Background
Parkinson’s disease (PD) is characterized by non-motor and motor symptoms, the latter caused by a depletion of striatal dopamine (DA) due to neurodegeneration of dopaminergic neurons in the
substantia nigra (SN). Numerous studies show that mitochondrial dysfunction plays an important role in the pathophysiology of PD and that complexes of the mitochondrial electron transport chain are affected [
1,
2].
Different laboratories have analyzed the activity of the mitochondrial respiratory chain complexes in samples of PD, but the results were variable. Complex I activity was decreased in the SN, platelets and muscles of affected individuals [
3] and SN dopaminergic neurons from PD patients harbored high levels of mtDNA deletions and cytochrome c oxidase dysfunction [
4,
5]. Some groups found impaired complex II, complex II + III, and complex IV activity in platelets [
6], whereas others reported a selective complex I inhibition [
3]. Different studies also showed impaired complex I, II + III, and IV in PD muscle [
7‐
9], whereas others reported no differences, and results on PD lymphocytes are even more contradictory.
A small percentage of PD cases is monogenic and investigations into the associated gene mutations confirmed the importance of mitochondria. PARK2 (encoding for Parkin) and PINK1, for example, are two genes mutated in rare forms of monogenic PD, and they cooperate for selective autophagy of damaged mitochondria (mitophagy) [
10,
11].
Because mitochondrial dysfunctions appear to have a role in the pathogenesis of PD, it has been suggested that an increase in neuronal mitochondrial content could compensate for the bioenergetic defects that lead to neurodegeneration [
12,
13].
Pioglitazone is an agonist of PPAR-γ (peroxisome proliferator-activated receptor γ), a receptor that regulates cellular functions such as lipid metabolism, cell growth, differentiation and inflammation. PPAR-γ is co-activated by PGC-1α a master regulator of mitochondrial biogenesis [
14]. Pioglitazone treatment has been shown to increase mitochondrial biogenesis in various tissues and to reduce neurodegeneration in different mouse models. In an X-linked adrenoleukodystrophy model, pioglitazone restored mitochondrial content and locomotor impairment [
15]. In Alzheimer’s disease models, it improved spatial learning, enhanced AKT signaling, and attenuated tau hyperphosphorylation and neuroinflammation [
16], probably by enhancing the microglial uptake of beta-amyloid [
17,
18]. Pioglitazone ameliorated the 3-nitropropionic acid-induced mitochondrial dysfunction in striatal neurons in a mouse model of Huntington’s disease [
19]. It has also been used on acute pharmacological models of PD, attenuating neurodegeneration in MPTP-treated mice, and monkeys [
20‐
23]. Here we describe that pioglitazone can ameliorate the motor symptoms of a novel genetic mouse model of PD not by preventing dopaminergic neuron loss but by reducing inflammation.
Discussion and conclusions
The creation of animal models is crucial to study the pathogenesis of PD and to develop new therapies. Although there is no perfect model that encompasses all characteristics of PD pathogenesis, each allow specific aspects of the disease to be analyzed. Drug-induced models (6-OHDA, MPTP, Rotenone) cause fast degeneration of dopaminergic neurons and can be used, for example, to test neuroprotective drugs, usually administered before the toxin itself. MPTP crosses the blood–brain barrier and its metabolite, MPP+, blocks Complex I of the electron transport chain, inducing rapid neurodegeneration of the dopaminergic neurons. However, MPTP has several disadvantages: 1) its extreme toxicity makes it difficult and dangerous to work with; 2) the reproducibility of the lesion depends on the mouse strain, gender, age and body weight [
35], as well as on the purity of the compound and on the administration protocol; 3) this treatment has a high death rate which occurs within the first 24 hours and is unrelated to the damage in the dopaminergic system [
36]. 6-OHDA is also effective in inducing dopaminergic lesions, but has to be focally injected into small brain regions [
37], causing variability in the results.
Although most cases of PD are sporadic, to study the pathophysiology of this disease mouse models have been created by knocking-out or knocking-in mutated forms of genes that are involved in the rare genetic forms of PD [
38]. Among them, mice overexpressing α-synuclein or mutated forms of LRRK2 [
39], and knockout for Parkin, PINK1 and DJ1 have been created. Unfortunately, they showed only mild motor coordination defects and impairment in striatal DA release with no dopaminergic neurodegeneration. Nonetheless, they have been useful models to increase our understanding of the endogenous role of these gene products [
40].
Because of the putative mitochondrial involvement in PD, transgenic mouse models have been created by disrupting mitochondrial functions and dynamics, two of them by reducing mtDNA levels: the “Mito-Park” mouse [
41] and the “PD-mito-
PstI” mouse [
28]. These mice show different rates of neurodegeneration, depletion of DA in the striatum, L-DOPA sensitive-defects in motor coordination, and abnormal mitochondrial aggregates, further indicating the importance of mitochondrial function in the dopaminergic system. Retrograde neurodegeneration has been induced also by knocking out Mfn2 in dopaminergic cells [
32].
However, some mitochondrial perturbations do not induce neurodegeneration. For example, even though
Ndufs4-KO animals (Complex I deficient) showed a severe phenotype and died before 7 weeks of age, knocking out
Ndufs4 only in dopaminergic neurons led to a mild reduction of dopamine in the striatum and did not induce neurodegeneration or motor defects [
42].
We created a novel mouse model of PD by disrupting cytochrome c oxidase (Complex IV) in dopaminergic neurons. Dopaminergic axons in the striatum were markedly decreased in Cox10/DAT-cre mice at 2 months of age resulting in almost complete depletion of striatal dopamine. Accordingly, the number of TH+ neurons was also dramatically decreased at this time point with Cox10/DAT-cre mice showing a severe motor phenotype with a loss of coordination and a decrease in voluntary movement.
Further decline in motor coordination was observed between 2 and 4 months of age, but despite these defects, the life span of the mice was not markedly reduced and KO mice could be identified only by the decreased body weight. Although it is surprising that mice can live relatively normal lives in cages with such a severe loss of dopaminergic neurons, motor phenotypes were clear. Still, it points to important differences between the mouse and human disorders. Also, compared to other mouse models of PD, such a dramatic dopaminergic neurodegeneration has been obtained only with drug-induced models.
Because in our model, a massive neurodegeneration is observed, we could test the efficacy of different compounds in a model of advanced disease. In fact, when PD patients are diagnosed, typically they already show a loss of approximately 70 % of their dopaminergic neurons, and therefore current treatments are administered only to attenuate and slow the progression of the motor symptoms. Even though L-DOPA administration remains the most effective treatment in alleviating the motor symptoms of PD, its long-term use leads to LID L-Dopa-Induced Dyskinesia (LID) [
43].
Pioglitazone is a PPARγ agonist traditionally used as an insulin-sensitizing drug for the treatment of type-II diabetes. Therefore, one of the advantages of this drug is that it is already used in a clinical setting. PPARγ is ubiquitously expressed in the CNS [
34], including in TH+ dopaminergic cells in the
substantia nigra and in the
ventral tegmental area [
44].
This transcription factor regulates gene-expression programs of metabolic pathways and its activation is also involved in increasing expression of mitochondrial proteins, enhancing mitochondrial function and OXPHOS capacity [
45]. PPARs also have a role in inflammation and in neurodegenerative disorders that have a inflammatory component, including PD, AD, stroke, ALS and spinal cord injury [
46]. In recent studies, it has been shown that pre-treatment of non-human primates with pioglitazone protects dopaminergic neurons from MPTP-induced neurodegeneration, modulating inflammation and increasing the expression of PPAR-γ [
20,
22]. However, it is important to note that a very recent phase II clinical trial of PD patients with pioglitazone, though inconclusive, did not suggest a major beneficial effect [
47].
Here we report the beneficial effects of pioglitazone on the motor phenotype of a mouse model of PD with early-onset dopaminergic neuron loss due to a mitochondrial respiratory chain defect. In contrast to the other studies, pioglitazone was administered when the vast majority of TH+ neurons had already degenerated, a scenario akin to PD patients at diagnosis. The positive effect of pioglitazone on motor behavior was not sufficient to restore the motor coordination to control levels, but this is not surprising, considering that the neurodegeneration in the Cox10/DAT-cre mice was already severe at 2 months of age.
One possible mechanism for an amelioration of the motor symptoms could be a decrease in apoptosis of dopaminergic neurons by decreasing the activity of caspase-3 [
48]. However, we did not detect any change in neuronal loss of treated mice compared to controls probably because of the already massive degeneration that affects
Cox10/DAT-cre mice at this young age. A second hypothesis could be that, even if the neurodegeneration is not decreased, there is a stimulation of mitochondrial biogenesis by the action of pioglitazone in activating PGC-1α and other mitochondrial genes in the remaining cells. However, we did not detect changes in mitochondrial biogenesis. As mentioned before, one of the mechanisms involved in pioglitazone-mediated neuroprotection is the decrease of neuroinflammation. Indeed, our results are in agreement with this model. We analyzed inflammation in the two brain regions mostly affected by dopaminergic neurodegeneration and we detected differences in neuroinflammation: Iba1+ cells were increased in the midbrain but not in the striatum of
Cox10/DAT-cre animals, while MHCII, a marker of activated microglia, was increased in both regions in 4-month-old animals. This suggests that consequential inflammation occurs in the midbrain but also, to a lesser extent, in the striatum by activation but not proliferation of the resident microglial cells. Moreover, the activation of microglia is concomitant with the worsening of the motor phenotype at 4 months. IL1β, CXCL10 and CCL2 were significantly increased in the midbrain of 2 month-old
Cox10/DAT-cre mice compared to DAT-cre animals. Even if not significant, we also detected a trend of all other cytokines to be increased in striatum and midbrain of 2 and 6-month-old
Cox10/DAT-cre animals compared to DAT-cre mice. Pioglitazone treatment specifically reduced both microglial cell number in the midbrain and microglial activation in midbrain and striatum. Moreover, CCL2 was significantly decreased in the striatum of
Cox10/DAT-cre mice treated with pioglitazone.
It is possible that pioglitazone may benefit only cases where neuroinflammation is prominent, which could be restricted to a sub-population of PD patients, or that the effect of pioglitazone may be stage-dependent, as neuroinflammation becomes more prominent as the disease progresses [
49]. This would explain the inconsistent results with patients [
47].
Our mouse model also showed another interesting phenotype. When we attempted to compare the improvement in motor coordination obtained with pioglitazone treatment with age-matched mice (6 months of age) treated with L-DOPA, we noticed hyperactivity, abnormal stereotypic movements and absence of rearing activity. Dysregulation of DA release and clearance resulting from a loss of nigrostriatal DA terminals [
50] is also involved in the development of LID. LID is one of the main side effects of L-DOPA treatment in PD patients, who develop LID in 90 % of cases after 8–10 years of L-DOPA treatment. The molecular mechanism responsible for the development of LID is still not fully characterized, and the most utilized animal model for this condition is the 6-OHDA-lesioned mouse, where the massive neurodegeneration is achieved by intracranial unilateral injection of 6-OHDA. Although further characterization would be necessary, the
Cox10/DAT-cre mice also appear to show LID, and they can be potentially used as a model of late stage of PD with dyskinesia-resembling phenotype. Moreover, LID was not observed in mice treated with pioglitazone, suggesting a potential co-adjuvant, which may allow for reduced doses of L-DOPA for long-term PD patients.
Methods
Mice procedures
All animals used in this work were males and had a pure C57Bl/6 J background, backcrossed for at least 10 generations. All experiments and animal husbandry were performed according to a protocol approved by the University of Miami Institutional Animal Care and Use Committee. Mice were housed in a virus-antigen-free facility of the University of Miami Division of Veterinary Resources in a 12-h light/dark cycle at room temperature and fed ad libitum.
Enzymatic activity assays
Striatum homogenates were prepared in PBS containing complete protease inhibitor cocktail (Roche diagnostics) in a volume of 10x the weight. The tissue was disrupted by 10–15 strokes, using a motor-driven pestle. Homogenates were centrifuged at 1000 g for 5 min and supernatants used for enzymatic assays. The activities of CIV and citrate synthase were measured spectrophotometrically as described [
51]. Protein concentrations were determined using the Bio-Rad Bradford Assay Kit with bovine serum albumin (BSA) as standard. Specific activity was determined and values represented as μmoles/min/mg protein.
Motor behavioral tests
Pole test
Pole test for motor coordination/nigrostriatal dysfunction of mice was previously described [
52]. Animals were hung upright on a vertical (8 mm diameter; 55 cm length) pole and were given three minutes to change orientation to descend. Animals were given three trials with an average taken of the latency to descent to the base. Failure to descend or fall from the pole was given a maximum time of three minutes.
RotaRod test
Motor coordination was evaluated with a RotaRod (IITC Life Sciences) designed for mice. Animals were tested with five runs on a given day with one run for practice. Three runs were recorded and combined to find the average latency to fall. A resting period of 120 seconds between each run was given. Animals were required to position limbs to stay on a rotating rod accelerating from 6 rpm-20 rpm over a 180 seconds time period. Mice that completed the task received a final latency time of 180 seconds.
Ambulatory movement measurement
Spontaneous self-initiated movement was recorded using an activity cage setup (Columbus Instruments) designed for mice. Animals were housed individually in a novel cage environment thirty minutes prior to their dark cycle and monitored for a twelve or twenty-four hour period undisturbed. Ambulatory movement was counted as the number of infrared beam breaks that occurred inside the cage.
Open field test
Open field (Med Associates Inc.) is a sensitive method for measuring gross and fine locomotor activity. It consists of a chamber and a system of 16 infrared transmitters that record the position of the animal in the three dimensional space. With this system not only the horizontal movement can be recorded but also the rearing activity. For our study, the animals were placed in the chamber 30 minutes before the test and the locomotor activities were recorded for 15 minutes.
Pharmacological treatment
Mice were treated with L-3,4-dihydroxyphenylalanine (L-DOPA) at different concentrations (see manuscript) and benserazide (3,125 mg/kg, Sigma) dissolved in saline, administrated via intraperitoneal injection (I.P.) one hour before behavioral testing.
Pioglitazone food was provided by Bio-Serv (Frenchtown, NJ). Pharmaceutical Actos tablets (pioglitazone) were incorporated into the mouse diet at a concentration of 120 mg/kg.
The Vanderbilt University CMN/KC Neurochemistry Core Lab using HPLC separation followed by fluorescent and/or electrochemical detection performed dopamine and metabolite quantification measurements. Freshly isolated striatum was harvested and quickly frozen in liquid nitrogen from mice sacrificed as previously described.
Western blots
Protein extracts were prepared from the striatal neuroanatomical regions and homogenized in PBS containing a protease inhibitor mixture (Roche). Upon use, SDS was added to the homogenate at the final concentration of 4 %. Homogenates were then centrifuged at 14,000 g at 4 °C, and the supernatant was collected for analysis. Protein concentration was determined by Lowry assay using the BCA kit (BioRad). Approximately 30–50 μg of protein was run on a 4-20 % gradient Tris–HCl gel (BioRad) and transferred to a PVDF or nitrocellulose membrane (BioRad).
Membranes were blocked in 1:1 Odyssey blocking solution (LI-COR Biosciences) for 1 h at room temperature. Primary antibodies, which were incubated overnight at 4 °C, were: anti-TH (tyrosine hydroxylase) 1:1000 dilution (Sigma), anti-DAT (Dopamine transporter) 1:1000 (Sigma), anti-porin/VDAC 1:2000 (MitoSciences), OXPHOS cocktail rodent mixture 1:1000 (MitoSciences), α-tubulin 1:2000 (Sigma), β-actin 1:5000 (Sigma), GFAP (glial fibrillary acidic protein) 1:1000 (Cell Signaling Technology), Iba1 1:500 (Wako), MHCII 1:1000 (Abcam), NeuN 1:1000 (Chemicon), TUJ1 1:1000 (abcam). Secondary antibodies used were infrared-conjugated antibodies anti-rabbit-700/anti-mouse-800 (Rockland) at 1:3000 to 1:5000 concentrations respectively, and incubated at room temperature for 1 h. Blots were visualized with Odyssey Infrared Imaging System (LI-COR Biosciences). Optical density measurements were taken by software supplied by LI-COR.
Immunostaining
Anesthetized mice were transcardially perfused with ice-cold PBS and 4 % PFA. The brain was isolated, regions of interest were dissected using a brain matrix, and cryoprotected in sucrose 30 % and frozen in OCT. Frontal sections were cut at a 20-μm thickness with a cryostat (Leica).
Sections were blocked with 10 % normal goat serum (NGS) for 30 min at RT, and then incubated with primary Ab for 16 h at 4 °C: anti-TH 1:500 (Sigma); GFAP 1:500 (Cell Signaling Technology), Iba1 1:500 (Wako) NeuN 1:500 (Chemicon), TUJ1 1:500 (abcam),. Slides were then incubated with secondary Ab biotin-conjugated goat anti-mouse (KPL) for 1 h at RT and Streptavidin-Peroxidase (KPL) for 30 min at RT. Staining was visualized using a solution of 0.05 % 3,3’-diaminobenzidine (DAB), 50 mM Tris–HCl pH 7.2, 0.02 % H2O2. Images were captured with an Olympus BX51 microscope.
For immunofluorescent staining, sections were blocked with 10 % normal goat serum (NGS) for 1 h at RT, and permeabilized with 1 % Triton X-100. Sections were incubated with primary Ab (anti-TH 1:500 (Sigma), anti Iba1 1:500 (Wako)) for 16 h at 4 °C. Slides were then incubated with Alexa-fluor secondary Ab for 1 h at RT and mounted with Vectashield mounting medium for fluorescence. Images were captured with an Olympus BX51 confocal microscope.
MtDNA content measurement by qPCR
DNA was isolated from the striatal homogenates using phenol:chloroform extraction. We designed a couple of primers to amplify mtDNA. To determine relative quantity of mtDNA in each sample, we used the comparative Ct method (Schmittgen and Livak, 2008), normalizing the amplification with primers that amplify a genomic DNA region. Relative quantity was corrected for relative PCR amplification efficiency using Biorad CFX Manager Software. Primers for mtDNA were as follows: ND1-3281 F, CAGCCTGACCCATAGCCATA; ND1-3364 B, ATTCTCCTTCTGTCAGGTCGAA. Primers for genomic DNA were as follows: β-actin F, GCGCAAGTACTCTGTGTGGA; β-actin B, CATCGTACTCCTGCTTGCTG. Targets were amplified using Maxima SYBR Green/ROX qPCR Master Mix (Fermentas) using CFX96 Realtime PCR system (Bio-Rad) under the suggested PCR conditions from the manufacturer.
Cell counting
Stereological counting was used to quantify Iba1+ cells. Immunohistochemical identification of Iba1+ cells is described above. Microglial cells were counted with ImageJ program in double blind. Four sections were randomly chosen through the midbrain and three specific midbrain regions (depicted with black squares in Fig.
7) were chosen for the counting, with a total of 12 counting sections per individual animal. Iba1+ cell numbers were determined on 20 μm immunohistochemical stained sections. A guard zone of 3 μm from the top and from the bottom of each section was not considered in the counting. The midbrain area was identified using the 4× objective and Iba1+ cells were counted at 20x magnification.
Multi-analyte profiling: determination of Cytokines in brain tissues
Concentrations of IL-1β, IL-6, IL-10, CXCL10 (IP10), CCL2 (MCP-1), CCL5 (RANTES), TNFα, were simultaneously quantified in each brain tissue sample using a MILLIPLEX®MAP mouse cytokine magnetic bead panel (Millipore). Briefly, brain tissues were collected and snap frozen in liquid nitrogen, homogenized in suggested lysis buffer (20 mMol TrisHCl pH7.4, 150 mMol NaCl, 1 mMol PMSF, 0.05 % Tween-20, protease inhibitor mixture), sonicated for 5” and centrifuged for 10’ at 10000 g. Supernatant was collected and protein concentration was adjusted to 1 mg/ml and diluted 1:1 with sample buffer. Plate was prepared following the manual instructions. Cytokine concentrations were then measured using MagPix Luminex100 reader and analyzed by Miami Center for AIDS Research (CFAR). Mean fluorescent intensities (MFI) were analyzed with MILLIPLEXTM Analyst Software (EMD Millipore) and cytokine levels expressed as pg/ml. CCL5 and TNFα were below detectable levels.
Statistical analysis
Two-tailed, unpaired Student t-test was used to determine the statistical significance between two different groups. Multiple groups were compared using a one-way ANOVA followed by a Bonferroni post-hoc comparison. Error bars represent SEM. P values are indicated by asterisks (*p < 0.05, **p < 0.01, ***p < 0.001).
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
MP designed and coordinated the study, carried out most of the pathology, behavior studies, and western blots. She also drafted the manuscript. NN assisted in some neuropathology analyses, performed some of the statistic analyses and edited the manuscript. SP performed the enzymatic assays, assisted with some of the western blots and edited the manuscript. FD created the Cox10 mouse, helped create the DAT conditional KO and edited the manuscript. RB assisted with the multiplex analysis for inflammatory cytokines and to revise the manuscript. CTM conceived of the study, and participated in its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript.