Increased mitochondrial calcium sensitivity and abnormal expression of innate immunity genes precede dopaminergic defects in Pink1-deficient mice

Abstract

Background: PTEN-induced kinase 1 (PINK1) is linked to recessive Parkinsonism (EOPD). Pink1 deletion results in impaired dopamine (DA) release and decreased mitochondrial respiration in the striatum of mice. To reveal additional mechanisms of Pink1-related dopaminergic dysfunction, we studied Ca²+ vulnerability of purified brain mitochondria, DA levels and metabolism and whether signaling pathways implicated in Parkinson's disease (PD) display altered activity in the nigrostriatal system of Pink1⁻/⁻ mice.

Methods and findings: Purified brain mitochondria of Pink1⁻/⁻ mice showed impaired Ca²+ storage capacity, resulting in increased Ca²+ induced mitochondrial permeability transition (mPT) that was rescued by cyclosporine A. A subpopulation of neurons in the substantia nigra of Pink1⁻/⁻ mice accumulated phospho-c-Jun, showing that Jun N-terminal kinase (JNK) activity is increased. Pink1⁻/⁻ mice 6 months and older displayed reduced DA levels associated with increased DA turnover. Moreover, Pink1⁻/⁻ mice had increased levels of IL-1β, IL-12 and IL-10 in the striatum after peripheral challenge with lipopolysaccharide (LPS), and Pink1⁻/⁻ embryonic fibroblasts showed decreased basal and inflammatory cytokine-induced nuclear factor kappa-β (NF-κB) activity. Quantitative transcriptional profiling in the striatum revealed that Pink1⁻/⁻ mice differentially express genes that (i) are upregulated in animals with experimentally induced dopaminergic lesions, (ii) regulate innate immune responses and/or apoptosis and (iii) promote axonal regeneration and sprouting.

Conclusions: Increased mitochondrial Ca²+ sensitivity and JNK activity are early defects in Pink1⁻/⁻ mice that precede reduced DA levels and abnormal DA homeostasis and may contribute to neuronal dysfunction in familial PD. Differential gene expression in the nigrostriatal system of Pink1⁻/⁻ mice supports early dopaminergic dysfunction and shows that Pink1 deletion causes aberrant expression of genes that regulate innate immune responses. While some differentially expressed genes may mitigate neurodegeneration, increased LPS-induced brain cytokine expression and impaired cytokine-induced NF-κB activation may predispose neurons of Pink1⁻/⁻ mice to inflammation and injury-induced cell death.

Figure 1. Inactivation of the mouse Pink1 locus by gene targeting in ES cells.

(A) Mouse Pink1 gene structure, (B) targeting vector and (C) mutated Pink1 gene lacking exons 4 and 5 after homologous recombination with the targeting vector. The PINK1 kinase domain is encoded within exons 2–8, with exons 4 and 5 specifying amino acids 257–374. Active site Asp362 and at least 15 familial PD-associated Pink1 mutations cluster in exons 4 and 5 .

Figure 2. Southern blot analysis of ES cell colonies and F2 generation mice.

(A) Genomic DNA from PCR-positive ES cell clones was double-digested with NcoI/SacI or NdeI/XbaI and hybridized with the indicated probes. The location of the two probes and expected sizes of the various DNA fragments for wildtype (WT) and Pink1^+/− ES cells are shown in Figure 1. (B) Tail DNA from offspring of Pink1^+/− breeder pairs was digested with NcoI and SacI and analyzed by Southern blot using “outside exon 8” probe. The 4513 bp band indicates the wildtype Pink1 allele and the 4152 bp band is diagnostic for the mutated Pink1 allele. Location of the probe and restriction enzyme cleavage sites are shown in Figure 1.

Figure 3. Analysis and quantification of Pink1 mRNA expression.

Total RNA isolated from the brain of mice was converted to cDNA. (A) PCR with a forward primer located in exon 3 and reverse primers located in exon 6, 7 or 8, generating expected PCR products of 480, 624 and 1001 bp for the Pink1 wildtype allele in wildtype and Pink1^+/− samples. In contrast, PCR products of 133, 149 and 289 that would arise if exon 3 were directly spliced to exon 6, 7, or 8 in transcripts derived from the mutant Pink1 allele were absent in both Pink1^+/− and Pink1^−/− samples. This shows that alternatively spliced transcripts are not generated from the disrupted Pink1 allele. (B) Quantitative real-time PCR with primers located in exon 1 (forward) and exon 3 (reverse) showing that Pink1^−/− mice express at most 6.8% of a truncated Pink1 mRNA encompassing exons 1–3, compared to wildtype mice. However, such a truncated mRNA would not give rise to any PINK1 protein with kinase activity (see main text).

Figure 4. Pink1^−/− brain mitochondria have significantly lower Ca²⁺ load capacity.

Brain mitochondria were isolated from 2 month-old male wildtype (WT) and Pink1^−/− (PINK1-KO) mice to measure calcium load capacity. Mitochondrial protein (200 µg) from each genotype was incubated in a 2-ml reaction mixture containing 125 mM KCl buffer (37°C) along with fluorescence indicators: 100 nM Calcium Green 5-N (for measurement of extra-mitochondrial calcium concentration; excitation 506 nm, emission 532 nm) and 100 nM TMRE (for measurement of membrane potential; excitation 550 nm, emission 575 nm). Mitochondrial bioenergetic coupling was assessed by following changes in membrane potential (TMRE fluorescence, data not shown) following the addition of pyruvate + malate (PM), ADP (A) and oligomycin (O) at 1, 2 and 3 min respectively. Calcium infusion began at 5 min (rate of 160 nmol/mg protein/min) and changes in extra-mitochondrial Ca²⁺ were assessed (CaG5N fluorescence). In Panel A, representative traces are shown to depict the calcium uptake and storage capacity before the opening of the mitochondrial permeability transition (mPT) pore. Mitochondria from PINK1-KO mice had decreased calcium load capacity, indicated by the earlier onset of mPT (sharp rise in extra-mitochondrial Ca²⁺) as compared to WT mice. Calcium load capacity in PINK-KO mice was increased by incubating mitochondria with cyclosporine A (CSA, 1 µM), a specific inhibitor of the PTP, showing that the reduced calcium load capacity was due to enhanced mPT. In Panel B, quantitative measurements of maximal calcium load capacity before mPT for mitochondria isolated from WT and PINK1-KO mice are shown and expressed as nmol Ca²⁺ infused/mg protein. The graph in panel B illustrates a significant loss of calcium load capacity in PINK1-KO mice (*P<0.05, n = 4 mice per genotype).

Figure 5. Accumulation of phospho-c-Jun in the substantia nigra of Pink1^−/− mice.

Cryosections of the substantia nigra from wildtype (A–B) and Pink1^−/− mice (C–F) were stained with antibodies against phospho-c-Jun (nickel-DAB staining) and subsequently TH (fluorescent) as described in the Methods. Arrows in panels C–F point to cells expressing nuclear phospho-c-Jun that is surrounded by TH-positive cytosol, suggesting that these cells are dopaminergic neurons. AP coordinates of sections according to the mouse stereotaxic atlas (Franklin and Paxinos, The Mouse Brain in Stereotaxic Coordinates, Third Edition 2007) are indicated to demonstrate that phospho-c-Jun was expressed in distantly spaced sections of Pink1^−/− mice (C–F), including sections anatomically matched to wildtype control mice (compare A–B and C–D). All three Pink1^−/− mice but none of the wildtype mice showed expression of phospho-c-Jun in a subpopulation of TH-positive neurons.

Figure 6. Dopamine levels, dopamine turnover and dopamine neuron counts.

(A) Decreased DA levels in the striatum of Pink1^−/− mice aged 6 months and older. (B) Normal counts of dopaminergic neurons in the substantia nigra pars compacta (SNc) of 1-year old Pink1^−/− mice. (C) Increased DA turnover in Pink1^−/− mice. Eight mice per genotype were used for catecholamine analysis (A and C). Five wildtype and six Pink1^−/− mice were used to determine nigral DA neuron numbers by unbiased stereology (B). * P<0.05, ** P<0.01, *** P<0.001.

Figure 7. Basal and inflammatory signal-induced NF-κB activity is reduced in Pink1^−/− embryonic fibroblasts.

Wildtype and Pink1^−/− MEF were transfected with plasmid pNF-κB-luc (Clontech). Twenty-four hours after transfection the cells were incubated for 8 hours with 30 ng/ml TNF-α, 10 ng/ml IL-1β, 100 ng/ml LPS or remained untreated (control) and luciferase activity was measured as described in the Methods. (A) NF-κB-dependent luciferase activity, expressed as relative light units (RLU) per mg protein. Data represent pooled values from two independent experiments with similar results. In each experiment luciferase activity was measured in five wells per condition. Non-transfected cells (NT) showed no luciferase activity. (B) Wildtype and Pink1^−/− fibroblasts were transfected with the same plasmid/lipofectamine mixture to ensure equal transfection efficiency, which was confirmed to be the case with an EGFP expression plasmid as described in the Methods. ** P<0.01.

Figure 8. Cytokine expression in the striatum and isolated microglial cultures.

(A) Wildtype and Pink1^−/− mice (n = 4 mice/genotype) were injected ip with 0.33 µg LPS/g body weight. Cytokines in striatal homogenates (corresponding to 100 µg total protein) were measured eight hours later by ELISA as described in the Materials and Methods. * P<0.05, compared to wildtype mice. Basal cytokine levels, measured in a separate experiment, were not statistically different between wildtype and Pink1^−/− mice. (B) Microglial cultures derived from the forebrain of neonatal wildtype and Pink1^−/− mice were incubated with 100 ng/ml LPS for 24 hours in 24-well plates and the cytokines were measured with an ELISA as described in the Methods (n = 8 wells per condition). (C) Strong induction of IL-6, TNF-α and G-CSF in wildtype and Pink1^−/− microglia cells demonstrates that the cells were capable of responding to the LPS stimulus. In panels B and C, background-corrected absorbance is plotted (OD450 minus OD570). (D) Real-time PCR expression analysis of CD3 mRNA (specific T cell marker) in the striatum of control and LPS-treated wildtype and Pink1^−/− mice, showing that the T cell marker is barely detectable (Ct values of 39.67 and 38.46) and not increased by LPS treatment.