Nov 20, 2015 - cular diseases,6 as well as skeletal dysplasias.7â9 For cell survival, .... The acquired MS/MS data were converted to Mascot Generic ...... with iTRAQ ratios obtained for Exp 2. (116/113). (XLSX). Supplemental method. (PDF).
Article pubs.acs.org/jpr
Label-Free Quantitative Proteomics Reveals Survival Mechanisms Developed by Hypertrophic Chondrocytes under ER Stress Mateusz Kudelko,† Cecilia W. L. Chan,† Rakesh Sharma,†,§ Qing Yao,† Edward Lau,‡ Ivan K. Chu,‡ Kathryn S. E. Cheah,† Julian A. Tanner,† and Danny Chan*,† †
Department of Biochemistry and ‡Department of Chemistry, The University of Hong Kong, Pokfulam, Hong Kong, China Department of Biology and Chemistry, City University of Hong Kong, Tat Chee Avenue, Hong Kong, China
§
S Supporting Information *
ABSTRACT: Emerging evidence implicates ER stress caused by unfolded mutant proteins in chondrocytes as the underlying pathology of chondrodysplasias. ER stress is triggered in hypertrophic chondrocytes (HCs) in a mouse model (13del) of metaphyseal chondrodysplasia type Schmid (MCDS) caused by misfolded mutant collagen X proteins, but the HCs do not undergo apoptosis; rather chondrocyte differentiation is altered, causing skeletal abnormality. How 13del HCs can escape from apoptosis and survive ER stress is not understood. Here we compared the proteomes of HCs isolated from 13del growth plates with normal HCs using a label-free quantitative mass spectrometry approach. Pathway enrichment analyses of differentially expressed proteins showed significant changes in glycolysis and ER-mitochondria pathways in 13del HCs as well as in ATDC5 cell lines expressing wt and 13del collagen X. In vivo, we showed expression of mitochondrial calcium channels was reduced while mitochondrial membrane polarity was maintained in 13del chondrocytes, while in vitro, glucose uptake was maintained. We propose 13del HCs survive by a mechanism whereby changes in ERmitochondria communication reduce import of calcium coupled to maintenance of mitochondrial membrane polarity. These findings provide the initial insights into our understanding of growth plate changes caused by protein misfolding in the pathogenesis of chondrodysplasias KEYWORDS: glycolysis, mitochondrial membrane polarity, calcium signaling, metaphyseal chondrodysplasia type Schmid
■
anomaly.15 Among them, metaphyseal chondrodysplasia-type Schmid (MCDS) is a disease associated with heterozygous mutations in the COL10A1 gene encoding the α1(X) chain of collagen X, affecting folding and trimeric assembly of the collagen X molecule. MCDS is characterized by a mild short stature and hypertrophic zone elongation of the growth plate,16−18 and evidence indicates activation of ER stress as the pathogenesis for MCDS.8,9 Comprehending how cells respond to ER stress has important implications in developmental processes and disease mechanisms. Previous studies have used mainly in vitro cell systems and chemical agents such as tunicamycin and thapsigargin to activate ER stress,19−21 with the limitation that ER stress in vitro normally leads to apoptosis, and changes observed in dying cells may not apply in an in vivo context. Here we use a mouse model (13del) for MCDS, a skeletal dysplasia characterized by endogenous activation of ER-stress in hypertrophic chondrocytes (HCs) of the growth plate.8 13del mice express Col10a1 with a 13-base pair deletion within the NC1 encoding domain of collagen X in hypertrophic chondrocytes. This mutation alters collagen X folding, triple helix assembly, and secretion from the ER resulting in the natural activation of ER-stress in HCs.22,23 Surprisingly,
INTRODUCTION A cell experiences both intrinsic and extrinsic stress conditions during its life. Increased protein folding load in the endoplasmic reticulum (ER stress) is one of the most studied and characterized types of stress that occurs during developmental processes and disease states such as type II diabetes,1,2 neurodegenerative diseases,3,4 Parkinson’s disease,5 cardiovascular diseases,6 as well as skeletal dysplasias.7−9 For cell survival, specific or integrated stress pathways, collectively termed the unfolded protein response (UPR), are activated to alleviate the stress and adapt to the new condition; otherwise, the cell will die. Three mechanistically distinct branches of the UPR have been described with IRE1, PERK, and ATF6 as sensors that are kept inactive by the binding of BiP but are activated when BiP is sequestered away in the presence of unfolded proteins.10,11 Thus, the UPR is a cell-adaptive mechanism that helps cells to survive by restoring ER homeostasis but can also lead to cell death if the ER dysfunction cannot be mitigated and homeostasis cannot be reestablished.12,13 While the general mechanisms that link UPR and apoptosis are well characterized and described today,14 the events that occur during prolonged ER stress remain largely unknown. Skeletal dysplasias are a group of more than 450 clinically distinct pathologies that are caused by abnormal cartilage or bone development and commonly result in dwarfism and bone © 2015 American Chemical Society
Received: June 9, 2015 Published: November 20, 2015 86
DOI: 10.1021/acs.jproteome.5b00537 J. Proteome Res. 2016, 15, 86−99
Article
Journal of Proteome Research
1 h and then alkylated by the addition of 1 μL of 200 mM iodoacetamide (30 min at 25 °C in the dark) following manufacturer’s instruction (AB Sciex, CA). Protein concentrations were estimated using the linear Bradford assay32 after the removal of detergent using Pierce detergent removal resin and columns (Thermo Scientific). 20 μg of dried proteins per sample was reconstituted into 25 mM ammonium bicarbonate, followed by overnight trypsin (Promega) digestion.
activation of ER stress did not lead to apoptosis of the HCs in 13del mice; instead they survive, redifferentiate, and proliferate,8 providing a unique model to address the natural consequences of ER stress in vivo. It has been reported that growth plate chondrocytes survive hypoxic stress during development, probably through the cytoprotective actions of HIF-1α, PERK, and eIF2α.24−26 Furthermore, recent findings showed that some mutations causing ER stress and cell death in vitro may not necessarily induce cell death in vivo, indicating that the in vivo environment provides specific pro-survival signals such as the hypoxic stress pathway.25 Thus, hypertrophic chondrocytes in the growth plate appear to have a distinct response to ER stress that allows survival instead of triggering apoptosis. In the present study, we utilize the 13del mouse model as well as an in vitro cell culture system to investigate the adaptive mechanism(s) developed by HCs to facilitate their survival under an ER stress environment. We undertook an in vivo proteomic approach to unravel the mechanisms of the chondrocytic ER stress response. Cartilage proteomics is challenging due to the limited available tissue and dominance of poorly soluble matrix components.27 Thus, previous studies were based on microarray analysis of microdissected mouse cartilage zones28,29 despite the fact that ER stress also affects translation30 and therefore protein expression. Furthermore, we extend from a study that uses a “temporal” approach based on the analysis of normal mouse femoral head cartilage in different development stages31 that was unable to address changes in the differentiation program of chondrocytes within the growth plate. Using a “spatial” strategy through specific labeling of hypertrophic chondrocytes with GFP in mice and dissection of hypertrophic cartilage, we performed comparative proteomics revealing altered glucose utility and mitochondrial function as part of the survival mechanism.
■
LC−MS/MS Analysis
Each sample was analyzed in triplicate by LC−MS/MS using an UltiMate 3000 HPLC system (Dionex) connected online with an LTQ-Orbitrap Velos (Thermo Scientific, Bremen, Germany). Tryptic peptides were loaded onto a self-packed PicoTip column (360 μm outer diameter, 75 μm inner diameter, 15 μm tip, New Objective) packed with 10 cm length of C18 material (ODS-A C18 5 μm beads, YMC) with a high-pressure injection pump (Next Advance) at 1 μL/min in 98% solvent A (0.1% (v/v) formic acid) and 2% solvent B (0.1% (v/v) formic acid in acetonitrile) for 6 min and subsequently eluted with a linear gradient B from 2 to 40% for 120 min at a flow rate of 300 nL min−1. The LTQ-Orbitrap Velos was controlled using Xcalibur, version 2.0.7 (Thermo Fisher Scientific) and operated in data-dependent acquisition mode, whereby the survey scan was acquired in the Orbitrap with a resolving power set to 60 000 (at 400 m/z). MS/MS fragment ions were detected in LTQ mass analyzer; for each duty cycle, the 20 most intense precursor ions from a survey scan were selected for MS/MS. Ion injection times for the MS and MS/MS scans were 350 and 150 ms, respectively. The automatic gain control targets for MS (FT) and MS/MS (LTQ) were set to 1 million and 10 000, respectively. The mass range for precursor ion selection was selected between 350− 1700 m/z. Fragmentation was carried out using CID mode with 35% normalized collision energy, with activation of 0.25 and 10 ms activation time. Ions selected for fragmentation were dynamically excluded for a period of 30s from MS/MS analysis. For MS/MS analysis, monoisotopic precursor mass was selected and singly charged ions were rejected from analysis. The lock mass was enabled for accurate mass measurements using polydimethylcyclosiloxane (m/z, 445.12) ions.
EXPERIMENTAL DETAILS
Hypertrophic Chondrocyte Dissection and Protein Extraction
13del transgenic mice were crossed to Col10a1-eGFP mice to generate 13del;Col10a1-eGFP. HCs from proximal and distal tibia were obtained from 10-day postnatal (P10) wt-Col10a1eGFP and 13del;Col10a1-eGFP mice in accordance with Institutional Animal Ethics guidelines. The dissected cartilage was rinsed in PBS, frozen immediately into liquid nitrogen, and stored at −80 °C. The P10 HCs were obtained by dissecting the green fluorescent zone from tibia and removing the bone collar under fluorescent microscope. We used five biological replicates per genotype, where each replicate sample comprised HCs from proximal and distal tibia pooled from three to four P10 mice. Each replicate was then cryosectioned and transferred to Eppendorf tubes containing 100 μL of triethylammonium bicarbonate (0.5 M) and 2% (w/v) SDS. Cell disruption was performed by a combination of 10 freeze− thaw cycles followed by sonication (Sonics, Vibra-Cell, USA) on ice for 30 min (10s bursts with intensity ∼40% and 10 s breaks). The suspension was centrifuged at 1000g for 10 min and supernatants were collected.
Criteria for Protein Identification and Validation
The acquired MS/MS data were converted to Mascot Generic format with Proteome Discoverer 1.3 software (Thermo Fisher Scientific). All MS/MS data were analyzed using MASCOT version 2.2.06 (Matrix Science) against UniProtKB mouse database of canonical sequences (Oct 2014; 16 480 entries) and sequences for common contaminants appended (downloaded from the Max Planck Institute; http://maxquant.org). By default, Mascot decoy database search was performed for the entire data set. Enzyme specificity was set to trypsin with a maximum of two missed cleavages: S-carboxamidomethylation of cysteine residues specified as a fixed modification and oxidation of methionine specified as variable modification. Parent ion and fragment ion mass tolerance were set to 10 ppm and 0.5 Da, respectively. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium33 via the PRIDE partner repository with the data set identifier PXD002125. The Mascot search results (.dat files) were incorporated in Scaffold 4 (version 4.4.3, Proteome Software) for protein validation and to assign probabilities to peptide and protein matches.34 Peptide-spectrum matches were accepted if the
Protein Reduction, Alkylation, and In-Solution Trypsin Digestion
Protein samples for LC−MS/MS analysis were sequentially reduced by adding 2 μL of 50 mM tris(2-carboxyethyl) phosphine (TCEP) to each sample and incubating at 60 °C for 87
DOI: 10.1021/acs.jproteome.5b00537 J. Proteome Res. 2016, 15, 86−99
Article
Journal of Proteome Research
Figure 1. Strategy for spatial extraction of tibia proximal and distal hypertrophic cartilage and identification of differentially expressed proteins. Hypertrophic cartilage were extracted and perichondrium removed from wt-Col10a1-eGFP and 13del; Col10a1-eGFP proximal and distal tibia, respectively. Immunofluorescence image of extracted HCs (eGFP+) and transmission electron micrographs of representative HCs from wt and 13del are presented (* = polysomes; scale bars represent 20 (panels a and d), 10 (panel e), 5 (panel b), and 1 μm (panels c and f)). Proteins were extracted and trypsin digested from wt-Col10a1-eGFP and 13del; Col10a1-eGFP-labeled HCs before being loaded and analyzed by tandem mass spectrometer (LC−MS/MS). Five biological replicates (n = 5) for each condition (wt and 13del) were analyzed in triplicate. Each technical triplicate was recombined in silico using the MudPIT function of Scaffold for the analysis. Data were searched using Mascot for protein identification. Protein quantitation and validation was performed by spectral counting using Scaffold software. Criteria used to select differentially expressed proteins: Fold change (FC) FC ≥ 2 and Fisher’s Exact test (p ≤ 0.05).
peptide was assigned a probability >0.95, as specified by the Peptide Prophet algorithm.35 Protein identifications were accepted if the protein contained at least two unique peptide counts, and the protein was assigned a probability >0.99 by the Protein Prophet algorithm and the FDR was
