14-3-3 Expression in Denervated Hippocampus after Entorhinal Cortex Lesion Assessed by Culture-Derived Isotope Tags in Quantitative Proteomics Carina Sihlbom,*,†,‡ Ulrika Wilhelmsson,† Lizhen Li,† Carol L. Nilsson,§ and Milos Pekny*,† Center for Brain Repair and Rehabilitation (CBR), Department of Clinical Neuroscience and Rehabilitation, Institute of Neuroscience and Physiology, and Department of Medical Chemistry and Cell Biology, Institute of Biomedicine, Sahlgrenska Academy, Go¨teborg University, Box 440, SE-405 30 Go¨teborg, Sweden, and National High Magnetic Field Laboratory, Florida State University, Tallahassee, Florida 32306 Received February 26, 2007
Activation of astrocytes accompanies many brain pathologies. Reactive astrocytes have a beneficial role in acute neurotrauma but later on might inhibit regeneration. 2D-gel electrophoresis and mass spectrometry were applied to study the proteome difference in denervated hippocampus in wildtype mice and mice lacking intermediate filament proteins glial fibrillary acidic protein (GFAP) and vimentin (GFAP-/-Vim-/-) that show attenuated reactive gliosis and enhanced posttraumatic regeneration. Proteomic data and immunohistochemical analyses showed upregulation of the adapter protein 143-3 four days postlesion and suggested that 14-3-3 upregulation after injury is triggered by reactive gliosis. Culture-derived isotope tags (CDIT) and mass spectrometry demonstrated that 14-3-3 epsilon was the major isoform upregulated in denervated hippocampus and that its upregulation was attenuated in GFAP-/-Vim-/- mice and thus most likely connected to reactive gliosis. Keywords: quantitative proteomics • neuroregeneration • neurodegeneration • hippocampus • astrocytes • brain injury • 14-3-3 protein • isoform • FTICR mass spectrometry
Introduction Astrocytes are important for the maintenance of the CNS homeostasis and recycling of neurotransmitters. They also play a role in the plasticity of CNS, e.g., by controlling the number and function of neuronal synapses1,2 or generation of new neurons.3,4 In many CNS pathologies, astrocytes become reactive (known as reactive gliosis) and this is accompanied by an altered expression of many genes.5,6 Upregulation of intermediate filament proteins GFAP and vimentin by astrocytes is the hallmark of reactive gliosis. Reactive gliosis may facilitate the healing process but may also inhibit regeneration and thus reduce the extent of functional recovery. Several experimental studies have shown that improved regeneration and functional recovery after neurotrauma can be achieved by reducing reactive gliosis.7-9 We and others have previously reported improved regeneration in mice deficient for GFAP and vimentin (GFAP-/-Vim-/- mice, which exhibit attenuated reactive gliosis after neurotrauma), specifically better axonal regeneration,10,11 survival and integration of neural grafts and neural progenitor cells in the retina,12,13 reduced photoreceptor degeneration after retinal detachment,13 or improved synaptic regeneration in the denervated dentate gyrus of the hippocampus15 (for review, see ref 16). * To whom correspondence should be addressed. E-mail, carina.sihlbom@ medkem.gu.se or
[email protected]; Phone, +46-31-7863269; Fax, +46-31-416108. † Department of Clinical Neuroscience and Rehabilitation. ‡ Department of Medical Chemistry and Cell Biology. § Florida State University. 10.1021/pr070108e CCC: $37.00
2007 American Chemical Society
14-3-3 is an adapter protein implicated in the regulation of a large spectrum of both general and specialized signaling pathways. 14-3-3 proteins bind to specific phosphorylated sites on a number of diverse target proteins and force conformational changes or influence protein-protein interactions. For example, the Src homology 2 like domain of 14-3-3 binds to pTyr and the WD40 domain binds to sites containing pThr and pSer residues.17,18 There are seven closely related genes, encoding 14-3-3 protein isoforms β (beta, MW 28.1 kDa), (epsilon, MW 29.2 kDa), η (eta, MW 28.2 kDa), γ (gamma, MW 28.3 kDa), τ (theta, MW 27.8 kDa, also called tau), ζ (zeta, MW 27.8 kDa), and σ (sigma, MW 27.7 kDa) in humans. Phosphorylated isoforms of β and ζ are called R (alpha) and δ (delta), respectively. 14-3-3 proteins were proposed to have a central role in cell proliferation control, survival, cellular trafficking, or actin dynamics.19 Recently, a tandem affinity purification study in transgenic mice characterized protein complex binding to 14-3-3. A total of 147 proteins were identified, including proteins involved in neuronal development, neurotransmitter release, exocytosis, glutamate receptor signaling, and cytoskeleton rearrangements.20 The highest level of 14-3-3 proteins in the body is found in the brain, where it represents around 1% of the total soluble proteins.21 Most 14-3-3 isoforms are present in the adult CNS and show heterogeneous patterns of expression in different cell types and different anatomical locations,22 suggesting specific functions for each isoform. Importantly, the expression pattern is altered in CNS pathologies, e.g., increased expression of 14Journal of Proteome Research 2007, 6, 3491-3500
3491
Published on Web 07/31/2007
research articles 3-3 proteins in reactive astrocytes was found in brain samples from patients with multiple sclerosis,22,23 Creutzfeldt-Jacob’s disease25 or stroke.26 14-3-3 proteins may also be released into the cerebrospinal fluid in patients with neurodegenerative diseases and have recently been used as a diagnostic marker for sporadic Creutzfeldt-Jacob’s disease.27-29 14-3-3 proteins have been identified in neurofibrillary tangles in Alzheimer’s disease30,31 and through mediating the phosphorylation of tau, they may promote their formation.32-34 Similarly, 14-3-3 proteins may contribute to the pathology of Parkinson’s disease: 14-3-3 interacts and colocalizes with alpha-synuclein35-37 in the intracellular inclusions characteristic for the disease and 143-3 eta has been shown to inhibit the activity of parkin ubiquitin ligase.38 Here we have used both classical differential proteomics and a novel application of quantitative proteomics with CDIT to study the protein expression levels in denervated hippocampus after entorhinal cortex lesion in GFAP-/-Vim-/- and wildtype mice. GFAP-/-Vim-/- mice exhibit attenuated reactive gliosis after neurotrauma. We performed classical proteomics by combining 2D-gel electrophoresis (2D-GE) and mass spectrometric analysis to identify differentially expressed proteins. To allow quantitative analysis of specific proteins, mass spectrometry-based quantification with stable isotope labeling was applied. A novel approach based on stable isotope labeling by amino acids in cell culture (SILAC),39-41 uses culture-derived isotope labeled proteins as internal standards for quantitative tissue proteomics.42 Quantitative changes of protein expression in tissue can be estimated by comparing the calculated average ratio of peptide areas from CDIT and tissue peptides as determined with mass spectromery. By combining the CDITbased method and immunohistochemical analysis of contralateral and lesioned brains, we show that 14-3-3 protein is upregulated in denervated hippocampus and that 14-3-3 epsilon is the isoform most responsible for this upregulation. Moreover, we show that upregulation of 14-3-3 is connected with the response of astrocytes to neurotrauma, i.e., with reactive gliosis.
Experimental Section Mice and Surgical Procedures. Mice carrying a null mutation in the GFAP and vimentin gene and the wildtype controls have been described,16,43-46 were on C57Bl/6-129Sv-129Ola genetic background and were 7-12 months old, mixed females and males in both GFAP-/-Vim-/- and wildtype groups of mice, and maintained in a barrier animal facility. All experiments were approved by the ethical committee at Go¨teborg University. Unilateral entorhinal cortex lesion was performed as described previously.15,47-49 Briefly, anesthetized mice were placed in a stereotactic frame, and a hole was drilled through the skull. A retractable wire knife (David Kopf Instruments, Tujunga, CA) was lowered 1 mm down from the dura +3.6 mm laterally and -0.2 mm posterior to lambda. The wire knife was expanded 2 mm horizontally and then lowered 2 mm twice at +30 and -135° to avoid the hippocampal formation. The mice were killed 4 days after the lesion. For proteomic analysis, the complete hippocampus, according to the anatomy described by Franklin and Paxinos,50 was dissected out and stored at -80 °C until analysis. For immunohistochemical analysis, the mice were perfused transcardially with phosphate buffered 4% paraformaldehyde. Hippocampus Sample Preparation for 2D Gel Electrophoresis. Frozen mouse hippocampus, from 6 wildtype mice, 3492
Journal of Proteome Research • Vol. 6, No. 9, 2007
Sihlbom et al.
6 GFAP-/-Vim-/- mice, 6 wildtype mice after injury, and 6 GFAP-/-Vim-/- mice after injury, was homogenized in 460 µL lysis buffer containing 5 M urea, 2 M thiourea, 4% CHAPS, 40 mM Tris base, 100 mM DTT, 1 mM EDTA, 2% immobilized pH gradient (IPG) buffer, and proteinase inhibitor (Complete Mini cocktail tablet, Roche Diagnostics) using a motor-driven grinder pestle. The sample was then sonicated during 1 min and agitated for 1 h. The crude protein extract was centrifuged at 13 000 rpm for 15 min at room temperature. Total protein assays were performed with the Bradford method using Coomassie Plus (Pierce) and measuring absorbance at 595 nm. Bromophenol blue solution (1-2 µL) was added to the hippocampus protein sample. 2D Gel Electrophoresis. Isoelectric focusing was performed with 24 cm immobilized pH gradient (IPG) strips (Immobiline DryStrip pH 3-10 nonlinear, Amersham Biosciences/GE Healthcare) in the Ettan IPGphor IEF System (Amersham Biosciences/ GE Healthcare). Active rehydration was performed at 30 V for 12 h in 450 µL sample solution described above. The focusing step was completed at 55 000 Vhrs with a program of 500 V for 1 h, 1000 V for 1 h and a gradient up to 8000 V during 1 h at 20 °C. The IPG strips were incubated with gentle shaking in equilibration buffer (50 mM Tris-HCl pH 8.8, 6 M urea, 30% glycerol, 2% SDS, bromphenol blue) containing 1% DTT for 15 min and 2.5% iodoacetamide for additional 15 min. The second dimension separation was carried out by use of in-house made 12% polyacrylamide (Bio-Rad) slab gels with the strip laid on the top in the presence of 0.5% agarose prepared in running buffer. Gels were run in the Protean plus Dodeca Cell apparatus (Bio-Rad) at 18 °C and 10 mA/gel for 1 h and then 25 mA/gel overnight in 125 mM Tris, 960 mM glycine, 0.5% SDS. After washing for 1 h in 10% v/v ethanol, 7% (v/v) acetic acid in water, staining was carried out overnight in 350 mL Sypro Ruby Protein stain (Molecular Probes). Gels were destained for 30 min in a 10% v/v ethanol and 7% (v/v) acetic acid solution and maintained in water. Images were acquired with a 2D 2920 Master Imager (Amersham Biosciences/GE Healthcare). The scan parameters were: excitation at 540 nm, emission at 630 nm during 3 s with 16-bit pixel density and image resolution of 150 µm. The optical density of protein spots is proportional to protein concentration. 2D-Gel Image Analysis. Image analysis was carried out with the PD-Quest software (Bio-Rad). Protein levels were evaluated as volumes (spot area optical density) for the protein spots matched among gels representing all groups. Spot volume was normalized for each gel on total density in gel image. Data were log transformed and analyzed with Mann-Whitney statistics tools included in the PD-Quest software. Altered spots were selected using the average of six gels within one group and compared to another group at confidence levels 90 and 99%, respectively. Spots which gave significant results were verified visually to exclude artifacts. Immunohistochemistry. The brains were dissected out, post-fixed in phosphate buffered 4% paraformaldehyde overnight at 4 °C, and immersed in 30% sucrose in phosphate buffer for several days at 4 °C. Horizontal 25 µm sections were cut with a cryostat and stored in cryoprotectant at -20 °C until used for immunohistochemistry as described.15 Primary antibodies used were goat anti 14-3-3 eta (1:50; Santa Cruz Biotechnology), rabbit anti-14-3-3 epsilon (1:100; Santa Cruz Biotechnology), mouse anti-GFAP (1:100; Sigma-Aldrich), and biotin-conjugated mouse anti-NeuN (1:100, Chemicon Europe Ltd), followed by secondary antibodies conjugated with Alexa
14-3-3 Expression in Denervated Hippocampus
Fluor 488 or 568 (Invitrogen) and TO-PRO-3 (Invitrogen) for visualization of nuclei. In two sections or more from each of 10 lesioned mice the cellular distribution of 14-3-3 epsilon immunoreactivity was investigated. Representative images of the findings were obtained using a Leica TCS SP2 laser scanning confocal microscopy (Leica Microsystems GmbH). Mouse Hippocampus Protein Preparation for 1D Gel Electrophoresis. Frozen mouse hippocampus, from 5 wildtype mice, 3 GFAP-/-Vim-/- mice, 4 wildtype mice after injury and 4 GFAP-/-Vim-/- mice after injury, was homogenized in 400 µL lysis buffer containing 50 mM DTT, 25 mM Tris HCl, 35 mM Tris base, 0.5% lithium dodecyl sulfate (LDS), 2.5% glycerol, 12.5 mM EDTA, and proteinase inhibitor using a motor-driven grinder pestle. The sample was then sonicated for 1 min, agitated for 2 h, and heated to about 80 °C during 3 min. The crude protein extract was treated in the same way as in preparation for 2D-GE. Culture-Derived Isotope Tagging (13C-Labeled Astrocyte Culture). Primary astrocyte cultures were prepared from a 2-day old wildtype mouse as described51 and maintained in Dulbecco’s modified Eagle’s medium (Specialty media, Millipore,) deficient in L-leucine, supplied with 13C6-leucine (Cambridge Isotope Laboratories), 10% fetal calf serum (Gibco BRL), L-glutamine (2 mM), and penicillin/streptomycin (Gibco BRL). The cultures containing about 95% astrocytes were grown to confluency in 10 cm Petri dishes and washed 10 times with ice-cold phosphate buffered saline (PBS) prior to harvesting. Cells were scraped in PBS and centrifuged at 500× g for 5 min. After removal of the buffer, the cells were frozen and stored at -80 °C until analysis. The astrocyte pellet was suspended in 350 µL 1D gel lysis buffer as described. 1D Gel Electrophoresis for Quantification of Hippocampal Proteins. The sample volume of total hippocampal protein was adjusted (15-25 µL) to gain an equal loading of 30 µg protein and 8 µL of 13C6-leucine labeled astrocyte lysate was added to all tubes resulting in a 1:1 mixture. 1D gel electrophoresis was performed on NuPAGE Novex Bis-Tris 10% acrylamide gel using MOPS SDS running buffer (Invitrogen) at 200 V, constant voltage. In-Gel Trypsin Digestion. Each selected spot or band was cut with a biopsy punch and processed as described51 with minor modifications. Briefly, the gel pieces were washed twice in 100 µL of H2O/CH3CN (1:1 v/v) for 30 min, dried in a vacuum centrifuge, rehydrated in 10 µL of digestion buffer (25 mM NH4HCO3) containing 12.5 ng/µL sequencing grade trypsin (Promega), and incubated at 37 °C overnight. Any remaining digestion was stopped with 10 µL 1% HCOOH (Merck) in water. The supernatant was collected, and the peptides were extracted twice with 30 µL of 0.1% HCOOH in water/CH3CN (1:1). The combined supernatants were evaporated to dryness in a vacuum centrifuge. Tryptic digests were reconstituted in 18 µL 0.1% HCOOH. Mass Spectrometry and Database Searches. NanoLC-MS/ MS was performed on a C18-fused silica column packed inhouse and connected to a hybrid linear ion trap-Fourier Transform Ion Cyclotron mass spectrometer (FTICR-MS) (LTQ-FT, Thermo Electron), equipped with a 7 T magnet. Sample injections (2 µL) were made (HTC-PAL auto sampler, CTC Analytics AG), and the tryptic peptides were trapped on a C18-precolumn (4.5 cm × 100 µm i.d.) and separated in a 20 cm × 50 µm i.d. fused silica column packed with ReproSil-Pur C18-AQ 3 µm porous particles (Dr. Maisch GmbH). After 3 min linear run, loading the precolumn, the gradient was 0-50%
research articles
Figure 1. Schematic drawing of a horizontal section through hippocampus and the entorhinal cortex (EC) after EC lesion. EC lesion disrupts the perforant path innervating the outer molecular layer (OML, light gray) of the ipsilateral dentate gyrus of the hippocampus. This leads to axonal and synaptic degeneration in the OML, astrocyte activation in the same area and limited synaptic regeneration later on.
CH3CN, starting with HCOOH 0.2% in water for 40 min (flow rate 100 nL/min, Agilent 1100 binary pump), and the eluent was electrosprayed (+1.4 kV) from the emitter tip (20 µm i.d., 150 µm o.d., Polymicro) into the heated capillary of the mass spectrometer. The mass spectrometer was operated in the datadependent mode to automatically switch between MS and MS/ MS acquisition. Survey MS spectra (m/z 400-1600) were acquired with the FTICR-MS, and the three most abundant doubly or triply protonated ions in each FT-scan were selected for isolation, fragmentation and detection in the linear ion trap. The exclusion time for isobaric precursor ions was 6 s. In the analysis of 1D gel samples for protein quantification, an inclusion list of peptides was used to select peptides of special interest. Mass calibration was performed with a mixture, including caffeine, the peptide MRFA, and Ultramark 1621 (m/z 195 to 1822), dissolved in 50:50 water/acetonitrile solution, according to the manufacturer’s recommendation (ESI calibration solution, Thermo Electron). The typical mass accuracy is
