DJ-1 is indispensable for the S-nitrosylation of Parkin, which maintains function of mitochondria

The DJ-1 gene, a causative gene for familial Parkinson’s disease (PD), has been reported to have various functions, including transcriptional regulation, antioxidant response, and chaperone and protease functions; however, the molecular mechanism associated with the pathogenesis of PD remains elusive. To further explore the molecular function of DJ-1 in the pathogenesis of PD, we compared protein expression profiles in brain tissues from wild-type and DJ-1-deficient mice. Two-dimensional difference gel electrophoresis analysis and subsequent analysis using data mining methods revealed alterations in the expression of molecules associated with energy production. We demonstrated that DJ-1 deletion inhibited S-nitrosylation of endogenous Parkin as well as overexpressed Parkin in neuroblastoma cells and mouse brain tissues. Thus, we used genome editing to generate neuroblastoma cells with DJ-1 deletion or S-nitrosylated cysteine mutation in Parkin and demonstrated that these cells exhibited similar phenotypes characterized by enhancement of cell death under mitochondrial depolarization and dysfunction of mitochondria. Our data indicate that DJ-1 is required for the S-nitrosylation of Parkin, which positively affects mitochondrial function, and suggest that the denitrosylation of Parkin via DJ-1 inactivation might contribute to PD pathogenesis and act as a therapeutic target.


Two-dimensional fluorescence difference gel electrophoresis (2D-DIGE)
After measuring the weight of the mouse brain samples, we added ten volumes of lysis buffer (10 mM Tris (pH 8.0), 7 M urea, 2 M thiourea, 5 mM magnesium acetate, 4% (w/v) CHAPS, 4 mM protease inhibitor) and homogenized and sonicated the samples. After centrifugation at 8000 × g for 10 min, the supernatant was used for 2D-DIGE analyses.
Proteomic comparison of brain tissues between three pairs of wild-type and DJ-1 -/mice was performed by 2D-DIGE analyses using a pooled sample, the equivalent mixture of all samples, as the internal standard. The samples were labeled using 400 pmol of fluorescent dye reagents (IC3-OSu or IC5-OSu) per 50 µg of protein extract. Individual samples and the internal standard were labeled with IC5 and IC3, respectively. The labeling reaction was performed at 37 ºC for 20 min and quenched by the addition of 1 mM ethanolamine at room temperature (RT) for 10 min in the dark.
Samples containing 15 µg of protein each were loaded in analytical gels, and samples containing 200 µg of protein each were loaded in preparative gels for protein identification. 2D-PAGE was conducted using immobilized dry strip gels with a pH range of 4-7, and SDS-PAGE (%T = 7.5 and %C = 3.0) was performed as described previously with slight modifications S1 on duplicates of each individual sample.
After 2D-PAGE, gels were scanned on a Typhoon FLA 9500 laser scanner (GE Healthcare), and IC3-and IC5-labeled protein spots were detected at excitation/emission wavelengths of 550/570 nm and 649/670 nm, respectively. Duplicate images of each individual sample and the standard sample were uploaded into Progenesis SameSpots gel image analysis software ver. 4.5 (Totallab, Newcastle, UK), and automated inter-gel alignment, spot detection, normalization, and comparative determination were performed. The computer-detected spots were listed by fold change and statistical significance for the comparison of the normalized spot volumes between the wild-type and knockout groups. We considered a spot to be significantly altered between the groups when the fold change was greater than 1.5 or if p < 0.05 by ANOVA.

Spot picking and in-gel trypsin digestion
Differentiated protein spots were mechanically picked from 2 preparative gels using Ettan Spot Picker (GE Healthcare).
In-gel digestion was performed on the picked gel spots using an XL-TrypKit according to the manufacturer's instructions, and the tryptic peptides from each spot were subjected to liquid chromatography with matrix-assisted laser desorption ionization (LC-MALDI) analysis and protein identification.

LC-MALDI analysis
LC-MALDI analyses were carried out using the DiNa-MaP direct nanoLC and MALDI fraction system (KYA Technologies, Tokyo, Japan) and a MALDI-TOF/

Protein identification
Protein identification was performed via an MS/MS ion search using ProteinPilot™ software (ver.4.5; SCIEX). The search parameters were as follows: database, uniprot_sprot_can+iso_20100622; species, mouse; Cys alkylation, iodoacetamide; digestion, trypsin; and special factors, Gel-based-ID and max missed cleavage, 1. Protein identifications were considered to be correct based on the following selection criteria: proteins with a protein score (ProtScore) of > 1.3 (unused, p < 0.05, 95% confidence) and proteins having at least 2 peptides with an ion score above the 95% confidence threshold.

Gene ontology (GO) analysis
Differentially expressed proteins between WT and KO were further analyzed for gene ontology analysis using the PANTHER classification system (http://www.pantherdb.org). S2

Pathway analysis
To extract molecular networks biologically relevant to differentially expressed proteins, pathway analysis was performed using KeyMolnet ver. 6.2 (KM Data, Tokyo, Japan). S3 The "Protein ID" list of differentially expressed proteins was uploaded into KeyMolnet, and the "Interaction search" algorithm was used to generate the network of molecular interactions in two paths from the starting points, including direct activation/inactivation, transcriptional activation/repression, and complex formation. The canonical pathways associated with differentially expressed proteins in the brain tissues of DJ-1 -/mice were extracted from the KeyMolnet knowledge base.
Another pathway analysis was performed using the Kyoto Encyclopedia of Genes and Genomes (KEGG) (https://www.genome.jp/kegg/pathway.html), which is a publicly accessible knowledge base that covers a wide range of pathway maps on metabolic, genetic, environmental, and cellular processes, and human diseases S4 . By uploading the list of "Protein ID" of differentially expressed proteins, KEGG extracts relevant pathways composed of the proteins enriched in the given list, followed by statistical evaluation.

Confocal microscopy
Cells were fixed with 4% paraformaldehyde in PBS and permeabilized with 0.2% Triton X-100 in PBS. Cells stained with the appropriate antibodies were imaged using a laser scanning microscope (FV1000, Olympus, Tokyo, Japan) fitted with a UPlanSApo 40×/0.9NA lens (Olympus, Tokyo, Japan). To quantify the compaction index of mitochondria, midplane images of cells immunostained for Tom20 were obtained as described previously S5 . In ImageJ (NIH), data in the Tom20 channel were converted into binary data, and the area and perimeter of mitochondria within the cell of interest (selected using the "region of interest" tool) were measured using the "analyze particles" function. The compaction index (which is the perimeter of a circle with the same area as the object of interest divided by the actual perimeter of the object of interest) was calculated from the perimeter (P) and the area (A) using the following formula: (2π × ((A/π) 1/2 ))/P.

Construction of Plasmids
To construct targeting vectors, fragments were amplified from genomic DNA purified from SH-SY5Y cells by Polymerase chain reaction (PCR) and inserted into pMulti-ND 1.0 vector (a kind gift from Dr. J. Takeda and K. Horie at Osaka University and Nara Medical University) S6 . Construct for ingle cysteine mutant of DJ-1 were generated using the PrimeSTAR Mutagenesis Basal Kit (Takara, Otsu, Shiga, Japan) and appropriate primers following manufacturer's instructions. TALEN plasmids were constructed as described previously S7 .

Establishment of cell lines
SH-SY5Y cells were transfected with targeting vector and TALEN plasmids as described in the Materials and Methods.
Selection and maintenance of transfectants were performed in the presence of hygromycin and/or puromycin for DJ-1 -/cells and zeocin and/or G418 for Parkin C323S cells. Colonies were resuspended and grown in 96-well plates at a density of about one cell per well. Genotyping of each cell line was performed by PCR and Southern blotting of genomic DNA.
After genotyping, cells were transfected with plasmids expressing cre recombinase to remove the selection marker, and then were resuspended in 96-well plates at a density of about 1/10 cell per well to be cloned.

Southern blot analysis
Southern blot analysis was performed using the DIG system (Merck KGaA, Darmstadt, Germany) according to the manual.