Integrated Proteomic Analysis Reveals a ... - ACS Publications

Nov 16, 2009 - Department of Biomedical Sciences, University of Antwerp, B-2610 Antwerp, Belgium, Department of Medical Protein Research, VIB, B-9000 ...
0 downloads 0 Views 2MB Size
Integrated Proteomic Analysis Reveals a Substantial Enrichment of Protein Trafficking Processes in Hippocampus Tissue after Hypoxic Stress Roos Van Elzen,† Bart Ghesquie`re,‡,§ Evy Timmerman,‡,§ Stefaan Vandamme,| Luc Moens,† Kris Gevaert,‡,§ and Sylvia Dewilde*,† Department of Biomedical Sciences, University of Antwerp, B-2610 Antwerp, Belgium, Department of Medical Protein Research, VIB, B-9000 Ghent, Belgium, Department of Biochemistry, Ghent University, B-9000 Ghent, Belgium, and Department of Biology, University of Antwerp, B-2610 Antwerp, Belgium Received June 12, 2009

Acute and chronic hypoxic episodes of the brain have been generally recognized as a common denominator of several neuropathologies of which cerebral ischemic stroke represents one of the leading causes of mortality and morbidity. In an attempt to clarify the plethora of molecular events elicited by ischemic or hypoxic stress, several studies have reported before on large-scale expression analysis; however, only a minority have put focus on proteome based changes. To further enrich our knowledge, we investigated the differences in protein levels following prolonged exposure of mice to global hypoxic stress in the hippocampus, one of the most susceptible regions of the brain. This was accomplished using the conventional 2-DE approach and peptide-centered quantitative methionyl-COFRADIC. Together both methods resulted in the identification of 110 unique hypoxia regulated proteins, and 34 posthypoxic reoxygenation regulated proteins based on 2-DE analysis alone. The integration and comparison of the implicated biological functions with other large-scale analyses of similar hypoxia and ischemic stroke models gave an overall resemblance of implicated biological processes apart from model specific alterations in distribution. Nevertheless, further examination of the data clearly depicted a substantial enrichment of protein trafficking and targeting processes in our data which could be related to synaptic plasticity and remodeling events. Keywords: Brain • hypoxia • proteomics • two-dimensional electrophoresis • COFRADIC • peptide-centered

Introduction Hypoxia is a widely occurring condition necessary for regulating developmental processes of organs and tissues, but detrimental to cells and organisms in the absence of proper antagonizing regulatory processes. Several pathologic and environmental conditions exist that give birth to acute or chronic depletion in oxygen, such as pulmonary disease, cardiac failure, orthologic hypotension, stroke, obstructive sleep apnea, and living on high altitude. All of these conditions can have adverse repercussions on the structural and functional integrity of the brain, which is generally viewed as one of the most susceptible organs to insufficiencies in oxygen supply. The lack in energy substrates generally leads to the slow down and finally complete arrest of ATP synthesis by oxidative phosphorylation. In combination with a high ATP consumption rate and low alternative energy supplies, this leads to a rapid decline in cellular ATP concentration.1 Simultaneously ongoing anaerobic glycolysis leads to intracellular acidification and * To whom correspondence should be addressed. E-mail: Sylvia.dewilde@ ua.ac.be. † Department of Biomedical Sciences, University of Antwerp. ‡ Department of Medical Protein Research, VIB. § Department of Biochemistry, Ghent University. | Department of Biology, University of Antwerp.

204 Journal of Proteome Research 2010, 9, 204–215 Published on Web 11/16/2009

transient increased ROS generation, which are both stressful factors for most cells. Of these, it is the drastic loss in ATP and the ensuing malfunction of ATP-dependent processes that makes brain tissue so sensitive to oxygen deficits.2 In answer to this, the affected tissue and cells react with a complex and multifaceted cascade of physiological and biochemical events which determine the fate of the affected cells on short-term and long-term. Within seconds to minutes of reduced oxygen supply, a drastic reorganization of channel activities, rearrangement of metabolic pathways, massive neurotransmitter release, and overall suppression of energy consuming processes has been observed.3,4 On a longer term, cells react by modulating the expression of genes and proteins implicated in metabolism, structural, pro- as well as antiapoptotic, inflammatory, and stress related events. Important modulators of these miscellaneous molecular cascades are HIF-1R, CREB, and NFkB transcription factors.5–7 Together, these cellular damaging and protective mechanisms determine the fate of the cell. The exact sequence of events and vulnerability to the energy deficit depends greatly on the intrinsic resistance of a certain brain region and even cell type. Brain structures like hippocampus and frontal cortex belong to the most vulnerable regions of the 10.1021/pr900517m

 2010 American Chemical Society

Expression Analysis of the Hypoxic Response brain, each with a specific dynamic response of protective and harmful effects.8 Despite this knowledge, none of the efforts to provide an efficient treatment have been proven unequivocally efficacious. This is in part due to the limited knowledge on the activated signaling pathways. Given the large number of events occurring during an ischemic cascade of brain injury, it is likely that multiple targets should be placed in focus for an efficient treatment.9 The simultaneous monitoring of altered expressions in answer to a hypoxic deficit of whole sets of genes and proteins has the power to provide an overview on the implicated molecular pathways and potential targets. Since the pioneering work of Soriano and co-workers,10 several groups have endeavored a similar approach to further disentangle the complex spatially and temporally regulated molecular response on transcriptional,11–26 translational27–34 and even posttranslational level.35–37 Each of these studies has contributed on some level to the overall elucidation of the major implicated molecular pathways in the hypoxic response. Nonetheless, a careful interpretation of these results is necessary since a wide heterogeneity of in vivo and in vitro experimental models was found between different large-scale expression studies, which mainly focused on gene transcription levels. Together with the fact that there is no direct correlation between gene transcription and downstream translational and protein post-translational events, the previous described observations clearly stress the need for proteomic studies on well-defined tissue-regions or cells of robust experimental models.38 We here report on a differential proteome analysis of the hippocampus-specific effects following a sustained global hypoxic treatment of mice. To enhance the comprehensiveness of the analysis, alterations in protein expression following hypoxic stress were mapped by use of two-dimensional electrophoresis (2-DE) and the peptide-centered quantitative COmbined FRActional DIagonal Chromatography (COFRADIC) technique.39 To assess the similarity of the observed alterations with previously characterized molecules, a comparison was done with 34 reported transcriptomic and proteomic studies of the cerebral hypoxic/ischemic response. In a second phase, the expression of a subset of “consensus molecules”, likely to be significantly regulated upon ischemic/hypoxic stress, was monitored using Western blot. These results from proteomic and computational analyses allowed us to envision the regulatory network that may mediate oxygen regulation of protein expression in mouse hippocampus.

Experimental Procedures In Vivo Hypoxic Conditions. Female mice littermates (age ∼12 weeks) were placed in a hypoxic chamber perfused with a 7% O2 gas mixture (premixed gases: 7% O2, 93% N2, Air Liquide, Belgium). Mice were provided with food and water ad libitum and allowed to adapt to the hypoxic environment over a 1 h period. Animals were sacrificed by cervical dislocation either immediately after 48 h hypoxia (hypoxia) or 24 h after readapting to normal atmospheric conditions (posthypoxic reoxygenation). Mice exposed to a normal atmosphere (21% O2) were chosen as a control group. Immediately after sacrificing the mice, hippocampal tissue was dissected out of the brains, frozen in liquid nitrogen and stored at -70 °C. All procedures were approved by the Local Ethics Committee of the University of Antwerp and conformed to European Community regulations. Extraction of Hippocampal Proteins. Prior to protein extraction, the brain tissue was crushed in liquid nitrogen and

research articles desalted with 10% trichloroacetic acid and three subsequent acetone precipitation steps. Dry samples were homogenized with a glass pestle in (A) 2-DE lysis buffer (8 M urea, 2 M thiourea, 2% CHAPS, and complete protease inhibitor cocktail (Roche Diagnostics GmbH, Penzberg, Germany)) for the 2-DE analysis or (B) COFRADIC compatible lysis buffer (4 M of guanidinium hydrochloride in 50 mM sodium phosphate buffer pH 7.5 supplemented with a protease inhibitor cocktail) and left for 1 h on an orbital shaker. The extracts were clarified by 30 min centrifugation at 10 000g. This procedure was done twice after which the clarified supernatant fractions were pooled and quantified with the 2-DE compatible RC DC protein assay (Bio-Rad) or the BCA assay kit (Thermo Fisher Scientific) for the COFRADIC samples. For 2-DE analysis, a total of 15 samples were analyzed (5 normoxia, 5 hypoxia, and 5 reoxygenation treated mice). The quantitative methionyl-COFRADIC analysis was done on the pooled extracts of 3 normoxia and 3 hypoxia treated mice in order to allow swap labeling without introducing additional variability. 2-DE Analysis. A total of 600 µg of soluble proteins was supplemented with 260 mM DTT and 2% carrier ampholytes (pH 4-6.5 and pH 3-10; GE Healthcare). Completed samples were loaded on 24 cm IPG strips with a 3-10 pH nonlinear gradient (GE Healthcare) by overnight rehydration. Five IPG strips, representing proteome preparations from 5 individual mice, were analyzed for each condition. The first separation was performed on a Multiphor II system (GE Healthcare) at 20 °C using a gradual increase in voltage to reach 72 kVh. After IEF, the IPG strips were equilibrated 2 × 15 min, first in a reducing buffer (50 mM Tris-HCl pH 8.8, 6 M urea, 30% glycerol, 2% SDS and 130 mM DTT), then alkylated in the same buffer containing 135 mM iodoacetamide instead of DTT. The second electrophoretic separation was done on 12.5% SDSPAGE gels in the Ettan Dalt six electrophoresis system (GE Healthcare) and took approximately 5 h to completion. Visualization of the separated proteins was established through colloidal Coomassie blue staining,40 and analysis of digitized images with Image Master 2D Platinum v.5 software (GE Healthcare). Automatic spot detection and matching of the gels was done, followed by manual rechecking of the matched and unmatched protein spots. The intensity volumes of individual spots were normalized with the total intensity volume of all spots present in each gel (%V). Differences of g1.5 in expression (ratio %V) between matched spots were considered significant whenever a spot group passed statistical analysis (t test, p e 0.05) and a second manual verification of the spots on the gel images. In-Gel Digestion and Protein Identification. Protein spot identification was carried out using an AB4800 Proteomics Analyzer (Applied Biosystems, Foster city, CA). Differentially expressed spots were excised from the 2-DE gels with a sterile filter tip. The gel pieces were washed sequentially with water, 50% acetonitrile (ACN) (Mallinckrodt Baker, Deventer, Netherlands) in 25 mM ammonium bicarbonate until they were completely destained. After dehydration with 100% ACN, the dry gel pieces were rehydrated on ice with 12.5 ng/µL proteomics grade trypsin (Roche Diagnostics GmbH) in digest buffer (25 mM ammonium bicarbonate pH 8.5, 2% ACN), and after 45 min, the remaining fluid was replaced by 10 µL of digest buffer and digestion was carried out at 37 °C for 16 h. Trypsin activity was stopped by acidifying samples to pH