Research data supporting "Gsk3β and Tomm20 are substrates of the SCFFbxo7/PARK15 ubiquitin ligase associated with Parkinson's disease"

Abstract

Supplementary Table S1: Protoarray® v5.0 total protein content. Supplementary Table S2: Negative controls on the Protoarray® v5.0 .Each slide has 3004 negative control spots. Supplementary Table S3: Positive controls on the Protoarray® v5.0. Each slide has 576 positive control spots. Supplementary Table S4: Analysis of the SCF-Fbxo7 E3 ubiquitin ligase: Samples of Fbxo7 ubiquitin ligase were purified from cells and resolved ∼1 cm ona pre-cast SDS–polyacrylamide gel, and the entire lane was excised and cut into four equal slices. Proteins were reduced and alkylated, then digested in-gel using trypsin. The resulting peptides were analyzed by LC–MS/MS using an Orbitrap XL (Thermo) coupled to a nanoAcquity UPLC (Waters). Data were acquired in a DDA fashion with MS/MS in the LTQ triggered at 1000 counts. Raw files were converted into mzML using MSconvert (ProteoWizard) and submitted to MASCOT 2.3.0 to search a human Uniprot database (20 264 entries, downloaded on 09 June 14). Carbamidomethyl cysteine was set as a fixed modification with oxidized methionine and deamidation of asparagine and glutamine as potential variable modifications. Peptide and protein validation were performed using Scaffold 4.3.2. Peptides required a minimum of 95% probability and proteins required a minimum of 90% probability and two peptides in order to be counted. Supplementary Table S5: Substrates of the SCF-Fbxo7 E3 ubiquitin ligase. Protoarray® v5.0 was obtained from Life Technologies (catalog number PAH0525101). Protocols were followed according to the manufacturer's instructions (Protoarray® v5.0, Invitrogen, MA, USA). Slides were incubated in Protoarray® Synthetic Block for 1 h at 4°C with shaking at 50 rpm. During this time, reactions were prepared in a volume of 120 µl as follows: 25 or 50 nM of the purified SCFFbxo7 or Fbxo7(ΔF-box) in combination with ubiquitin mix [E1 (100 nM), UbcH5a (500 nM), Mg-ATP (2 mM), and biotin-ubiquitin 0.1 mg/ml in ubiquitination buffer; Boston Biochem]. The slides were washed with assay buffer (AB; 50 mM Tris, pH 7.5, 50 mM NaCl, 5 mM MgSO4, 0.1% Tween 20, 1% BSA, and 1 mM DTT) and 110 µl of the reaction was added to the slide and overlaid with a coverslip followed by incubation for 1.5 h at 30°C in a humidified chamber. Slides were washed in 0.5% SDS and AB and then incubated with 1 µg/µl of streptavidin–AlexaFluor 647 for 45 min at 4°C with shaking. The arrays were washed with AB, once with distilled water, and finally dried by centrifugation at 1000 × g for 2 min, before being scanned on a GenePix Personal 4100A (Axon–Molecular Devices). Data acquisition and analysis: Software used for Protoarray® image acquisition was GenePix Pro 4.1 (Molecular Devices). The experimental design comprised two biological replicates [25 or 50 nM ligase of wild type (WT) or mutant F-box protein, ΔF] with two intraslide technical replicates. The intensity value of each array feature was considered as the average raw intensity of all pixels in the delimited spot region minus the median intensity of pixels immediately surrounding the spot region (local background). Background-subtracted intensities were subjected to normalization to make them directly comparable among different Protoarray® slides (replicates and conditions). Assuming that the manufacturer-produced positive controls should present the same theoretical intensities among replicates, their background-corrected values were obtained and a single centering (normalization) value per slide was defined as the average of all known positive control spots. An interslide, study-wide overall positive control value was obtained by averaging all the intraslide centering/normalization factors. Finally, all array features had their background-subtracted intensities corrected by this overall factor to end up with comparable normalized intensities (IWT,25 nM, IWT,50 nM, IΔF,25 nM, and IΔF,50 nM). To determine spots with statistically significant signals, we used all the negative control spots to estimate non-parametrically the null density distribution (NC) for each slide using a Gaussian kernel density estimator [35]. The intensity value that encompasses the vast majority of null probability mass was chosen as cutoff (c) for each slide: P(NC > c) = 0.005. This yielded four cutoff values: cWT,25 nM; cWT,50 nM; cΔF,25 nM, and cΔF,50 nM. Selected spots presented intensities IWT > cWT and IΔF log2(5). To increase stringency, only proteins which met these criteria simultaneously on both concentrations were defined as possible substrates. Mean FC values of possible substrates were calculated and used to rank the proteins. The selected list of proteins was subjected to Gene Enrichment analysis using DAVID Bioinformatics Resources v6.7 [36] using Protoarray® v5.0 content (Supplementary Table S1) or the human proteome as the background. DAVID's P-values up to 0.05 were considered significant.