Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Jul 1;26(13):2412-2425.
doi: 10.1093/hmg/ddx132.

SLP-2 Interacts With Parkin in Mitochondria and Prevents Mitochondrial Dysfunction in Parkin-deficient Human iPSC-derived Neurons and Drosophila

Affiliations
Free PMC article

SLP-2 Interacts With Parkin in Mitochondria and Prevents Mitochondrial Dysfunction in Parkin-deficient Human iPSC-derived Neurons and Drosophila

Alessandra Zanon et al. Hum Mol Genet. .
Free PMC article

Erratum in

Abstract

Mutations in the Parkin gene (PARK2) have been linked to a recessive form of Parkinson's disease (PD) characterized by the loss of dopaminergic neurons in the substantia nigra. Deficiencies of mitochondrial respiratory chain complex I activity have been observed in the substantia nigra of PD patients, and loss of Parkin results in the reduction of complex I activity shown in various cell and animal models. Using co-immunoprecipitation and proximity ligation assays on endogenous proteins, we demonstrate that Parkin interacts with mitochondrial Stomatin-like protein 2 (SLP-2), which also binds the mitochondrial lipid cardiolipin and functions in the assembly of respiratory chain proteins. SH-SY5Y cells with a stable knockdown of Parkin or SLP-2, as well as induced pluripotent stem cell-derived neurons from Parkin mutation carriers, showed decreased complex I activity and altered mitochondrial network morphology. Importantly, induced expression of SLP-2 corrected for these mitochondrial alterations caused by reduced Parkin function in these cells. In-vivo Drosophila studies showed a genetic interaction of Parkin and SLP-2, and further, tissue-specific or global overexpression of SLP-2 transgenes rescued parkin mutant phenotypes, in particular loss of dopaminergic neurons, mitochondrial network structure, reduced ATP production, and flight and motor dysfunction. The physical and genetic interaction between Parkin and SLP-2 and the compensatory potential of SLP-2 suggest a functional epistatic relationship to Parkin and a protective role of SLP-2 in neurons. This finding places further emphasis on the significance of Parkin for the maintenance of mitochondrial function in neurons and provides a novel target for therapeutic strategies.

Figures

Figure 1
Figure 1
Parkin interacts with mitochondrial SLP-2. (A) Whole cell lysates of untransfected SH-SY5Y cells were subjected to co-immunoprecipitation (IP) with antibodies against Parkin (left panel) and SLP-2 (right panel), followed by Western blotting (WB) of input and IP fractions with the indicated antibodies (the blots were probed consecutively with the antibodies). Cells with knockdown (KD) constructs against Parkin and SLP-2 validate the sensitivity and specificity of the anti-Parkin and anti-SLP-2 antibodies, respectively. IgG was used as negative control for the IPs. Molecular mass markers are in kilodaltons (kDa). (B) SH-SY5Y cells, wild type (WT) and with stable Parkin KD, were processed using the PLA to quantitatively assess the Parkin-SLP-2 interaction under normal culture conditions and after CCCP treatment (3h, 10 µM). The PLA signal is visualized in red, while DAPI-stained nuclei are shown in blue. Exposure to CCCP increased the PLA signal indicating an augmented interaction between the two proteins. The amount of the increase was higher in WT cells (2.8x) compared to Parkin KD cells (1.6x). Two-tailed Students t-test *P < 0.05; **P < 0.01. n.s. = not significant. Scale bar = 7.5 μm. (C) A PLA experiment for Parkin and SLP-2 was co-stained with a mitochondrial marker (green fluorescent protein attached to a mitochondrial leading sequence, mito-GFP) showing a co-localization of the PLA signal with the mitochondria. The specificity of the PLA interaction results was confirmed by performing the experiments with only one of the two primary antibodies. Mitochondrial fractions of SH-SY5Y cells stably overexpressing Parkin, untreated and treated with CCCP, were subjected to IP with anti-Parkin, anti-SLP-2, and anti-IgG antibodies, respectively, followed by WB with the indicated antibodies.
Figure 2
Figure 2
SLP-2 overexpression restores mitochondrial dysfunction in Parkin-deficient SH-SY5Y cells. Complex I activity was determined in mitochondrial fractions of SH-SY5Y cells expressing lentivirally transduced shRNA constructs for Parkin and SLP-2 as well as Parkin-deficient SH-SY5Y cells transduced with a lentiviral construct containing SLP-2. (A) Whole cell lysates of SH-SY5Y cells: scrambled shRNA control, Parkin KD, SLP-2 KD, Parkin + SLP-2 KD, and Parkin KD with SLP-2 overexpression (O/E), were analyzed by WB with the indicated antibodies. Tubulin served as loading control. (B) Percentage of complex I activity in mitochondrial fractions isolated from SH-SY5Y cells. SLP-2 O/E restored the reduced complex I activity in Parkin KD cells. The horizontal red line indicates the reference activity of complex I in scrambled shRNA control cells and was set to 100%; values are means ± SEM. Global P-value Kruskal-Wallis test = 0.011. P-values of the pairwise comparisons between each group and the control are shown (adjusted P-value for significance = 0.006). N = 4, except for SLP-2 KD and Parkin + SLP-2 KD n = 3. (C) Relative percentages of ATP synthesis rate in SH-SY5Y cells with stable Parkin KD, SLP-2 KD, and Parkin KD with SLP-2 O/E cells compared to control cells. SLP-2 O/E increased the reduced ATP synthesis rate in Parkin KD cells. The horizontal line indicates the reference activity of ATP synthesis in control cells; values are means ± SEM. Global P-value Kruskal-Wallis test = 0.076; n = 3. (D) Mitochondrial network morphology in control, stable Parkin KD, SLP-2 KD, and Parkin KD with SLP-2 O/E SH-SY5Y cells. Global P-value of Kruskal-Wallis test < 0.0001 (for pairwise comparisons, adjusted P-value for significance = 0.004). The degree of mitochondrial branching (form factor) was significantly lower in SLP-2 and Parkin-deficient cells compared to the scrambled shRNA control and compared to the Parkin KD with SLP-2 O/E cells (*** P-value Dunn’s test < 0.0001). No statistically significant difference was detected between controls vs Parkin KD with SLP-2 O/E cells and between SLP-2 KD vs Parkin KD. The box plot represents the raw data. Images of 30 randomly selected cells per condition were analyzed. (E–H) Representative examples of images that were analysed are shown. Scale bar = 7.5 μm.
Figure 3
Figure 3
Generation of iPSCs and DA neurons. (A–F) Reprogramming of patient-derived skin fibroblasts harboring a Parkin mutation into iPSC line iPS-B125. (A) Immunofluorescence analysis shows the presence of pluripotency markers OCT4, Tra-1-60, NANOG, and SSEA-4 in iPS-B125. (B) Expression levels of endogenous pluripotency markers NANOG, GDF3, OCT4, and SOX2 in fibroblasts and iPS-B125 relative to β-actin (as loading control) as assessed by quantitative RT-PCR. The values for fibroblasts were set to 1. The error bars indicate the standard deviation. (C) Residual expression levels of transgenes OCT4, SOX2, cMYC, and KLF4 (relative to β-actin) were examined by quantitative RT-PCR. The values for the infected fibroblasts (isolated 7 days after infection) were set to 1. (D) Karyotype of iPS-B125 was normal. (E) Direct sequencing confirmed the Parkin mutation c.1072delT in iPS-B125. (F) RT-PCR analyses of various differentiation markers for the three germ layers (endoderm: AFP, SOX17; mesoderm: MSX1, MHY6; ectoderm: NCAM, PAX6) in iPSCs (iPS) and after 4 days in suspension culture to form embryoid bodies (EB) followed by 7 days in adherent culture. The values are relative to β-actin as assessed by quantitative RT-PCR and normalized to the parental iPSC line (set to 1). (G) Immunofluorescence staining of neuronal cultures derived from iPS-B125, an additional Parkin mutant iPSC line, iPS-L3244 (Supplementary Material, Fig. S4), and the two iPSC control lines, iPS-HFF-wt and iPS-L7659-wt (Supplementary Material, Fig. S4). Staining shows the neuronal marker TUJ1 (red), the DA marker TH (green), and nuclear DAPI (blue). Quantifications were performed for iPS-HFF-wt and iPS-B125 (Supplementary Material, Fig. S5). (H) Neurons from iPS-B125 and iPS-HFF-wt were tested for their electric activity (Supplementary Material, Table S1). They had pronounced voltage-gated Na+/K+ currents, fired evoked multiple and spontaneous action potentials (APs), and showed spontaneous post-synaptic currents (PSC, n = 24 and n= 9 cells analysed, respectively).
Figure 4
Figure 4
Complex I activity and mitochondrial network morphology in iPSC-derived neurons from patients harboring Parkin mutations. (A) Complex I activity in mitochondrial fractions of neuronal cultures derived from three controls, iPS-B125 and iPS-B125 overexpressing SLP-2. SLP-2 O/E restored the reduced complex I activity in Parkin mutant neurons. The horizontal red line indicates the reference activity of complex I in control neurons; values are means ± SEM (P-value Kruskal-Wallis = 0.010; number of observations: control n = 5, Parkin mutant n = 5, Parkin mutant + SLP-2 O/E n = 3). Adjusted P-value for significance = 0.0125. (B) Mitochondrial network morphology is rescued after SLP-2 overexpression in neurons derived from iPS-B125 and iPS-L3244. The degree of mitochondrial branching (mitochondrial form factor) was significantly higher in TH-positive neurons of three healthy control individuals and in the patients’ neurons after SLP-2 overexpression compared to the patients harboring Parkin mutations (controls vs iPS-B125: +57.5%, P < 0.0001; controls vs iPS-L3244: +47.3%, P = 0.002; iPS-B125 after SLP-2 overexpression vs iPS-B125:+59.2%, P < 0.0001; iPS-L3244 after SLP-2 overexpression vs iPS-L3244: +55.8%, P = 0.004). ***P < 0.0001, **P < 0.01. Data were derived from three pooled independent experiments and analysed with a random intercept model. Total number of TH-positive neurons analyzed was: controls = 58, iPS-B125 = 46, SLP-2 overexpression in iPS-B125 = 25, iPS-L3244 = 23, SLP-2 overexpression in iPS-L3244 = 20. The box plot represents the raw data. (C–E) Immunofluoresence staining of DA neurons for the mitochondrial marker GRP-75 (green), TH (red), and SLP-2 (blue). To clearly visualize differences in the morphology by eye, example images shown were taken from third quartile (upper half of the data) for healthy controls and Parkin mutants overexpressing SLP-2, and from the first quartile (lower half of the data) for the Parkin mutants. Scale bar = 10 μm.
Figure 5
Figure 5
SLP-2 genetically interacts with Parkin and rescues parkin mutant phenotypes when overexpressed in the Drosophila model. (A) Mitochondrial morphology visualized by mitoGFP transgene driven by IFM-Gal4 in adult indirect flight muscles of wild type (w1118) and parkin null (park25) flies alone or in combination with Drosophila SLP-2 (dSLP-2) or human SLP-2 (hSLP-2) overexpression transgenes. Indirect flight muscles from thoraces of 3-days old flies were dissected to record mitoGFP pattern. Scale bar = 50 μm. (B) Measurements of ATP levels from thoracic muscles of 5-days old flies expressing transgenes in all tissues under the control of tub-Gal4. The relative levels of ATP were determined by dividing the luminescence by the total protein concentration and normalized to wild-type flies. (C) Flight ability of wild type (w1118), parkin and SLP-2 RNAi fly lines with or without SLP-2 transgenes expressed under the control of tub-Gal4. 1-week old flies were used to perform the flight assay. A minimum of 100 flies were tested each time in two independent experiments. Flight score of 6 indicates a wild type flight function, score of 0 represents a flightless phenotype. (D) Quantification of abnormal wing posture phenotype in wild type, parkin and SLP-2 RNAi flies with or without SLP-2 transgenes. The percentage of flies displaying abnormal wing posture was determined in 2-weeks old F1 males. (E) Quantification of DA neurons in PPL1 cluster. The numbers of DA neurons in the PPL1 cluster visualized by TH > StingerGFP on the posterior side of the brain were counted. (F) Measurement of motor function using climbing assay. 4-weeks old males aged at 25 °C were subjected to climb an 8 cm mark, and the percentage of flies that crossed this mark in 10sec was determined. A minimum of 100 flies were tested in repeated trails as described in the methods section. Transgenes were expressed under the control of TH-Gal4. Statistical differences were calculated by one-way ANOVA followed by Tukey’s post hoc test to correct for multiple comparisons. Data is plotted as ± SEM. ***P ≤ 0.001, **P ≤ 0.01, *P ≤ 0.05.

Similar articles

See all similar articles

Cited by 12 articles

See all "Cited by" articles

MeSH terms

LinkOut - more resources

Feedback