survival, but also protected the neuritic network in vitro and the striatal axonal innervation in vivo. This is of special pathophysiological relevance because degeneration of the neuronal soma and its processes has been demonstrated to use different molecular pathways, both of which seem to be beneficially influenced by rho kinase inhibition. Although numerous successful strategies to counteract cell death in models of neurodegenerative diseases have been proposed (Vila and Przedborski, 2003), simple cellular protection has not proven to be sufficient as a therapeutic option in the clinic (Olanow et al., 2008). Rho kinase inhibitors such as fasudil may thus represent therapeutic agents that are dually effective in this context. Most importantly, fasudil has already been successfully implemented into clinical practice and is used in Japan for the treatment of patients suffering from subarachnoid haemorrhage-induced vasospasms due to its vasodilator effect on cerebral blood vessels (Zhao et al., 2006). Its favourable safety profile may allow a reassessment of this drug and facilitate its evaluation in a pilot clinical study in patients with Parkinson’s disease in the near future.
Michael J. Fox Foundation for Parkinson’s Research (Target Validation Spring 2010 Program) (to P.L.); DFG Research Centre for Molecular Physiology of the Brain (CMPB), Göttingen, Germany (to P.L. and M.B.); Ernst-und-Berta-Grimmke-Stiftung (to L.T.).
Conflict of interest
The authors thank Elisabeth Barski and Sabine Ceramella for expert technical assistance.
high-performance liquid chromatography
mitogen-activated protein kinase
signal transducer and activator of transcription